Proefschrift Hop

Improving burn care efficiency

M. Jenda Hop


M.J. H

op

Uitnodiging

voor het bijwonen van de openbare verdediging

van het proefschrift

‘Improving burn care efficiency’

door M. Jenda Hop

donderdag 12 november

11:45

in de aula Vrije Universiteit Amsterdam

De Boelelaan 1105

na afloop bent u van harte welkom

op de receptie ter plaatse

paranimfen: Rende Hop

[email protected]

en

Marianne [email protected]

Ook wil ik u graag uitnodigen voor het promotiefeest vanaf 20:30 in Pont 13

Haparandadam 501013 AK Amsterdam


Martine Jenda Hop

Financial support for the printing of this thesis was kindly provided by:Stichting Brandwonden Research Instituut, Stichting Plastische Chirurgie Onderzoek Friesland MCL, Plastische, reconstructieve- en handchirurgie VUmc, Nederlandse Vereniging voor Plastische chirurgie, Nederlandse Brandwonden Stichting, and the Maasstad Academie.

ISBN: 978-94-6233-095-5

Lay out and printing by: Gildeprint

VRIJE UNIVERSITEIT


ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad Doctor aande Vrije Universiteit Amsterdam,

op gezag van de rector magnificusprof.dr. F.A. van der Duyn Schouten,

in het openbaar te verdedigenten overstaan van de promotiecommissie

van de Faculteit der Geneeskundeop donderdag 12 november 2015 om 11.45 uur

in de aula van de universiteit,De Boelelaan 1105

door

Martine Jenda Hop

geboren te Zwolle

promotor: prof.dr. E. Middelkoopcopromotoren: dr. M.E. van Baar dr. S. Polinder

Overige leden promotiecommisie:Prof. Dr. R.S. BreederveldProf. Dr. A. BurdorfProf. Dr. S. MonstreyDr. C.M. Moues-VinkProf. Dr M.W. van TulderProf. Dr. P. M.N. Werker

Contents

Chapter 1 General introduction and outline of the thesis 9

Part I. Overview of burn care costsChapter 2 Costs of burn care: a systematic review 23 Wound Repair Regen. 2014 Jul-Aug;22(4):436-50. Chapter 3 Economic burden of burn injuries in the Netherlands: 51 a 3 months follow-up study Injury. Accepted. Chapter 4 Reconstructive surgery after burns: a 10-year follow-up study 69 Burns. 2014 Dec;40(8):1544-51.

Part II. Improving burn care: diagnostics and costsChapter 5 Cost-effectiveness of laser Doppler imaging in burn care in 89 the Netherlands Study protocol BMC Surg. 2013 Feb 1;13:2.Chapter 6 Cost-effectiveness of laser Doppler imaging in burn care in 103 the Netherlands; a randomised controlled trial PRS. In press.Chapter 7 Photographic assessment of burn size and depth: reliability and validity 121 J Wound Care. 2014 Mar;23(3):144-5, 148-52.

Part III. Improving burn care: therapeutics and costsChapter 8 Early excision and grafting for burns 139 Cochrane review. In revision.Chapter 9 Cost study of dermal substitutes and topical negative pressure 175 in the surgical treatment of burns Burns. May;40(3):388-96.

Chapter 10 General discussion 191

Addendum Summary 207 About the author 211 Bibliography 213 Dankwoord 215

Chapter 1General introduction

R1R2R3R4R5R6R7R8R9

R10R11R12R13R14R15R16R17R18R19R20R21R22R23R24R25R26R27R28R29R30R31R32R33R34R35R36R37R38R39

10 | Chapter 1

R1R2R3R4R5R6R7R8R9R10R11R12R13R14R15R16R17R18R19R20R21R22R23R24R25R26R27R28R29R30R31R32R33R34R35R36R37R38R39

General introduction | 11

General introduction

Health economicsHealthcare costs are high and are increasing rapidly in the last decades. In 2012, healthcare expenditures were over €90 billion in the Netherlands, which was more than 10% of our gross domestic product1,2. It is expected that by 2040 healthcare costs will be 19-31% of our gross domestic product2. Next to aging of the population and raising cost prices (e.g. personnel costs), the development of new technologies (e.g. surgical techniques and medication) plays an important role in the rise of healthcare costs. Because of high costs and limited budgets, policy makers are highly interested in the effectiveness and efficiency of new technologies in healthcare: are new interventions more effective than the others, and if so, what are the cost consequences2? In case of higher costs, is improvement of care by the new interventions considered to be worth the extra costs? Economic studies provide the necessary insights into the optimal use of resources to maximize efficiency in healthcare3,4. To convince policy makers to introduce new and better interventions, adequate economic evaluation studies are needed. In this thesis, we will focus on improving the efficiency of burn care.

Burn injuriesEach year more than 11 million people suffer from burn injuries worldwide. Burns and fires still cause an estimated 265 000 deaths yearly throughout the world5, the vast majority occur in low- and middle-income countries. In the Netherlands, approximately 6300 people per year are treated at an emergency department because of burns, 1800 people per year are admitted to a hospital6 because of burns including a total of 750 people treated in one of the three Dutch burn centres, yearly7. The morbidity and mortality of patients with burn injuries decreased in the past decades, due to prevention programmes and several improvements in treatment. Examples of major improvements in burn care of the last decades are the introduction of silver-containing topical antimicrobials, early excision and grafting, shock prevention and the multidisciplinary approach to burn care8-10. Overall burn centre mortality rate in the Netherlands was 6.9% between 1996-2006 and decreased to 3.2%11 between 2006-201112. Still, burn injuries cause significant morbidity. A literature review on functional outcome after burns concluded that up to 37-42% of the patients had problems with their appearance, over 10% had problems with activities of daily life (ADL) and a permanent incapacity for work was reported in 0.7-15.2% of patients13. We can conclude that burn injuries represent an important health care problem. Despite the improvements in burn care over the past decades, survivors of burn injuries experience still major physical, mental and social problems. Therefore, there is an on-going need for quality improvement of burn care.

R1R2R3R4R5R6R7R8R9


12 | Chapter 1

Economic burden of burns Burn care is traditionally considered expensive care. Costs are high because patients with burns often need specialized burn centre treatment during a substantial length of stay, including time- and material-intensive surgical and non-surgical wound care, intensive care and long periods of rehabilitation14,15. Also non-healthcare costs, like productivity loss can be significant after burn injuries16,17. Described predictors for high costs are fire/flame burns and a high total body surface area (TBSA) burned18,19. In order to answer questions on cost-effectiveness of new burn care technologies, first of all a good insight into the extent of burn care costs in general should be obtained. To interpret the existing literature on burn care costs, a systematic evaluation of the quality of methodology and reporting of results is important. Several narrative literature reviews on the costs and quality of burn care have been published in recent years14,20,21. The reviews discussed various options for achieving burn care cost reductions, including the use of less expensive dressing materials, and early excision and grafting14,20. These reviews were narrative and did not systematically review study quality and outcome. To evaluate the true extent of burn care costs and the quality of study methods and presentation of results, we performed an extensive systematic review in which we aimed to evaluate all cost studies and economic evaluations in burn care published from 1950 to 2012.To complement existing knowledge gaps, new cost studies in burn care should be designed. The existing literature on burn care costs14,20,21 does not give satisfying answers on the extent of burn care costs, on predictors for high costs or expensive cost items. A complicating factor in cost studies in burn care is the long-term follow-up necessary to describe costs of rehabilitation, reconstructive surgery, psychosocial care and productivity loss. In this thesis, we attempted to give an extensive overview of the total health care and productivity costs due to burn injuries by both a prospective and retrospective cohort study. The prospective cohort study aimed to give a detailed overview of short-term and long-term health care and productivity costs due to burn care in the Netherlands. Since no detailed information was available of the most important unit costs, detailed burn centre unit prices for both ICU and non-ICU burn centre days were determined by the bottom-up approach, in which an extensive inventory of all resource used per patient is made22. The retrospective study was performed with a ten-year follow-up to identify the epidemiology of reconstructive surgery after burn injuries and the long-term costs of reconstructive surgery. In current literature, epidemiologic information on the prevalence of reconstructive surgery after burns is scarce. One relatively old retrospective study, from the USA in 1991, described that reconstructive surgery was performed in 19.9% of patients admitted to a burn centre23. A recently published article from our research group presented a prevalence of reconstructive surgery of 5.3% in patients admitted to a burn centre with facial burns24. In this thesis we performed a comprehensive study on the epidemiology of reconstructive surgery after burn injuries, to create a baseline for future studies that focus on improvement of scar quality and a decreased need for reconstructive surgery.



Burn care diagnostics An early and accurate diagnosis of burn size and depth is important in determining the most optimal burn treatment. In daily practice, burn size is clinically assessed using the Lund and Browder Charts, rules of nine and/or hand rule25-27. Inexperienced assessors often estimate burn size incorrectly, which leads to significant errors in fluid resuscitation and expensive over-triage28-32. Burn depth is generally assessed clinically as well; the diagnosis relies on an evaluation of several wound features such as appearance, capillary refill and sensibility33. Clinical assessment of burn depth, even by an experienced assessor, is accurate in only 2/3 of the cases33. Inaccurate burn depth assessment can lead to suboptimal choice of topical treatment and to unnecessary surgery or an unnecessary delay in surgery29,34-38. Because of the serious consequences of inadequate burn size and depth assessment, it is important to improve burn care diagnostics.In this thesis we will focus on two upcoming methods to improve burn diagnostics; laser Doppler imaging (LDI) and telemedicine. LDI is a method to assess burn depth early and accurately. In a literature review of Monstrey et al., on current modalities for burn depth diagnosis33 several techniques were presented, varying from biopsy and histology to perfusion measurement techniques such as thermography, vital dyes, video angiography, video microscopy and laser Doppler techniques. Laser Doppler imaging is the technique with the largest weight of evidence in the assessment of burn depth: in recent years, several studies presented laser Doppler imaging to be an accurate predictor of burn depth: clinical assessment combined with laser Doppler imaging has an accuracy of >95%36,39-42. In daily practice, LDI will be used in combination with standard clinical assessment39, as a so-called add-on test43. Up to now, the therapeutic impact of the introduction of LDI in burn care was only presented in one retrospective cohort study37 and one prospective non-randomised study38. In these studies a shorter time to surgical decision, a lower rate of surgical interventions and a reduced length of hospital stay were presented. No prospective studies are available on the costs and the possible cost reductions of LDI in burn care, nor are cost-effectiveness studies. Therefore, we can conclude that is it still unclear whether LDI actually influences diagnostic and therapeutic decisions, patient outcomes and costs, and thus adds to the quality of care. Before we decided whether LDI should be introduced in Dutch burn care, a randomised controlled trial (RCT) on its effectiveness and cost-effectiveness was desired and is presented in this thesis.Telemedicine, in other words, the practice of medicine at distance by using photographs, telephone, or live interactive video, offers the possibility to improve the burn assessment, both size and depth, of inexperienced clinicians (e.g. clinicians at emergency departments or general practitioners). The accuracy of the initial assessment of a burn wound depends on the experience of the assessor33,44. By using telemedicine, inexperienced clinicians are able to receive advise from burn experts. Therefore, telemedicine is a technique that has the potential to improve the initial burn diagnosis and care outside the walls of the burn centres.

R1R2R3R4R5R6R7R8R9


14 | Chapter 1

Unfortunately, literature on the diagnostic abilities of telemedicine in the assessment of burn size and depth is incomplete30. In this thesis we evaluated the reliability and validity of using photographs of burns to assess both their size and depth, to verify whether it is a useful tool in burn care.

Treatment of burnsNext to improvement of burn diagnostics, also the search for improvements of burn treatment is on-going. Burn wound treatment can be roughly divided in non-operative and operative care. Superficial burns heal with few or no functional and aesthetical problems and can be adequately treated non-operatively with various dressings45. Deep partial and full thickness burns, however, can result in aesthetical and functional problematic scars. In deep partial thickness burns, the epidermis and most of the dermis is destroyed, hypertrophic scarring may occur if re-epithelialisation (wound healing) does not occur within two to three weeks46,47. These burns are often treated surgically. In full thickness wounds, all layers of the skin are destroyed, leaving little chance of healing form the epithelial elements in the bottom of the wound. Unless the burn is very small, full thickness burns need to be treated surgically. In practice, burn wounds are often of mixed depth, which complicates the determination of optimal timing and treatment choices.First, we will focus on the optimal timing of surgery. In the United States, there seems to be consensus on early excision and grafting of deep partial thickness and full thickness burns48, while in Europe, usually, a more conservative approach is applied; surgery is postponed until inevitable. Only a few years ago, two burn experts, dr. G.I.J.M. Beerthuizen from the Netherlands and dr. A. Kay from the UK, debated on this subject on an international burn congress. Arguments for early excision and grafting were: a reduced mortality, and a shorter length of hospital stay, and arguments against: a greater volume of blood loss and the risk of overzealous excision49. So, despite the progress in burn care, there is still no worldwide consensus on the best timing of surgery in burns, furthermore, there is no insight in the effects of early excision and grafting on costs. In this thesis we aimed to provide the best evidence on early excision and grafting of burns by performing a Cochrane literature review.The development of new technologies, like surgical techniques, plays an important role in the rise of healthcare costs. The most applied surgical technique to treat burn wounds is tangential excision and split skin grafting. With tangential excision all non-viable tissue is removed layer by layer, preserving as much viable tissue as possible48. After tangential excision follows grafting. In burn care often split skin grafts are used, which consist of epidermis and some, but not the whole, dermis. It is hypothesised that the lack of dermal tissue, plays an important role in the scar formation after burn surgery. The possible solution for this problem is the use of dermal substitutes in combination with skin grafts, which is recommended increasingly in the treatment of acute burns and in burn scar reconstruction50,51. Our research group also



investigated the effectiveness of the use of dermal substitutes in recent years52. In daily practice, reluctance was noticed in the use of dermal substitutes, because of the high costs of this product. But we did not have insight in the overall cost consequences of the use of dermal substitutes in burn care. Therefore, we aimed to comprehensively analyse the medical and non-medical costs of patients with burns that were surgically treated with dermal substitutes and split skin grafts in combination with topical negative pressure. The results of this economic evaluation are presented in this thesis.

Cost-effectiveness analyses in burn careNew interventions in burn care can be costly, but the cost price of an intervention should not be the only point of consideration, also effectiveness has to be taken into account. An improved quality of care can be worth the extra costs. Next to that, initially expensive diagnostics and treatments can reduce costs on the long term by e.g. reducing hospital length of stay or enable an earlier return to work. Because in our economic climate budgets are limited, we cannot permit to focus on a high quality of care against any cost. Therefore, we should constantly search for a better quality of care in a cost-effective manner14,15. In health economics, balancing costs and effects is performed in cost-effectiveness analyses, using Incremental Cost Effectiveness Ratio’s (ICER’s) or cost minimisation analyses in case of equal effectiveness between the different interventions4. Economic evaluation studies provide information to policy makers on those interventions with the most favourable balance between costs and healthcare, because resources- people, time, facilities, equipment, and knowledge are scarce3,4. To be able to implement new technologies and interventions, data on cost-effectiveness are crucial. Also in burn care, traditionally classified as expensive, there is a growing interest in the costs and cost effectiveness of care53. Surprisingly, to our knowledge, only 3 cost-effectiveness studies in burn care were published until this thesis54-56. All these studies compared different topical treatment strategies. The clinical outcomes used were number of patients healed at day 21 post burn54,55 and time to 50% wound healing56. Calculated costs included mean total hospital costs56 or mean total dressing costs54,55. All studies found that the intervention was cost-effective. The studies were of moderate methodological quality, with only limited costs analysis and short-term outcomes involved. In the design of cost-effectiveness analyses in burn care, both costs and the effect outcome measure should be carefully chosen. The search for a ‘core set’ of outcomes after burn injuries is ‘a live issue’ in the Dutch burn centres, several burn experts/researchers are currently examining which outcomes should be included in the core set of outcomes. Different effect outcome measure can be used in economic evaluations, e.g. the primary clinical outcome, quality of life, or quality adjusted life years (QALY). A highly relevant clinical outcome parameter after burn injuries is scar quality. Scar quality, can be measured with several

R1R2R3R4R5R6R7R8R9


16 | Chapter 1

patient and/or observer reported measures, like the POSAS and Vancouver Scar Scale, and objective measures like the dermaspectrometer to measure skin colour and pigmentation or the cutometer® to measure skin elasticity57-60. An indirect way to asses scar quality is the need for reconstructive surgery after burns. Reconstructive surgery is an important part of the rehabilitation phase after burns61,62. When we are able to reach an improved scar quality by optimal care in both the acute and rehabilitation phase of burn care, we will hopefully notice a decreased requirement for reconstructive surgery in the rehabilitation phase.

Aims of this thesis To improve burn care efficiency we should aim for optimal burn care, by improving our diagnostics and treatment. Improvements in burn care, as in all health care, must be balanced against total costs of burn care, therefore, a good insight in costs and the performance of economic evaluations is of paramount importance. The aim of this thesis was to search for methods to optimise efficiency in burn care, by:

- Identifying the healthcare and non-healthcare costs of burn care, important costs items and predictors for high costs in burn care

- Evaluation of the effectiveness and cost-effectiveness of (new developments in) burn diagnostics and therapeutics

Outline

Part I. Overview of burn care costsTo elucidate the current knowledge on burn care costs, an extensive systematic literature review was designed, which is presented in Chapter 2. The objectives of this review were 1) to assess the methodological quality of economic studies in burn care, 2) to present the range of medical costs, and the non-medical costs of burn care, and 3) to present economic evaluation studies of burn care.Armed with the information of our review, a thorough prospective cohort cost study was conducted, which aimed to augment existing knowledge gaps by giving a detailed overview of all costs of patients with burns admitted in a burn centre in the Netherlands covering the period of acute care and the first period of rehabilitation in patients with burns. This study has a follow-up of 2 years. In Chapter 3 of this thesis the first results of this study, from injury to 3 months post burn, are described. In our search for identification of important cost items, also long-term follow-up costs needed to be included. Chapter 4 describes the results of a 10-year follow-up study on reconstructive surgery in burn patients admitted to the three Dutch burn centres. The objectives of this study were to analyse 1) the prevalence for reconstructive surgery after burn injuries, 2) the predictors



for reconstructive surgery, 3) the indications and techniques of these reconstructions and 4) the medical costs of reconstructive surgery after burns.

Part II. Improving burn care efficiency: diagnostics and costs Facilitated by the results presented in Part I, several new cost studies and economic evaluations in burn care were designed. Part II of this thesis focuses on new technologies in the initial diagnosis of burn wounds and related costs. The technique ‘laser Doppler imaging’, used for an early, accurate diagnosis of burn depth, is the main connection between the chapters in Part II.To decide whether LDI should be implemented in Dutch burn care, we designed a randomised controlled trial. The aim of this trial was to analyse the effectiveness and cost-effectiveness of LDI in burn care. The study design of this RCT is explained in Chapter 5, in Chapter 6 the results are presented. Photographs obtained from the LDI trial were used for a study on telemedicine presented in Chapter 7. The aim of this study was to examine the reliability and validity of the simplest form of telemedicine, photographic assessment of burn size and burn depth by both burn centre experts and referring physicians, in order to determine whether this type of telemedicine can be used as a diagnostic tool and supports therapeutic decisions.

Part III. Improving burn care efficiency: therapeutics and costsIn the final part of this thesis two studies on improving surgical treatment of burns are presented. In Chapter 8 the results of a systematic Cochrane review are presented, the objective of this review was to assess the effects of early excision and grafting on scar quality in people with burns of all depths.New interventions in burn care can be costly, but can be cost reducing on long term by e.g. reducing rehabilitation costs, next to that costly interventions can be judged to be worth the extra costs when improving the quality of burn care. The aim of the study presented in Chapter 9 was to comprehensively analyse the medical and non-medical costs of patients with burns that were surgically treated with dermal substitutes and split skin grafts in combination with topical negative pressure, in a multicentre randomised controlled trial.

In the last chapter, Chapter 10, all results are summarised and discussed and recommendations for future research directions are proposed.

R1R2R3R4R5R6R7R8R9


18 | Chapter 1

References

1. Centraal bureau voor de statistiek. http://statline.cbs.nl/Statweb/. Dec 2014.2. Van der Horst A, van Erp R, de Jong J. Trends in gezondheid en zorg. Nov, 2011.3. Husereau D, Drummond M, Petrou S, Carswell C, Moher D, Greenberg D, Augustovski F, Briggs

AH, Mauskopf J, Loder E; CHEERS Task Force. Consolidated Health Economic Evaluation Reporting Standards (CHEERS) statement BMJ. 2013 Mar 25;346:f1049.

4. Drummond MF, Sculpher MJ, Torrance GW, O’Brien BJ, Stoddart GL. Methods for the Economic Evaluation of Health Care Programmes. New York: Oxford University Press; 2005.

5. World Health Organization. WHO fact sheet on burns. www.who.int/mediacentre/factsheets/fs365/en/updated. Apr 2014

6. Draisma JA Brandwonden Ongevalscijfers. Amsterdam Veiligheid NL, May, 2014.7. Dokter J, Vloemans AF, Beerthuizen GI, van der Vlies CH, Boxma H, Breederveld R, Tuinebreijer

WE, Middelkoop E, van Baar ME; Dutch Burn Repository Group. Epidemiology and trends in severe burns in the Netherlands. Burns. 2014 Nov;40(7):1406-14.

8. Barrow RE, Herndon DN. History of treatments of burns. Total Burn Care, 3rd ed.Philadelphia, PA: Saunders-Elsevier 2007:1-8.

9. Herndon DN, Barrow RE, Rutan RL, Rutan TC, Desai MH, Abston S. A comparison of conservative versus early excision. Therapies in severely burned patients. Ann.Surg. 1989;5:547.

10. Herndon DN, Barrow RE, Kunkel KR, Broemeling L, Rutan RL. Effects of recombinant human growth hormone on donor-site healing in severely burned children. Ann.Surg. 1990;4:424.

11. Bloemsma GC, Dokter J, Boxma H, Oen IM. Mortality and causes of death in a burn centre. Burns. 2008 Dec;34(8):1103-7.

12. Dokter J, Felix M, Krijnen P, Vloemans JF, Baar ME, Tuinebreijer WE, Breederveld RS; Dutch Burn Repository Group. Mortality and causes of death of Dutch burn patients during the period 2006-2011. Burns. 2015 Mar;41(2):235-40.

13. van Baar ME, Essink-Bot ML, Oen IM, Dokter J, Boxma H, van Beeck EF. Functional outcome after burns: a review. Burns. 2006 Feb;32(1):1-9.

14. Stavrou D, Weissman O, Winkler E, Millet E, Nardini G, Tessone A, Zmora N, Mushin OP, Haik J. Managing the relationship between quality and cost-effective burn care. Burns. 2011 May;37(3):367-76.

15. Pereira C, Murphy K, Herndon D. Outcome measures in burn care. Is mortality dead? Burns. 2004 Dec;30(8):761-71. Review.

16. Sánchez JL, Perepérez SB, Bastida JL, Martínez MM. Cost-utility analysis applied to the treatment of burn patients in a specialized center. Arch Surg. 2007 Jan;142(1):50-7.

17. Sanchez JL, Bastida JL, Martínez MM, Moreno JM, Chamorro JJ. Socio-economic cost and health-related quality of life of burn victims in Spain. Burns. 2008 Nov;34(7):975-81.

18. Latenser BA, Miller SF, Bessey PQ, Browning SM, Caruso DM, Gomez M, Jeng JC, Krichbaum JA, Lentz CW, Saffle JR, Schurr MJ, Greenhalgh DG, Kagan RJ. National Burn Repository 2006: a ten-year review. J Burn Care Res. 2007 Sep-Oct;28(5):635-58.

19. Shields BJ, Comstock RD, Fernandez SA, Xiang H, Smith GA. Healthcare resource utilization and epidemiology of pediatric burn-associated hospitalizations, United States, 2000. J Burn Care Res. 2007 ov-Dec;28(6):811-26.

20. Bass MJ, Phillips LG. Economics of pediatric burns. J Craniofac Surg. 2008 Jul;19(4):888-90. 21. Mandal A. Quality and cost-effectiveness--effects in burn care. Burns. 2007 Jun;33(4): 414-7. 22. Gold MR. Siegel JE, Russel LB, Weinstein MC. Cost-effectiveness in health and medicine. New

York: Oxford University Press; 1996.23. Prasad JK, Bowden ML, Thomson PD. A review of the reconstructive surgery needs of 3167

survivors of burn injury. Burns. 1991 Aug;17(4):302-5.24. Hoogewerf CJ, van Baar ME, Hop MJ, Bloemen MC, Middelkoop E, Nieuwenhuis MK.Burns to the

head and neck: Epidemiology and predictors of surgery. Burns. 2013 Sep;39(6):1184-92. 25. Nagel TR, Schunk JE. Using the hand to estimate the surface area of a burn in children. Pediat

Emerg Care. 1997; 13:254–5.



26. Lund CC, Browder NC. The estimation of areas of burns. Surg Gynecol Obst. 1944;79:352-8.27. Wallace AB. The exposure treatment of burns. The Lancet. 1951;257(6653):501–4.28. Baartmans MG, van Baar ME, Boxma H, et al. Accuracy of burn size assessment prior to arrival

in Dutch Burn centres and its consequences in children: a nationwide evaluation. Injury. 2012 Sep;43(9):1451-6.

29. Saffle JR. Telemedicine for acute burn treatment: the time has come. J Telemed Telecare. 2006; 12(1):1-3.

30. Wallace DL, Hussain A, Khan N, Wilson YT. A systematic review of the evidence for telemedicine in burn care: with a UK perspective. Burns. 2012 Jun;38(4):465-80.

31. Freiburg C, Igneri P, Sartorelli K, Rogers F. Effects of differences in percent total body surface area estimation on fluid resuscitation of transferred burn patients. J Burn Care Res. 2007;28:42– 48.

32. Hagstrom M, Wirth GA, Evans GR, Ikeda CJ. A review of emergency department fluid resuscitation of burn patients transferred to a regional, verified burn centre. Ann Plast Surg 2003;51:173–6.

33. Monstrey S, Hoeksema H, Verbelen J, et al. Assessment of burn depth and burn wound healing potential. Burns. 2008 Sep;34(6):761-9.

34. Shokrollahi K, Sayed M, Dickson W and Potokar T. Mobile phones for the assessment of burns: we have the technology. Emerg Med J. 2007 November; 24(11): 753–755.

35. Niazi ZB, Essex TJ, Papini R, et al. New laser Doppler scanner, a valuable adjunct in burn depth assessment. Burns 1993; 19(6):485–9.

36. Jeng JC, Bridgeman A, Shivnan L, et al. Laser Doppler imaging determines need for excision and grafting in advance of clinical judgment: a prospective blinded trial. Burns 2003; 29(7):665–70.

37. Petrie N, Norbury W, Fogarty B, et al. The use of laser Doppler imaging to reduce operative intervention in the treatment of paediatric burns. In Abstract book, 12th congress of the International Society for Burn Injuries. Yokahama: ISBI; 2004.

38. Kim LH, Ward D, Lam L, Holland AJ. The impact of laser Doppler imaging on time to grafting decisions in pediatric burns. J Burn Care Res 2010; 31(2):328–332.

39. Hoeksema H, Van de Sijpe K, Tondu T, Hamdi M, Van Landuyt K, Blondeel P, Monstrey S: Accuracy of early burn depth assessment by laser Doppler Imaging on different days post burn. Burns 2009, 35(1):36–45.

40. Pape SA, Skouras CA, Byrne PO. An audit of the use of laser Doppler imaging(LDI) in the assessment of burns of intermediate depth. Burns. 2001 May;27(3):233-9.

41. Riordan CL, McDonough M, Davidson JM, Corley R, Perlov C, Barton R, Guy J, Nanney LB. Noncontact laser Doppler imaging in burn depth analysis of the extremities. J Burn Care Rehabil. 2003 Jul-Aug;24(4):177-86.

42. Monstrey SM, Hoeksema H, Baker RD, Jeng J, Spence RS, Wilson D, Pape SA. Burn wound healing time assessed by laser Doppler imaging. Part 2: validation of a dedicated colour code for image interpretation. Burns. 2011 Mar;37(2):249-56.

43. van den Bruel A, Cleemput I, Aertgeerts B, Ramaekers D, Buntinx F: The evaluation of diagnostic tests: evidence on technical and diagnostic accuracy, impact on patient outcome and cost- effectiveness is needed. J Clin Epidemiol 2007, 60(11):1116–22.

44. Berkebile BL, Goldfarb IW, Slater H. Comparison of burn size estimates between prehospital reports and burn centre evaluations. J Burn Care Rehabil. 1986 Sep-Oct;7(5):411-2.

45. Wasiak J, Cleland H, Campbell F, Spinks A. Dressings for superficial and partial thickness burns. Cochrane Database Syst Rev. 2013 Mar 28;3:CD002106.

46. Cubison TCS, Pape SA, Parkhouse N. Evidence for the link between healing time and the development of hypetrophic scars (HTS) in paediatrich burns due to scald injury. Burns 2006;32 (8):992-9.

47. Deitch EA. Wheelahan TM, Rose MP, Clothier J, Cotter J. Hypertrophic burn scars: analysis of variables. Journal of Trauma 1983;23 (10): 895-8.

48. Herndon 2007. Total Burn Care. 3rd edition. Philadelphia: Saunders Elsevier, 2007. 49. Hop MJ, Hoogewerf CJ, van Baar ME, van der Vlies CH, Middelkoop E. A call for evidence: timing

of surgery in burns. Burns. 2012 Jun;38(4):617-8. 50. Bloemen MC, van Leeuwen MC, van Vucht NE, van Zuijlen PP, Middelkoop E. Dermal substitution

in acute burns and reconstructive surgery: a 12-year follow-up. Plast Reconstr Surg. 2010 May;125(5):1450-9.

R1R2R3R4R5R6R7R8R9


20 | Chapter 1

51. Ryssel H, Germann G, Kloeters O, Gazyakan E, Radu CA. Dermal substitution with Matriderm(®) in burns on the dorsum of the hand. Burns. 2010 Dec;36(8):1248-53.

52. Bloemen MC, van der Wal MB, Verhaegen PD, Nieuwenhuis MK, van Baar ME, van Zuijlen PP, Middelkoop E. Clinical effectiveness of dermal substitution in burns by topical negative pressure: a multicenter randomized controlled trial. Wound Repair Regen. 2012 Nov-Dec;20(6):797-805.

53. Hemington-Gorse SJ, Potokar TS, Drew PJ, Dickson WA. Burn care costing: the Welsh experience. Burns 2009;35(May (3)):378–82

54. Caruso DM, Foster KN, Blome-Eberwein SA, Twomey JA, Herndon DN, Luterman A, Silverstein P, Antimarino JR, Bauer GJ. Randomized clinical study of Hydrofiber dressing with silver or silver sulfadiazine in the management of partial-thickness burns. J Burn Care Res. 2006 May-Jun;27(3):298-309.

55. Silverstein P, Heimbach D, Meites H, Latenser B, Mozingo D, Mullins F, Garner W, Turkowski J, Shupp J, Glat P, Purdue G. An open, parallel, randomized, comparative, multicenter study to evaluate the cost-effectiveness, performance, tolerance, and safety of a silver-containing soft silicone foam dressing (intervention) vs silver sulfadiazine cream. J Burn Care Res. 2011Nov-Dec;32(6):617-26

56. Carayanni VJ, Tsati EG, Spyropoulou GC, Antonopoulou FN, Ioannovich JD. Comparing oil based ointment versus standard practice for the treatment of moderate burns in Greece: a trial based cost effectiveness evaluation. BMC Complement Altern Med. 2011 Dec 1;11:122.

57. Draaijers LJ, Botman YA, Tempelman FR, Kreis RW, Middelkoop E, van Zuijlen PP. Skin elasticity meter or subjective evaluation in scars: a reliability assessment. Burns. 2004 Mar;30(2):109-14.

58. Draaijers LJ, Tempelman FR, Botman YA, Kreis RW, Middelkoop E, van Zuijlen PP. Colour evaluation in scars: tristimulus colorimeter, narrow-band simple reflectance meter or subjective evaluation? Burns. 2004 Mar;30(2):103-7

59. van der Wal MB, Tuinebreijer WE, Bloemen MC, Verhaegen PD, Middelkoop E, vanZuijlen PP. Rasch analysis of the Patient and Observer Scar Assessment Scale(POSAS) in burn scars. Qual Life Res. 2011 May 20

60. Sullivan T, Smith J, Kermode J, McIver E, Courtemanche DJ. Rating the burn scar. J Burn Care Rehabil 1990;11:256–60.

61. Barret JP. Burns reconstruction. BMJ. 2004 Jul 31;329(7460):274-6.62. Serghiou M, Cowan A, Whitehead C. Rehabilitation after a burn injury. Clin Plast Surg. 2009

Oct;36(4):675-86.

Part I. Overview of burn care costs

Chapter 2Costs of burn care: a systematic review

M. Jenda HopSuzanne Polinder

Cornelis H. van der Vlies Esther Middelkoop

Margriet E. van Baar

Wound Repair Regen. 2014 Jul-Aug;22(4):436-50.

R1R2R3R4R5R6R7R8R9


24 | Chapter 2

Abstract

Burn care is traditionally considered expensive care. However, detailed information about the costs of burn care is scarce, despite the increased need for this information and the enhanced focus on healthcare cost control. In this study, economic literature on burn care was systematically reviewed to examine the problem of burn-related costs.Cost or economic evaluation studies on burn care that had been published in international peer-reviewed journals from 1950-2012 were identified. The methodology of these articles was critically appraised by two reviewers, and cost results were extracted. A total of 156 studies met the inclusion criteria. Nearly all of the studies were cost studies (n=153) with a health care perspective (n=139) from high-income countries (n=127). Hospital charges were often used as a proxy for costs (n=44). Three studies were cost-effectiveness analyses. The mean total healthcare cost per burn patient in high-income countries was $88,218 (range $704- $717,306, median $44,024).A wide variety of methodological approaches and cost prices was found. We recommend that cost studies and economic evaluations employ a standard approach to improve the quality and harmonisation of economic evaluation studies, optimise comparability and improve insight into burn care costs and efficiency.


Costs of burn care: a systematic review | 25

Introduction

Burn care is traditionally considered expensive care. This assumption is supported by a recent study by Sanchez et al., who found that the mean annual cost of burn patient treatment in Spain was $99,773 compared with $13,826 for the mean annual cost of treatment for stroke survivors during their first year post-stroke and $13,823 for annual care for HIV/AIDS patients1. In today’s economic climate, it is important to attempt cost reductions in healthcare. Burn care costs are thought to be high because patients with burns often need specialised burn centre treatment during a substantial length of stay, including time- and material-intensive surgical and non-surgical wound care, intensive care and long periods of rehabilitation2. Good insight into the extent of burn care costs should be obtained prior to attempts to improve burn care cost-effectiveness. Health economics provide the necessary insights regarding community resource allocation to maximise efficiency in healthcare3,4. Different types of economic studies can be performed. Two types of studies are generally identified: cost studies and economic evaluation studies. Cost studies can provide valuable insight into the distribution of costs and allow the development of specific cost-reducing measures. Cost-of-illness (COI) studies, a specific type of cost study, aim to measure and identify the societal costs associated with a particular disease5. Economic evaluation studies primarily aim to evaluate the outcomes and costs of different interventions that are designed to improve health. Economic evaluation studies provide information on the interventions that provide the most favourable balance between costs and health effects4. The optimal use of resources is important because health care is increasingly expensive and budgetary pressures are rising.It is challenging to clearly report the results of health economic evaluations. This spring, the Health Economic Evaluation Reporting Standards (CHEERS) statement guidelines were published as a tool for optimising health economic evaluation reporting. This document attempted to update the previous health economic evaluation guidelines and consolidate them into one current, useful source of reporting guidance3 that could be applied to economic studies of burn care. Central issues include the systematic reporting of methodology (e.g., study perspective, time horizon, and cost calculations) and results (incremental costs and outcomes and the characterization of uncertainty and heterogeneity). Several narrative literature reviews on the costs and quality of burn care have been published in recent years2,6,7. One review focused on paediatric patients7, whereas the other reviews discussed the relationship between quality and cost-effective burn care2,6. The available reviews did not follow the CHEERS recommendation that the methodology of the included studies should be systematically assessed. The reviews did discuss several options for achieving burn care cost reductions without compromising the quality of care, including the use of ambulances instead of helicopters, the use of less expensive dressing materials, and early excision and grafting2,6.

R1R2R3R4R5R6R7R8R9


26 | Chapter 2

Again, no systematic information discussed the quality of the results of the included studies.There is a lack of systematic information on the methodological quality of cost studies and economic evaluations in the area of burn care. No overview of the extent or distribution of burn care costs or of the global variation in these costs is available. To define the problem of burn-related costs, a systematic literature review was performed in this study to evaluate all of the existing economic literature on burn care. The objectives of this review were 1) to assess the methodological quality of economic studies on burn care, 2) to present the range of medical costs and non-medical costs of burn care and 3) to present economic evaluation studies of burn care. Costs per treatment and per day compared with non-burn care costs were included in the analyses.

Methods

Search strategySearches of eligible studies were conducted in Medline (PubMed) and EMBASE. The search terms used in Medline included burns (MesH Terms), burn* (Title/Abstract), scald* (Title/Abstract), thermal NEXT injur* (Title/Abstract), cost and cost analysis (MesH Terms), and cost* (Title/Abstract). The search terms used for EMBASE were burn*, scald*, and cost* or cost allocation*. Reference lists of included articles (published from 2002-2012) were screened for relevant articles.

Selection criteriaThe included studies met the following criteria:

• Burn care cost studies or economic evaluation studies in which costs per patient were reported in the methods or results section

• Empirical studies in all market economies that were published in international peer-reviewed journals during the period from Jan 1950 to May 2012 (reviews were excluded)

• Studies published in English, French, German or Dutch

Data extractionRelevant papers were selected by screening the titles (first step), abstracts (second step) and entire articles (third step) that were retrieved using the database and manual searches. During each respective step, the title, abstract or entire article was screened to ensure that the article met the selection criteria. The screening was conducted independently by two researchers (MJH and MEB). Reviewer disagreements about article eligibility were resolved through discussion with a third researcher (SP). The full articles were extracted by two researchers



(MJH and MEB). The assessment of the quality of the economic evaluations was based on the cost-utility analysis reporting checklist recommended by the Panel on cost-effectiveness in health and medicine5 and based on the guidelines developed for economic submissions to the British Medical Journal (BMJ).Furthermore, cost information in the following categories was identified:

• Total healthcare costs per burn patient • Unit price per hospital per inpatient day• Healthcare costs for burn patients versus other patients• Economic evaluation studies: cost, effect, and incremental cost-effectiveness

ratioIn addition, detailed information on costs in high-income countries (www.worldbank.org) was identified:

• Costs related to injury characteristics: aetiology and Total Body Surface Area (TBSA) burned

• The cost distribution (%) of different healthcare cost componentsThe two independent reports were compared, and disagreements were resolved through discussion with a third researcher (SP). To enable the comparison of the costs presented in the included studies, all costs were converted to 2012 US dollars using currency and inflation correction. The mean and median total healthcare costs per patient and the hospital day prices were calculated using the data from all of the studies that indicated total burn care costs or hospital day prices, respectively. The costs prices per 1 % TBSA burned were calculated by dividing the total health care costs per patient by the mean TBSA burned; figures were taken from studies with both TBSA and total costs.

Results

Literature searchThe Medline search yielded 1072 articles, and 702 additional articles were found in Embase. Finally, two articles were retrieved by manual searching for a total of 1776 unique articles. Title screening resulted in the selection of 780 articles that appeared to meet all of the selection criteria. In the second phase, abstract screening resulted in the exclusion of 489 articles. Of the remaining 291 articles, 135 did not meet the inclusion criteria after the papers had been fully read; thus, the final sample includes 156 articles. Three full texts could not be retrieved. The main reasons for exclusion were failure to address burn care or burn patients and the absence of a description of the study’s cost calculations in the methods or results.

R1R2R3R4R5R6R7R8R9


28 | Chapter 2

Study characteristics: patients and methodologyThe first study included in this review was published in 1967. The studies were performed in various market economies, but predominantly in high-income countries (n=127). Most studies were conducted in burn centres in North America. The focus was primarily on general burn patients (n=122). However, some studies focused on paediatric burn patients (n=23) or compared burn patients with other patients (n=11). Some studies identified the TBSA, and the mean TBSA burned (n=91) ranged from 1-95%; most of these studies focused on more severe burns (Table 1, and Supplemental Information).Cohort studies (n=107) were the most common in this review, followed by randomized controlled trials (n=24). The most frequently performed type of economic study was the cost study (n=153). Although the most extensive and preferred economic approach is that of the societal study that takes into account costs to all members of a society4, only two studies employed this perspective1,8. Most studies employed the healthcare perspective (n=139). The time horizon of most cost calculations was <1 year, and only acute burn care costs were generally examined; the costs during the rehabilitation phase were generally neglected.The relevant costs in these studies can be sorted into four categories: direct medical costs (e.g., the cost of hospital stays), indirect medical costs (e.g., the cost of care during life years gained), direct non-medical costs (e.g., traveling costs), and indirect non-medical costs (e.g., productivity loss). The majority of studies (n=143) only included direct medical costs. The included cost components varied widely; some articles calculated dressing costs only9-24, whereas other studies calculated total medical and non-medical burn care costs1-8. The most studied cost component was hospitalisation (n=96), followed by dressing (n=34), medication (n=28) and surgery (n=22). Preferably, detailed information on cost calculations should be available. The cost price data of specific cost units, primarily hospital day prices (n=34), were presented in 44 studies. The data source, e.g., hospital charges (n=44) or the hospital finance department, was presented in 101 studies. Most studies presented the total costs per patient without subdividing these costs into more detailed healthcare sources. Information on resource use (the volume of healthcare consumption) was presented in 36 studies. The data source, resource use and cost price data were presented in only 17 studies. The cost price calculation methods of “top-down” and “bottom-up” indicate important choices regarding unit cost calculation. In top-down cost calculations, the financial administration data of the healthcare provider is the primary source that is used to determine unit costs, e.g., the price of one hospital day is calculated by dividing the annual hospital costs by the number of hospital days. In bottom-up calculations, which are appropriate for heterogeneous groups such as groups of burn patients, unit costs are determined by measuring the actual use of personnel, materials and equipment for a single patient25. Eighteen studies used a bottom-up approach to determine cost prices. Four studies used a top-down approach, and the approach was not described in most studies.



Table 1. Study characteristics: patients and methodology (n=156) 1,8-24, 26-109, 114-167

Study characteristics studies (n) study characteristics studies (n)Study continent Time horizon cost analysis

North America 91 < 1 year 151South America 1 1-5 years 1

Europe 25 6-10 years 2Asia 24 Lifelong 2

Africa 9 Cost categoriesAustralia 6 Direct medical costs 143

Setting Direct medical and non-medical costs 5Burn centre/unit 116 Direct medical and non-medical costs and

indirect non-medical costs2

Trauma centre 2 Direct medical and indirect non-medical costs

6

ICU 2 Cost components*General Hospital 27 Initial transport to hospital 4

Others 9 Intramural careType of patients Hospitalization** 96

General burn patient 122 Surgical treatment 22Paediatric patient 23 Dressings 34

Burn patient vs. other patient 11 Medication 28Mean TBSA burned Pressure garments 3

0-10% 27 Allied health professionals 611-20% 21 Diagnostics 14

>20% 43 Reconstructive surgery 2Not described 65 Emergency department 4

Study design Outpatient hospital care 13Randomized clinical trial 24 Productivity loss

Insurance claimsPatient costs, direct and productivity loss

224

Controlled clinical trial 4 Availability information calculation costs

Case control 12 Price data 44Cohort study 107 Data sources 101

Case study 9 Resources (healthcare consumption) 36Type of economic study Data source and resources 31

Economic evaluation: cost-effectiveness analysis

3 Data source and resource and price data 17

Cost study 153 Approach: bottom up/ top downEconomic perspective Bottom up 18

Societal 4 Top-down 4Healthcare 139 Approach not described 134Insurance 8Individual 5

*Diverse components per study are possible**With various definitions: often including all hospital healthcare costs/charges, sometimes excluding physician fees, sometimes excluding surgery/dressings etc. A complete version of this table, including references per category can be found as Supplemental Information online.

R1R2R3R4R5R6R7R8R9


30 | Chapter 2

Cost results A total of 34 studies calculated the cost prices of a single hospital day (Table 2), and included costs were frequently not defined. The mean cost of burn centre days was $2,627 (range of $111-$11,607; median of $2,018). Only 6 studies separately described burn centre intensive care unit (ICU) days, and the mean cost was $3,164 (range of $1,590-$4657; median of $2,969). US studies showed an increase in burn centre day costs over the years after inflation correction; the cost was approximately $2,000 per day in the 1980s but rose to $5,000 per day in the past decade (Table 2.a). Furthermore, the costs of burn care per day in general hospitals were described. The mean cost of burn care per day in general hospitals was $1,159 (range of $25-$4,314; median of $585) (Table 2.b.). Only one unit price was provided for an ICU day in a general hospital: $4,356. One study (Wheeler et al., USA26) presented both burn centre day costs and general hospital day costs in patients with burns; the costs were $1,485 and $585, respectively. The mean total healthcare cost per burn patient (85 studies) was $76,497 (range of $102-$717,306; median of $36,696)1,8,26-108. The range includes costs in different market economies and healthcare settings, and the majority of studies calculated acute burn hospitalisation costs only. The mean total healthcare cost per burn patient in high-income countries (73 studies) was $88,218 (range of $704-$717,306; median of $44,024), whereas the corresponding figure was $5,196 (range of $102-$15,555; median of $3,559) for the 12 studies of low- and middle-income countries. The costs of burn patients per treatment and/or day were higher in most studies (7/11) than those of other patients (e.g., other injury or other ICU patients) (Table 3).



Tabl

e 2.

a. P

rices

of u

nit c

ost:

burn

cen

tre

hosp

ital d

ays

and

burn

cen

tre

ICU

day

s (c

onve

rted

to U

S$, 2

012)

*Fi

rst a

utho

r, ye

ar,

coun

try

patie

nt ty

pepa

tient

s (n

)m

ean

%TB

SA

reso

urce

/ da

ta s

ourc

e/bo

ttom

up-

to

p do

wn

defin

ition

hos

pita

l day

unit

cost

bu

rn c

entr

e ho

spita

l day

($

)

unit

cost

bu

rn c

entr

e IC

U d

ay($

)

tota

l tre

atm

ent

cost

/n

($)

Jans

en, 2

012,

C

anad

a (5

2)ad

ult a

nd

paed

iatri

c26

4na

+/+/

nadi

em ra

te; i

ncl.

over

head

, nur

sing

, alli

ed h

ealth

&

supp

ort s

taff,

dre

ssin

g, e

xcl.

phar

mac

y, p

hysi

cian

fe

es, o

pera

tive

cost

s

1,84

6na

55,0

30

Car

ayan

ni, 2

011,

G

reec

e (1

04)

adul

t and

pa

edia

tric

211

na-/-

/na

hosp

ital d

ay, n

s11

1na

704

Sahi

n, 2

011,

Tur

key

(80

adul

t and

pa

edia

tric

4336

-/+/n

aho

spita

l day

cha

rges

: bas

ed o

n co

sts

per o

ccup

ied

bed

day,

dre

ssin

gs, m

edic

ine,

inje

ctio

ns, l

ab

test

s, b

lood

and

blo

od p

rodu

cts,

sur

gery

, mea

ls,

beve

rage

s, s

peci

al n

utrit

ion

435

na15

,555

Kas

tenm

eier

, 201

0,

USA

(56)

ad

ult a

nd

paed

iatri

c 15

,219

11+/

+/na

hosp

ital d

ay c

harg

es, n

s4,

982

na48

,823

Patil

, 201

0,

Aus

tral

ia (8

8)ad

ult

1323

-/+/b

uIC

U d

ay: n

ursi

ng a

nd m

edic

al s

taff,

med

icat

ion,

flu

ids,

labo

rato

ry, r

adio

logi

cal i

nves

tigat

ion,

ph

ysio

ther

apy,

wou

nd d

ress

ings

na2,

426

10,4

27

Ber

ger,

2010

, Sw

itzer

land

(14

5)ad

ult

4627

+/+/

naho

spita

l day

, ns

/ IC

U d

ay, n

s65

12,

575

na

Stra

nd, 2

010,

Sw

eden

(84)

paed

iatri

c1,

060

6+/

+/na

hosp

ital d

ay, n

s89

8na

11,2

66

Pella

tt, 2

010,

U

K (4

3)pa

edia

tric

337

+/+/

buw

ard

day:

incl

udin

g dr

ugs,

50%

med

ical

sta

ff bu

dget

, nu

rsin

g st

aff,

cate

ring

/ pae

diat

ric in

tens

ive

care

uni

t (in

clud

ing

over

head

)

636

3,36

410

4,81

3

high

dep

ende

ncy

unit

day

(incl

udin

g ov

erhe

ad)

1,41

610

4,81

3D

unca

n, 2

009,

UK

(1

46)

adul

t and

pa

edia

tric

nana

-/+/b

ust

anda

rd w

ard

bed,

ns/

occ

upie

d IC

U b

ed, n

s1,

281

4,37

1na

Hem

ingt

on-G

orse

, 20

09, U

K (2

7)ad

ult a

nd

paed

iatri

c 3

38+/

+/td

low

dep

ende

ncy

war

d, n

s/ IC

U w

ard,

ns

750

4,65

771

7,30

6

paed

s w

ard,

ns

1,81

2na

naPa

pp, 2

008,

Fi

nlan

d (5

8)pa

edia

tric

4526

-/+/n

aho

spita

l day

Kuo

pio

Bur

n U

nit (

incl

udin

g re

gula

r and

IC

U d

ays

and

oper

atio

ns)

2,37

7na

47,5

37

hosp

ital d

ay H

elsk

inki

Bur

ns U

nit (

incl

udin

g re

gula

r an

d IC

U d

ays

and

oper

atio

ns)

2,62

1na

47,1

95

R1R2R3R4R5R6R7R8R9


32 | Chapter 2

Firs

t aut

hor,

year

, co

untr

ypa

tient

type

patie

nts

(n)

mea

n %

TBSA

re

sour

ce/

data

sou

rce/

botto

m u

p-

top

dow

n

defin

ition

hos

pita

l day

unit

cost

bu

rn c

entr

e ho

spita

l day

($

)

unit

cost

bu

rn c

entr

e IC

U d

ay($

)

tota

l tre

atm

ent

cost

/n

($)

Late

nser

, 200

7,

USA

(50)

adul

t and

pa

edia

tric

26,1

76na

-/+/n

aho

spita

l day

cha

rges

, ns

5,63

7na

68,4

84

Fong

, 200

5,

Aus

tral

ia (7

3)ad

ult

709

-/-/n

aho

spita

l day

, ns

2,01

8na

26,5

41

Wils

on, 2

004,

USA

(3

9)ad

ult

217

37+/

+/na

hosp

ital d

ay, n

s5,

343

na13

6,96

7

Suzm

an, 2

001,

USA

(1

27)

adul

t 76

13+/

-/na

hosp

ital d

ay (e

xclu

ding

phy

sici

an fe

es)

4,79

4na

na

Sjöb

erg,

200

0,

Swed

en (1

51)

adul

t and

pa

edia

tric

226

na-/-

/na

hosp

ital d

ay, T

BS

A 20

-30%

3,72

9na

na

hosp

ital d

ay, n

s, T

BS

A >

30%

4,43

8na

naYa

rbro

ugh,

200

0,

USA

(42)

adul

t and

pa

edia

tric

5440

-/+/n

aIC

U d

ay c

harg

es, e

stim

atio

nna

1,59

069

,884

Bar

ret,

2000

, USA

(3

2)pa

edia

tric

2092

-/-/n

aho

spita

l day

(obt

aine

d fro

m s

tudy

Her

ndon

199

0)3,

180

na28

3,02

0

Saffl

e, 1

997,

USA

(3

0)ad

ult a

nd

paed

iatri

c 49

35-/+

/na

hosp

ital d

ay c

harg

es, n

s11

,607

na44

9,97

0

Wel

ls, 1

995,

C

anad

a (1

00)

adul

t and

pa

edia

tric

261

-/-/n

aho

spita

l day

, ns

1,35

8na

1,90

1

Saffl

e, 1

995,

USA

(5

1)ad

ult a

nd

paed

iatri

c 6,

417

25-/+

/na

hosp

ital d

ay c

harg

es, n

s4,

410

na64

,834

Wu,

199

2,

Sing

apor

e (6

1)ad

ult a

nd

paed

iatri

c 10

48+/

+/na

hosp

ital d

ay, n

s1,

032

na44

,413

Prui

tt, 1

990,

USA

(1

40)

adul

t and

pa

edia

tric

na38

-/-/n

aho

spita

l day

(rou

tine

serv

ice

cost

s: p

erso

nnel

43%

, m

edic

al s

uppl

ies

17%

, adm

inis

tratio

n 19

%, h

ouse

st

aff 4

%, c

entra

l ser

vice

s 3%

, lau

ndry

3%

, oth

ers

15%

)

2,10

1na

na

Fran

k, 1

987,

USA

(4

8)ad

ult a

nd

paed

iatri

c 8,

069

na+/

+/na

hosp

ital d

ay (e

xclu

ding

phy

sici

an fe

es)

3,33

7na

79,4

82

Dim

ick,

198

6, U

SA

(65)

adul

t1,

553

33+/

+/na

hosp

ital d

ay c

harg

es (p

hysi

cal t

hera

py, n

utrit

ion,

ph

arm

acy

and

othe

r ser

vice

in b

urn

care

)2,

306

na40

,716

Tabl

e 2.

a. C

ontin

ued



Firs

t aut

hor,

year

, co

untr

ypa

tient

type

patie

nts

(n)

mea

n %

TBSA

re

sour

ce/

data

sou

rce/

botto

m u

p-

top

dow

n

defin

ition

hos

pita

l day

unit

cost

bu

rn c

entr

e ho

spita

l day

($

)

unit

cost

bu

rn c

entr

e IC

U d

ay($

)

tota

l tre

atm

ent

cost

/n

($)

Whe

eler

, 198

3, U

SA

(26)

**

adul

t and

pa

edia

tric

416

6+/

+/bu

hosp

ital d

ay (i

npat

ient

car

e in

clud

ing

oper

atin

g ro

om

cost

s)1,

488

na30

,401

Vand

enbu

ssch

e,

1981

, Fra

nce

(76)

adul

t and

pa

edia

tric

285

na-/+

/na

hosp

ital d

ay (h

ousi

ng, m

edic

atio

n, m

ater

ial,

pers

onne

l, ov

erhe

ad)

626

na18

,385

Sore

nsen

, 196

8,

Den

mar

k (9

2)ad

ult a

nd

paed

iatri

c 20

8na

-/-/n

aho

spita

l day

(bot

h un

it da

ys/o

utpa

tient

clin

ic v

isit)

369

na7,

296

*All

stud

ies

wer

e co

st s

tudi

es, e

xcep

t for

one

: Car

yann

i et a

l; C

EA

. In

all s

tudi

es a

hea

lthca

re p

ersp

ectiv

e w

as a

dapt

ed, i

nclu

ded

cost

s w

ere

only

dire

ct

cost

s in

hea

lthca

re, a

nd s

tudy

per

iod

was

<1

year

. ** g

ener

al h

ospi

tal d

ay p

rices

of t

his

stud

y ar

e pr

esen

ted

in ta

ble

2.b.

Na,

not

ava

ilabl

e; n

s, n

ot s

peci

fied;

bu

, bot

tom

up;

td, t

op d

own.

R1R2R3R4R5R6R7R8R9


34 | Chapter 2

Tabl

e 2.

b. P

rice

of u

nit c

ost:

gene

ral h

ospi

tal d

ays

and

ICU

day

s (c

onve

rted

to U

S$,2

012)

Firs

t aut

hor,

year

, co

untr

ypa

tient

ty

pepa

tient

s (n

)m

ean

%TB

SAre

sour

ce/d

ata

sour

ce/b

otto

m

up- t

op d

own

defin

ition

hos

pita

l day

unit

cost

ho

spita

l day ($

)

unit

cost

IC

U d

ay ($)

tota

l tre

atm

ent

cost

s/n ($)

Allo

rto,

201

1,

Sout

h A

fric

a (9

1)ad

ult a

nd

paed

iatri

c45

013

+/+/

buho

spita

l day

, ns

155

na8,

164

Aha

chi,

2011

, N

iger

ia (9

8)

adul

t and

pa

edia

tric

27na

-/-/n

aho

spita

l day

, ns

95na

2,24

5

Mile

nkov

ic, 2

007,

U

SA (8

5)ad

ult a

nd

paed

iatri

c32

,500

na+/

+/na

hosp

ital d

ay, n

s2,

318

na21

,106

Forju

oh, 1

998,

U

SA (6

0)

adul

t and

pa

edia

tric

3,13

7na

+/+/

tdho

spita

l day

, cha

rges

, ns

4,31

4na

45,8

18

Cou

rtrig

ht, 1

993,

Et

hiop

ia (1

05)

adul

t and

pa

edia

tric

271

na-/-

/na

hosp

ital d

ay, e

stim

atio

n25

na66

7

Lofts

, 199

1, N

ew

Zeal

and

(46)

adul

t and

pa

edia

tric

26>3

0+/

+/na

hosp

ital d

ay: i

nclu

ding

all

treat

men

ts a

nd in

terv

entio

ns,

sala

ries

and

capi

tal e

xpen

ditu

re,

excl

udin

g de

prec

iatio

n of

eq

uipm

ent /

ICU

uni

t day

: pr

ovid

ed b

y he

ad o

f dep

artm

ent

618

4,35

692

,361

Whe

eler

, 198

3,

USA

(26)

adul

t and

pa

edia

tric

54na

+/+/

buho

spita

l day

, ns

58

5na

6,81

6

Na,

not

app

licab

le; n

s, n

ot s

peci

fied;

bu,

bot

tom

up.



Tabl

e 3.

Hea

lthca

re c

osts

of b

urn

patie

nts

vs. o

ther

pat

ient

s (c

onve

rted

to U

S$, 2

012)

Fi

rst a

utho

r, ye

ar, c

ount

ryse

tting

type

of p

atie

nts

part

icip

ants

(b

urn

patie

nt:

othe

r pat

ient

)

reso

urce

/da

ta s

ourc

e/

botto

m u

p-to

p do

wn

resu

ltsTo

tal c

osts

bur

n pa

tient

/oth

er

patie

nt

($)

Cos

t per

day

bu

rn p

atie

nt/

vs. o

ther

pa

tient

($)

Alin

ia, 2

011,

Ira

n (1

07)

gene

ral

hosp

ital

burn

s/ o

ther

fire

wor

k re

late

d in

jury

72

3/1,

094

-/-/n

am

ean

tota

l ho

spita

l cos

ts/n

239/

120

na

Buc

her,

2010

, U

SA (1

15)

gene

ral

hosp

ital

ambu

lato

ry b

urns

/ oth

er w

ound

s?/

7,80

6-/-

/na

cost

of a

n ou

tpat

ient

vis

it 77

6/1,

628

na

Jian

g, 2

010,

C

hina

(106

)ge

nera

l ho

spita

lsc

alds

/ tot

al u

nint

entio

nal i

njur

y 1,

279/

6,21

3-/+

/na

mea

n to

tal

hosp

ital c

osts

/n24

8/20

8na

Patil

, 201

0,

Aus

tral

ia (8

8)IC

Ubu

rns

ICU

/ ot

her I

CU

(mat

ched

for

LOH

S)

13/1

3-/+

/bu

med

ian

tota

l co

sts/

n9,

863/

9,71

41,

003/

988

Kas

tenm

eier

, 20

10, U

SA (5

6)bu

rn c

entre

burn

/ non

-bur

n ad

mis

sion

s 15

,219

/ 3,0

27+/

+/na

mea

n to

tal

hosp

ital c

harg

es/n

48,8

23/1

28,3

124,

982/

7,63

8

Mile

nkov

ic,

2007

, USA

(85)

gene

ral

hosp

ital

burn

adm

issi

ons/

all

hosp

ital s

tays

32,5

00/

29,4

99,8

00+/

+/na

mea

n to

tal

hosp

ital c

osts

/n21

,106

/10,

980

2,37

1/2,

153

Pres

sley

, 200

7,

USA

(69)

*ge

nera

l ho

spita

lpa

edia

tric

burn

s/ m

otor

veh

icle

dr

iver

s

paed

iatri

c bu

rns/

poi

soni

ng

12,5

58/3

85,3

85-/+

/na

mea

n ho

spita

l ch

arge

s/n

36,5

53/4

2,16

4

36,5

53/1

0,36

3

na

Mill

er, 2

006,

U

SA (6

3)bu

rn c

entre

fire-

flam

e/ s

kin

dise

ases

14,5

02/4

96-/+

/na

mea

n ho

spita

l ch

arge

s/n

104,

869/

148,

572

na

scal

d/ s

kin

dise

ases

9,05

3/49

643

,260

/148

,572

naSi

riton

gtaw

orn,

20

06, T

haila

nd

(83)

traum

a ce

ntre

burn

s/ o

ther

inju

ry (n

ervu

s/ey

e/sk

in)

33/1

89-/+

/na

mea

n ho

spita

l co

sts/

n11

,747

/3,5

94na

Kag

an, 2

004,

U

SA (3

3)

burn

cen

tre/

univ

ersi

ty

hosp

ital

Bur

n pa

tient

s un

derg

oing

trac

heo-

stom

y/ n

on b

urn

patie

nts

unde

rgoi

ng

trach

eost

omy

(e.g

. poi

soni

ng,

circ

ulat

ory/

resp

irato

ry d

isor

der)

8/24

0+/

+/na

mea

n to

tal

hosp

ital c

osts

/n24

2,87

9/10

6,28

94,

762/

3,68

4

Cor

nish

, 200

3,

Can

ada

(102

)bu

rn c

entre

burn

s (a

cute

) / T

oxic

Epi

derm

al

Nec

roly

sis

19/2

+/+/

nam

ean

inpa

tient

dr

ug c

osts

/n1,

485/

1,04

5na

*Cos

ts o

f sev

eral

inju

ries

wer

e de

scrib

ed in

this

stu

dy, t

he lo

wes

t and

hig

hest

cos

ts a

re p

rese

nted

in th

e ta

ble.

Na,

not

ava

ilabl

e; b

u, b

otto

m u

p.

R1R2R3R4R5R6R7R8R9


36 | Chapter 2

Cost results high-income countries Detailed cost results for high-income countries are presented in Table 4. The mean cost of burn centre days in high-income countries was $2,705 (range of $111-11,607; median of $2,060). The mean cost of burn care per day in general hospitals in high-income countries was $1,959 (range of $585-$4,314; median of $1,468). The mean total healthcare cost of flame burns ($87,139) was generally higher than the cost of scalds ($33,960)50,60,63,72. The mean cost of electric burns was $55,28150,63,72. Several studies calculated the costs of workplace-related burn injuries, indicating mean medical and or claim costs of $16,23281,94,120,122,153. A higher TBSA burned was associated with increased costs30,49,50,63,72,109, but no further increase was reported above 80% TBSA50,63,49. The costs per 1% TBSA burned were calculated to be $723 and $1,863 in two studies8,58. We calculated cost prices per 1% TBSA burned in studies that presented both mean TBSA and mean total burn care costs. The mean total burn care cost per 1% TBSA in these 46 studies was $4,097 (range of $162-$20,663; median of $2,633). This mean was relatively consistent in different subgroups of mean TBSA subgroups (0-10%, 11-20%, >20%), with mean figures of $3,883, $3,879, and $4,312 per 1% TBSA, respectively. Mean total burn care cost per 1% TBSA in two studies on massive burns (>80%TBSA) was $2,729 (range of $2,374-3,086).

Table 4. Healthcare costs of burn patients in high-income countries (converted to US$, 2012)Mean ($) range ($) median ($) references

Costs per burn centre day

2,705 111-11,607 2,060 26,27,30,32,39,42,43,48,50-52,56,58,61,65,73,76, 84,88, 100,104,127,140,145,146,151,153

Costs per burn centre ICU day

3,164 1,590-4,657 2,969 27,42,43,88,145,146

Costs per general hospital day

1,959 585-4,314 1,468 26,46,60,85

Costs per general ICU day

4,356 4,356 4,356 46

Total healthcare costs/pt

88,218 704 -717,306 44,024 1,8,26-29,30,31-39,40-69,70,72,73-77-79,81,82,84-86,88-90,92,94,95,97,99,100,102-104

Flame 87,140 50,508-109,469 94,291 50,60,63,72Scald 33,960 15,882- 32,526 33,981 50,60,63,72

Electric 55,281 26,076-70,311 69,457 50,63,72Costs per 1% TBSA

4,159 162-20,663 2,633 1,8,26,27,29,30,31,33,35, 36,37,39,40-44,46, 51,53,55,56-58, 59,61,64 - 68,71,73,74,75,79,84,86,88,90,94,97,100, 103

In high-income countries, hospitalisation in the ICU and general hospital days were an important cost component in most studies that presented the distribution of acute burn care hospital



costs into different cost components (7 studies). Hospitalisation costs (including personnel costs) in studies from a healthcare perspective amounted to 82% of the total burn care costs per patient 46. Surgery was another important cost category in most studies. Although they were often studied, medication (e.g., medication for pain or itching or antimicrobial medication) and dressings appeared not to be major cost categories in most studies (Figure 1).In addition to hospital costs, pre-hospital costs and healthcare and non-healthcare costs in the rehabilitation phase had to be calculated in relation to burn injuries. Only 4 articles described pre-hospital costs, and 3 of these articles calculated helicopter flight costs, with a mean of $14,01340, 114, 128. Furthermore, only two studies presented both healthcare and non-healthcare costs, such as caregiver costs and social and labour costs. Medical costs represented only 10-11% of the total costs per patient1,8; the other 90% of the costs primarily related to productivity losses and informal care. No recent studies describing rehabilitation costs, e.g., the costs of reconstructive procedures, were found.

0% 20% 40% 60% 80% 100%

Lofts, '91, New Zealand (46)

Wu, '92, Singapore (61)

Eldad, '93, Israel (34)

Khoo, '94, Singapore, <15% TBSA (94)

Khoo, '94, Singapore, >15% TBSA (94)

Griffiths, '06, UK (97)

Hemington-Gore, '09, UK (27)

Pellat, '10, UK (43) % hospital days

% ICU

% surgery

% dressing

%others: medication, fluids,diagnostics, allied healthprofessionals, etc.

Figure 1. Distribution of total acute burn care hospital costs in different cost components Only studies calculating distribution of mean total burn care costs were included

Economic evaluationsThis study identified 3 full economic evaluations in burn care, which were all cost-effectiveness analyses that compared different dressings (Table 5). The clinical outcomes used were the number of patients healed at day 21 post-burn10,12 and time to 50% wound healing104. The calculated costs included mean total hospital costs104 and mean total dressing costs10,12. To calculate cost-effectiveness, intervention cost differences were divided by outcome differences to generate an incremental cost effectiveness ratio (ICER). All studies found a positive ICER in favour of intervention (Table 4). Two studies included a sensitivity analysis, which is necessary to study the effects of measurement uncertainty. None of the studies performed discounting, i.e., correction for costs in different years.

R1R2R3R4R5R6R7R8R9


38 | Chapter 2

Tabl

e 5.

Eco

nom

ic e

valu

atio

n: 3

cos

t-effe

ctiv

enes

s an

alys

es (c

onve

rted

to U

S$, 2

012)

Firs

t au

thor

, yr,

coun

try

nst

udy

desi

gnre

sour

ce/

data

so

urce

/ bo

ttom

up-

top

dow

n

inte

rven

tion:

con

trol

cost

ou

tcom

esco

st

resu

lts

($)

clin

ical

ou

tcom

escl

inic

al

resu

lts

ICER

nu

mer

ator

/de

nom

inat

or

ICER

resu

ltsse

nsiti

vity

an

alys

isdi

scou

nt-

ting

Car

ayan

ni,

2011

, G

reec

e (8

4)

211

RC

T-/-

/na

I: m

oist

exp

osed

bur

n oi

ntm

ent (

ME

BO

)C

: pov

idin

e io

dine

+

bepa

nthe

nol

mea

n co

st o

f ho

spita

l sta

y/N

I:805

C:8

60p

= 0.

10

time

of 5

0%

wou

nd h

ealin

g in

day

s

I: 8.

70

C: 1

0.75

p= 0

.00

mea

n ho

spita

l co

sts/

days

of

hosp

italiz

atio

n

$90

per d

ay o

f ho

spita

lizat

ion

gain

ed

yes

no

Silv

erst

ein,

20

11, U

SA

(12)

100

RC

T+/

-/na

I: si

lver

con

tain

ing

soft

silic

one

foam

dre

ssin

gC

: silv

ersu

lfadi

azin

e

mea

n to

tal

wou

nd

man

agem

ent

cost

s/N

I: 33

1

C:5

50p<

0.0

01

% N

with

full

epith

elia

lizat

ion

at 2

1 da

ys

I: 78

.3

C: 6

6.2

p: n

a

mea

n to

tal

cost

s of

wou

nd

man

agem

ent/

rate

of f

ull

reep

ithel

ialis

atio

n

$1,8

06 in

fa

vour

of

silic

one

foam

dr

essi

ng

nono

Car

uso,

20

06, U

SA

(10)

82R

CT

+/+/

naI:

hydr

ofibe

r (A

quac

el

AG

)

C: S

ilver

sulfa

diaz

ine

mea

n dr

essi

ng

(labo

ur a

nd

mat

eria

l) co

st/N

I: 1,

269

C: 1

,440

p =:

na

% N

with

full

epith

elia

lizat

ion

at 2

1 da

ys

I: 74

:

C: 6

0p

= 0.

222

mea

n to

tal

dres

sing

cos

ts/

rate

of f

ull

epith

elia

lisat

ion

in

21 d

ays

$1,2

43 in

fa

vour

of

hydr

ofibe

r

yes

no

Na,

not

ava

ilabl

e; I,

Inte

rven

tion;

C, C

ontro

l; p,

p-v

alue



Discussion

This review systematically assessed the methodology, costs and economic evaluations of 156 articles on the costs of burn care. Nearly all studies were cost studies in high-income countries and provided limited information on cost calculations and components. Charges were often used as a proxy to calculate costs. Broad ranges of costs were presented for burn centre days, ICU days and total burn treatment. Burn care was generally more expensive than other forms of healthcare. There were few available economic evaluations comparing the costs and effects of specific interventions; only 3 studies from high-income countries satisfied the conditions of an economic evaluation. Until now, reviews on the costs of burn care have not focused on the methodology of the included cost studies2,6,7. This review provides the first overview of burn cost study methods and results. Because of methodological variation and incomplete information, it is difficult to determine the causes of the wide range of costs found in this review. For example, the components included in the prices of burn centre days greatly varied; ICU and non-ICU costs were presented separately in some studies, whereas others studies presented one price for burn centre days without identifying the included costs. The quality and comparability of future economic studies of burn care may be substantially enhanced by the development of an extensive common core of basic methodological choices. We recommend the use of a societal perspective, common cost categories, and real cost calculations instead of the use of charges. We also recommend that data sources for costs, price data for cost units and data for healthcare consumption be disclosed. A clear definition should also be used and disclosed for hospital day costs. For example, the Dutch manual on the standardisation of costs defines a hospital day as including personnel (physicians/nurses), material, equipment, food and laundry, medication, and overhead25. Additionally, the recently published CHEERS statement can be used to ensure a more transparent and complete approach to reporting methods and findings3.Despite the insufficient methodological information in most articles, an attempt was made to answer some questions about the costs of burn care. First, is burn care expensive? Burn care was indeed more expensive in most studies that compared the healthcare costs of burn patients to those of other patients. Additionally, we made a detailed analysis of data from high-income countries to overcome the differences in price levels between low-, middle- and high-income countries. In high-income countries, the costs of burn care per burn centre day ($2,705) were higher than those of burn care per day in a general hospital ($1,959) and were higher than the standard price ($608) of a general hospital day for all patient types in a high-income country (The Netherlands) (US$ 2012)25. However, burn centres deliver multidisciplinary care and have achieved a significant reduction in mortality for patients with major burns during the last 50 years2. Furthermore, patients who require this high level of care are primarily referred to burn centers110. The mean total healthcare cost per burn patient

R1R2R3R4R5R6R7R8R9


40 | Chapter 2

in high-income countries ($88,218) was higher than the inpatient costs of trauma and acute surgery patients in one study ($17,245 and $26,468, respectively111) and was also higher than the inpatient cost of trauma patients in another study (mean $10,603112). Healthcare costs for burn patients during the rehabilitation phase and non-healthcare costs such as productivity losses are likely to be high. Unfortunately, these costs were rarely studied; only Sanchez et al. presented costs from a societal perspective, including indirect costs and direct non-healthcare costs along with direct healthcare costs. This study confirmed our assumption and showed that direct healthcare costs represented only 10% of total costs. We can conclude that burns have a high burden of illness and that burn care is indeed expensive. Comparison of cost levels between high- and low- and middle-income countries showed substantial differences in mean costs per patient ($88,218 versus $5,196). Obviously, price levels are different in low- and middle-income countries. In addition, treatment protocols will deviate substantially, because of different therapeutic possibilities and limited resources. Future studies in low- and middle-income countries using an optimal design, are necessary to gain more insight in the costs and cost-effectiveness of local burn care. In these studies, attention for local treatment protocols including for instance the different windows for resuscitation and active treatment versus compassionate care should be included, to understand differences in burn care costs between various market economies.

What factors make burn care expensive? A subgroup analysis of the studies conducted in high-income countries was performed to answer this question. The number of included studies from low- and middle-income countries was too small for a subgroup analysis. Patient characteristics that predicted high costs were a high % TBSA and flame burns. The percentage of TBSA burned seemed to be a stable predictor of burn costs with a mean price of $4,097 per 1% TBSA burned. In massive burns (TBSA>80%), this predictor cannot be used. Probably due to a high mortality, the price per 1% TBSA in these studies (n=2) was lower ($2,729 versus $4,097). The most expensive burn care component in our review was hospital stay. The included studies, however, often focused on medication or dressings. We recommend that future research on cost-effective burn care focuses on reducing hospital stay length without compromising the quality of care. Our research group is currently conducting a trial of the cost-effectiveness of Laser Doppler Imaging in burn care113 to analyse whether LDI will lead to earlier excision and grafting and less expensive hospital stays74.Based on this systematic review, we conclude that numerous economic studies in burn care in especially high-income countries have been previously performed. We propose a standard approach to cost studies and economic evaluations to improve quality, enhance the harmonisation of economic studies, optimise comparability and improve insight into burn care costs and efficiency. Furthermore, more studies on medical and non-medical costs in



the rehabilitation phase are needed to gain better insight into burn burden from a societal perspective. Cost studies in low-income and middle-income countries are limited in number; additional studies are needed to gain better insight into costs and cost-effectiveness in these contexts. This research is urgently needed due to the high incidence of burn and the limited resources in low-income countries. Finally, because our resources are scare and burn care in both low- middle- and high-income countries always can be improved, future burn care RCTs should include economic evaluation to reach an optimal balance between costs and health effects in global burn care.

AcknowledgementsThe authors would like to thank J. van Meel, Maasstad Hospital Rotterdam librarian, for his efforts to collect several full texts for this literature review.This research was financially supported by a grant from the Dutch Burns Foundation (11.102).None of the authors has a conflict of interest.

R1R2R3R4R5R6R7R8R9


42 | Chapter 2

References


2. Stavrou D, Weissman O, Winkler E, Millet E, Nardini G, Tessone A, Zmora N, Mushin OP, Haik J. Managing the relationship between quality and cost-effective burn care. Burns. 2011 May;37(3):367-76.

3. Husereau D, Drummond M, Petrou S, Carswell C, Moher D, Greenberg D, Augustovski F, Briggs AH, Mauskopf J, Loder E; CHEERS Task Force. Consolidated Health Economic Evaluation Reporting Standards (CHEERS) statement BMJ. 2013 Mar 25;346:f1049.

4. Drummond MF, Sculpher MJ, Torrance GW, O’Brien BJ, Stoddart GL. Methods for the Economic Evaluation of Health Care Programmes. New York: Oxford University Press; 2005.

5. Gold MR, Siegel JE, Russel LB, Weinstein MC. Cost-effectiveness in health and medicine. New York: Oxford University Press; 1996.

6. Mandal A. Quality and cost-effectiveness--effects in burn care. Burns. 2007 Jun;33(4):414-7. 7. Bass MJ, Phillips LG. Economics of paediatric burns. J Craniofac Surg. 2008 Jul;19(4):888-90. 8. Sánchez JL, Perepérez SB, Bastida JL, Martínez MM. Cost-utility analysis applied to the treatment

of burn patients in a specialized centre. Arch Surg. 2007 Jan;142(1):50-7. 9. Lang EM, Eiberg CA, Brandis M, Stark GB. Biobrane in the treatment of burn and scald injuries in

children. Ann Plast Surg. 2005 Nov;55(5):485-9. 10. Caruso DM, Foster KN, Blome-Eberwein SA, Twomey JA, Herndon DN, Luterman A, Silverstein

P, Antimarino JR, Bauer GJ. Randomized clinical study of Hydrofiber dressing with silver or silver sulfadiazine in the management of partial-thickness burns. J Burn Care Res. 2006 May-Jun;27(3):298-309.

11. Borsuk DE, Gallant M, Richard D, Williams HB. Silver-coated nylon dressings for paediatric burn victims. Can J Plast Surg. 2007 Spring;15(1):29-31.

12. Silverstein P, Heimbach D, Meites H, Latenser B, Mozingo D, Mullins F, Garner W, Turkowski J, Shupp J, Glat P, Purdue G. An open, parallel, randomized, comparative, multicenter study to evaluate the cost-effectiveness, performance, tolerance, and safety of a silver-containing soft silicone foam dressing (intervention) vs silver sulfadiazine cream. J Burn Care Res. 2011Nov-Dec;32(6):617-26.

13. Cuttle L, Naidu S, Mill J, Hoskins W, Das K, Kimble RM. A retrospective cohort study of Acticoat versus Silvazine in a paediatric population. Burns. 2007 Sep;33(6):701-7.

14. Silver GM, Robertson SW, Halerz MM, Conrad P, Supple KG, Gamelli RL. A silver-coated antimicrobial barrier dressing used postoperatively on meshed autografts: a dressing comparison study. J Burn Care Res. 2007 Sep-Oct;28(5):715-9.

15. Demling RH, DeSanti L. Closure of partial-thickness facial burns with a bioactive skin substitute in the major burn population decreases the cost of care and improves outcome. Wounds. 2002;14(7).

16. Torrati FG, Rossi LA, Ferreira E, Dalri MC, de Carvalho EC, dos Santos Barbeira CB. Analysis of cost of dressings in the care of burn patients. Burns. 2000 May;26(3):289-93.

17. Busche MN, Herold C, Schedler A, Knobloch K, Vogt PM, Rennekampff HO. [The Biobrane glove in burn wounds of the hand. Evaluation of the functional and aesthetic outcome and comparison of costs with those of conventional wound management]. Handchir Mikrochir Plast Chir. 2009 Dec;41(6):348-54.

18. Gerding RL, Imbembo AL, Fratianne RB. Biosynthetic skin substitute vs. 1% silver sulfadiazine for treatment of inpatient partial-thickness thermal burns. J Trauma. 1988 Aug;28(8):1265-9.

19. Mabrouk A, Boughdadi NS, Helal HA, Zaki BM, Maher A. Moist occlusive dressing (Aquacel(®) Ag) versus moist open dressing (MEBO(®)) in the management of partial-thickness facial burns: a comparative study in Ain Shams University. Burns. 2012 May;38(3):396-403.

20. Muangman P, Nitimonton S, Aramwit P. Comparative Clinical Study of Bactigras and Telfa AMD for Skin Graft Donor-Site Dressing. Int J Mol Sci. 2011;12(8):5031-8.

21. Cockington RA. Ambulatory management of burns in children. Burns. 1989 Aug;15(4):271-3.



22. Summer GJ, Hansen FL, Costa BA, Engrav LH, Sharar SR. The Unna “cap” as a scalp donor site dressing. J Burn Care Rehabil. 1999 Mar-Apr;20(2):183-8,discussion 182.

23. Muangman P, Muangman S, Opasanon S, Keorochana K, Chuntrasakul C. Benefit of hydrocolloid SSD dressing in the outpatient management of partial thickness burns. J Med Assoc Thai. 2009 Oct;92(10):1300-5.

24. Cox SG, Cullingworth L, Rode H. Treatment of paediatric burns with a nanocrystalline silver dressing compared with standard wound care in a burns unit: a cost analysis. S Afr Med J. 2011 Sep 27;101(10):728-31.

25. Hakkaart-van Roijen L, Tan SS,Bouwmans CAM. [Standardisation of costs: the Dutch manual for costing in economic evaluations]. Diemen: College voor zorgverzekeringen; 2010.

26. Wheeler JR, Harrison RV, Wolfe RA, Payne BC. The effects of burn severity and institutional differences on the costs of care. Med Care. 1983 Dec;21(12):1192-203.

27. Hemington-Gorse SJ, Potokar TS, Drew PJ, Dickson WA. Burn care costing: the Welsh experience. Burns. 2009 May;35(3):378-82.

28. Carr JA, Phillips BD, Bowling WM. The utility of bronchoscopy after inhalation injury complicated by pneumonia in burn patients: results from the National Burn Repository. J Burn Care Res. 2009 Nov-Dec;30(6):967-74.

29. Kucan J, Bryant E, Dimick A, Sundance P, Cope N, Richards R, Anderson C. Systematic care management: a comprehensive approach to catastrophic injury management applied to a catastrophic burn injury population--clinical, utilization, economic, and outcome data in support of the model. J Burn Care Res. 2010 Sep-Oct;31(5):692-700.

30. Saffle JR, Wiebke G, Jennings K, Morris SE, Barton RG. Randomized trial of immune-enhancing enteral nutrition in burn patients. J Trauma. 1997 May;42(5):793-800; discussion 800-2.

31. Munster AM, Weiner SH, Spence RJ. Cultured epidermis for the coverage of massive burn wounds. A single centre experience. Ann Surg. 1990 Jun;211(6):676-9; discussion 679-80.

32. Barret JP, Wolf SE, Desai MH, Herndon DN. Cost-efficacy of cultured epidermal autografts in massive paediatric burns. Ann Surg. 2000 Jun;231(6):869-76.

33. Kagan RJ, Gamelli R, Kemalyan N, Saffle JR. Tracheostomy in thermally injured patients: does diagnosis-related group 483 adequately estimate resource use and hospital costs? J Trauma. 2004 Oct;57(4):861-6.

34. Eldad A, Stern Z, Sover H, Neuman R, Ben Meir P, Wexler MR. The cost of an extensive burn survival. Burns. 1993 Jun;19(3):235-8.

35. Cianci P, Williams C, Lueders H, Lee H, Shapiro R, Sexton J, Sato R. Adjunctive hyperbaric oxygen in the treatment of thermal burns. An economic analysis. J Burn Care Rehabil. 1990 Mar-Apr;11(2):140-3.

36. Clayton MC, Solem LD, Ahrenholz DH. Pulmonary failure in geriatric patients with burns: the need for a diagnosis-related group modifier. J Burn Care Rehabil. 1995 Jul-Aug;16(4):451-4.

37. Alden NE, Bessey PQ, Rabbitts A, Hyden PJ, Yurt RW. Tap water scalds among seniors and the elderly: socio-economics and implications for prevention. Burns. 2007 Aug;33(5):666-9.

38. Takayanagi K, Kawai S, Aoki R. The cost of burn care and implications for efficient care. Clin Perform Qual Health Care. 1999 Apr-Jun;7(2):70-3.

39. Wilson SJ, Knipe CJ, Zieger MJ, Gabehart KM, Goodman JE, Volk HM, Sood R. Direct costs of multidrug-resistant Acinetobacter baumannii in the burn unit of a public teaching hospital. Am J Infect Control. 2004 Oct;32(6):342-4.

40. Saffle JR, Edelman L, Morris SE. Regional air transport of burn patients: a case for telemedicine? J Trauma. 2004 Jul;57(1):57-64; discussion 64.

41. Ganzoni N. Intensive care of patients with burns--investment, costs and clinical aspects. Helv Chir Acta. 1972 Dec;39(5):607-12.

42. Yarbrough DR 3rd, Philbeck TE Jr, Simmons FM, Finnell K, Smith P, Warren JB,Burleson DG. Therapeutic and financial outcomes using a pulsating low-air-losssurface for patients surviving severe posterior burns. Ostomy Wound Manage. 2000 Mar;46(3):64-9.

43. Pellatt RA, Williams A, Wright H, Young AE. The cost of a major paediatric burn. Burns. 2010 Dec;36(8):1208-14.

R1R2R3R4R5R6R7R8R9


44 | Chapter 2

44. Danks RR, Wibbenmeyer LA, Faucher LD, Sihler KC, Kealey GP, Chang P, Amelon M, Lewis RW 3rd. Methamphetamine-associated burn injuries: a retrospective analysis. J Burn Care Rehabil. 2004 Sep-Oct;25(5):425-9.

45. Davidson SB, Blostein PA, Walsh J, Maltz SB, Elian A, Vandenberg SL. Resurgence of methamphetamine related burns and injuries: A follow-up study. Burns. 2013 Feb;39 (1):118-25.

46. Lofts JA. Cost analysis of a major burn. N Z Med J. 1991 Nov 27;104(924):488-90. 47. Liebau J, Pallua N, Machens HG, Berger A. [Budget analysis in a burn centre for severely burned

patients]. Unfallchirurg. 1996 Nov;99(11):875-9. 48. Frank HA, Berry C, Wachtel TL, Johnson RW. The impact of thermal injury. J Burn Care Rehabil.

1987 Jul-Aug;8(4):260-2. 49. Jeng JC. Growth rings of a tree: progression of burn care charges abstracted from a decade of the

National Burn Repository. J Burn Care Res. 2007 Sep-Oct;28(5):659-60. 50. Latenser BA, Miller SF, Bessey PQ, Browning SM, Caruso DM, Gomez M, Jeng JC, Krichbaum

JA, Lentz CW, Saffle JR, Schurr MJ, Greenhalgh DG, Kagan RJ. National Burn Repository 2006: a ten-year review. J Burn Care Res. 2007 Sep-Oct;28(5):635-58.

51. Saffle JR, Davis B, Williams P. Recent outcomes in the treatment of burn injury in the United States: a report from the American Burn Association Patient Registry. J Burn Care Rehabil. 1995 May-Jun;16(3 Pt 1):219-32; discussion 288-9.

52. Jansen LA, Hynes SL, Macadam SA, Papp A. Reduced Length of Stay in Hospital for Burn Patients Following a Change in Practice Guidelines: Financial Implications. J Burn Care Res. 2012 Nov-Dec;33(6):e275-9.

53. Shalom A, Friedman T, Wong L. Burns and diabetes. Ann Burns Fire Disasters. 2005 Mar 31;18(1):31-3.

54. Mathews JJ, Supple K, Calistro A, Gamelli RL. A burn centre cost-reduction program. J Burn Care Rehabil. 1997 Jul-Aug;18(4):358-63; discussion 357.

55. Nicolau DP, White RL, Friedrich LV, Kays MB. The costs of burn care: an analysis with an emphasis on the use of parenteral antimicrobials. J Burn Care Rehabil. 1994 May-Jun;15(3):244-50.

56. Kastenmeier A, Faraklas I, Cochran A, Pham TN, Young SR, Gibran NS, Gamelli R, Halerz M, Light TD, Kealey GP, Wibbenmeyer L, Latenser B, Jenabzadeh K, Mohr WJ, Ahrenholz DH, Saffle JR. The evolution of resource utilization in regional burn centers. J Burn Care Res. 2010 Jan-Feb;31(1):130-6.

57. Weinsier RL, Heimburger DC, Samples CM, Dimick AR, Birch R. Cost containment: a contribution of aggressive nutritional support in burn patients. J Burn Care Rehabil. 1985 Sep-Oct;6(5):436-41.

58. Papp A, Rytkönen T, Koljonen V, Vuola J. Paediatric ICU burns in Finland 1994-2004. Burns. 2008 May;34(3):339-44.

59. Hunt JL, Purdue GF. Cost containment/cost reduction: the economic impact of burn DRGs. J Burn Care Rehabil. 1986 Sep-Oct;7(5):417-21.

60. Forjuoh SN. The mechanisms, intensity of treatment, and outcomes of hospitalized burns: issues for prevention. J Burn Care Rehabil. 1998 Sep-Oct;19(5):456-60.

61. Wu WT, Ngim RC. Anatomy of a burns disaster: the Miri Bank explosion. AnnAcad Med Singapore. 1992 Sep;21(5):640-8.

62. Anderson NJ, Bonauto DK, Adams D. Psychiatric diagnoses after hospitalization with work-related burn injuries in Washington State. J Burn Care Res. 2011 May-Jun;32(3):369-78.

63. Miller SF, Bessey PQ, Schurr MJ, Browning SM, Jeng JC, Caruso DM, Gomez M, Latenser BA, Lentz CW, Saffle JR, Kagan RJ, Purdue GF, Krichbaum JA. National Burn Repository 2005: a ten-year review. J Burn Care Res. 2006 Jul-Aug;27(4):411-36.

64. Burke BA, Lewis RW 2nd, Latenser BA, Chung JY, Willoughby C, Kealey GP, Wibbenmeyer LA. Methamphetamine-related burns in the cornbelt. J Burn Care Res. 2008 Jul-Aug;29(4):574-9.

65. Dimick AR, Potts LH, Charles ED Jr, Wayne J, Reed IM. The cost of burn care and implications for the future on quality of care. J Trauma. 1986 Mar;26(3):260-6.

66. Munster AM, Smith-Meek M, Sharkey P. The effect of early surgical intervention on mortality and cost-effectiveness in burn care, 1978-91. Burns.1994 Feb;20(1):61-4.

67. Gray DT, Pine RW, Harnar TJ, Marvin JA, Engrav LH, Heimbach DM. Early surgical excision versus conventional therapy in patients with 20 to 40 percent burns. A comparative study. Am J Surg. 1982 Jul;144(1):76-80.



68. Warden G, Kravitz M, Schnebly A. The outpatient management of moderate and major thermal injuries. J Burn Care Rehabil. 1981 May-Jun;2(3):159-162.

69. Pressley JC, Trieu L, Kendig T, Barlow B. National injury-related hospitalizations in children: public versus private expenditures across preventable injury mechanisms. J Trauma. 2007 Sep;63(3 Suppl):S10-9.

70. Saffle JR, Larson CM, Sullivan J, Shelby J. The continuing challenge of burn care in the elderly. Surgery. 1990 Sep;108(3):534-43.

71. Hussmann J, Kucan JO, Russell RC, Bradley T, Zamboni WA. Electrical injuries--morbidity, outcome and treatment rationale. Burns. 1995 Nov;21(7):530-5.

72. Shields BJ, Comstock RD, Fernandez SA, Xiang H, Smith GA. Healthcare resource utilization and epidemiology of paediatric burn-associated hospitalizations, United States, 2000. J Burn Care Res. 2007 ov-Dec;28(6):811-26.

73. Fong J, Wood F, Fowler B. A silver coated dressing reduces the incidence of early burn wound cellulitis and associated costs of inpatient treatment: comparative patient care audits. Burns. 2005 Aug;31(5):562-7.

74. Engrav LH, Heimbach DM, Reus JL, Harnar TJ, Marvin JA. Early excision and grafting vs. nonoperative treatment of burns of indeterminant depth: a randomized prospective study. J Trauma. 1983 Nov;23(11):1001-4.

75. Gore DC. Outcome and cost analysis for outpatient skin grafting. J Trauma. 1997 Oct;43(4):597-600.

76. Vandenbussche F, Krizek M, Robbe M, Besse MD, Truffaut A, Clavel HP. [Actual estimate of the cost of severe burns. Results for the year 1979 - Lille]. Lille Med. 1981 Mar;26(3):118-22. French.

77. Onarheim H, Jensen SA, Rosenberg BE, Guttormsen AB. The epidemiology of patients with burn injuries admitted to Norwegian hospitals in 2007. Burns. 2009 Dec;35(8):1142-6.

78. Foglia RP, Moushey R, Meadows L, Seigel J, Smith M. Evolving treatment in a decade of paediatric burn care. J Pediatr Surg. 2004 Jun;39(6):957-60; discussion 957-60.

79. Hayes-Lundy C, Ward RS, Saffle JR, Reddy R, Warden GD, Schnebly WA. Grease burns at fast-food restaurants. Adolescents at risk. J Burn Care Rehabil. 1991 Mar-Apr;12(2):203-8.

80. Sahin I, Ozturk S, Alhan D, Açikel C, Isik S. Cost analysis of acute burn patients treated in a burn centre: the Gulhane experience. Ann Burns Fire Disasters. 2011 Mar 31;24(1):9-13.

81. Fordyce TA, Kelsh M, Lu ET, Sahl JD, Yager JW. Thermal burn and electrical injuries among electric utility workers, 1995-2004. Burns. 2007 Mar;33(2):209-20.

82. Wibbenmeyer LA, Amelon MA, Loret de Mola RM, Lewis R 2nd, Kealey GP. Trash and brush burning: an underappreciated mechanism of thermal injury in a rural community. J Burn Care Rehabil. 2003 Mar-Apr;24(2):85-9.

83. Siritongtaworn P, Peunchompoo N. Economic problem of referred trauma cases in Siriraj Hospital. J Med Assoc Thai. 2006 Oct;89(10):1621-9.

84. Strand O, San Miguel L, Rowan S, Sahlqvist A. Retrospective comparison of two years in a paediatric burns unit, with and without acticoat as a standard dressing. Ann Burns Fire Disasters. 2010 Dec 31;23(4):182-5.

85. Milenkovic M, Russo CA, Elixhauser A. Hospital Stays for Burns, 2004. HCUP Statistical Brief #25. Rockville, MD: Agency for Health Care Research and quality; 2007 Jan.

86. Klein MB, Hollingworth W, Rivara FP, Kramer CB, Askay SW, Heimbach DM, Gibran NS. Hospital costs associated with paediatric burn injury. J Burn Care Res. 2008 Jul-Aug;29(4):632-7.

87. Zhou HC, Wu YC, Jiang Y, Zhang SH. [Management of extensive deep partial thickness burn wounds by dermabrasion during early postburn shock stage]. Zhonghua Shao Shang Za Zhi. 2003 Aug;19(4):219-22.

88. Patil V, Dulhunty JM, Udy A, Thomas P, Kucharski G, Lipman J. Do burn patients cost more? The intensive care unit costs of burn patients compared with controls matched for length of stay and acuity. J Burn Care Res. 2010 Jul-Aug;31(4):598-602.

89. Bassett GR. Childhood burns, causes and economic consequences. Northwest Med. 1968 Feb;67(2):146-8.

90. Linn BS, Stephenson SE Jr, Bergstresser P, Smith J. Do dollars spent relate to outcomes in burn care? Med Care. 1979 Aug;17(8):835-43.

R1R2R3R4R5R6R7R8R9


46 | Chapter 2

91. Allorto NL, Clarke DL, Thomson SR. A cost model case comparison of current versus modern management of burns at a regional hospital in South Africa. Burns. 2011 Sep;37(6):1033-7.

92. Sorensen B, Thomsen M. Cost of burns in Denmark. Scand J Plast Reconstr Surg. 1968;2(1):32-5.93. Nursal TZ, Yildirim S, Tarim A, Caliskan K, Ezer A, Noyan T. Burns in southern Turkey: electrical

burns remain a major problem. J Burn Care Rehabil. 2003 Sep-Oct;24(5):309-14. 94. Khoo AK, Wee JT, Ngim RC, Wong MK. Occupational burns in the Burns Centre at the Singapore

General Hospital. Ann Acad Med Singapore. 1994 Sep;23(5):680-3.95. Paddock HN, Fabia R, Giles S, Hayes J, Lowell W, Adams D, Besner GE. A silver-impregnated

antimicrobial dressing reduces hospital costs for paediatric burn patients. J Pediatr Surg. 2007 Jan;42(1):211-3.

96. Steenkamp WC, van der Merwe AE, de Lange R. Burn injuries by paraffin stoves.S Afr Med J. 2002 Jun;92(6):445.

97. Griffiths HR, Thornton KL, Clements CM, Burge TS, Kay AR, Young AE. The cost of a hot drink scald. Burns. 2006 May;32(3):372-4.

98. Ahachi CN, Fadeyibi IO, Abikoye FO, Chira MK, Ugburo AO, Ademiluyi SA. The direct hospitalization cost of care for acute burns in Lagos, Nigeria: a one-year prospective study. Ann Burns Fire Disasters. 2011 Jun 30;24(2):94-101.

99. Bishai D, McCauley J, Trifiletti LB, McDonald EM, Reeb B, Ashman R, Gielen AC. The burden of injury in preschool children in an urban medicaid managed care organization. Ambul Pediatr. 2002 Jul-Aug;2(4):279-83.

100. Wells NJ, Boyle JC, Snelling CF, Carr NJ, Courtemanche DJ. Lower extremity burns and Unna paste: can we decrease health care costs without compromising patient care? Can J Surg. 1995 Dec;38(6):533-6.

101. Kai-Yang L, Shi-Hui Z, Hong-Tai T, Yi-Tao J, Zhao-Fan X, Dao-Feng B, Wei L, Guang-Qing W, Guang-Yi W, Shi-Chu X, Da-Sheng C, Bing M, Jian-Rong Z. The direct hospitalisation costs of paediatric scalds: 2-year results of a prospective case series. Burns. 2009 Aug;35(5):738-45.

102. Cornish P, Mittmann N, Gomez M, Cartotto RC, Fish JS. Cost of medications in patients admitted to a burn center. Am J Clin Dermatol. 2003;4(12):861-7.

103. Kagan RJ, Warden GD. Care of minor burn injuries: an analysis of burn clinic and emergency room charges. J Burn Care Res. 2001 Sep-Oct;22(5):337-40.

104. Carayanni VJ, Tsati EG, Spyropoulou GC, Antonopoulou FN, Ioannovich JD. Comparing oil based ointment versus standard practice for the treatment of moderate burns in Greece: a trial based cost effectiveness evaluation. BMC Complement Altern Med. 2011 Dec 1;11:122.

105. Courtright P, Haile D, Kohls E. The epidemiology of burns in rural Ethiopia.J Epidemiol Community Health. 1993 Feb;47(1):19-22.

106. Jiang X, Zhang Y, Wang Y, Wang B, Xu Y, Shang L. An analysis of 6215 hospitalized unintentional injuries among children aged 0-14 in northwest China. Accid Anal Prev. 2010 Jan;42(1):320-6.

107. Alinia S, Rezaei S, Daroudi R, Hadadi M, Akbari Sari A. Extent, nature and hospital costs of fireworks-related injuries during the Wednesday Eve Festival in Iran. J Inj Violence Res. 2013 Jan;5(1):11-6.

108. Oncul O, Yüksel F, Altunay H, Açikel C, Celiköz B, Cavuşlu S. The evaluation of nosocomial infection during 1-year-period in the burn unit of a training hospital in Istanbul, Turkey. Burns. 2002 Dec;28(8):738-44.

109. Warden GD, Saffle JR, Kravitz M. Potential DRG reimbursement vs actual cost for burn care. II. Referral distance. J Burn Care Rehabil. 1986 Jan-Feb;7(1):48-51.

110. Vloemans AF, Dokter J, van Baar ME, Nijhuis I, Beerthuizen GI, Nieuwenhuis MK, Kuijper EC, Middelkoop EM. Epidemiology of children admitted to the Dutch burn centres. Changes in referral influence admittance rates in burn centres. Burns. 2011 Nov;37(7):1161-7.

111. Fakhry SM, Martin B, Al Harakeh H, Norcross ED, Ferguson PL. Proportional Costs in Trauma and Acute Care Surgery Patients: Dominant Role of Intensive Care Unit Costs. J Am Coll Surg. 2013 Feb 12.

112. Curtis K, Lam M, Mitchell R, Black D, Taylor C, Dickson C, Jan S, Palmer CS, Langcake M, Myburgh J. Acute costs and predictors of higher treatment costs of trauma in New South Wales, Australia. Injury. 2012 Oct 20.



113. Hop MJ, Hiddingh J, Stekelenburg CM, Kuipers HC, Middelkoop E, Nieuwenhuis MK, Polinder S, van Baar ME; LDI study group. Cost-effectiveness of laser Doppler imaging in burn care in the Netherlands. BMC Surg. 2013 Feb 1;13:2.

114. Vercruysse GA, Ingram WL, Feliciano DV. Overutilization of regional burn centers for pediatric patients--a healthcare system problem that should be corrected. Am J Surg. 2011 Dec;202(6):802-8; discussion 808-9.

115. Bucher BT, Seigel J, Rosenblum E, Nesslein C, Keswani S, Foglia RP, Dillon PA, Warner BW, Keller MS. A pediatric acute wound service: a novel approach in wound management. J Pediatr Surg. 2010 Jan;45(1):167-70; discussion 170.

116. Armour AD, Shankowsky HA, Swanson T, Lee J, Tredget EE. The impact of nosocomially-acquired resistant Pseudomonas aeruginosa infection in a burn unit. J Trauma. 2007 Jul;63(1):164-71.

117. Neely AN, Mayes T, Gardner J, Kagan RJ, Gottschlich MM. A microbiologic study of enteral feeding hang time in a burn hospital: can feeding costs be reduced without compromising patient safety? Nutr Clin Pract. 2006 Dec;21(6):610-6.

118. Palmieri TL, Lee T, O’Mara MS, Greenhalgh DG. Effects of a restrictive blood transfusion policy on outcomes in children with burn injury. J Burn Care Res. 2007 Jan-Feb;28(1):65-70.

119. Patel NP, Brandt CP, Yowler CJ. A prospective study of the impact of a critical care pharmacist assigned as a member of the multidisciplinary burn care team. J Burn Care Res. 2006 May-Jun;27(3):310-3.

120. Horwitz IB, McCall BP. An analysis of occupational burn injuries in Rhode Island: workers’ compensation claims, 1998 to 2002. J Burn Care Rehabil. 2005 Nov-Dec;26(6):505-14.

121. Corpron CA, Martin AE, Roberts G, Besner GE. The pediatric burn unit: a profit center. J Pediatr Surg. 2004 Jun;39(6):961-3.

122. Horwitz IB, McCall BP. Quantification and risk analysis of occupational burns: Oregon workers’ compensation claims, 1990 to 1997. J Burn Care Rehabil.2004 May-Jun;25(3):328-36.

123. Nguyen LT, Massman NJ, Franzen BJ, Ahrenholz DH, Sorensen NW, Mohr WJ 3rd,Solem LD. Telemedicine follow-up of burns: lessons learned from the first thousand visits. J Burn Care Rehabil. 2004 Nov-Dec;25(6):485-90.

124. Redlick F, Roston B, Gomez M, Fish JS. An initial experience with telemedicine in follow-up burn care. J Burn Care Rehabil. 2002 Mar-Apr;23(2):110-5.

125. Baggs J, Curwick C, Silverstein B. Work-related burns in Washington State,1994 to 1998. J Occup Environ Med. 2002 Jul;44(7):692-9.

126. Wahl WL, Franklin GA, Brandt MM, Sturm L, Ahrns KS, Hemmila MR, Arbabi S. Does bronchoalveolar lavage enhance our ability to treat ventilator-associated pneumonia in a trauma-burn intensive care unit? J Trauma. 2003 Apr;54(4):633-8.

127. Suzman MS, Sobocinski K, Himel H, Yurt RW. Major burn injuries among restaurant workers in New York City: an underappreciated public health hazard. J Burn Care Rehabil. 2001 Nov-Dec;22(6):429-34.

128. De Wing MD, Curry T, Stephenson E, Palmieri T, Greenhalgh DG. Cost-effective use of helicopters for the transportation of patients with burn injuries. J Burn Care Rehabil. 2000 Nov-Dec;21(6):535-40.

129. Miller PL, Matthey FC. A cost-benefit analysis of initial burn cultures in the management of acute burns. J Burn Care Rehabil. 2000 Jul-Aug;21(4):300-3.

130. Mathews JJ, Aleem RF, Gamelli RL. Cost reduction strategies in burn nutrition services: adjustments in dietary treatment of patients with hyponatremia and hypophosphatemia. J Burn Care Rehabil. 1999 Jan-Feb;20(1 Pt1):80-4.

131. Evasovich M, Klein R, Muakkassa F, Weekley R. The economic effect of child abuse in the burn unit. Burns. 1998 Nov;24(7):642-5.

132. Best T, Lobay G, Moysa G, Tredget E. A prospective randomized trial of absorbable staple fixation of skin grafts for burn wound coverage. J Trauma. 1995 Jun;38(6):915-9.

133. Lujan HJ, Dries DJ, Gamelli RL. Comparative analysis of bedside and operating room tracheostomies in critically ill patients with burns. J Burn Care Rehabil. 1995 May-Jun;16(3 Pt 1):258-61.

134. Barone CM, Mastropieri CJ, Peebles R, Mitra A. Evaluation of the Unna Boot for lower-extremity autograft burn wounds excoriated by pruritus in pediatric patients. J Burn Care Rehabil. 1993 May-Jun;14(3):348-9.

R1R2R3R4R5R6R7R8R9


48 | Chapter 2

135. Rue LW 3rd, Cioffi WG, McManus WF, Pruitt BA Jr. Wound closure and outcome in extensively burned patients treated with cultured autologous keratinocytes. J Trauma. 1993 May;34(5):662-7.

136. Maxwell G, Silverstein P. Burn center reimbursement analysis. J Burn Care Rehabil. 1992 Sep-Oct;13(5):597-9.

137. Bauman LW, Hawkins ML, Howdieshell TR, Till AH, Treat RC. Bilaminate synthetic dressing for partial thickness burns. Lack of cost reduction for inpatient care. Am Surg. 1991 Mar;57(3):131-3.

138. Gerding RL, Emerman CL, Effron D, Lukens T, Imbembo AL, Fratianne RB. Outpatient management of partial-thickness burns: Biobrane versus 1% silver sulfadiazine. Ann Emerg Med. 1990 Feb;19(2):121-4.

139. Kealey GP, Jensen KL, Laubenthal KN, Lewis RW. Prospective randomized comparison of two types of pressure therapy garments. J Burn Care Rehabil. 1990 Jul-Aug;11(4):334-6.

140. Pruitt BA, Mason AD, Goodwin CW. Epidemiology of burn injury and demography of burn care facilities. Probl Gen Surg. 1990 Apr-Jun;7(2):235-51.

141. Warden GD, Saffle JR, Kravitz M. Potential DRG reimbursement vs actual cost of burn care. I. Patients with burns of less than 25% TBSA. J Burn Care Rehabil. 1986 Jan-Feb;7(1):45-8.

142. McMillan SS, Parker CJ, Winkler JB, Hilton JG, Herndon DN. Prediction of operating room costs. J Burn Care Rehabil. 1985 Sep-Oct;6(5):444-6.

143. Sadove AM, Creekmur TA, Brueggemann SG, McGregor K, Bennett JE. Early excision: a financial assessment. J Burn Care Rehabil. 1985 Sep-Oct;6(5):442-3.

144. Lotter O, Jaminet P, Amr A, Chiarello P, Schaller HE, Rahmanian-Schwarz A. Reimbursement of burns by DRG in four European countries: an analysis. Burns. 2011 Nov;37(7):1109-16.

145. Berger MM, Davadant M, Marin C, Wasserfallen JB, Pinget C, Maravic P, Koch N, Raffoul W, Chiolero RL. Impact of a pain protocol including hypnosis in major burns. Burns. 2010 Aug;36(5):639-46.

146. Duncan RT, Dunn KW. Cost of burn care in the British isles and service remuneration options. Burns. 2009 Sep;35(S): S10.

147. Taghizadeh R, Gilbert PM. Comparison of commonly used mesher types in burns surgery revisited. Burns. 2008 Feb;34(1):109-10.

148. Windle EM. Nutrition support in major burn injury: case analysis of dietetic activity, resource use and cost implications. J Hum Nutr Diet. 2008 Apr;21(2):165-73.

149. Gravvanis AI, Tsoutsos DA, Iconomou TG, Papadopoulos SG. Percutaneous versus Conventional Tracheostomy in Burned Patients with Inhalation Injury. World J Surg. 2005 Dec;29(12):1571-5.

150. Favaro P. Aspects pharmaco-economiques du traitement des brulures. Ann Pharm Fr. 2002 Jul. 60 (4): 260-267.

151. Sjöberg F, Danielsson P, Andersson L, Steinwall I, Zdolsek J, Ostrup L,Monafo W. Utility of an intervention scoring system in documenting effects of changes in burn treatment. Burns. 2000 Sep;26(6):553-9.

152. Carsin H, Ainaud P, Le Bever H, Rives JM, Le Coadou A, Stephanazzi J. [Objectives, results and future prospects of burn treatment in 1997]. Bull Acad Natl Med. 1997 Oct;181(7):1307-19.

153. de Roche R, Lüscher NJ, Debrunner HU, Fischer R. Epidemiological data and costs of burn injuries in workers in Switzerland: an argument for immediate treatment in burn centres. Burns. 1994 Feb;20(1):58-60.

154. Mashreky SR, Rahman A, Khan TF, Rahman F. Consequences of non-fatalelectrical injury: findings of community-based national survey in Bangladesh. Injury. 2012 Jan;43(1):109-12.

155. Petkar KS, Dhanraj P, Kingsly PM, Sreekar H, Lakshmanarao A, Lamba S, Shetty R, Zachariah JR. A prospective randomized controlled trial comparing negative pressure dressing and conventional dressing methods on split-thickness skin grafts in burned patients. Burns. 2011 Sep;37(6):925-9.

156. Muangman P, Pundee C, Opasanon S, Muangman S. A prospective, randomized trial of silver containing hydrofiber dressing versus 1% silver sulfadiazine for the treatment of partial thickness burns. Int Wound J. 2010 Aug;7(4):271-6.

157. Mashreky SR, Rahman A, Chowdhury SM, Giashuddin S, Svanström L, Khan TF, Cox R, Rahman F. Burn injury: economic and social impact on a family. Public Health. 2008 Dec;122(12):1418-24.

158. Mashreky SR, Rahman A, Chowdhury SM, Giashuddin S, Svanström L, Linnan M, Shafinaz S, Uhaa IJ, Rahman F. Consequences of childhood burn: findings from the largest community-based injury survey in Bangladesh. Burns. 2008 Nov;34(7):912-8.



159. Venakatachalapathy TS, Mohan Kumar S, Saliba MJ. A comparative study of burns treated with topical heparin and without heparin. Ann Burns Fire Disasters. 2007 Dec 31;20(4):189-98.

160. Al-Kaisy AA, Salih Sahib A, Al-Biati HA. Effect of zinc supplement in the prognosis of burn patients in iraq. Ann Burns Fire Disasters. 2006 Sep 30;19(3):115-22.

161. Al-Kaisy AA, Salih Sahib A. Role of the antioxidant effect of vitamin e with vitamin C and topical povidone-iodine ointment in the treatment of burns. Ann Burns Fire Disasters. 2005 Mar 31;18(1):19-30.

162. Ang ES, Lee ST, Gan CS, See PG, Chan YH, Ng LH, Machin D. Evaluating the role of alternative therapy in burn wound management: randomized trial comparing moist exposed burn ointment with conventional methods in the management of patients with second-degree burns. MedGenMed. 2001 Mar 6;3(2):3.

163. Danchivijitr S, Chokloikaew S, Chantrasakul C, Trakoolsomboon S. An outbreak of methicillin-resistant Staphylococcus aureus (M.R.S.A.) in a burn unit. J Med Assoc Thai. 1995 Jul;78 Suppl 1:S11-4.

164. Ogundipe KO, Adigun IA, Solagberu BA. Economic burden of drug use in patients with acute burns: experience in a developing country. J Trop Med. 2009;2009:734712.

165. Allam AM, Mostafa W, Zayed E, El-Gamaly J. Management of the Acute Partial-thickness Burned Hand; Moist Exposed Burn Ointment or Silver Sulphadiazine Cream both Combined with a Polyethylene Bag. Ann Burns Fire Disasters. 2007 Sep 30;20(3):144-8.

166. Atiyeh BS, Dham R, Kadry M, Abdallah AF, Al-Oteify M, Fathi O, Samir A.Benefit-cost analysis of moist exposed burn ointment. Burns. 2002 Nov;28(7):659-63.

167. Williams F, Knapp D, Wallen M. Comparison of the characteristics and features of pressure garments used in the management of burn scars. Burns. 1998 Jun;24(4):329-35.

Chapter 3Economic burden of burn injuries in the Netherlands:

a 3 months follow-up study

M. Jenda Hop Ben F.M. Wijnen,

Marianne K. NieuwenhuisJan Dokter

Esther MiddelkoopSuzanne Polinder

Margriet E. van BaarThe Dutch Burn Repository group

Injury. Accepted

R1R2R3R4R5R6R7R8R9


52 | Chapter 3

Abstract

Introduction: Burn care has rapidly improved in the past decades. However, healthcare innovations can be expensive, demanding careful choices on their implementation. Obtaining knowledge on the extent of the costs of burn injuries is an essential first step for economic evaluations within burn care. The objective of this study was to determine the economic burden of patients with burns admitted to a burn centre and to identify important cost categories until three months post-burn.Patients and Methods: A prospective cohort study was conducted in the burn centre of Maasstad Hospital Rotterdam, the Netherlands, including all patients with acute burn related injuries from August 2011 until July 2012. Total costs were calculated from a societal perspective, until three months post injury. Subgroup analyses were performed to examine whether the mean total costs per patient differed by age, aetiology or percentage total body surface area (TBSA) burned. Results: In our population, with a mean burn size of 8%, mean total costs were €26,540 per patient varying from €742 to €235,557. Most important cost categories were burn centre days (62%), surgical interventions (5%) and work absence (20%). Flame burns were significantly more costly than other types of burns, adult patients were significantly more costly than children and adolescents and a higher percentage TBSA burned also corresponded to significantly higher costs. Discussion and Conclusion: Mean total costs of burn care in the first three months post injury were estimated at €26,540 and depended on age, aetiology and TBSA. Mean total costs in our population probably apply to other high-income countries as well, although we should realize that patients with burn injuries are diverse and represent a broad range of total costs. To reduce costs of burn care, future intervention studies should focus on a timely wound healing, reducing length of stay and enabling an early return to work.


Economic burden of burn injuries in the Netherlands: a 3 months follow-up study | 53

Introduction

Burn care has rapidly improved in the past decades, which is reflected by the fact that even patients with extensive burns can survive nowadays 1,2. Examples of major improvements in burn care of the second half of the 20th century are the introduction of silver-containing topical antimicrobials, early excision and grafting, shock prevention and the multidisciplinary approach to burn care 3-5. Further advances in wound healing, rehabilitation and psychological care are desirable to help burn survivors to reach an optimal quality of life. Unfortunately, healthcare innovations can be very expensive and in the current economic climate authorities have to make careful choices on the implementation of (new) treatments. Therefore the costs of new interventions should be calculated and balanced against their effectiveness. In other words, in the development of burn care improvements, cost-effectiveness analysis should also be taken into account.A first step in the economic evaluation of burn care is obtaining knowledge on the extent of burn care costs. We recently conducted a systematic literature review on the costs of burn care 6. A surprisingly high number of studies (n=156) could be included, unfortunately, the methodology of most studies was poor. Studies often provided limited information on cost calculations and components, and often used charges as a proxy to calculate costs. This impeded the presentation of total costs and cost distribution within burn care. Our review showed that the costs of burn patients per treatment or per day were often higher than other injuries 7-12. Total burn care costs in high-income countries varied widely in the different studies with a mean of $88,218 (€64,112) per patient and a range of $704-$717,306 (€512 – €521,298). Burn centre days and burn centre intensive care days proved to be major cost components 13-15 and amounted to 82% of the total burn care costs per patient. Surgery was another important cost category. The majority of studies calculated acute burn hospitalisation costs only. In two studies, from Sanchez et al., costs were described from a societal perspective, thus including caregiver, social and labour costs: medical costs represented only 10-20% of the total costs per patient, the other 80-90% of the costs related primarily to productivity losses and informal care 16,17. Patient characteristics predicting high costs were both flame burns 18-20

and extensive burns, but no further increase was reported above 80% total body surface area (TBSA) burned 18-22.Although existing literature gives a useful first impression of burn care costs, several questions remain. In the literature broad ranges of total burn care costs and burn centre days were presented, often based on hospital charges instead of real costs analyses. Therefore, it is questionable whether the mean burn care costs based on these study results correspond to burn care in a high-income country, such as the Netherlands. Furthermore, it is interesting to determine the extent of non-medical costs in patients with burns, and to see whether they are as high as predicted by Sanchez et al. 16. Therefore, the objective of this study was to

R1R2R3R4R5R6R7R8R9


54 | Chapter 3

give a detailed overview of all costs of patients with burns admitted in a burn centre in the Netherlands, with a three months follow-up.

Patients and Methods

The study was conducted in the burn centre of the Maasstad Hospital Rotterdam, the Netherlands, including all patients admitted with a burn related injury between 1 August 2011 and 31 July 2012. All children with burns over 5% and adults with burns over 10% TBSA could be referred. Additional referral criteria included burns of special areas (i.e. face, major joints), full thickness burns> 5% TBSA, and burns with associated inhalation injury. During a part of the study period (26 February until 31 July) the burn centre was limitedly available due to renovation activities: admission was restricted to adults ≤40% TBSA burned and children <10% TBSA burned only. Data regarding patients’ baseline characteristics and healthcare use were obtained from hospital patient records. Patients received a questionnaire three months post-burn with questions regarding extramural medical costs, and non-healthcare costs. Patients gave consent to participate in the study, for patients <18 years old, written informed consent was signed by the (both) parents/caregivers. Approval for the study was obtained from the ethics committee of the Maasstad Hospital (protocol 2011/34).

Costs analysis The cost analysis was performed in accordance with the Dutch guidelines 23. Costs were calculated from a societal perspective including direct healthcare costs (burn centre stay, other specialized burn care costs, and other healthcare costs), direct non-healthcare costs (patient costs and travel costs) and indirect non-healthcare costs (productivity loss). Real medical costs were calculated by multiplying the volumes of healthcare use with the corresponding unit prices. The costs applied to the financial year 2012. The costs of burn centre stay consisted of personnel (including burn physicians), material (excluding wound care), equipment, food, laundry and medication costs per days including a 41.9% increase for housing and overhead. We included a minimum of categories in the burn centre day costs (mainly fixed costs). All costs, which could be calculated separately per patient, were left out of the day price and are described in the paragraph below. Unit prices of ICU and non-ICU burn centre days were calculated according to the bottom up approach, following the micro costing method of Gold et al. with prices derived from our financial department 24. The unit price of day-care was determined using a proportion (37%) (according to Hakkaart et al. 23) of the unit price for a non-ICU hospital day.



Other specialized burn care costs consisted of diagnostics, surgical treatment, wound care and blood products.

• Costs of diagnostic procedures were calculated based on charges derived Hakkaart et al. 23

• The unit price of surgery was determined by micro-costing, taking into consideration the initial investment of equipment, investments during use, maintenance, number of years of use and discounting the number of procedures per year, material costs, personnel costs (per hour) and a 41.9% increase for housing and overhead in accordance with the guidelines. Personnel costs were determined by the average number of physicians, surgical and anaesthetic assistants per surgery, which was registered in a previous study 25.

• The unit prices of wound care were determined by microcosting; separate unit prices were calculated for patients with 1) hydrofiber wound dressings (often used in scalds <10% TBSA burned), 2) TBSA ≤20% (excluding patients defined as hydrofiber-patients), and 3) TBSA >20%. Cost estimation was based on a subset of patients (n=30), with prices derived from our financial department.

• The unit price of reconstructive surgery was determined before by micro-costing 26

• Unit prices of blood products (erythrocytes), pressure garments, silicone therapy and splints were derived from our financial department and Dutch guidelines 23.

• The unit prices of inpatient and outpatient consultations were derived from Dutch guidelines 23.

Other healthcare costs, including nursing-home care, rehabilitation centre care, visits to general practitioners and allied health professionals outside the hospital, were assessed by questionnaires, which were sent three months post-burn. Unit prices were derived from the Dutch guidelines 23.

Patient costs, including loss of economic productivity due to absence from work (by both patients and parents in case of patients < 18 years of age) and travel costs, were assessed by questionnaires three months post-burn. Unit prices were derived from the Dutch guidelines, hourly wages of €32 were used for our population 23. Productivity loss was estimated using the friction cost method, this method accounts for the fact that everybody is replaceable within a certain period of time, which is related to the unemployment rate and mobility of the labour market (friction period of 160 days) 23.

R1R2R3R4R5R6R7R8R9


56 | Chapter 3

Statistical analyses Data analysis was performed using SPSS Statistics 20.0. As cost data are typically highly skewed, non-parametric bootstrapping (1000 times) was used to calculate 95% confidence intervals (CI).A non-response analysis, regarding patients and other healthcare costs, was conducted to determine whether responders differed from non-responders regarding age, gender, burn size, length of stay (LOS), aetiology, ICU (yes/no) and surgery (yes/no). To correct for non-response bias, extrapolation to the total cohort was done using multiple imputation 27,28. Missing values for designated (imputed) variables were estimated from an appropriate distribution based on several predictor factors, which was repeated 5 times and pooled afterwards 27. Imputation was based on age, gender, aetiology, percentage TBSA burned, LOS and surgery, using the Markov chain Monte Carlo technique 29. To account for non-normality of the cost data, predictive mean matching was used in which “real” observed values from similar cases are imputed instead of imputing regression estimates 30,31. Subgroup analyses were performed to examine whether the total costs per patient differed by age, aetiology and percentage TBSA burned.

Results

Patient, injury and treatment characteristics A total of 249 patients were admitted to the burn centre of Rotterdam between 1 August 2011 and 31 July 2012. Their mean age was 29 years (range 0-91 years); 64.3% of the patients were male. Fire/flame burns (38.6%) and scalds (41.0%) were most common, the mean burn size was 8% (range 0.2-95%). The mean LOS was 12.2 days (range 0-92 days). A total of 20.5% of the patients required intensive care and 39.0% needed surgery (Table 1).



Table 1. Patient, injury and treatment characteristicsCharacteristics Total cohort

% (N = 249)

Responders %

(N=134)

Non-responders %

(N=115)

Gender (m/f)*Male 64 61 68Female 36 39 32

Mean age at injury (range)*† 29 (0-91) 31 (0-88) 27 (1-91)0 – 4 22.9 19.4 275 – 17 12.9 16.4 8.718 – 40 33.7 29.9 38.341 - 60 21.7 20.9 22.6>60 8.8 13.4 3.4

Aetiology*Fire/Flame 96 51 45Scald 102 53 49Fat/Hot oil 22 17 5Contact 15 9 6Other** 14 4 10

Mean % TBSA (range)* 8.0 (0.2-95) 8.0 (0.4-90) 7.9 (0.2-95)0-5% 54.6 47.0 63.5>5-10% 26.1 32.1 19.1>10-20% 13.3 15.7 10.4>20% 6.0 5.2 7.0

Body location burned***Arm 70.7 - -Leg 39.4 - -Head/Neck 43.0 - -Trunk 49.0 - -

Mean length of stay in days (range)*† 12.2 (0-92) 15 (0-92) 8.9 (0-92)ICU admission* 20.5 19 23Mean ICU admission in days (range) 1.8 (0-46) - -Surgery*† 39 50 26Mean number of surgeries (range) 0.7 (0-8) - -Deceased 1.6 - -Paid job before injury+ -mean hours worked/week

64 34

* Variable used for non-response analysis; ** Consists of chemical and electrical burns;,*** More than one location per patient is possible; †Significant difference between responders and non-responders at 5% level; + 11 missing

R1R2R3R4R5R6R7R8R9


58 | Chapter 3

Specialized burn care costs (until 3 months post-burn) Mean total specialized burn care costs were €20,549 (CI €16,922-€24,782) per patient, these costs consisted of burn centre stay (ICU and non ICU), diagnostics, treatment, consults and outpatient care. The costs of burn centre stay of €16,575 (€13,448-€20,016) accounted for 81% of the total specialized burn care costs (Table 2). The detailed composition of ICU (€2,966/day) and non-ICU burn centre days (€948/day) can be found in Figure 1; personnel costs represented the largest part of costs in both. Furthermore, surgery was an important cost driver in specialized burn care costs, with a mean of €1,298 per patient (overall, n=249) and €3,298 per surgically treated patient (n=98). Maximal surgery costs were €23,004 in a patient requiring 12 surgical procedures.

Table 2. Mean total costs per patient within three months post-burn (€, 2012) (N=249) Total costs (€)

(N=134)Total costs (€) (N=249)

Burn centre stayTransport hospital 461Non-ICU burn centre days 10,083ICU burn centre days 5,300Readmittance days 701Day-care 31Total burn centre stay [95% CI] 16,575 [13,448-20,016]

Diagnostic proceduresSwabs 394Lab tests 138Others 117Total diagnostic procedures [95% CI] 650 [482-831]

TreatmentSurgical treatment 1298Wound care 906Blood products (erythrocytes)* 55Pressure garments 212Silicon therapy 59Splints 2Reconstructive surgery 11Total treatment [95% CI] 2,544 [2,027 - 3,098]

Clinical consultations Physiotherapist 126Occupational therapist 27Social worker 33Dietitian 13Psychologist 5Skin therapist 2Psychiatrist 22



Speech therapist 8Internist 9Plastic surgeon 4Rehabilitation physician 10Others 69Total clinical consultations [95% CI] 328 [266 - 403]Outpatient burn careOutpatient wound care 295Outpatient scar care 83Occupational therapy 31Plastic surgeon 5Physiotherapist 10Rehabilitation physician 1After care nurse 18Others 8Total outpatient burn care [95% CI] 452 [410 - 497]

Total costs specialized burn care [95% CI] 20,549 [16,922 – 24,782]

Other healthcare costsRehabilitation centre (days) 145Nursing home (days) 438General practitioner 10Home (nursing) care (days) 114Extramural physiotherapy 39Total other healthcare costs [95% CI] 746

[292 - 1,380]619 ***

[331 - 964]

Patient costsWork absence (hours) 5,255Travel costs (km) 303Total patient costs [95% CI] 5,558

[4,174 – 6,857]5,372 ***

[4,581 – 6,151]

Total (mean) costs per patient [95% CI] 26.853 26,540 *** [22,320 – 31,238]

* Including, ophthalmologist, pulmonologist, cardiologist, (child) psychologist, skin therapist, speech therapist and internist** Including: skin therapist, social worker and dietician*** Extrapolation to n=249 based on age, gender, aetiology, percentage TBSA, LOS, and surgery (y/n)

Other healthcare costs and patient costs (until three months post-burn)The questionnaire was completed by 134 patients (54%). Responders had a longer LOS and more surgery than non-responders, elderly and teenagers were over-represented. Total non-burn centre related medical costs and non-medical costs were extrapolated to the total sample (N=249). Mean other healthcare costs per patient were €619 (CI €331-964), accounting for 2.3% of the total costs per patient and mainly consisted of nursing home days

R1R2R3R4R5R6R7R8R9


60 | Chapter 3

and rehabilitation centre days. Mean patient costs were €5,372 (CI € 4,581-6,151) and largely consisted of productivity loss. Patient costs accounted for 20% of the total costs per patient (Table 2). Among responders, 69% of patients with a paid job had returned to their work within 3 months (n=84). Figure 1. Composition of ICU burn centre days (€2,966) and non-ICU burn centre days (€948)

56%

7%

7%

30%

ICU hospital days

Personnel costs

Equipment/material/nutrition costs

Medication costs

Housing and overhead costs

64%4%

2%

30%

Non-ICU hospital days

Personnel costs


Medication costs


Figure 1. Composition of ICU burn centre days (€2,966) and non-ICU burn centre days (€948)

56%

7%

7%

30%

ICU hospital days

Personnel costs


Medication costs


64%4%

2%

30%

Non-ICU hospital days

Personnel costs


Medication costs


Figure 1. Composition of ICU burn centre days (€2,966) and non-ICU burn centre days (€948)

Total costs per patient and subgroups Mean total (healthcare and non-healthcare) costs per patient were €26,540 (CI €22,320 - €31,238), 77% of these costs consisted of specialized burn care. Costs varied from €742 to €235,557. In younger patients (age <18 years), total costs were lower, with a mean of



€12,671 (CI €10,167- €15,496), than in adult patients (age ≥18 years; Figure 2) with a mean of €342,258 (CI €28,689 - €40,521). Highest mean costs were seen in patients > 60 years old; €40,268 (CI €25,433 – €58,215), mainly due to high burn centre related costs (Figure 2). Other healthcare costs (including nursing home) were also high in the elderly (>60 years). Patient costs were relatively low in elderly, they were the highest in the working age (18-65 years), and consisted mainly of productivity loss. Mean costs of flame burns were significantly higher compared to other burns, except for patients with chemical burns (Figure 3). A significant increase in costs was shown with increasing burn size, mainly caused by burn centre related costs (Figure 4). Mean burn centre related costs per 1% TBSA were €4,245. This mean was relatively consistent across different TBSA subgroups (0-10%, 11-20%, >20%). In the four included patients with a TBSA >80%, no further increase in costs was seen: mean costs were €44.033 (SD: €80.949, range €3,501-€165,457). Three of these patients deceased shortly after hospitalisation. Figure 2. Mean costs after burn injuries (€) and 95% CI per age category

*Vertical bars representing the 95%CI ** Other specialized burn centre costs includes costs of diagnostic procedures, treatment costs, clinical consultation costs and outpatient burn care costs.

0-4 year(N=57)

5-17year

(N=32)

18-40year

(N=84)

41-60year

(N=54)

>60 year(N=22)

Patient costs € 3.344 € 2.547 € 7.200 € 7.793 € 1.809 Other healthcare costs € 176 € 26 € 552 € 837 € 2.355 Other specialized burn care

costs € 1.909 € 2.840 € 4.391 € 5.099 € 6.527

Burn centre stay costs € 6.960 € 10.529 € 17.888 € 22.633 € 29.576 Total € 12.388 € 15.942 € 30.031 € 36.361 € 40.268

€ -

€ 10.000

€ 20.000

€ 30.000

€ 40.000

€ 50.000

€ 60.000

€ 70.000

€, 2

012

Figure 2. Mean costs after burn injuries (€) and 95% CI per age category*Vertical bars representing the 95%CI** Other specialized burn centre costs includes costs of diagnostic procedures, treatment costs, clinical consultation costs and outpatient burn care costs.

R1R2R3R4R5R6R7R8R9


62 | Chapter 3

Figure 3. Mean costs after burn injuries (€) and 95% CI per aetiology

* Vertical bars representing the 95%CI ** Other specialized burn centre costs includes costs of diagnostic procedures, treatment costs, clinical consultation costs and outpatient burn care costs.

Scald(N=102

)

Flame(N=96)

Fat(N=22)

Contact(N=15)

Chemical

(N=11)

Electric(N=3)

Patient costs € 4.178 € 6.379 € 6.299 € 5.046 € 5.963 € 6.352 Other healthcare costs € 728 € 754 € 231 € 159 € 1 € 47 Other specialized burn care

costs € 2.985 € 5.376 € 3.557 € 3.052 € 3.887 € 911

Burn centre stay costs € 12.301 € 24.729 € 11.475 € 7.305 € 10.979 € 5.267 Total € 20.192 € 37.238 € 21.562 € 15.562 € 20.829 € 12.577

€ -

€ 5.000

€ 10.000

€ 15.000

€ 20.000

€ 25.000

€ 30.000

€ 35.000

€ 40.000

€ 45.000

€ 50.000 €,

201

2

Figure 3. Mean costs after burn injuries (€) and 95% CI per aetiology* Vertical bars representing the 95%CI** Other specialized burn centre costs includes costs of diagnostic procedures, treatment costs, clinical consultation costs and outpatient burn care costs. Figure 4. Mean burn care costs (€) and 95% CI categorized burn size

* Vertical bars representing the 95%CI ** Other specialized burn centre costs includes costs of diagnostic procedures, treatment costs, clinical consultation costs and outpatient burn care costs.

0-5%TBSA

(N=136)

5-10%TBSA(N=65)

10-20%TBSA(N=33)

20%TBSA(N=15)

Patient costs € 4.870 € 5.548 € 6.932 € 5.725 Other healthcare costs € 199 € 677 € 1.240 € 2.816 Other specialized burn care

costs € 1.697 € 3.648 € 9.354 € 14.237

Burn centre stay costs € 5.682 € 15.063 € 45.689 € 57.845 Total € 12.447 € 24.936 € 63.215 € 80.623

€ -

€ 20.000

€ 40.000

€ 60.000

€ 80.000

€ 100.000

€ 120.000

€, 2

012

Figure 4. Mean burn care costs (€) and 95% CI categorized burn size* Vertical bars representing the 95%CI** Other specialized burn centre costs includes costs of diagnostic procedures, treatment costs, clinical consultation costs and outpatient burn care costs.



Discussion

The aim of this study was to assess the economic burden of burn injuries. The mean total costs of burn injuries in our burn centre were €26,540 per patient within the first 3 months post-burn, with as most important cost drivers respectively burn centre days, work absence and surgery. Flame burns, adult patients and extensive burns corresponded to significantly higher costs. Our study used real costs instead of charges for cost calculations, important cost categories, like (ICU) burn centre days and surgery, were determined by micro-costing (bottom up approach). Furthermore, a societal perspective was adopted, meaning also non-burn centre related medical costs and patient costs were included to determine the total costs of burn injuries in a dedicated burn centre in the Netherlands. Burn care has often been identified as expensive care, especially compared to other well-known health-related problems in high-income countries 17,32-34. In this study, mean total burn centre related costs were €20,549 per patient and total costs were €26,540, which is notably lower than the mean total healthcare costs per burn patient as estimated in our costs review (€56,009). However, we should take into account that costs of burn injuries are strongly associated with TBSA. The mean TBSA in our study population was 8%, whereas most studies included in the review were on more extensive burns. Other factors explaining differences in costs between studies are methodological discrepancies, such as the use of charges instead of real costs derived from financial departments, and differences in time horizon and included cost categories 6. Recent studies on costs of burn injuries in high-income countries presented burn care costs varying from €6.436 to €73.398 per patient 16,35-37. A study from a high-income country similar to the Netherlands (Finland) presented comparable mean costs (€25,000) to our study, distribution per TBSA group was comparable as well 37. Our cost study resulted in lower mean costs than previous cost calculations of the financial department of our hospital. This can be attributed to different methods of cost calculations; e.g. the bottom up approach and the use of Dutch guidelines for in and outpatient consultations in our study.In our study, costs of burn centre stay, with non-ICU and ICU burn centre days being the most important cost drivers, accounted for 81% of the total specialized burn care costs, which is in accordance with a number of other studies 13-15. Costs per day in a burn centre (€948, non-ICU day, €2,966, ICU day, this study) are higher than costs of care for all patients types in a general hospital ward (€462, year 2012 23), which is mainly due to higher personnel costs. In previous literature mean total costs of other trauma patients were often estimated lower (€10,603-26,468) than of burn patients 6. All together, we can state that burn care indeed is expensive.

R1R2R3R4R5R6R7R8R9


64 | Chapter 3

Patient costs, which were hardly studied before, accounted for a substantial part (20%) of the total costs per patient and mainly consisted of productivity loss. Among our responders, 31% of patients with a paid job had not returned to their work within 3 months The only other study, to our knowledge, that included patients costs was that of Sanchez et al. who estimated sociolabor costs as being 80-90% of the total (mean) costs per patient in burn care 16,17. This substantially higher estimate can partly be attributed to a longer time horizon (12 months) and a more comprehensive inclusion of indirect costs components in the latter study, such as informal care.When looking at subgroups of patients, mean total healthcare costs of adult patients and patients with flame burns were in general higher than costs of other burns, which is consistent with previous literature 35. Scalds are often associated with lower costs as they usually cause superficial burns and therefore need fewer operations and a shorter LOS than some other types of burns 34. Moreover, a higher TBSA burned was associated with increased costs, which was also presented by other studies 33. Mean costs of 1% TBSA burned were €4,245, which is somewhat higher than the €2,974 as estimated in our literature review, and the estimated €3.000 in the recent study from Finland 37. A higher age was shown to be associated with higher costs. The highest costs were seen in patients >60. This was mainly caused by a longer LOS, and a higher frequency of other healthcare use such as nursing home days and rehabilitation centre days. Because our population is ageing, in near future, burn care costs are probably rising too. Therefore, also from a financial perspective, prevention of burns in elderly needs attention. This study has some limitations. First, part of the burn centre was limitedly available from 26th February 2011 onwards due to renovation activities, which had several implications for the admission of patients. Admission was restricted to adults ≤40% TBSA burned and children <10% TBSA burned only. As burn size proved to be a strong predictor for high costs, it is likely that the presented mean total costs per patient are an underestimation of the true costs per patient admitted in a Dutch burn centre. However, with a mean burn size of 8% and 70% of patients with a burns size below 10%; burn size was similar to patients in other Dutch burn centres as well as to patients admitted in US burn centres38. Second, as patients were asked to self-report their use of other health care resources, travel costs and productivity losses, recall-bias is likely. However, we have to rely on data as reported by respondents. Thirdly, our cost analyses were confined to burn patient related cost. We did not include hospital costs for 24/7 availability of general surgeons, who support the actual burn centre admission in out of office hours. Lastly, we did not take into account informal care. Given the short time horizon of this study this is not likely to have a major impact on total patients cost, however, on the long-term it could be an important factor as shown by Sanchez et al. 16. Notwithstanding this, our approach is in line with other studies in the field of costs analyses 39,40.



To conclude, mean total costs of burn care in the first three months post injury were estimated at €26,540 varying from €12.388 to €80,623 per subgroup, depending on age, aetiology and TBSA. Patient costs, accounted for a substantial part (20%) of the total costs per patient. Mean total cost in our population probably apply to other high-income countries as well, although we should realize that patients with burn injuries are diverse and represent a broad range of total costs. Further cost studies in burn care should focus on the long-term economic costs to determine the extent of specialized burn care costs (such as reconstructive surgery), of non-burn centre related medical costs (such as rehabilitation centre) and non-medical costs (such a permanent work absence). In addition, burn injuries have been shown to significantly impact quality of life in subgroups of patients 16,41. Additional insight in these patients could contribute to our understanding of the long-term consequences of burns. The results from this study can support us in the design of new studies on interventions in burn care. To reduce costs of burn care, future intervention studies should focus on a timely wound healing, reducing LOS and enabling an early return to work.

R1R2R3R4R5R6R7R8R9


66 | Chapter 3

References

1. Fratianne RB, Brandt CP. Improved survival of adults with extensive burns. Journal of Burn Care & Research 1997;4:347-52.

2. Wolf SE, Rose JK, Desai MH, Mileski JP, Barrow RE, Herndon DN. Mortality determinants in massive paediatric burns. An analysis of 103 children with> or= 80% TBSA burns (> or= 70% full-thickness). Ann.Surg. 1997;5:554.

3. Barrow RE, Herndon DN. History of treatments of burns. Total Burn Care, 3rd ed.Philadelphia, PA: Saunders-Elsevier 2007:1-8.

4. Herndon DN, Barrow RE, Rutan RL, Rutan TC, Desai MH, Abston S. A comparison of conservative versus early excision. Therapies in severely burned patients. Ann.Surg. 1989;5:547.

5. Herndon DN, Barrow RE, Kunkel KR, Broemeling L, Rutan RL. Effects of recombinant human growth hormone on donor-site healing in severely burned children. Ann.Surg. 1990;4:424.

6. Hop MJ, Polinder S, van der Vlies CH, Middelkoop E, van Baar ME. Costs of burn care: A systematic review. Wound Repair Regen. 2014 Jul;22(4):436-50.

7. Alinia S, Rezaei S, Daroudi R, Hadadi M, Sari AA. Extent, nature and hospital costs of fireworks-related injuries during the Wednesday Eve Festival in Iran. Journal of injury and violence research 2013;1:11.

8. Jiang X, Zhang Y, Wang Y, Wang B, Xu Y, Shang L. An analysis of 6215 hospitalized unintentional injuries among children aged 0–14 in northwest China. Accident Analysis & Prevention 2010;1:320-6.

9. Milenkovic M, Russo CA, Elixhauser A. Hospital Stays for Burns, 2004 2007;Available from http://www.ncbi.nlm.nih.gov/books/NBK63486/;.

10. Patil V, Dulhunty JM, Udy A, Thomas P, Kucharski G, Lipman J. Do burn patients cost more? The intensive care unit costs of burn patients compared with controls matched for length of stay and acuity. Journal of Burn Care & Research 2010;4:598-602.

11. Siritongtaworn P, Peunchompoo N. Economic problem of referred trauma cases in Siriraj Hospital. Journal of the Medical Association of Thailand 2006:1621-9.

12. Kagan RJ, Gamelli R, Kemalyan N, Saffle JR. Tracheostomy in thermally injured patients: Does diagnosis-related group 483 adequately estimate resource use and hospital costs?. The Journal of Trauma and Acute Care Surgery 2004;4:861-6.

13. Pellatt R, Williams A, Wright H, Young A. The cost of a major paediatric burn. Burns 2010;8:1208-14.

14. Hemington-Gorse SJ, Potokar TS, Drew PJ, Dickson WA. Burn care costing: the Welsh experience. Burns 2009;3:378-82.

15. Griffiths HR, Thornton K, Clements CM, Burge T, Kay A, Young A. The cost of a hot drink scald. Burns 2006;3:372-4.

16. Sanchez JLA, Bastida JL, Martínez MM, Moreno JMM, Chamorro JJ. Socio-economic cost and health-related quality of life of burn victims in Spain. Burns 2008;7:975-81.

17. Sanchez JA, Perepérez SB, Bastida JL, Martínez MM. Cost-utility analysis applied to the treatment of burn patients in a specialized center. Archives of Surgery 2007;1:50.

18. Latenser BA, Miller SF, Bessey PQ, Browning SM, Caruso DM, Gomez M, et al. National Burn Repository 2006: a ten-year review. Journal of Burn Care & Research 2007;5:635-58.

19. Miller SF, Bessey PQ, Schurr MJ, Browning SM, Jeng JC, Caruso DM, et al. National Burn Repository 2005: a ten-year review. Journal of burn care & research 2006;4:411-36.

20. Shields BJ, Comstock RD, Fernandez SA, Xiang H, Smith GA. Healthcare resource utilization and epidemiology of paediatric burn-associated hospitalizations, United States, 2000. Journal of Burn Care & Research 2007;6:811-26.

21. Schafer JL. Analysis of incomplete multivariate data.: CRC press; 2010.22. Jeng JC. Growth rings of a tree: progression of burn care charges abstracted from a decade of the

National Burn Repository. Journal of Burn Care & Research 2007;5:659-60.



23. Hakkaart-van Roijen L, Tan S, Bouwmans C. Handleiding voor kostenonderzoek. Methoden en standaard kostprijzen voor economische evaluaties in de gezondheidszorg. Geactualiseerde versie 2010.

24. Gold MR. Siegel JE, Russel LB, Weinstein MC. Cost-effectiveness in health and medicine. New York: Oxford University Press; 1996

25. Hop MJ, Bloemen MCT, van Baar ME, Nieuwenhuis MK, van Zuijlen PPM, Polinder S, et al. Cost study of dermal substitutes and topical negative pressure in the surgical treatment of burns. Burns 2014;3:388-96.

26. Hop MJ, Langenberg LC, Hiddingh J, Stekelenburg CM, van der Wal MBA, Hoogewerf CJ, et al. Reconstructive surgery after burns: a 10-year follow-up study. Burns 2014. DOI: 10.1016/j.burns.2014.04.014.

27. Taylor JM, Cooper KL, Wei JT, Sarma AV, Raghunathan TE, Heeringa SG. Use of multiple imputation to correct for nonresponse bias in a survey of urologic symptoms among African-American men. Am.J.Epidemiol. 2002;8:774-82.

28. Rubin DB. Multiple imputation for nonresponse in surveys.: John Wiley & Sons; 2009.29. Acock AC. Working with missing values. Journal of Marriage and Family 2005;4:1012-28.30. Grittner U, Gmel G, Ripatti S, Bloomfield K, Wicki M. Missing value imputation in longitudinal

measures of alcohol consumption. International journal of methods in psychiatric research 2011;1:50-61.

31. Horton NJ, Lipsitz SR. Multiple imputation in practice: comparison of software packages for regression models with missing variables. The American Statistician 2001;3:244-54.

32. Lopez Bastida J, Serrano Aguilar P, Monton Alvarez F, Duque González B. The economic burden of stroke in Spain. Value in Health 2003;6:615.

33. Lopez-Bastida J, Oliva Moreno J, Perestelo Perez L, Serrano Aguilar P. Determinants of health care costs in ambulatory patients living with HIV/AIDS. Value in Health 2005;6:A61.

34. Sahin I, Ozturk S, Alhan D, Açikel C, Isik S. Cost analysis of acute burn patients treated in a burn centre: the Gulhane experience. Annals of burns and fire disasters 2011;1:9.

35. Ahn CS, Maitz PK. The true cost of burn. Burns 2012;7:967-74.36. Jeevan R, Rashid A, Lymperopoulos N, Wilkinson D, James M. Mortality and treatment cost

estimates for 1075 consecutive patients treated by a regional adult burn service over a five year period: The Liverpool experience. Burns 2013;2:214-22.

37. Haikonen K, Lillsunde PM, Vuola J. Inpatient costs of fire-related injuries in Finland. Burns 2014.38. Latenser BA, Miller SF, Bessey PQ, Browning SM, Caruso DM, Gomez M, Jeng JC, Krichbaum

JA, Lentz CW, Saffle JR, Schurr MJ, Greenhalgh DG, Kagan RJ. National Burn Repository 2006: a ten-year review. J Burn Care Res. 2007 Sep-Oct;28(5):635-58

39. Polinder S, Heijnen E, Macklon N, Habbema J, Fauser B, Eijkemans M. Cost-effectiveness of a mild compared with a standard strategy for IVF: a randomized comparison using cumulative term live birth as the primary endpoint. Human reproduction 2008;2:316-23.

40. Damen T, Wei W, Mureau M, Tjong-Joe-Wai R, Hofer S, Essink-Bot M, et al. Medium-term cost analysis of breast reconstructions in a single Dutch centre: a comparison of implants, implants preceded by tissue expansion, LD transpositions and DIEP flaps. Journal of Plastic, Reconstructive & Aesthetic Surgery 2011;8:1043-53.

41. van Loey NE, van Beeck EF, Faber BW, van de Schoot R, Bremer M. Health-related quality of life after burns: a prospective multicenter cohort study with 18 months follow-up. J.Trauma.Acute Care.Surg. 2012;2:513-20.

Chapter 4Reconstructive surgery after burns:

a 10-year follow-up study

M. Jenda Hop Lisette C. Langenberg

Jakob Hiddingh Carlijn M. Stekelenburg

Martijn B.A. van der Wal Cornelis J. Hoogewerf

Maaike L.J. van Koppen Suzanne Polinder

Paul P.M. van Zuijlen Margriet E. van Baar

Esther Middelkoop

Burns. 2014 Dec;40(8):1544-51

R1R2R3R4R5R6R7R8R9


70 | Chapter 4

Abstract

Background: There is minimal insight into the prevalence of reconstructive surgery after burn injuries. The objective of this study was to analyse the prevalence, predictors, indications, techniques and medical costs of reconstructive surgery after burn injuries.Methods: A retrospective cohort study was conducted in the three Dutch burn centres. Patients with acute burns, admitted from January 1998 until December 2001, were included. Data on patient and injury characteristics and reconstructive surgery details were collected in a 10-year follow-up period.Results: In 13.0% (n=229/1768) of the patients with burns, reconstructive surgery was performed during the 10-year follow-up period. Mean number of reconstructive procedures per patient were 3.6 (range 1-25). Frequently reconstructed locations were hands and head/neck. The most important indication was scar contracture and the most applied technique was release plus random flaps/ skin grafting. Mean medical costs of reconstructive surgery per patient over 10-years were € 8342.Conclusions: With this study we elucidated the reconstructive needs of patients after burn injuries. The data presented can be used as reference in future studies that aim to improve scar quality of burns and decrease the need for reconstructive surgery.


Reconstructive surgery after burns: a 10-year follow-up study | 71

Background

The significant reduction in mortality in patients with burns in the past decades has resulted in a shift in attention from mortality to improving functional outcome, including scar quality1. Burn scars may cause severe functional and aesthetic problems. Since the problem of burn scars is complex, different treatment options are used. Next to pressure garments, splinting, silicones, laser therapy, massage and corticosteroids, reconstructive surgery may be necessary in problematic burn scars2-8. The need for reconstructive surgery is an important long term outcome indicator for patients with burn scars. In our search for an optimal scar quality, a decreased need for reconstructive surgery is one of the endpoints. Fortunately, not all patients with burns undergo reconstructive surgery. However, the question is, which patients require reconstructive surgery, and what type of reconstructive surgery is performed in burn scars?In literature a broad overview of possible reconstructive techniques is presented. Nevertheless, epidemiologic information on prevalence, indications, locations and techniques of reconstructive surgery after burns is scarce. Only two articles were found on the epidemiology of reconstructive surgery after burns. One relatively old retrospective study, from the USA in 1991, described the prevalence of reconstructive surgery in patients who were initially admitted to a burn centre for acute burn care. In 19.9% of the patients reconstructive surgery was performed. Patients requiring reconstructive surgery were younger and had a larger burn size9. The most applied technique was release and grafting, and the most reconstructed locations were arms, hands and neck. A recently published article from our research group presented the prevalence and predictors of reconstructive surgery to the head and neck after burn centre admission: facial reconstructive surgery was performed in 5.3% of all facial burn patients, significant predictors were burns to the neck, fire/flame burns and number of facial surgeries in the acute phase of the wound healing process10. Nowadays, insight in costs of healthcare is becoming increasingly important because healthcare costs are rising and pressure on budgets too. To our knowledge, literature on the costs of reconstructive surgery after burns in high-income countries is lacking. We noted one article describing healthcare costs of reconstructive surgery after burns, in one year, in Nigeria. Mean reconstructive surgery costs per patient represented the highest cost category (35%) of the total calculated medical costs in rehabilitation phase ($1301). Hospital stay was the second highest cost category11. Thus, in current research, there is limited knowledge on the prevalence, predictors, indications and costs of reconstructive surgery after burn injuries. Insight in prevalence and predictors of reconstructive surgery is essential to improve our understanding of the reconstructive needs of burn patients9. In addition, the need for reconstructive surgery is an important long term outcome parameter of burn care, representing the quality of the burn scar. However, in order to measure a decline in reconstructive surgery we need a baseline. Therefore, the

R1R2R3R4R5R6R7R8R9


72 | Chapter 4

prevalence, high risk populations and anatomical locations for reconstructive surgery after burns should be identified. Furthermore, the costs involved in reconstructive surgery after burns must be examined, to establish which potential savings can be achieved when we will be able to decrease the need for reconstructive surgery. The objectives of this study were to analyse 1) the prevalence for reconstructive surgery after burn injuries, 2) the predictors for reconstructive surgery, 3) the indications and techniques of these reconstructions and 4) the medical costs of reconstructive surgery after burns.

Methods

Study design and populationA retrospective study was conducted, including all patients with acute burns admitted to one of the Dutch burn centres (Beverwijk, Groningen, Rotterdam) from January 1998 until December 2001. Acute burns were defined as all burns before initial wound closure. Patients were excluded from the study if they died within 6 months post burn, further treatment was continued abroad after discharge or information on the need for reconstructive surgery was not available. Data were collected on gender, age, aetiology, body location burned, percentage of Total Body Surface Area (TBSA) burned, percentage of full thickness TBSA burned, number of surgeries in acute phase and the need for reconstructive surgery in a 10-year follow-up period. Reconstructive surgery was defined as all surgical procedures performed by a plastic surgeon or burn physician on burn scars, i.e. after initial wound closure. Detailed data on the reconstructions were collected, including date of surgery, location of surgery and indications and technique per location. In addition, data on healthcare consumption were collected: duration of surgical treatment, surgical personnel (including physicians) and length of hospital stay (both hospital days and day care). Data were derived from historical databases of the Dutch burn centres and patient records. This study was approved by the medical ethical board of the Maasstad Hospital (protocol 2012/16) and local hospitals.

Referral criteria and treatment protocolsUp tot 1999 referral to a Dutch burn centre was advised in burns >25% TBSA in adults or deep burns >10% TBSA, and in burns >10% TBSA in children and elderly, irrespectively the depth. In addition, referral was advised in minor burns associated with another injury or pre-existent disease that may increase the risk for complications. In 1998 the course Emergency Management of Severe Burns (EMSB) was introduced, with new referral criteria: from then all children with burns over 5% and adults with burns over 10% TBSA were advised to be referred12. Between 1998-2001, the most used topical agents were silver sulphadiazine, and cerium nitrate-silver sulphadiazine. Hydrofiber dressings were introduced in 1999. Early



tangential excision was not performed regularly: only in obviously full thickness wounds excision and autografting was performed within one week post burn, in indeterminate depth wounds decision for surgery was made approximately two weeks post burn. In our burn centres, general surgeons and burn physicians were responsible for the treatment of acute burn wounds. In severe hand of facial burns a plastic surgeon was involved in the acute phase as well. After wound closure, silicones and custom made pressure garments were applied, depending on location and scar activity. Patients with problematic scars were referred to a plastic surgeon13,14.

Statistical analysisData analysis included univariable and multivariable logistic regression (forward stepwise LR) to identify predictors of reconstructive surgery. Relative risks were estimated by odds ratios (ORs) with 95% confidence intervals (CIs). In the multivariable analyses, variables were checked for multicollinearity (Spearmans ρ>0.75). In the multivariable analyses possible predictors were included based on theory and p-values (<0.20) from the univariable analyses. Multivariable analyses were performed with at least 10 cases for each estimated parameter. Additionally, risk profiles were calculated using the β’s derived from the multivariable analyses. Data analysis on details of reconstructive procedures was limited to descriptive statistics.The cost analysis was conducted from a healthcare perspective during the follow-up period of 10-years and included direct medical costs of the reconstructive procedures and length of hospital stay (plastic surgery ward) for reconstructive procedures. Because of the retrospective character of this study, other medical costs, such as outpatient treatment, consultations of paramedics, pressure garments, splints, etc., and non-medical costs, like productivity loss, were not included. Direct medical costs of the acute burn centre admission were also not included. Costs were calculated by multiplying healthcare consumption with corresponding cost prices. A bottom up approach was applied (according to the micro-costing method of Gold et al. 1996) based on a detailed inventory of all resources used to calculate the surgery costs15. Cost prices of the reconstructive procedures were inventoried in the financial department of one burn centre and were used for all centres. Cost prices of hospital days were derived from the Dutch guideline for standardization of costs16. Inflation adjustment was applied to the year 2011 and costs were reported in Euro (€).Data were analysed using PASW (Predictive Analytics SoftWare) Statistics 18.0 (IBM, New York City).

R1R2R3R4R5R6R7R8R9


74 | Chapter 4

Results

InclusionsA total of 1907 patients had been admitted in the three Dutch burn centres between Jan 1998-Dec 2001. In total 7.3% (n=139) of the patients was excluded, because of mortality within 6 months post burn (n=114), lost to follow-up after discharge (n=17) or lack of information on the possibility of reconstructive surgery (n=8). This resulted in 1768 included patients.

Prevalence of reconstructive surgery and predictorsIn 13.0% (n=229) of the included patients reconstructive surgery was performed during the 10-year follow-up period. The highest prevalence of reconstructive surgery was seen in the patients admitted in 2001: 15.0%. The mean prevalence of 1998, 1999 and 2000 combined was 12.2% (Table 1).

Table 1. Prevalence of reconstructive surgery in patients admitted with acute burns in a 10-year follow-up period

% (n/total n)1998 14.1% (57/403)1999 12.3% (51/414)2000 10.5% (50/477)2001 15.0% (71/474)Overall mean 13.0% (229/1768)

The extremes of age had lower odds of requiring reconstructive surgery. Patients with fire/flame burns had higher odds, and patients with scalds had lower odds of requiring reconstructive surgery. Patients with higher percentage TBSA burned and patients with higher percentage full thickness TBSA burned more frequently required reconstructive surgery. Patients that required surgery in the acute phase of the burn had higher odds of reconstructive surgery. In addition, patients with a higher number of surgical interventions in the acute phase of the burn more frequently required reconstructive surgery (Table 2, Univariable odds).



Significant independent predictors for reconstructive surgery, in multivariable regression analysis, were: burns to the arms (including hands and shoulders) (OR 3.16), fire/flame burns (OR 1.50), number of surgical interventions in acute phase (OR 1.74) and a higher percentage TBSA burned (OR 1.02) (Table 2, Multivariable odds). Based on these results, 2 risk profiles were calculated:

• The risk of requiring reconstructive surgery for a patient with burns caused by fire/flames, 20% TBSA burned, burns of the arms, who required 2 surgical procedures before initial wound closure was 28.9%.

• In contrast, the risk of requiring reconstructive surgery for a patient with a none fire/flame burn of 5% TBSA, excluding the arms, who required no surgery for initial wound closure was 1.9%.

Prevalence of reconstructive surgery per locationSubgroup analyses in patients with hand burns showed a prevalence of reconstructive surgery of the hands of 15.1% (85/562). Sub-analyses in patients with head and neck burns showed a prevalence of reconstructive surgery of head and neck of 8.9% (64/652). Predictors for reconstructive surgery in both locations were comparable to the complete group: fire/flame burns, number of surgical interventions and a larger percentage TBSA burned (significant OR univariable analysis, data not shown).

Timing and number of reconstructionsThe mean time to the first reconstructive procedure was 1.6 years post burn (range 0.1-9.8 years, n=227, 2 missing). The mean time post burn of all reconstructions was 2.9 years (range 0.1-9.9 years, n=805, 12 missing). Most of the reconstructions were conducted within two years post burn: 26.3% within one year and 25.6% between 1-2 years post burn (Figure 1.a/b). The majority of patients (62.9%) underwent more than one reconstructive procedure (Figure 2). The mean number of reconstructions per patient was 3.6 (range 1-25) and a mean of 1.7 (range 1-7) body locations were operated on, per reconstruction.

R1R2R3R4R5R6R7R8R9


76 | Chapter 4

Tabl

e 2.

Pat

ient

, inj

ury

and

trea

tmen

t cha

ract

eris

tics

for p

atie

nts

with

or w

ithou

t rec

onst

ruct

ive

surg

ery

incl

udin

g pr

edic

tors

for r

econ

stru

ctiv

e su

rger

yR

econ

stru

ctiv

e su

rger

y%

(n=

229)

No

reco

nstru

ctiv

e su

rger

y%

(n=

1539

)U

niva

riabl

e od

ds

ratio

(95%

CI)

Mul

tivar

iabl

e od

ds ra

tio

(95%

CI)

****

Tota

l 13

.087

.0G

ende

r (m

/f)M

ale

60.8

63.8

0.88

(0.6

6-1.

17)

Fem

ale

39.2

36.2

ref.

Mea

n ag

e at

inju

ry (S

D)

28.4

(18.

7)24

.8 (2

3.6)

0-4

12.2

33.4

1.03

(0.4

4-2.

41)

5-17

21.8

12.9

4.78

(2.1

0-10

.85)

18-4

035

.828

.03.

59 (1

.62-

7.97

)41

-60

27.1

17.1

4.46

(1.9

9-10

.02)

>60

3.1

6.0

ref.

Etio

logy

Fire

/Fla

me

74.4

42.3

3.97

(2.8

6-5.

50)*

**1.

50 (1

.00-

2.25

)S

cald

12.3

41.1

0.20

(0.1

3-0.

31)

Oth

ers

13.3

16.6

0.77

(0.5

1-1.

17)

Mea

n %

TB

SA

(SD

)20

.6 (1

8.7)

8.1

(8.2

)P

er 1

%1.

08 (1

.06-

1.09

)1.

02 (1

.01-

1.04

)0-

1037

.775

.6>1

0-20

27.6

17.6

>20-

3011

.44.

3>3

023

.22.

5M

ean

% F

ull T

hick

ness

TB

SA

(SD

)11

.3 (1

3.6)

2.2

(4.4

)P

er 1

%1.

14 (1

.12-

1.17

)0

10.6

46.0

>0-1

052

.749

.5>1

0-20

19.5

3.6

>20-

308.

00.

5>3

09.

30.

5



Bod

y lo

catio

n bu

rned

*H

ead

and

neck

69.7

50.8

2.23

(1.5

9-3.

13)

Trun

k64

.059

.81.

20 (0

.86-

1.66

)A

rms*

*92

.166

.95.

80 (3

.31-

10.1

4)3.

16 (1

.74-

5.74

)H

ands

74.6

37.2

4.96

(3.4

6-7.

10)

Legs

46.6

37.7

1.44

(1.0

5-1.

98)

Sur

gica

l tre

atm

ent i

n ac

ute

phas

e88

.645

.79.

22 (6

.04-

14.0

6)

Mea

n nu

mbe

r of s

urge

ries

in a

cute

pha

se (S

D)

2.8

(3.5

)0.

6 (1

.3)

1.85

(1.6

6-2.

06)

1.74

(1.5

1-2.

00)

Bol

d nu

mbe

rs w

ere

stat

istic

ally

sig

nific

ant (

p<0.

05),

ref.

= re

fere

nce

grou

pM

issi

ng v

alue

s: g

ende

r 18

, age

16,

etio

logy

164

, TB

SA

23, F

ull T

hick

ness

TS

BA

23, h

ead

and

neck

, tru

nk, a

rms

and

legs

424

, han

ds 4

34, s

urgi

cal

treat

men

t 167

, num

ber o

f sur

gerie

s 28

5. *m

ore

than

one

loca

tion

per p

atie

nt is

pos

sibl

e, **

incl

udin

g ha

nds,

***r

efer

ence

gro

up: a

ll ot

hers

, ***

* the

follo

win

g va

riabl

es w

ere

incl

uded

in th

e m

ultiv

aria

ble

odds

: flam

e, T

BS

A, h

ead

and

neck

, arm

s, le

gs, n

umbe

r of s

urge

ries

in a

cute

pha

se, i

n to

tal 4

49 m

issi

ng v

alue

s (m

ultic

ollin

earit

y ag

e an

d fla

me,

mul

ticol

linea

rity

FT T

BS

A an

d nu

mbe

r of s

urge

ries,

mul

ticol

linea

rity

arm

s an

d ha

nds)

R1R2R3R4R5R6R7R8R9


78 | Chapter 4

Figure 1.a Time to first reconstruction (N=227, 2 missing)

Figure 1.a Time to first reconstruction (N=227, 2 missing)Figure 1.b Time to all reconstructions (N=805, 12 missing)

Figure 1.b Time to all reconstructions (N=805, 12 missing)



Figure 2. Number of reconstructions per patient (N=229)

Figure 2. Number of reconstructions per patient (N=229)

Indication, technique and location of reconstructionsThe most frequent locations of reconstruction were hands (36.0%), followed by head and neck (32.0%) (Table 3). The majority of reconstructed locations, 91.0% (1237/1360, 42 missing), was also surgically treated in the acute phase. Contractures were the most dominant indication for reconstructive surgery. Several contractures were further specified in the medical records, these consisted mainly of webspace contractures and flexion contractures, respectively 19.0% and 7.3% of all reconstructions. Other frequent reconstruction indications were hypertrophic scars (5.7%) and instable scars (5.4%). The most applied technique was: release/excision and flaps (50.5%), followed by release/excision and skin grafts (32.4%). Random flaps were the most frequently used flaps: 94.7% of all flaps. Split skin grafts were the most used grafts: 62.0% of all grafts (Table 3).The earliest reconstructive procedures (within four months post burns, n=362) were mainly conducted on eyes 21.3%, hands 21.3%, elbows 17.3% and neck 16.0%; most often because of contractures: 81.6% (including ectropion 12.7%).

R1R2R3R4R5R6R7R8R9


80 | Chapter 4

Table 3. Details of reconstructions: location, indication and technique of surgeryLocation %, n= 1402*Head and neck, Total 32.0

Lower eyelids 4.0Face and scalp n.o.s. 18.1

Neck 9.9Arms, Total 57.4

Axilla 9.2Elbow 5.5

Arm, n.o.s. 6.8Hands 36.0Digits 12.8

Hand and wrist, n.o.s.** 23.2Trunk, Total 6.2Legs, Total 4.4IndicationContractures, Total 72.3

Flexion contracture 7.3Extension contracture 1.0

Webspace contracture 19.0Contracture n.o.s. 41.5

Ectropion 3.5Other scar problems 27.7

Instable scar 5.4Hypertrophic scar 5.7

Pigment/relief/colour/contour n.o.s. 9.3Others *** 7.3

TechniqueRelease/excision + skin grafting 32.4

Split skin graft 20.1Full thickness graft 12.0

Dermal substitute + graft 0.4Release/excision + flaps 50.5

Random flaps, Z-plasty 10.0Random flaps, others 29.8

Random flap + skin grafting 8.0Other flaps **** 2.7

Others ***** 17.1

* location, indication and techniques, resp. 9, 79 and 76 missing **including webspaces ***including ao: bouttoniere deformity, nailbed deformity, microstomia and alopecia **** including axial and perforator flaps ***** including ao: excison and primary closure, tenolyse/artrholyse and laser/dermabrasion

CostsThe mean costs per surgical procedure were € 1294 (range €591-3387). The mean costs of hospital stay (regular plastic surgery ward) per reconstructive procedure were €922 (range €0-14 509). With a mean number of reconstructions per patient of 3.6, mean total direct medical



costs were €8342 (range €928-70 067) per patient, in a ten year follow-up period (Table 4.a). In Table 4.b mean costs per sub-groups of patients are shown, mean medical costs in patient with one reconstructive procedure were €2080 (range €928-7610), rising to €35 548 (range €15 350-70 067) in patients with 11-25 reconstructive procedures.

Table 4.a Mean direct medical costs of reconstructive surgery (€, 2011) per patient, in a 10-year follow-up period

Cost price Mean costs/procedure

Mean costs/patient

N=817 N= 229Reconstructive procedures (N) 1 3.6 (1-25)

Personnel* 515/hour 762 2719 Equipment 89 89 317

Material 104 104 371Housing and overhead 35.5% 339 1209

Total (range) 1294 (591-3387) 4617 (674-43 025)Hospital days (range) **Day care (range) **

440254

922 (0-14 509)122 (0-254)

3289 (0-33 809) 436 (0-3552)

Total per patient (range) 2338 (850-16 381) 8342 (928-70 067)

* mean personnel price of all OR personnel (surgeon, anesthetist, OR assistant, anesthetist assistant, recovery nurse, pre-operative consuls anesthetist), ** only costs of reconstruction related days were calculated

Table 4.b. Mean costs per patient in subgroups (€, 2011), in a 10-year follow-up periodTotal number or reconstruction per patient

1 procedure 2-3 procedures 4-6 procedures 7-10 procedures 11-25 procedures

N=85 N=67 N=43 N=21 N=12Surgical procedure 1167 2858 6058 11 262 20 734Hospital daysDay care

767146

2282336

5177491

6836959

13 0001813

Total per patient(range)

2080 (928-7610)

5476 (1864-36 501)

11 726 (4627-28 536)

19 057 (8260-42 590)

35 548 (15 350-70 067)

Discussion

This is the first study since many years on the epidemiology of reconstructive surgery after burns. To our knowledge, this is the only study that extensively described predictors, indications, techniques and costs of reconstructive surgery after burns. The prevalence of reconstructive surgery after burn injuries in a 10-year follow-up period was 13.0%. Predictors for reconstructive surgery were burns to the arms, fire/flame burns, number of surgical

R1R2R3R4R5R6R7R8R9


82 | Chapter 4

interventions in the acute phase and a larger burn size. The majority had more than one reconstruction, most often within 2 years post burn. Frequently reconstructed locations were hands, head and neck. The most important indications for surgery were scar contractures. The most frequently used techniques were release/excision plus random flaps followed by release/excision plus skin grafting. Mean medical costs of reconstructive surgery per patient were €8342 (range €928-70 067) in a ten year follow-up period.The prevalence of reconstructive surgery (13.0%) in our burn population can be seen as a long term outcome measure of patients with burns. In the ideal situation of perfect burn care, reconstructive surgery would not be necessary. Studies that aim for an improved scar quality in burns, for example studies on the use of dermal substitution in acute burn surgery17-19, can use our prevalence as a baseline to compare their outcomes to, regarding the need of reconstructive surgery. The prevalence of reconstructive surgery in our study was lower than the prevalence shown in the study of Prasad et al. (19.9%)9. Possibly, improvements in acute burn care are responsible for the lower prevalence found in our study compared to Prasad et al. in which acute burn patients were studied between 1970-1985. Also the more frequent use of non-surgical therapies in the rehabilitation phase, like silicones and laser therapy may have caused a decrease in need for reconstructive surgery. Another reason for these differences may be attributed to dissimilarities in patient characteristics study samples: mean burn size and mean full thickness burn size was higher in the study of Prasad et al. (16.4% and 5.8%) than in our study (9.8% and 3.4%). Furthermore, follow-up time differed: 10-years in our study vs. 5-20 years in the study of Prasad et al.. The study of Hoogewerf et al. (limited to head and neck burns)10, on the other hand, presented a lower prevalence (5.3%) of head and neck reconstructions than our study 8.9%. This was probably due to the limited follow-up time in the study of Hoogewerf et al. (2-7 years). Similar to our study, Hoogewerf et al. showed that fire/flame and number of surgeries in acute phase were independent predictors for reconstructive surgery. Flame burns are often more severe (in size and depth) than, for example, scalds, and more often require surgical treatment in the acute phase. This is the most likely reason that patients with flame burns required more reconstructive surgery. Prasad et al. observed more reconstructions in patients with a larger burn size, a larger full thickness burn size, aged 18-49 years and in burns of the arms followed by head and neck burns, which was also similar to the findings in our study. Possible explanations for the lower percentage of reconstructive surgery in patients at extremes of age were: postponing of reconstructive surgery until growth was completed, being not ‘ fit’ enough for surgery and fewer scar contractures in the elderly skin.In several studies, delaying of reconstructive surgery is advised until scar maturation is complete (for one year or more). Many scars do not require surgery once the phase of scar maturation is achieved2. Furthermore, the risk of new scar problems is reduced20. However, in some cases surgery should not be postponed; for example when vital structures are exposed or functions



are hampered (e.g. ectropion, microstomia), in severe contractures (e.g. contractures causing growth disturbance, or neck contractures causing airway problems) or in instable scars 2,4,6,7. In our study population, approximately 25% of the reconstructions were performed within one year post burn; waiting for scar maturation in these cases was apparently considered inappropriate. In accordance with literature9, we observed that most patients needed more than one reconstruction (mean 3.6) and most frequently because of functional problems, i.e. scar contractures. Whereas the most recommended reconstructive techniques used to be release and skin grafting or random flaps2,5,21,22, currently also other techniques like perforator based flaps22,23, dermal substitutes2,4,22 and lipofilling24 are applied increasingly. In our follow-up period, release and skin grafting (32.4% of all reconstructions) still appeared to be a ‘workhorse’ in reconstructive surgery, as previously described in literature2,21,22. Furthermore, even more random flaps (47.8%), than skin grafts were used for burn reconstructions. This was also previously advised by Barret et al. and Sniezek et al.2,5. In our study, more complex reconstructions were hardly applied. This minimal application can probably be explained by the limited indications for more complex flaps and the broad time frame (1998-2012) of the follow-up, including reconstructive habits of 15 years ago.Although acute burn care improved in past decades, there still are several questions to answer in order to further improve care: e.g. what is the optimal timing for excision and grafting22, what is the role of dermal substitution in acute burns, and what is the role of devices like splints and pressure garments in improving scar quality? Answers to these questions possibly lead to a further decrease in the need for reconstructive surgery and the related healthcare costs. Until now, there was minimal insight into the costs involved in reconstructive surgery after burns. Ahachi et al. (Nigeria) were the only to present costs of patients presenting with post-burn ulcers or other complications that required reconstructive surgery, pressure garments or other agents. In their prospective cost analysis during one year, not only hospital days and surgery costs (like in our retrospective study), but also costs of splints, pressure garments, physiotherapy, and investigations were taken into account. Patient and injury characteristics of their population were unclear. A mean price of $1308 per patient was presented11. Our study was the first to present costs of reconstructive surgery in a high-income country. The mean costs of reconstructive surgery including hospital admittance were $9273 per patient in a ten year follow-up, 64.3% of all costs were made within 2 years post burn. Our results indicate that not only acute burn care is expensive, but that healthcare costs can still be significant several years post burn. In a recent review, the mean costs of acute burn care in high-income countries per patient were $88 218 (submitted review Hop et al.). Therefore, the highest medical costs of patients with burns are made in the acute phase.Some limitations of this study should be mentioned. First of all, due to the retrospective design, some patients were probably lost to follow-up, but we do not expect this to significantly affect prevalence and reconstruction details, since we assumed that most of the reconstructive

R1R2R3R4R5R6R7R8R9


84 | Chapter 4

surgery was performed in one of the three nationwide burn centres and not in other hospitals. Our prevalence of reconstructive surgery was influenced by The Volendam café fire in 200126. However this effect was small with a prevalence of 15.0% in the cohort of 2001, versus 10.5-14.1% in other years. Likely, also other retrospectively non-detectable factors, such as logistics (number of plastics surgeons/OR time available, referral pattern to plastic surgeon), influenced the prevalence and timing of reconstructive surgery in our population. Furthermore, we do not have insight in the number of patients that declined reconstructive surgery. However, we do not think that this was a high number, because reconstructive surgery is covered by health insurance in our country. In addition, we do not have data on skin type and thus have not gained any insight in the relation between skin type and scar quality. However, the majority of included patients was Caucasian, with skin types 1 or 2. For future studies, it would be good to include skin type or ethnicity to elucidate the relation between skin type and scar quality/contractures and the subsequent need for reconstructive surgery. Finally, due to the retrospective design and limited data in patient records, it was not possible to calculate all medical costs during follow-up. Outpatient treatment, including conservative scar treatment (silicones, pressure garments, etc.) was not clearly documented. Besides that, we calculated healthcare costs only, total societal costs would have been even higher if additional non-healthcare costs such as productivity loss had been included. However, as in earlier studies, length of hospital stay and surgical procedures represent major cost items in burn care27,28, therefore this study included the most important medical costs items.With this study we elucidated the reconstructive needs of patients after an admission due to burn injuries in the Dutch burn centres. To decrease the need for reconstructive surgery in burn patients and the subsequent economic burden, we should keep searching for further improvements in burn care. Our data can be used as reference in future studies that aim to improve scar quality of burns.

Conflict of interest statement The authors declare that they have no conflict of interest to disclose.

AcknowledgementsThe authors thank G.I.J.M. Beerthuizen, MD PhD, J. Dokter, MD, E.C. Kuijper, MD and A.F.M.P. Vloemans, MD PhD, for their contribution to the historical database. We also thank R. Tjong, MD and S.J.M. Jongen, MD, for their permission to use the reconstructive surgery data. Furthermore, we would like to thank H. Eshuis, RN and S. de Waard, MD, for collecting part of the data. This research was financially supported by a grant of the Dutch Burns Foundation (11.102).



References

1. van Baar ME, Essink-Bot ML, Oen IM, Dokter J, Boxma H, van Beeck EF. Functional outcome after burns: a review. Burns. 2006 Feb;32(1):1-9.

2. Barret JP. Burns reconstruction. BMJ. 2004 Jul 31;329(7460):274-6.3. Serghiou M, Cowan A, Whitehead C. Rehabilitation after a burn injury. Clin Plast Surg. 2009

Oct;36(4):675-86.4. Grevious MA, Paulius K, Gottlieb LJ. Burn scar contractures of the pediatric neck. J Craniofac Surg.

2008 Jul;19(4):1010-5.5. Sniezek J, Sabri A, Burkey B, David B (2000). Reconstruction after burns of the face and neck. Curr

Opin Otolaryngol Head Neck Surg.6. Huang T. Overview of burn reconstructions. In: Herndon DN. Total burn care. 3th edition. Galveston,

USA: Saunders;2007.7. Armour AD, Billmire DA. Pediatric thermal injury: acute care and reconstruction update. Plast

Reconstr Surg. 2009 Jul;124(1 Suppl):117e-127e.8. Bloemen MC, van der Veer WM, Ulrich MM, van Zuijlen PP, Niessen FB, Middelkoop E. Prevention

and curative management of hypertrophic scar formation. Burns. 2009 Jun;35(4):463-75.9. Prasad JK, Bowden ML, Thomson PD. A review of the reconstructive surgery needs of 3167

survivors of burn injury. Burns. 1991 Aug;17(4):302-5.10. Hoogewerf CJ, van Baar ME, Hop MJ, Bloemen MC, Middelkoop E, Nieuwenhuis MK.Burns to the

head and neck: Epidemiology and predictors of surgery. Burns. 2013 Sep;39(6):1184-92. 11. Ahachi CN, Fadeyibi IO, Abikoye FO, Chira MK, Ugburo AO, Ademiluyi SA. The =direct

hospitalization cost of care for acute burns in Lagos, Nigeria: a one-year prospective study. Ann Burns Fire Disasters. 2011 Jun 30;24(2):94-101.

12. Vloemans AF, Dokter J, van Baar ME, Nijhuis I, Beerthuizen GI, Nieuwenhuis MK, et al. Epidemiology of children admitted to the Dutch burn centres. Burns 2011;37 (November (7)):1161-7.

13. Dokter J, Boxma H, Oen IM, van Baar ME, van der Vlies CH. Reduction in skin grafting after the introduction of hydrofiber dressings in partial thickness burns: a comparison between a hydrofiber and silver sulphadiazine. Burns 2013;39(1):130-5.

14. van der Wal MB, Vloemans JF, Tuinebreijer WE, van de Ven P, van Unen E, van Zuijlen PP, et a. Outcome after burns: an observational study on burn scat maturation and predictors for severe scarring. Wound Repair Regen 2012;20(5):676-87.

15. Gold MR, Siegel JE, Russel LB, Weinstein MC: Cost-effectiveness in health and medicine. New York: Oxford University Press; 1996.

16. Hakkaart-van Roijen L, Tan SS, Bouwmans CAM: Handleiding voor kostenonderzoek; methoden en standaard kostprijzen voor economische evaluaties in de gezondheidszorg. Diemen: College voor zorgverzekeringen; 2004.

17. Bloemen MC, van der Wal MB, Verhaegen PD, Nieuwenhuis MK, van Baar ME, van Zuijlen PP, Middelkoop E. Clinical effectiveness of dermal substitution in burns by topical negative pressure: a multicenter randomized controlled trial. Wound Repair Regen. 2012 Nov-Dec;20(6):797-805.

18. Lagus H, Sarlomo-Rikala M, Böhling T, Vuola J. Prospective study on burns treated with Integra(®), a cellulose sponge and split thickness skin graft: Comparative clinical and histological study-Randomized controlled trial. Burns. 2013 Jul 20.

19. Selig HF, Keck M, Lumenta DB, Mittlböck M, Kamolz LP. The use of a polylactide-based copolymer as a temporary skin substitute in deep dermal burns: 1-year follow-up results of a prospective clinical noninferiority trial. Wound Repair Regen. 2013 May-Jun;21(3):402-9.

20. Schwanholt CA, Ridgway CL, Greenhalgh DG, Staley MJ, Gaboury TJ, Morress CS, Walling SJ, Warden GD. A prospective study of burn scar maturation in pediatrics: does age matter? J Burn Care Rehabil. 1994 Sep-Oct;15(5):416-20.

21. Germann G, Cedidi C, Hartmann B. Post-burn reconstruction during growth and development. Pediatr Surg Int. 1997 Jul;12(5-6):321-6.

22. Orgill DP, Ogawa R. Current methods of burn reconstruction. Plast Reconstr Surg. 2013 May;131(5):827e-36e.

R1R2R3R4R5R6R7R8R9


86 | Chapter 4

23. Verhaegen PD, Stekelenburg CM, van Trier AJ, Schade FB, van Zuijlen PP. Perforator-based interposition flaps for sustainable scar contracture release: a versatile, practical, and safe technique. Plast Reconstr Surg. 2011 Apr;127(4):1524-32.

24. Brongo S, Nicoletti GF, La Padula S, Mele CM, D’Andrea F. Use of lipofilling for the treatment of severe burn outcomes. Plast Reconstr Surg. 2012 Aug;130(2):374e-376e.

25. Hoogewerf CJ, Hop MJ, Nieuwenhuis MK, Middelkoop E, Van Baar ME. Early excision and grafting for burns (Protocol). Cochrane Database of Systematic Reviews 2012, Issue 3.

26. Welling L, van Harten SM, Patka P, Bierens JJ, Boers M, Luitse JS, Mackie DP, Trouwborst A, Gouma DJ, Kreis RW. The café fire on New Year’s Eve in Volendam, the Netherlands: description of events. Burns. 2005 Aug;31(5):548-54.

27. Hemington-Gorse SJ, Potokar TS, Drew PJ, Dickson WA. Burn care costing: the Welsh experience. Burns. 2009 May;35(3):378-82.

28. Ahn CS, Maitz PK. The true cost of burn. Burns. 2012 Nov;38(7):967-74.

Part II. Improving burn care:

diagnostics and costs

Chapter 5Cost-effectiveness of laser Doppler imaging

in burn care in the Netherlands; Study protocol

M. Jenda HopJakob Hiddingh

Carlijn M. StekelenburgHedwig C. KuipersEsther Middelkoop

Marianne K. NieuwenhuisSuzanne Polinder

Margriet E. van BaarLDI study group

BMC Surg. 2013 Feb 1;13:2

R1R2R3R4R5R6R7R8R9


90 | Chapter 5

Abstract

Background: Early accurate assessment of burn depth is important to determine the optimal treatment of burns. The method most used to determine burn depth is clinical assessment, which is the least expensive, but not the most accurate.Laser Doppler imaging (LDI) is a technique with which a more accurate (>95%) estimate of burn depth can be made by measuring the dermal perfusion. The actual effect on therapeutic decisions, clinical outcomes and the costs of the introduction of this device, however, are unknown. Before we decide to implement LDI in Dutch burn care, a study on the effectiveness and cost-effectiveness of LDI is necessary. Methods/Design: A multicentre randomised controlled trial will be conducted in the Dutch burn centres; Beverwijk, Groningen and Rotterdam. All patients treated as outpatient or admitted to a burn centre within 5 days post burn, with burns of indeterminate depth (burns not obviously superficial or full thickness) and a total body surface area burned of ≤ 20% are eligible. A total of 200 patients will be included.Burn depth will be diagnosed by both clinical assessment and laser Doppler imaging between 2-5 days post burn in all patients. Subsequently, patients are randomly divided in two groups: ‘new diagnostic strategy’ versus ‘current diagnostic strategy’. The results of the LDI-scan will only be provided to the treating clinician in the ‘new diagnostic strategy’ group. The main endpoint is the effect of LDI on wound healing time.In addition we measure: a) the effect of LDI on other patient outcomes (quality of life, scar quality), b) the effect of LDI on diagnostic and therapeutic decisions, and c) the effect of LDI on total (medical and non-medical) costs and cost-effectiveness. Discussion:This trial will contribute to our current knowledge on the use of LDI in burn care and will provide evidence on its cost-effectiveness.


Cost-effectiveness of laser Doppler imaging in burn care in the Netherlands; Study protocol | 91

Background

In patients with burns an early accurate diagnosis of burn depth is essential to determine the most appropriate treatment. Monstrey et al. recently reviewed all current modalities to diagnose burn depth. Bedside clinical examination is the most widely used and least expensive method for burn depth assessment. This technique is effective when diagnosing burns at the extreme end of the spectrum: superficial or full thickness. In partial thickness burns, however, clinical examination is not very accurate. Clinical burn depth assessment is accurate in about 2/3 of the cases, the most reported error is overestimation of depth1.Overestimation of burn depth can lead to unnecessary excision and grafting2,3. On the other hand, underestimation of burn depth may lead to an unnecessary delay in surgery, with a longer length of hospital stay and higher hospital costs as a consequence4,5. In burn care, traditionally classified as expensive care, there is a growing interest in costs and cost control6,7. In order to provide effective and cost-effective burn care, there is a need for an accurate method for burn depth estimation.

Laser Doppler imaging (LDI) is the only technique that has been shown to accurately predict wound depth with a large weight of evidence1. Laser Doppler imaging is based on the Doppler principle. Laser light that is directed at moving blood cells in sampled tissue exhibits a frequency change that is proportional to the amount of perfusion in the tissue. Laser Doppler imaging combines the advantages of laser Doppler and scanning techniques: the whole burn can be sampled and no direct contact with the burn surface is necessary2. In daily practice, LDI will be used in combination with standard clinical assessment8, as a so-called add-on test9. Several prospective studies on the diagnostic accuracy of laser Doppler imaging have demonstrated an accuracy varying between 95-100%3,8,10-12. Timing of LDI is important: only scanning between 48 hours and 5 days results in a high accuracy (>95%)8,11,13,14.

To decide whether a new diagnostic strategy, like LDI, should be implemented, assessment of diagnostic accuracy should be followed by assessment of diagnostic and therapeutic impact, effectiveness and cost-effectiveness of the new technology9,15. The literature on the accuracy of LDI in burn depth assessment is convincing. However, most studies only report on the accuracy of this technique. The diagnostic and therapeutic impact of the introduction of LDI is often only speculated upon. There is, to our knowledge, only one retrospective cohort study4, and one prospective non-randomised study5 that investigated the therapeutic impact of the introduction of laser Doppler imaging. Petrie et al. reported a lower rate of operative interventions (6.8% before and 2.2% after, p=0.029) in a paediatric burn population after the introduction of LDI and a reduced

R1R2R3R4R5R6R7R8R9


92 | Chapter 5

length of hospital stay of the surgical treated patients (15.1 days before and 9.8 days after, no p-value). The overall length of stay was 3.4 days in 235 patients before and 2.1 days in 270 patients after the introduction of the LDI4. Because of the retrospective nature of this study, other factors than the introduction of LDI alone could be responsible for the therapeutic changes9. In the study of Kim et al.5 the impact of LDI on surgically treated paediatric patients with burns was investigated. The mean time to decision making for grafting procedures was shorter in the LDI group compared to the clinically assessed group (8.9 vs. 11.6 days, p= 0.01). Because of the non-randomised design, it is unclear whether this can be contributed to the LDI or to other differences between the groups. Thus, current research gives some indications on the diagnostic and therapeutic impact of the LDI in burm care. However, randomised studies in both paediatric and adult patients with burns are lacking. The introduction of LDI possibly leads to a cost reduction in burn care, by preventing unnecessary surgery3,4, and reducing length of hospital stay4. However, no prospective studies are available on the costs and the possible cost reduction of LDI in burn care, nor are cost-effectiveness studies. Therefore, we can conclude that is it still unclear whether LDI actually influences diagnostic and therapeutic decisions, patient outcomes and costs, and thus adds to the quality of care.

The aim of our study is to analyse the effectiveness and cost-effectiveness of LDI in burn care. The effect of the use of LDI in burn care on decision-making, on clinical outcomes, on costs, and on cost effectiveness will be assessed. The current diagnostic strategy in burn depth assessment (clinical assessment) is compared with the new diagnostic strategy: LDI in combination with clinical assessment. We expect that LDI in combination with clinical assessment can lead to earlier excision and grafting in Dutch burn care. With the results of this cost-effectiveness study, we aim to provide a guidance to decide whether this instrument should be implemented in Dutch burn care.

Methods/Design

Study designA multicentre, randomised controlled trial will be conducted in the three Dutch burn centres: Beverwijk, Groningen and Rotterdam.

ParticipantsAll consecutive patients of any ages with acute burns of indeterminate depth (assessed by the treating clinician), who are seen within 5 day post burn at one of the three burn centres, are eligible.



Inclusion criteria:• Patients with acute burns of indeterminate depth (=intermediate depth; the burn

wound is not obviously superficial or obviously full thickness)• Outpatient treatment or admission in one of the three Dutch burn centres• Presentation within 5 days post burn

Exclusion criteria:• The presence of full thickness wounds, next to intermediate wounds• Topical treatment/dressings that impair scanning (e.g. hydrocolloid dressings)• Patients with peri-orbital facial burns, in which the eyes are unable to shield• Patients or their next of kin if they are under aged or temporary incompetent who can

not be expected to give informed consent e.g. because of cognitive dysfunction or poor Dutch proficiency.

• Patients with a TBSA burned > 20%

Intervention We will include a total of 200 patients and randomly divide them in two groups: new diagnostic strategy versus current diagnostic strategy. In the first few days the burns will be treated with regular topical antimicrobials or dressings. LDI is performed between 48 hours and 5 days post burn in this study. After removal of the topical agent (during regular wound care), all wounds of indeterminate depth are scanned by a trained research physician or nurse, who is not involved in the patient treatment. In case of clearly superficial wounds next to intermediate wounds in a study patient, the superficial wounds will not be scanned; similar to what would happen in daily practice.Results of LDI are (figure 1):12

• Red/pink represents a healing potential within 14 days post burn• Yellow/green represents a healing potential between 14-21 days• Blue represents a healing potential > 21 days

In the group ‘new diagnostic strategy’, the results of the LDI scan will be revealed to the treating clinician. Subsequent treatment and outcome of this group will be compared to that of the group with ‘current diagnostic strategy’ in which the treating clinician is blinded for the results of the LDI scan.

R1R2R3R4R5R6R7R8R9


94 | Chapter 5

Figure 1 Clinical appearance of a three-days-old flame burn in a woman of 21. The LDI scan shows a burn with a healing potential < 14 days of the right hand palm and a healing potential > 21 days of the right lower arm. Written consent was obtained from the participant to publish this figure.

Figure 1. Clinical appearance of a three-days-old flame burn in a woman of 21. The LDI scan shows a burn with a healing potential < 14 days of the right hand palm and a healing potential > 21 days of the right lower arm. Written consent was obtained from the participant to publish this figure.

OutcomesPrimary outcome measure: wound healing time Wound healing time is defined as the number of days between LDI (randomisation) and the day on which reepithelialisation of >95% is achieved. Next to that, for generalisation and LDI accuracy checks, we will present the time to wound healing from the day of injury. During wound care an experienced burn specialist will assess reepithelialisation; this is a reliable and valid technique (compared to digital image analysis) (16). Wounds in admitted patients are assessed daily and wounds in outpatients two or three times a week. Differences in wound healing between the two randomisation groups will be analysed.

Secondary outcomes measures: Quality of life and scar qualityQuality of life will be measured as soon as possible after injury and to determine pre-injuryfunctioning (at least within one month post burn) and after 3 months with the EuroQol-6D (EuroQol-5D + cognitive functioning; validated in patients aged ≥ 5 years) or the ItQol-47 questionnaire (validated in children < 5 years of age)17-19.



The EuroQol-5D questionnaire can result in 243 different health states, which can be converted in a summary (utility) score between 0 (death) and 1 (perfect health)18. The EuroQol (EQ-5D) is an extensively used general health questionnaire to measure quality of life. It is recommended for the assessment of HRQoL in injury patients, especially for economic assessments20.

Scar quality will be measured 3 months post-injury. Scar quality will be measured as follows;• Elasticity will be measured with the Cutometer® Skin Elasticity Meter 575 (Courage and

Khazaka Electronic GmbH, Cologne, Germany) The cutometer provides several elasticity parameters. In this study, maximal skin extension (Uf) (in mm) will be used, because this has been demonstrated to be the most reliable parameter21.

• Vascularity and pigmentation will be measured with the Dermaspectometer, which is a reliable narrowband spectrometer that computes an erythema and melanin index (Cortex Technology, Hadshund, Denmark)22.

• Subjective scar assessment will be performed by means of the Patients Observer Scar Assessment Scale (POSAS), a reliable and valid scar assessment tool that consists of a patient and observer scale22,23.The patient scores the scar characteristics colour, pliability, thickness, relief, itching and pain. The observer scale contains the items vascularisation, pliability, pigmentation, thickness, and relief. All items are scored numerically on a 10-point rating scale. In addition, the two observers and the patient give a general opinion on the scare quality on a 10-point rating scale.

Differences in quality of life and scar quality 3 months post burn between the two diagnostic groups will be analysed.

Diagnostic decisions: diagnostic effect of LDI and accuracyIn the ‘new diagnostic strategy’ group the diagnostic effect of the introduction of the LDI will be assessed on the scanning day by comparing diagnostic decisions of burn clinicians, just before and after the use of LDI. Possible diagnostic decisions before LDI are:

• Superficial partial thickness burn (no study wound) • Intermediate burn, the burn wound is not obviously superficial or full thickness (if

parts of a burn wound are obviously full thickness, this will be recorded also) • Deep partial thickness or full thickness burn (exclusion patient from study)

Possible diagnostic decisions after LDI are12:• Superficial partial thickness burn, will heal within 14 days• Intermediate partial thickness burn, will heal between 14-21 days (the spectrum of an

‘intermediate burn’ is smaller after the use of LDI compared to before the use of LDI)• Deep partial thickness or full thickness burn, will not heal within 21 days

R1R2R3R4R5R6R7R8R9


96 | Chapter 5

Accuracy of LDI in our study patients will be checked also. Results of laser Doppler imaging are compared with the reference tests: time of wound healing (from burn date until >95% reepithelialisation) and biopsies. In our study protocol, wounds with a healing potential ≤ 21 (assessed by the LDI) are considered as superficial or intermediate partial thickness and wounds with a healing potential > 21 days are considered as deep partial thickness or full thickness. Only in case of surgery a 3 mm-punch biopsy will be taken as a reference test to assess burn depth1: in a superficial burn the basal membrane is partially destroyed; in a deep partial thickness burn the basal membrane is entirely destroyed and the dermis partially destroyed; in a full thickness burn, the dermis is also completely destroyed.

Therapeutic decisions: timing of surgery indication, timing of surgery and length of hospital stayThe effect of LDI on therapeutic decisions by the burn clinician will be assessed. The first therapeutic decision is on admittance, the second monitored decision is after 2-5 days (based on clinical assessment or clinical assessment in combination with LDI result). In case of postponement of decision, reason of postponement will be investigated and decision making is followed until wound healing.The possible therapeutic decisions are:

• Surgery• Postponement of decision• No surgery

Differences in therapeutic decisions on the day of randomisation will be analysed. In case of surgery, differences in timing of surgical decision and timing of surgery (in days after randomisation) between the two diagnostic groups will be analysed. Differences in length of hospital stay between the two diagnostic groups will be analysed as well.

Economic evaluation: total costs and cost-effectivenessThe economic evaluation will be performed from a societal perspective in accordance with the Dutch guidelines for economic evaluation studies24. A cost-effectiveness analysis will assess the balance between costs and effects of the new diagnostic strategy compare to the current diagnostic strategy Total costs in this study represent direct healthcare costs (inpatient and outpatient medical costs), direct non-healthcare costs (travel costs) and indirect non-healthcare costs (productivity loss). Real medical costs will be calculated by multiplying the volumes of health care use with the corresponding unit prices. Costs will be calculated from randomisation until 3 months post burn. A bottom up approach will be applied (following the micro-costing method of Gold et al. 1996) based on a detailed inventory of all resource used during admittance in one of the Dutch burn centres25.



Cost prizes will be inventoried in the financial department of one burn centre and will be used for all centres in order to prevent measuring cost differences between burn centres. The costs apply to the financial year 2012.

Specialized burn care costs include intervention costs, hospital days, and other variable costs:• The costs per unit of the LDI scanning will be determined by taking into account the

initial investment of equipment, investments during use, maintenance, numbers of years of use, discounting, number of procedures per year and personnel costs.

• Hospital day (both ICU and non ICU) costs will be calculated by multiplying the length of hospital stay by the costs of one day of admission. The cost of re-admittance during follow-up will also be calculated. The costs of a day of admittance are based on a detailed inventory (following the micro-costing method of Gold et al. 1996) of fixed costs only: staffing costs, accommodation, equipment, overhead, food, laundry, and medication25.

• Variable hospital costs will be identified per patient and calculated by multiplying the volumes of health care use with the corresponding unit prices. The included items are: dressing costs, surgical procedures (material, equipment and personnel costs), diagnostic procedures (bronchoscopy, swabs, laboratory and radiology costs), treatment by allied health professionals, splints, pressure garments, and outpatient visits (staffing and material costs).

Data registration will start on the day of randomisation. Data regarding patients’ baseline characteristics and health-care use will be obtained from patient records and the electronic hospital administration.

Other healthcare costs and non healthcare costs will be calculated based on charges as a proxy of real costs24 with the help of data obtained from patient questionnaires after 3 months. Other healthcare costs include nursing-home and rehabilitation centre, homecare, visits to general physicians and allied health professional outside the hospital. Indirect non-healthcare costs include productivity loss in patients and partners or parents (if applicable), and direct non-healthcare costs include patient travel costs.

To measure the economic impact of LDI in burn care cost-effectiveness will be assessed by calculating the incremental cost-effectiveness ratio, defined here as the costs for the ‘new diagnostic strategy’ (minus savings) divided by the difference in wound healing time (measured from randomisation) between the ‘new diagnostic strategy’ and ‘current diagnostic strategy’. Secondary, a cost-utility analysis will be performed, i.e., as cost per Quality Adjusted Life Years (QALY). The cost-utility ratio can be calculated in patients ≥ 5 years only. In children under the age of 5 years, the EuroQol-5D is not validated and no questionnaires are available

R1R2R3R4R5R6R7R8R9


98 | Chapter 5

that allow the calculation of utility scores18,19. Overall utility scores for population-based quality of life (derived from the EuroQol-5D) will be obtained and expressed as QALY’s. QALY’s will be calculated by multiplying the utility of a health state by the time spent in this health state using the Dutch valuation tariff 26. Policymakers and health economists have proposed that costs varying from €25,000 up to €75,000 per QALY may be considered as acceptable27. In accordance with guidelines for differential discounting, effects will be discounted at a rate of 1.5% and costs at 4% per year24.In case no differences will be found in wound healing and quality of life, the economic evaluation will be based on a cost-minimisation analysis, that consist of a comparison of total costs in both diagnostic strategies.

Sample sizeA total of 57.5% of admitted burn patients in the Netherlands is treated conservatively (unpublished data, Dutch Burn Repository R3, 2011). The time to wound healing in patients without surgery is approximately 13 days, this will not change after the introduction of LDI. In the 42.5% of surgically treated patients we do expect an effect on wound healing time, because of an earlier operation. The mean time to first transplantation is 14.0 days post burn, the mean time to wound healing is 5 days after surgery28, resulting in a mean time to wound healing of 19.0 days. The overall mean time to wound healing for all patients combined is 15.6 days ((57.5%*13+ 42.5%*19)/100).Petrie et al describe a reduction of 5.3 days in LOS in patients undergoing surgery, after the introduction of the LDI. As a result, in our population, an overall effect of 2.25 (42.5%* 5.3) days can be expected, with a standard deviation of 5 days. A 2-sided test with an α level of 0.05 and a β level of 0.20 (power 0.80) indicates a required total sample size of 190 patients. We expect a low drop out rate because of a short time to follow up to assess primary outcome (time to wound healing). With a 5% drop out we need a total sample size of 200.

Randomisation Patients are randomly allocated to the group ‘new diagnostic strategy’ or the group ‘current diagnostic strategy’, on the day of the LDI scan. Stratification will be performed for the three different centres and the severity of the burns: total burned surface area ≤10% or >10%. A randomisation list is prepared for each stratum on the principle of random permuted blocks of patients, using a random number table29. The allocation sequence is concealed by using sequentially numbered, opaque sealed envelopes (SNOSE) prepared by the coordinating researcher (MJH) which are opened at day of after having obtained informed consent, by one of the local researchers30.The care provider, outcome assessor and patients are blinded to the results of the LDI in the group ‘current diagnostic strategy’. However, they are not blinded to the group in which patients are randomised.



Statistical analysisData will be primarily analysed according to the intention-to-treat principle. We assume that time to LDI will be equal in both diagnostic groups, as a result of randomisation. Thus, the differences in effects (e.g. time to wound healing) and costs will be assessed from randomisation onwards.Differences in time to wound healing (numbers of days between LDI and day of > 95% reepithelialisation) and timing of surgery in both diagnostic strategy groups will be analysed with Kaplan-Meier curves and log rank test.Quality of life (derived from the ItQol-47 and EuroQol-6D), scar quality (assessed by the Cutometer®, Dermaspectometer® and POSAS) and length of hospital stay will be analysed using the independent-sample-t-test (in case of normal distribution) or the Mann-Whitney test (non-normal distribution).Differences in diagnostic and therapeutic decisions will be analysed with the Chi-square test. Diagnostic accuracy will be assessed by calculating sensitivity and specificity of the LDI compared to the reference test: time to wound healing (in conservative treated patients) or biopsy (in surgical treated patients). In addition, receiver operating characteristic (ROC) curves will be calculated for both diagnostic groups.Differences in mean costs (after randomisation) will be analysed by the Mann-Whitney test, since healthcare costs are typically highly skewed. Non-parametric techniques (bootstrapping) will be used to derive a 95% confidence interval for the differences in distributions of the costs.The cost-effectiveness analysis and cost-utility analysis are performed by dividing the differences in average costs by the differences in average effects or utility. In a sensitivity analysis, the impact on cost-effectiveness of statistical uncertainty on the main study outcomes will be determined (uni- and multivariable).Data analysis will be performed using SPSS-software (Statistical Package for the Social Sciences).

Discussion

In this paper we described the design of our study into the effects and cost-effectiveness of laser Doppler imaging in Dutch burn care. This is the first randomised controlled study that analyses not only the accuracy, but also the effects and costs of the introduction of LDI in burn care. Strengths of our study are the extensive cost calculations (not only LDI costs are included) and the detailed analysis of the possible impact of the introduction of the LDI in terms of process changes. i. e. changes in diagnosis and therapy. A limitation of this study is that the cost calculations will be performed in Dutch burn centres only. Although the multicentre analysis improves the generalisability of the cost distribution, compared to a monocentre analysis, total cost and cost distribution in other countries are probably different31.

R1R2R3R4R5R6R7R8R9


100 | Chapter 5

Another limitation of this study is that it tests the effectiveness of the introduction of LDI, so there’s a possible learning effect in the study. In the first patients, professionals will less likely rely on LDI and adapt their therapy. As a result, differences between randomisation groups will probably be smaller, and subsequently our study underestimates the full effect of the introduction of LDI in burn care. Next to a time effect, there will probably also be an effect of differences in study sites. The usual care differs between study sites; this can influence the results of the introduction of LDI in the burn centres. These possible differences will be studied by providing sub-analyses per burn centre.This study will undoubtedly contribute to our knowledge about the use of LDI in burn care and will provide evidence on its cost-effectiveness. Inclusion of patients started in December 2011 and presentation of data will be expected at the end of 2013.

Ethical considerationsThis study has been approved by the Medical Research Ethics Committee of Rotterdam (NL37844.101.11).

Acknowledgements and fundingThe trial is financially supported by a grant of the Dutch Burns Foundation and the Nuts Ohra Foundation.

Competing interestsThe authors declare that they have no competing interests. The laser Doppler imagers were leased from Moor for this study.

Authors contributionsMJH made substantial contributions to conception and design, drafted the manuscript, and contributes currently to acquisition of data. JH made substantial contributions to conception and design, revised the manuscript critically, and contributes currently to acquisition of data. CS revised the manuscript critically and contributes currently to acquisition of data. HCK revised the manuscript critically and contributes currently to acquisition of data. EM made substantial contributions to conception and design and revised the manuscript critically. MKN made substantial contributions to conception and design and revised the manuscript critically. SP made substantial contributions to conception and design and revised the manuscript critically. MEB made substantial contributions to conception and design and revised the manuscript critically. All authors read and approved the final manuscript.



References

1. Monstrey S, Hoeksema H, Verbelen J, Pirayesh A, Blondeel P. Assessment of burn depth and burn wound healing potential. Burns. 2008 Sep;34(6):761-9.

2. Niazi ZB, Essex TJ, Papini R, Scott D, McLean NR, Black MJ. New laser Doppler scanner, a valuable adjunct in burn depth assessment. Burns. 1993 Dec;19(6):485-9.

3. Jeng JC, Bridgeman A, Shivnan L, Thornton PM, Alam H, Clarke TJ, Jablonski KA, Jordan MH. Laser Doppler imaging determines need for excision and grafting in advance of clinical judgment: a prospective blinded trial. Burns. 2003 Nov;29(7):665-70.

4. Petrie N, Norbury W, Fogarty B, Philp B, Baraat J, Dziewulski P. The use of the laser Doppler imaging to reduce operative intervention in the treatment of paediatric burns In: Abstract book, 12th congress of the International Society for Burn Injuries (ISBI);2004

5. Kim LH, Ward D, Lam L, Holland AJ. The impact of laser Doppler imaging on time to grafting decisions in paediatric burns. J Burn Care Res. 2010 Mar-Apr;31(2):328-32.

6. Ahn CS, Maitz PK. The true cost of burn. Burns. 2012 Nov;38(7):967-74. doi:10.1016/j.burns.2012.05.016. Epub 2012 Jul 13.

7. Hemington-Gorse SJ, Potokar TS, Drew PJ, Dickson WA. Burn care costing: The Welsh experience. Burns. 2009 March;35(2/3):378-82.

8. Hoeksema H, Van de Sijpe K, Tondu T, Hamdi M, Van Landuyt K, Blondeel P, Monstrey S. Accuracy of early burn depth assessment by laser Doppler Imaging on different days post burn. Burns. 2009 Feb;35(1):36-45.

9. Bruel A van den, Cleemput I, Aertgeerts B, Ramaekers D, Buntinx F. The evaluation of diagnostic tests: evidence on technical and diagnostic accuracy, impact on patient outcome and cost-effectiveness is needed. J Clin Epidemiol. 2007 Nov;60(11):1116-22.

10. Pape SA, Skouras CA, Byrne PO. An audit of the use of laser Doppler imaging(LDI) in the assessment of burns of intermediate depth. Burns. 2001 May;27(3):233-9.

11. Riordan CL, McDonough M, Davidson JM, Corley R, Perlov C, Barton R, Guy J,Nanney LB. Noncontact laser Doppler imaging in burn depth analysis of the extremities. J Burn Care Rehabil. 2003 Jul-Aug;24(4):177-86.

12. Monstrey SM, Hoeksema H, Baker RD, Jeng J, Spence RS, Wilson D, Pape SA. Burn wound healing time assessed by laser Doppler imaging. Part 2: validation of a dedicated colour code for image interpretation. Burns. 2011 Mar;37(2):249-56.

13. Kloppenberg FW, Beerthuizen GI, ten Duis HJ. Perfusion of burn wounds assessed by laser doppler imaging is related to burn depth and healing time. Burns. 2001 Jun;27(4):359-63.

14. Holland AJ, Martin HC, Cass DT. Laser Doppler imaging prediction of burn wound outcome in children. Burns. 2002 Feb;28(1):11-7.

15. Hunink MG, Krestin GP. Study design for concurrent development, assessment, and implementation of new diagnostic imaging technology. Radiology. 2002 Mar;222(3):604-14.

16. Bloemen MC, Boekema BK, Vlig M, van Zuijlen PP, Middelkoop E. Digital image analysis versus clinical assessment of wound epithelialization: a validation study. Burns. 2012 Jun;38(4):501-5.

17. Raat H, Landgraf JM, Oostenbrink R, Moll HA, Essink-Bot ML. Reliability and validity of the Infant and Toddler Quality of Life Questionnaire (ITQOL) in a general population and respiratory disease sample. Qual Life Res. 2007 Apr;16(3):445-60. Epub 2006 Nov 17.

18. Dolan P. Modeling valuations for EuroQol health states. Med Care. Nov 1997;35(11):1095-1108.19. Stolk EA, Busschbach JJ, Vogels T. Performance of the EuroQol in children with imperforate anus.

Qual Life Res. Feb 2000;9(1):29-38.20. Van Beeck EF, Larsen CF, Lyons RA, Meerding WJ, Mulder S, Essink-Bot ML: Guidelines for the

conduction of follow-up studies measuring injuryrelateddisability. J Trauma 2007, 62(2):534-550.21. Draaijers LJ, Botman YA, Tempelman FR, Kreis RW, Middelkoop E, van Zuijlen PP. Skin elasticity

meter or subjective evaluation in scars: a reliability assessment. Burns. 2004 Mar;30(2):109-14.22. Draaijers LJ, Tempelman FR, Botman YA, Kreis RW, Middelkoop E, van Zuijlen PP. Colour

evaluation in scars: tristimulus colorimeter, narrow-band simple reflectance meter or subjective evaluation? Burns. 2004 Mar;30(2):103-7

R1R2R3R4R5R6R7R8R9


102 | Chapter 5

23. van der Wal MB, Tuinebreijer WE, Bloemen MC, Verhaegen PD, Middelkoop E, vanZuijlen PP. Rasch analysis of the Patient and Observer Scar Assessment Scale(POSAS) in burn scars. Qual Life Res. 2011 May 20

24. Hakkaart-van Roijen L, Tan SS,Bouwmans CAM. Handleiding voor kostenonderzoek; methoden en standaard kostprijzen voor economische evaluaties in de gezondheidszorg. Diemen: College voor zorgverzekeringen; 2004.


26. Lamers LM, et al. Measuring the quality of life in economic evaluations: the Dutch EQ-5D tariff. Ned Tijdschr Geneeskd 2005;149:1574-8.

27. Touwse et al? Report of the Commission on Macroeconomics and Health: Macroeconomics and health: Investing in health for economic development. In. Geneva: WHO Commision on Macroeocnomics and Health.; 2001.

28 Bloemen MC, van der Wal MB, Verhaegen PD, Nieuwenhuis MK, van Baar ME, van Zuijlen PP, Middelkoop E. Clinical effectiveness of dermal substitution in burns by topical negative pressure: A multicentre randomized controlled trial. Wound Repair Regen. 2012 Nov;20(6):797-805. doi: 10.1111/j.1524-475X.2012.00845.x. Epub 2012 Oct 30. PubMed PMID: 23110478.

29. Altman DG. Practical statistics for medical research. London: Chapman and Hall; 1991.30. Doig GS, Simpson F. Randomization and allocation concealment: a practical guide for researchers.

J Crit Care. 2005 Jun;20(2):187-91; discussion 191-3.31. Wheeler JR, Harrison RV, Wolfe RA, Payne BC. The effects of burn severity and institutional

differences on the costs of care. Med Care. 1983 Dec;21(12):1192-203.

Chapter 6Cost-effectiveness of laser Doppler imaging in burn

care in the Netherlands; a randomised controlled trial

M. Jenda HopCarlijn M. Stekelenburg

Jakob HiddinghHedwig C. KuipersEsther Middelkoop

Marianne NieuwenhuisSuzanne Polinder

Margriet E. van BaarLDI study group. PRS.

In press.

R1R2R3R4R5R6R7R8R9


104 | Chapter 6

Abstract

Background: In patients with burns, an early accurate diagnosis of burn depth facilitates optimal treatment. Laser Doppler imaging (LDI) combined with clinical assessment leads to an accurate estimate of burn depth. However the actual effects of the introduction of LDI on therapeutic decisions, clinical outcomes and costs are unknown. Methods: To determine the effects of the introduction of LDI, a randomised controlled trial was conducted in all three Dutch burn centres, including 202 patients with burns of indeterminate depth. In the standard care (SC) group, estimation of burn depth was based on clinical assessment only, in the other group (LDI) clinical assessment and LDI were combined. Primary outcome was wound healing time. Furthermore, therapeutic decisions, costs and cost-effectiveness were analysed. Results: Mean time from randomization to wound healing was 14.3 days (95%CI=12.8-15.9) in the LDI group and 15.5 days (95%CI=13.9-17.2) in the SC group (p=0.258). On the day of randomization, clinicians decided significantly more often on operative or non-operative treatment in the LDI group (p<0.001), instead of postponing their treatment choice. Analyses in the subgroup of admitted patients, requiring surgery, showed an earlier decision for surgery and a shorter wound healing time in the LDI group (16.0 versus 19.9 days, p=0.022). Mean total costs per patient were €20 202 versus €20 504 (LDI versus SC, p=0.870). Conclusions: LDI improved therapeutic decisions. It resulted in a shorter wound healing time in the subgroup of admitted patients requiring surgery and has the potential for cost-savings of €875 per scanned patient. Clinical question/ level of evidence: Diagnostic, level I


Cost-effectiveness of laser Doppler imaging in burn care in the Netherlands | 105

Introduction

In patients with burns an early accurate diagnosis of burn depth facilitates optimal treatment. Bedside clinical examination is the most widely used and least expensive method for burn depth assessment. This technique is effective when diagnosing burns at the extreme ends of the spectrum: clearly superficial or clearly full thickness. In partial thickness burns, however, clinical examination is accurate in only 2/3 of cases1. The most reported error is overestimation of depth1, which may lead to unnecessary excision and grafting2,3. On the other hand, underestimation of burn depth may result in unnecessary delay in surgery, with a longer length of hospital stay and higher hospital costs as a consequence4,5. In burn care, traditionally considered expensive care, there is a growing interest in costs and cost control6-9. In order to provide cost-effective burn care, there is a need for an accurate and timely method for burn depth estimation.

Laser Doppler imaging (LDI) is the only technique that has been shown to accurately predict healing time with a large weight of evidence10. LDI should be used in combination with standard clinical assessment11, as a so-called add-on test12. In the past decade, several prospective studies on the diagnostic accuracy of LDI have demonstrated an accuracy varying between 95-100%2,3,10,11,14,15. Timing of LDI is important: only scanning between 48 hours and 5 days results in a high accuracy (>95%)10,11,14-17. The literature on the accuracy of LDI in burn depth assessment is convincing although, the therapeutic impact remains less clear. To our knowledge, only one retrospective cohort study4 and one prospective non-randomised study5 investigated the therapeutic impact of the introduction of LDI. Petrie et al.4 reported a lower rate of operative interventions (6.8% before and 2.2% after introduction of LDI, p= 0.029) in a paediatric burn population and a reduced length of hospital stay of the surgically treated patients (15.1 days before and 9.8 days after, no p-value), in a retrospective study design. The overall length of stay was 3.4 days before and 2.1 days after the introduction of the LDI. Kim et al.5 studied the impact of LDI on surgically treated paediatric patients with burns. The mean time to decision for surgery was shorter; 8.9 days in the LDI group compared with 11.6 days in the clinical assessment group (p= 0.01). No studies reported on health-related quality of life of patients after the introduction of LDI.

The British National Institute for Health and Care Excellence (NICE) recently published a guidance on LDI for the assessment of burns13. They concluded that there was good clinical evidence that LDI increases the accuracy of predicting burn wound healing and can be used to facilitate treatment plans. With LDI results, the decision whether surgery was needed could be made earlier, resulting in shorter hospital stay; it reduced the size of areas that were operated or avoided surgery completely in some patients who otherwise might have had to

R1R2R3R4R5R6R7R8R9


106 | Chapter 6

undergo operative treatment. According to NICE, the average cost saving was £1248 (€1504) per patient scanned, when using the MoorLDI-B1® in addition to clinical evaluation. This was based on a cost model, assuming a 17% reduction in the number of skin graft operations and a two-day reduction of hospital stay. The methods and results of the cost estimations however were incompletely described. For instance, the source of the numbers of reduction of surgery and hospital days remain unclear. Furthermore, it is unclear to what degree these cost-estimations apply to other healthcare settings, for instance in the Netherlands.

To summarize, current research gives indications on the diagnostic and therapeutic impact of the LDI in burn care. However, randomised cost-effectiveness studies in both paediatric and adult patients with burns are lacking. Therefore, it is still unclear whether LDI actually influences diagnostic and therapeutic decisions, patient outcomes and costs, and as a consequence adds to the quality of care.The aim of the present study was to analyse the effectiveness and cost-effectiveness of LDI in burn care.

Methods

Study design, participants and randomization A multicentre, randomised controlled trial was conducted in all three burn centres of the Netherlands from December 2011 to March 2013. The study protocol was approved by the Medical Research Ethics Committee of Rotterdam (NL37844.101.11) and registered at Clinical Trials (NCT01489540). Inclusion criteria were burns of indeterminate depth, admission or outpatient treatment, and presentation within 5 days post burn. Patients were excluded in case of presence of evident full thickness wounds (besides intermediate wounds), topical treatment/dressings that impaired scanning, peri-orbital facial burns, no informed consent and/or a total body surface area (TBSA) burned >20%. In randomization, stratification was performed for the three different centres and for burn size: TBSA burned ≤10% or >10%. The allocation sequence was concealed by using sequentially numbered, opaque sealed envelopes (SNOSE). The full trial protocol has been published elsewhere18.

Intervention Patients were randomly assigned to 1) clinical assessment + LDI (LDI group) or 2) clinical assessment alone (standard of care; SC group), at the day of the scan by a research physician/nurse. An LDI scan was made of all wounds (a wound was defined as an area bounded by non-burned skin) of indeterminate depth by a trained (2 months before start trial) research physician/nurse, between 48 hours and 5 days post burn. LDI results were revealed to the



treating clinician in the LDI group, and remained blinded in the SC group. The LDI results were presented as a healing potential (<14 days, between 14-21 days or >21 days), with a validated colour code for image interpretation10. Clinicians then decided on the preferred treatment, based on both the LDI and their clinical expertise (LDI group) or their clinical expertise alone (SC group).

OutcomesPrimary outcome measure: Time to wound healing (>95% reepithelialisation) from the day of randomization onwards. Reepithelialisation of all separate wounds within one patient was assessed in a bedside procedure by an experienced burn specialist, not blinded to randomization group19. Secondary outcomes measures: Diagnostic decisions of burn clinicians in the LDI group, before and after LDI. Sensitivity and specificity of the LDI were calculated compared to the respective reference test (time to wound healing (in conservative patients) or biopsy (in surgical patients)18. Therapeutic decisions of burn clinicians, including: surgery/no surgery/postponement decision, time to surgical decision, time to surgery and length of hospital stay (LOS). Quality of life, assessed at baseline and after 3 months with a questionnaire (patients 0 – 4 years old: ItQol-47 questionnaire (11 dimensions, range 0-100), 5 years and older: the EuroQol-6D (range 0.0-1.0)20-22.Scar quality, assessed after 3 months with the Cutometer® Skin Elasticity Meter 575 (Courage and Khazaka Electronic GmbH, Cologne, Germany)23, the Dermaspectrometer (vascularity and pigmentation) (Cortex Technology, Hadshund, Denmark)24 and the Patients Observer Scar Assessment Scale (POSAS)25,26.The economic evaluation was performed from a societal perspective in accordance with the Dutch guidelines for economic evaluation studies27. Direct healthcare costs, direct non-healthcare costs and indirect non-healthcare costs from randomization until three months post burn were included. Costs of LDI included costs of initial investment, investments during use and maintenance, personnel costs and overhead and housing.A bottom up approach was applied (following the micro-costing method of Gold et al.)28. Cost prices were inventoried in the financial department of one burn centre. Additionally, charges as a proxy of real costs27 were used with data from patient questionnaires. The costs applied to the financial year 2012.

StatisticsPower calculation was based on a reduction of 5.3 days in LOS in patients undergoing surgery, after the introduction of the LDI4, mean time to wound healing in surgically treated patients admitted to the Dutch burn centres, (42.5% of all admitted patients) was 19.0 days18. Thus an

R1R2R3R4R5R6R7R8R9


108 | Chapter 6

overall reduction of 2.25 (42.5%*5.3) days could be expected (SD=5). An α level of 0.05 and a β level of 0.20 (power 0.80) resulted in a calculated sample size of 95 per group. With a 5% drop out a total sample size of 200 was required.Data were analysed according to the intention-to-treat principle. Differences in time to wound healing, surgical decision and surgery in both groups were analysed with mixed model analysis (three levels: burn centre, patient, wound). Differences in therapeutic decision were analysed with multinomial logistic multilevel regression. Quality of life, scar quality and LOS were analysed using univariate linear regression, with a correction for burn centre and also for baseline measures in quality of life.Diagnostic accuracy was assessed by calculating sensitivity and specificity of the LDI compared to the reference standard (wound healing time in non-operative wounds and biopsies in operative treated wounds). Differences in mean costs were analysed with non-parametric techniques (bootstrapping, 1000 times). Effects of LDI in admitted patients (therefore excluding outpatients) were analysed separately in a sub-analysis. Data analysis was performed using SPSS PASW Statistics 18.0 (IBM, New York City) and Stata (StataCorp, College Station, Texas).

Results

Flowchart, patient and injury characteristicsIn Figure 1 the trials’ flowchart is presented. A total of 202 patients were included, 106 in the LDI and 96 in the SC group, of which baseline characteristics are shown in Table 1.

Wound healing timeMean time from injury to randomization was 3.4 days in the LDI and 3.5 days in the SC group. Mean time to wound healing from randomization was 14.3 days (95% CI 12.8-15.9) in wounds in the LDI group and 15.5 days (95%CI 13.9-17.2) in the SC group (p=0.258) (Table 2).A non-significant shorter time to wound healing was found in the operatively treated wounds, with 16.5 days (95%CI 14.0-18.9) in the LDI group (74 wounds) vs. 19.2 days (95%CI 16.8-21.5) in the SC group (63 wounds) (p=0.099). As expected, mean time to wound healing in the non-operatively treated wounds was equal in both diagnostic groups: 12.5 days (95%CI 10.4-14.5) in the LDI group (108 wounds) and 12.4 days (95%CI 10.3-14.5) (p=0.953) in the SC group (85 wounds).

The sub-analysis for admitted patients revealed a significantly shorter time to wound healing in operatively treated wounds, 16.0 days (95%CI 13.5-18.5) in the LDI group vs. 19.9 days (95%CI 17.5-22.3) in the SC group (p=0.022).



Excluded 141 not meeting inclusion criteria but otherwise eligible:

Admitted

107 Outpatient

34 - Refused to participate 19 7 - Poor Dutch proficiency 19 3 - Cognitive dysfunction 9 2 - TBSA burned >20% 36 0 - No one with LDI training available

0 5

- Others (missed) 24 17

Excluded 837 not eligible for this treatment:

Admitted

227 Outpatient

610 - Obviously superficial/full thickness

143 435

- Wound < euro coin 6 81 - Topical treatment that impairs scanning

39 40

- Peri-orbital facial burns 1 0 - Presentation > 5 days p.b.

29 54

- Burden patient too high 6 0 - Others 3 0

Assessed for eligibility (n=1184):

All burn patients: admitted (n=502) and outpatient (n=682) from 01-12-2011 to 10-12-2012

Randomized (n= 206) admitted (n=168) and

outpatient (n=38)

Enrolment

Allocated to new diagnostic strategy 106 Received allocated intervention 106

Allocated to current diagnostic strategy 100 Post randomization exclusions: 4

• Withdrawal informed consent 2 • Participation other study 1 • Departure abroad 1

Received allocated intervention 96

Included primary analysis: 106

Included in primary analysis: 96

Analyzed at 3 months follow-up: 98 Lost to follow-up 8

• Drop-outs 8

Analyzed at 3 months follow-up: 82 Lost to follow-up 14

• Drop-outs 13 • Departure abroad 1

Allocation

Analysis

Figure 1. Flow of patients through trial

Diagnostic decisionsThe wound depth according to LDI results was comparable in both diagnostic groups: 24.7% versus 25.2% had a predicted HP of <14 days, and 47.3% versus 51.7% had a predicted HP of >21 days. Accuracy was good with a sensitivity of 93.5% (95%CI 86.2-96.9) and a specificity of 88.6%(95%CI 70.6%-85.9%).

R1R2R3R4R5R6R7R8R9


110 | Chapter 6

Table 1. Patient and injury characteristicsLDI group Standard care group(n=106) (n=96)

Gender (m/f), % 62.3/37.7 69.8/30.2

Mean age in years at injury (range, SD) 33.6 (0-74, 23.3) 32.0 (0-89, 22.5)

Aetiology, %Fire/Flame 48.1 44.8

Scald 30.2 30.2Others 21.7 25.0

Mean % TBSA burned (range, SD) 5.6 (0.5-20.0, 4.3) 5.3 (0.2-18.0, 4.2)

Body location burned, % * Head/face 4.8 3.3

Trunk 13.4 16.6Arm 31.2 32.5

Hand 21.0 26.5Legs 22.0 17.9Feet 7.5 3.3

Total number of wounds 186 151Number of included wounds per patient (mean) 1.8 1.6

% of patients with: 1 wound 47.2 60.42 wounds 36.8 25.03 wounds 11.3 11.54 wounds 2.8 3.05 wounds 1.9 0.0

Admitted vs. outpatient treatment, %** 76.4/23.6 76.0/24.0

* more locations per patient are possible**at day of randomisation

Table 2. Time to wound healing, on wound level from randomization onwards*LDI group Standard care group

n(%) Time to wound healing (days)mean (95% CI)

n(%) Time to wound healing (days)mean (95% CI)

p-value

Wounds of all patientsAll wounds ** 182 14.3 (12.8-15.9) 148 15.5 (13.9-17.2) 0.258

Operative treated wounds 74 (40.7) 16.5 (14.0-18.9) 63 (42.6) 19.2 (16.8-21.5) 0.099Non-operative treated wounds 108 (59.3) 12.5 (10.4-14.5) 85 (57.4) 12.4 (10.3-14.5) 0.953

Wounds of admitted patients All wounds 154 14.0 (12.2-15.8) 122 16.0 (14.1-17.9) 0.117

Operative treated wounds 68 (44.2) 16.0 (13.5-18.5) 56 (45.9) 19.9 (17.5-22.3) 0.022Non-operative treated wounds 86 (55.8) 12.1 (9.7-14.4) 66 (54.1) 12.0 (9.4-14.6) 0.982

* adjustment for burn centre and patient level by multi level analysis **wound healing time missing in 7 wounds



Therapeutic decisionsOn the day of the LDI scan, clinicians decided significantly more often on operative or non-operative treatment (respectively in 20.4% and 44.1%) in the LDI group (p<0.001), instead of postponing their treatment decision (in 35.5%), which happened more often in the SC group (in 82.1%) (Figure 2). Reasons for postponement of treatment choice in the LDI group were: different predicted HP’s within one wound (65.6%) and a predicted HP of 14-21 days (23.0%), doubt on the LDI prediction (6.6%), wound to small for surgery (4.9%). Reasons for postponement in the SC group were: indeterminate burn depth (89.2%), or unclear (burn) size for surgery (10.8%).

LDI g

roup

, n=1

86 w

ound

s

No surgery,n= 82 (44.1%)

HP<14: n=43, HP 14-21: n=33, HP>21: n=6

No surgery, n= 76 (92.7%)HP<14: n=42, HP 14-21: n=28, HP>21: n=6

Surgery, n= 6 (7.3%)HP<14: n=1, HP 14-21: n=5, HP>21: n=0

Postponement decision,n= 66 (35.5%)

HP<14: n=3, HP 14-21: n=18, HP>21: n=45



Surgery,n= 38 (20.4%)

HP<14: n=0, HP 14-21: n=1, HP>21: n=37

No surgery, n=1 (2.6%)HP<14: n=0, HP 14-21: n=0, HP>21: n=1


Therapeutic decisionon randomization, basedon clinical assessment

+LDI*

Actual therapy

Figure 2.a Therapeutic decisions in LDI group*significantly different from SC group, p<0.001

SC

gro

up, n

=151

wou

nds

No surgery,n= 13 (8.6%)

HP<14: n=12, HP 14-21: n=0, HP>21: n=1



Postponement decision,n= 124 (82.1%)

HP<14: n=26, HP 14-21: n=30, HP>21: n=68



Surgery,n= 14 (9.3%)

HP<14: n=0, HP 14-21: n=5, HP>21: n=9



Therapeutic decision on randomization, based on

clinical assessment*Actual therapy

Figure 2.b Therapeutic decisions in Standard Care group

R1R2R3R4R5R6R7R8R9


112 | Chapter 6

The mean time to decision for surgical treatment was 4.6 days (95%CI 3.0-6.1) post randomization in wounds in the LDI group, compared to 6.3 days (95%CI 4.8-7.6) in wounds in the SC group (p=0.096). The time to actual surgery was 8.6 days (95%CI 7.1-10.1) post randomization in the LDI group versus 9.7 days (95%CI 8.3-11.1) in the control group (p=0.276) (Table 3). In a sub-analysis for the admitted patients only, time to the surgical decision was significantly shorter in the LDI group: time to surgical decision was 4.1 days (95%CI 2.5-5.7) in the LDI group vs. 6.5 days (95%CI 5.0-8.1) in the SC group (p=0.029). Time to actual surgery did not differ significantly, 8.5 days (95%CI 7.0-10.1) in the LDI group vs. 9.9 days (95%CI 8.5-11.4) (p=0.170) in the SC group. Mean LOS was 9.6 days (95%CI 7.5-11.7) in the LDI group, vs. 10.2 days (95%CI 7.9-12.4) in the SC group (p=0.696).

Unnecessary surgery, i.e. surgery performed in wounds with a predicted healing potential of <14 days, was observed in only 1 out of 38 wounds in the SC group (surgery was performed 5 days post randomization).

Table 3. Time to surgical decisions and time to actual surgery, on wound level from randomisation onwards

LDI group Standard care groupn mean (95% CI) n mean (95% CI) p-value

Wounds of all patients **Time to decision surgical treatment (days) *** 73 4.6 (3.0-6.1) 63 6.3 (4.8-7.8) 0.096Time to surgery (days) 74 8.6 (7.1-10.1) 64 9.7 (8.3-11.1) 0.276Wounds of admitted patients**Time to decision surgical treatment (days)*** 67 4.1 (2.5-5.7) 56 6.5 (5.0-8.1) 0.029Time to surgery (days) 68 8.5 (7.0-10.1) 57 9.9 (8.5-11.4) 0.170

* adjustment for burn centre and patient level by multi level analysis ** one outlier removed from the LDI group (one year old boy with complicated wound healing due to wound infection, time to decision surgical treatment was 52 day, which is more than 4 SD above the mean, TBSA operated was 0.3%.)*** two values missing (one in the LDI and one in the SC group)

Three months follow up: quality of life and scar qualityScar quality three months post burn was comparable between LDI and SC group (Table 4). In addition, quality of life three months post burn was also comparable between both randomisation groups (data not shown).



Table 4. Scar quality 3 months post burnLDI group Standard care group

n mean (95% CI) n mean (95% CI) p-valueScar elasticity (ratio Uf) 81 0.71 (0.62-0.80) 70 0.72 (0.62-0.82) 0.868Scar colour 97 82

Erythema 8.1 (6.1-10.1) 7.5 (5.3-9.8) 0.726Melanin 7.6 (6.2-9.1) 7.2 (5.5-8.8) 0.671

Subjective scar scale (POSAS)* Patient total score 97 4.2 (3.8-4.6) 77 4.1 (3.6-4.7) 0.910

Observer total score 98 3.2 (2.9-3.6) 80 3.3 (2.9-3.6) 0.903

Scar elasticity: ratio scar/control of maximal skin extension (Uf)Scar colour (scar-control) * The mean total score is calculated by adding the scores of the separate items of the POSAS and dividing this by the number of items

CostsThe mean price of LDI scan(s) per patient was €151 (Table 5.a). Mean total costs per patient were comparable in both groups: €20 202 (95%CI 17 341 - 23 063) in the LDI group and €20 504 (95%CI 17 614 – 23 394) in the SC group (p=0.870). Costs of specific parts of healthcare use and patient costs were also similar (Table 5.b).In sub-analysis, mean total costs for admitted patients were €22 854 (95%CI 19 664 – 26 044) in LDI group and €23 8168 (95%CI 20 661- 26 973) in SC group (p=0.627). In outpatients, mean total costs were respectively €8 056 (95%CI 5 551 – 10 562) in the LDI group and €6 7080 (95%CI 4 910 – 9 250) in the SC group (p=0.538).

Table 5.a. Full cost prize (€, 2012) of the LDI use per patient

LDI device* 73Personnel costs ** 33Overhead and housing (41.9%) 45Total costs per patient 151

*based on initial investment of equipment (€ 57 590), investments during use and maintenance (€2338), number of years of use (10 years), and number of procedures per year (131). ** based on 30 minutes scan time by a burn care nurse, and 10 minutes interpretation time

R1R2R3R4R5R6R7R8R9


114 | Chapter 6

Table 5.b. Mean costs of healthcare use, work absence and travel costs (€, 2012) per patient

Cost categoryLDI group Standard care

group(n=106) (n=96) Difference in costs,

p-values [95% CI]Hospital days (including readmittance)* 8 895 9481ICU days 565 216Day care 4 353 408Total burn centre stay 9 895 10105 0.881 [-2 538, 3 244]LDI **Other diagnostic procedures***Total diagnostic procedures

151385536

0276276 0.068 [-480, -117]

Surgical treatment 1 138 1054Wound care**** 660 605Blood products (erythrocytes) 39 13Scar care (pressure garments/silicones splints)

322 310

Total treatment 2 160 1982 0.621 [-918, 527]]Total clinical consultations 168 176 0.883 [-86, 107]

]

Total outpatient burn care530 537

0.885 [-91, 115]

Total costs specialized burn care 13 288 13 076 0.894 [-4 034, 3 434]

Total other healthcare costs (outside hospital)

374 575 0.461 [-281, 51]

Work absence (of patient/partner/parents)6 297 6459

Travel costs 243 394

Total patient costs 6 539 68540.706

[-1 294, 1 967]

Total costs per patient [95% CI]

20 202 [17 341, 23 063]

20 504 [17 614, 23 394] 0.870

[-3 384, 4 279]

* hospital day including personnel costs, accommodation, equipment, overhead, housing, food, laundry and medication** costs of LDI, including costs of initial investment, investments during use and maintenance, personnel costs and overhead and housing*** others: a.o. X-ray, CT scan, bronchoscopy**** wound care costs consist of material costs only, both in- and outpatient setting



Discussion

This is the first randomised controlled study that analysed the effects and costs of LDI as an additional test in the diagnosis of burn depth. It proved to be guidance for therapeutic decisions, for both non-operative and operative treated patients from the day of the LDI-scan onwards, which is valuable for both patients and professionals. On the day of the LDI-scan a definitive treatment choice (operative vs.non-operative) could be made significantly more often in patients in the LDI-group (p<0.001). Although we hypothesized that a significantly shorter time to wound healing in the LDI group was possible by earlier surgery for deep burns, overall no significant differences in time to wound healing were found. However, in the sub-group of admitted patients that required surgery, a significantly earlier decision for surgery and a shorter wound healing time (16.0 vs. 19.9 days, p=0.029) were seen. No differences in total costs were found between treatment groups. The introduction of LDI enabled an accurate burn depth diagnosis and early determination of therapy, against costs of €151 per scanned patient.

We conducted a pragmatic RCT, with broad inclusion criteria to mirror the use of LDI in our daily practice. We observed differences in the use of LDI in the clinical and outpatient setting. In the clinical setting, the majority of patients had an indication for LDI, in contrast to the outpatient setting, in which LDI was not frequently indicated. Furthermore, effectiveness of LDI was pre-dominantly seen in admitted patients, probably partly due to logistic reasons, such as more frequent wound inspection.11,30,31 An important finding of this trial was the major reduction in the ‘postponement of decision’, which is likely to improve patient and physician satisfaction and offers the possibility to perform surgery earlier and decrease LOS. We observed a significant reduction of 2.4 days in time to operative decision in admitted patients in the LDI group, similar to Kim et al.5 who observed a 1.7-day reduction. We were not able to reproduce the results of Petrie et al.4, who presented a large reduction of LOS (5.3 days) in operatively treated patients using the LDI, in a before-after cohort study. Possible reasons for these ‘less than expected’ results were the learning curve in the study, scarce OR time in our burn centres, and the inclusion of a relatively high number of outpatients.

To measure the real cost consequences of the introduction of a new diagnostic intervention for burn injuries, all costs involved should be calculated. Because no overall clinical effects were observed, we refrained from a cost–effective analysis and focused on comparison of cost on overall equally effective strategies. In our study, the cost price of LDI per patient was €151 (including equipment, personnel and overhead costs), which seems a reasonable price to improve our burn diagnosis and treatment prognosis. A reduction of one hospital day in surgically treated patients already undoes the extra costs of the LDI scans. Unfortunately,

R1R2R3R4R5R6R7R8R9


116 | Chapter 6

we were not able to detect such a cost reduction. Since we did detect a significantly earlier surgical decision, the potential for earlier surgery and cost reduction is certainly present. Next to the introduction of LDI, also hospital logistics need to be adapted to enable a reduction in hospital stay. With only one surgery day per week available (as in one of the participating burn centres) and scarce operating theatre time, it is difficult to operate earlier on patients. Of course, in large organizations like hospitals, logistics changes are complex and take time. If time to surgery can be reduced by 2.4 days, similar to the time to decision for surgery in our study, cost savings of €2124 per surgery patient can be reached, or €875 per scanned patient (this latter price also includes operatively treated patients).

To our knowledge, the only other manuscript, that discussed the costs and savings of LDI, was the British NICE guidance13. Their estimated average cost saving was £1248 (€ 1504) per patient scanned. This estimation was based on the assumption of a 17% reduction in the number of skin graft operations and a two-day reduction in LOS. In our treatment policy of delayed surgery (>10 days) in indeterminate burns, unnecessary surgery was not a pre-existing problem. In our centres cost savings can be reached by earlier surgery and a subsequently shorter LOS. However, in centres where early excision and grafting (within one week post-burn32) is daily practice, the potential cost savings can be reached by avoiding unnecessary surgery4. So, LDI can be of added value in both early and delayed excision.

The main limitation of this study is that a learning curve can be expected when a new diagnostic instrument is introduced. The LDI was already introduced 2 months before the start of the trial, but this was probably not long enough to eliminate a learning curve in interpretation of results. We suppose that effects of LDI would be larger, when this trial would be repeated in our centres in the near future. Although LDI is a desirable additional test in burn depth diagnosis, the assessment of burn depth and subsequent treatment choices remains difficult in some cases. We observed that treatment choice was often postponed in wounds with ’mixed’ LDI predictions. Earlier studies indicate that wounds which do not heal within 21 days have a higher risk of scar hypertrophy33. On the other hand, a recent study suggests that the risk for hypertrophic scarring in small burns (<5%TBSA) is low, also those that heal after 21 days34. Altogether, the best treatment for wounds with a mixed healing potential of 14-21 days and >21 days remains to be determined. To reach an optimal scar quality, should we wait for spontaneous healing of parts of these mixed wounds and perform late surgery on the remaining parts (like in our trial), or should we operate early? This would be a good subject for further study.



In conclusion, the introduction of LDI in burn care improved therapeutic decisions from the day of the LDI-scan, between 2-5 days post burn. No overall significant differences in time to wound healing or costs could be shown, but time to wound healing was improved in the subgroup of admitted patients requiring surgery. Further optimization of acute care in our burn centres, including logistics, can lead to potential cost-savings of €875 per scanned patient.

Acknowledgements We would like to thank the patients and their families for their contribution and participation in our study.

R1R2R3R4R5R6R7R8R9


118 | Chapter 6

References

1 Monstrey S, Hoeksema H, Verbelen J, Pirayesh A, Blondeel P. Assessment of burn depth and burn wound healing potential. Burns 2008; 34(6): 761-769.

2 Jeng JC, Bridgeman A, Shivnan L, Thornton PM, Alam H, Clarke TJ, et al. Laser Doppler imaging determines need for excision and grafting in advance of clinical judgment: a prospective blinded trial. Burns 2003; 29(7): 665-670.

3 Niazi ZB, Essex TJ, Papini R, Scott D, McLean NR, Black MJ. New laser Doppler scanner, a valuable adjunct in burn depth assessment. Burns 1993; 19(6): 485-489.

4 Petrie N, Norbury W, Fogarty B, Philip B, Baraat J, Dziewulski P. The use of the laser Doppler imaging to reduce operative intervention in the treatment of paediatric burns. Abstract Book, International Society for Burn Injuries 2004;(Abstract Book, International Society for Burn Injuries).

5 Kim LH, Ward D, Lam L, Holland AJ. The impact of laser Doppler imaging on time to grafting decisions in paediatric burns. J Burn Care Res 2010; 31(2): 328-332.

6 Hemington-Gorse SJ, Potokar TS, Drew PJ, Dickson WA. Burn care costing: the Welsh experience. Burns 2009; 35(3): 378-382.

7 Ahn CS, Maitz PK. The true cost of burn. Burns 2012;. 8 Hop JM, Polinder S, van der Vlies CH, Middelkoop E, van Baar ME. Costs of burn care; a systematic

review. Wound Repair Regen. 2014 Jul-Aug;22(4):436-50.9 Sanchez JL, Pereperez SB, Bastida JL, Martinez MM. Cost-utility analysis applied to the treatment

of burn patients in a specialized centre. Arch Surg 2007; 142(1): 50-7; discussion 57. 10 Monstrey SM, Hoeksema H, Baker RD, Jeng J, Spence RS, Wilson D, et al. Burn wound healing

time assessed by laser Doppler imaging. Part 2: validation of a dedicated colour code for image interpretation. Burns 2011; 37(2): 249-256.

11 Hoeksema H, Van de Sijpe K, Tondu T, Hamdi M, Van Landuyt K, Blondeel P, et al. Accuracy of early burn depth assessment by laser Doppler imaging on different days post burn. Burns 2009; 35(1): 36-45.

12 Van den Bruel A, Cleemput I, Aertgeerts B, Ramaekers D, Buntinx F. The evaluation of diagnostic tests: evidence on technical and diagnostic accuracy, impact on patient outcome and cost-effectiveness is needed. J Clin Epidemiol 2007; 60(11): 1116-1122.

13 National Institute for Health and Care Execllence (NICE). MoorLDI2-BI: a laser Doppler blood flow imager for burn wound assessment. NICE medical technology guidance 2 2011;.

14 Pape SA, Skouras CA, Byrne PO. An audit of the use of laser Doppler imaging (LDI) in the assessment of burns of intermediate depth. Burns 2001; 27(3): 233-239.

15 Riordan CL, McDonough M, Davidson JM, Corley R, Perlov C, Barton R, et al. Noncontact laser Doppler imaging in burn depth analysis of the extremities. J Burn Care Rehabil 2003; 24(4): 177-186.

16 Kloppenberg FW, Beerthuizen GI, ten Duis HJ. Perfusion of burn wounds assessed by laser Doppler imaging is related to burn depth and healing time. Burns 2001; 27(4): 359-363.

17 Holland AJ, Martin HC, Cass DT. Laser Doppler imaging prediction of burn wound outcome in children. Burns 2002; 28(1): 11-17.

18 Hop MJ, Hiddingh J, Stekelenburg C, Kuipers HC, Middelkoop E, Nieuwenhuis MK, et al. Cost-effectiveness of laser Doppler imaging in burn care in the Netherlands. BMC Surg 2013; 13: 2-2482-13-2.

19 Bloemen MC, Boekema BK, Vlig M, van Zuijlen PP, Middelkoop E. Digital image analysis versus clinical assessment of wound epithelialization: a validation study. Burns 2012; 38(4): 501-505.

20 Raat H, Landgraf JM, Oostenbrink R, Moll HA, Essink-Bot ML. Reliability and validity of the Infant and Toddler Quality of Life Questionnaire (ITQOL) in a general population and respiratory disease sample. Qual Life Res 2007; 16(3): 445-460.

21 Dolan P. Modelling valuations for EuroQol health states. Med Care 1997; 35(11): 1095-1108. 22 Stolk EA, Busschbach JJ, Vogels T. Performance of the EuroQol in children with imperforate anus.

Qual Life Res 2000; 9(1): 29-38.



23 Draaijers LJ, Botman YA, Tempelman FR, Kreis RW, Middelkoop E, van Zuijlen PP. Skin elasticity meter or subjective evaluation in scars: a reliability assessment. Burns 2004; 30(2): 109-114.

24 Draaijers LJ, Tempelman FR, Botman YA, Kreis RW, Middelkoop E, van Zuijlen PP. Colour evaluation in scars: tristimulus colorimeter, narrow-band simple reflectance meter or subjective evaluation? Burns 2004; 30(2): 103-107.

25 Draaijers LJ, Tempelman FR, Botman YA, Tuinebreijer WE, Middelkoop E, Kreis RW, et al. The patient and observer scar assessment scale: a reliable and feasible tool for scar evaluation. Plast Reconstr Surg 2004; 113(7): 1960-5; discussion 1966-7.

26 van der Wal MB, Tuinebreijer WE, Bloemen MC, Verhaegen PD, Middelkoop E, van Zuijlen PP. Rasch analysis of the Patient and Observer Scar Assessment Scale (POSAS) in burn scars. Qual Life Res 2012; 21(1): 13-23.

27 Hakkaart-van Roijen L, Tan SS, Bouwmans CAM. Handleiding voor kostenonderzoek; methoden en standaard kostprijzen voor economische evaluaties in de gezondheidszorg. Diemen: College voor zorgverzekeringen; 2010.

28 Gold MR. Siegel JE, Russel LB, Weinstein MC. Cost-effectiveness in health and medicine. New York: Oxford University Press; 1996.

29 Droog EJ, Steenbergen W, Sjoberg F. Measurement of depth of burns by laser Doppler perfusion imaging. Burns 2001; 27(6): 561-568.

30 Nguyen K, Ward D, Lam L, Holland AJ. Laser Doppler Imaging prediction of burn wound outcome in children: is it possible before 48 h? Burns 2010; 36(6): 793-798.

31 Park YS, Choi YH, Lee HS, Moon DJ, Kim SG, Lee JH, et al. The impact of laser Doppler imaging on the early decision-making process for surgical intervention in adults with indeterminate burns. Burns 2013; 39(4): 655-661.

32 Hoogewerf CJ, Hop MJ, Nieuwenhuis MK, Middelkoop E, van Baar ME. Early excision and grafting for burns. Protocol information. Cochrane database of systematic reviews (Online) 2012;.

33 Cubison TC, Pape SA, Parkhouse N. Evidence for the link between healing time and the development of hypertrophic scars (HTS) in paediatric burns due to scald injury. Burns 2006; 32(8): 992-999.

34. Hassan S, Reynolds G, Clarkson J, Brooks P. Challenging the dogma: relationship between time to healing and formation of hypertrophic scars after burn injuries. J Burn Care Res. 2014 Mar-Apr;35(2):e118-24.

Chapter 7Photographic assessment of burn size and depth:

reliability and validity

M. Jenda HopChantal M. Moues

Katja BogomolovaMarianne K. Nieuwenhuis

Irma M.M.H. OenEsther Middelkoop

Roelf S. Breederveld Margriet E. van Baar

J Wound Care. 2014 Mar;23(3):144-5, 148-52.

R1R2R3R4R5R6R7R8R9


122 | Chapter 7

Abstract

Objectives: Accurate burn size and depth assessment is important for optimal treatment choices. Telemedicine, using photographs of burns assessed by experts, has the potential to assist inexperienced physicians in burn assessment. The aim of this study was to examine the reliability and validity of using photographs of burns to assess both their size and depth.Methods: Fifty randomly selected photographs made on day 0-1 post burn were assessed by seven burn experts and eight referring physicians. Inter-rater reliability in both groups (experts vs. referrers) was calculated. Validity of burn size assessment was calculated using live assessment as gold standard, and of burn depth using clinical assessment in combination with laser Doppler imaging as gold standard. Validity of the photographically assessed indication for surgery was calculated using laser Doppler imaging and actual treatment as gold standard. Finally, agreement in referral indication was calculated.Results: Using photographs, burn size could be assessed reliably and validly by experts (ICCs of 0.83 and 0.87), but not by referrers (ICCs of 0.68 and 0.78). Photographic assessment of burn depth was neither reliable nor valid, with ICCs respectively of 0.38 and 0.28 for experts and 0.24 and 0.13 for referrers. The indication for surgery could also not be assessed validly. Agreement between assessors regarding referral indication was low. Conclusion: Burn size, but not burn depth, can be assessed reliably and validly by experts using photographs of the burn wound. We recommend exploring other forms of telemedicine, like live interactive video, to investigate if this leads to an improved burn depth assessment where clinical assessment is not possible.


Photographic assessment of burn size and depth: reliability and validity | 123

Background

The initial assessment of burn size and depth is important. First of all, accurate assessment of burn size, expressed as percentage of total body surface area (TBSA) burned, determines whether transfer to a burn centre is necessary as well as the need for and amount of initial intravenous fluid resuscitation. Secondly, early assessment of burn depth assessment determines the most optimal wound treatment: surgical or conservative. The initial assessment of a burn wound, however, is often difficult and its accuracy depends on the experience of the assessor1,2. Burn size is clinically assessed using the Lund and Browder Charts, rules of nine and/or hand rule3-5. Inexperienced assessors often estimate burn size incorrectly, which leads to significant errors in fluid resuscitation and expensive over-triage6-10. Burn depth is assessed clinically as well, the diagnosis relies on an evaluation of several wound features such as appearance, capillary refill and sensibility1. Inaccurate burn depth assessment can lead to suboptimal choice of topical treatment and to unnecessary surgery or an unnecessary delay in surgery7,11-15.Because the assessment of burn size and depth is primarily a visual skill8 and the number of burn experts is limited, the use of telemedicine in burn care is an attractive diagnostic tool to explore. Telemedicine has numerous definitions, relating to the practice of medicine at a distance. These past decades experience has been gained with the use of telemedicine in burn care,8 using three basic telemedicine systems: live interactive video, store-and-forward images/photographs and telephone16. The use of telemedicine in burn care has several aims, e.g. to improve burn diagnosis, but also to support treatment choices (including referral). The focus of this study was on the diagnostic abilities of telemedicine in burn care. As with all diagnostics, reliability and validity are important17. Reliability is the degree in which the assessment by telemedicine is free from measurement error18. Validity is defined as the degree to which an instrument truly measures the construct(s) it purports to measure18.Several studies already investigated reliability and validity of burn assessment by different systems of telemedicine11,19-22. However, the reliability of photographic assessment of burn size has not been studied yet, to our knowledge. Two studies analysed the criterion validity of respectively photographic and video burn size assessment: a high correlation was found between telemedicine assessment and live assessment by burn experts11,19. Regarding photographic burn depth assessment, Boccara et al. reported on the inter-rater reliability: 75% agreement was found between 3 experienced surgeons20. The criterion validity of photographic burn depth assessment was presented in several studies. Poor to good agreement (kappa’s ranging from 0.33-0.60) between photographic and live assessment by an experienced physician was found in two studies. The highest agreement was demonstrated in experienced burn surgeons21,22. Two other studies assessed burn depth on a two-point-scale, and demonstrated an accuracy of 94% in photographic vs. live assessment11 and a sensitivity

R1R2R3R4R5R6R7R8R9


124 | Chapter 7

of 77% and a specificity of 75% in photographic assessment vs. initial live diagnosis and actual treatment, by 3 experienced observers20.Summarizing, some experience with telemedicine in burn diagnosis has been gained today. However, reliability of burn size or burn depth assessment by telemedicine has not been studied conclusively. Furthermore, the number of observers was often small, inter-observer bias can be reduced by higher numbers of observers in these studies. Also, the gold standard in all studies that calculated validity of burn depth was not optimal: research shows that the accuracy of live assessment of burn depth is limited, even in experienced assessors (50-75%)1,12,13. None of the studies chose live assessment in combination with laser Doppler imaging as the gold standard, which has proven to be a far more accurate predictor of burn depth/wound healing time23. In other words, the evidence on the reliability and validity of the use of telemedicine in burn diagnosis is limited. Hence, the aim of this study was to examine the reliability and validity of photographic assessment of burn size and burn depth by both burn centre experts and referring physicians, in order to determine whether this type of telemedicine can be used as a diagnostic tool and supports therapeutic decisions. Reliability and validity were determined by calculating Intraclass Correlation Coefficients, a ICC of 0.70 is considered acceptable, ICCs of >0.80 are considered good24. Hypotheses were:

1) the inter-rater reliability and validity of photographic burn size assessment by burn experts are high (ICC>0.80) and higher than in referring physicians

2) the inter-rater reliability and validity of photographic burn depth assessment by burn experts are acceptable (ICC= 0.70) and higher than in referring physicians

Methods

Study design and population Photographs made on day 0 or 1 post-burn were used in this study. They were randomly selected from a recent trial25, using a random number table26 and represented a variety of burn patients with a TBSA ≤20%. Ethical approval was obtained to use the photographs for this study (protocol T2013/13). A total of fifty cases were selected, in accordance with previous clinimetric studies24. As in daily practice, photographs of lesser quality and of non-debrided wounds were also included. Each case was presented in printed form and included at least one photograph of the burn and basic clinical information (Fig. 1). All observers received a standard explanation on the assessment of burn depth and size, in which burn depth was categorized as superficial partial thickness, deep partial thickness, or full thickness. Superficial partial thickness burns have the potential to heal within 14 days, whereas deep partial thickness burns do not heal within 21 days, and full thickness burns do not heal spontaneously at al27.



Burn size was defined as percentage total body surface area (TBSA) burned, using the Lund and Browder chart4.

Figure 1. Example of photographic assessed case

Figure 1. Example of photographic assessed case

R1R2R3R4R5R6R7R8R9


126 | Chapter 7

The assessors in this study were 7 burn experts (burn physicians/surgeons with > 5 years of experience) from the Dutch Burn Centres, and 8 referring physicians (2 ER physicians, 2 plastic surgeons, and 4 plastic surgery/surgery residents) from one general hospital. The assessors scored (Fig. 1):

• burn size • burn depth, of a specific area • indication for surgery of the marked area• indication for referral of the patient to a burn centre

The gold standard used for burn size was the first live assessment of percentage TBSA burned by one experienced burn physician (5 different burn experts in total), using the Lund and Browder chart, rules of nine and/or palm rule3-5.The gold standard used for burn depth was live assessment by an experienced burn centre physician in combination with laser Doppler imaging (between 48h and 5 days post-burn) (moorLDI2-BI, Moor Instruments). Laser Doppler imaging in combination with clinical assessment (clin-LDI) is the most accurate predictor of wound healing time with an accuracy of ≥95%1,23. Possible LDI results are a healing potential (HP) <14 days, HP 14-21 days, and HP > 21 days. Furthermore, biopsy results, available from surgical burn wounds in our original trial, were used, to create a 3-point scale that equalled the photographic assessment. Burns with a HP of <14 days and 14-21 days, based on clin-LDI, were considered superficial partial thickness burns. In burns with a HP of > 21 days, the biopsy results were used to differentiate between deep partial thickness and full thickness wounds. The gold standard used for the indication for surgery were clin-LDI (HP ≤21 days and HP>21 days) results. Additionally, the final treatment choice (surgery/no surgery) was used as comparator; not all specialists consider a HP >21 days as a cut-off point and the size of a wound and the overall patient status influences the treatment choice as well.

Statistical analysis The inter-observer reliability of photographic assessment of burn size and depth was measured within the expert group and the referrer group by calculating Intraclass Correlation Coefficients for single assessments (ICCs) with a two way ANOVA. In burn size also standard error of measurement (SEM) agreement was calculated. To analyse the criterion validity in the expert and referrer group, results of photographic burn assessment were compared to the gold standards, calculating ICCs. The Bland and Altman method was used to evaluate the agreement between burns size assessment of observers and the gold standard. Positive predictive values and sensitivity and specificity of burn depth assessment were calculated. To analyse validity of treatment choices in the expert and referrer group, the



photographic assessment of treatment choice (surgery/ no surgery) was analysed calculating sensitivity and specificity. The agreement in photographically made referral choice in the expert and referrer group was measured by calculating ICCs.

Results

Patient and injury characteristicsFifty-two percent of the included patients was male and mean age was 29 (range 0-87). The body regions most frequently burned were hand and arm (48% and 22%, respectively). Main burn causes were scalds and fire/flame (34% and 38%). The mean percentage TBSA burned, according to live assessment at presentation, was 3.2 (range 0.5-10.0). Fifty-two percent of the photographed burns had a HP<14 day and 46 percent a HP>21 days according to the clinical assessment in combination with the laser Doppler imager (Table 1).

Table 1. Patient and injury characteristics, including gold standardsn=50 (%) n=50 (%)

Male/ female 26/24 (52/48) % TBSA burned (live assessment), mean (SD; range)

3.2 ( 2.47; 0.5-10.0)

Age mean (SD; range) 29 (24; 0-87) 0.5-1.0 12 (24)0-4 11 (22) >1-2.0 12 (24)5-17 7(14) >2-3.0 5 (10)18-40 16 (32) >3-4.0 8 (16)41-60 11 (22) >4-5.0 6 (12)61-80 4 (8) >5-10.0 7 (14)>80 1 (2)Aetiology Depth according to clinical assessment +LDI Fire/Flame 19 (38) Healing potential within 14days 26 (52)Scald 17 (34) Healing potential between 14-21 days 1 (2)Oil 11 (22) Healing potential >21 days 23 (46)Contact 2 (4)Steam 1 (2) Depth according to biopsy *Location Deep partial thickness 9 (18.0)Head/Neck 3 (6) Full thickness 8 (16.0)Trunk 5(10) No biopsy 33 (66)Arm 24(48)Hand 11(22)Leg/Foot 7 (14)

*biopsy results were available in surgical treated patients only

R1R2R3R4R5R6R7R8R9


128 | Chapter 7

Burn sizeThe inter-rater reliability of photographic burn size assessment by the 7 burn experts was good; ICC of 0.83 (95% confidence interval (CI) 0.76-0.89), SEM of 0,93% TBSA. The inter-rater reliability of photographic burn size assessment by the 8 referring physicians was considered not acceptable; ICC of 0.68 (95% CI 0.56-0.78) and the SEM 1,24%. The validity of photographic burn size assessment, compared to live assessment by a burn expert, was good for the burn experts with an ICC of 0.87 (95% CI 0.84-0.89) and significantly higher than for referring physicians; ICC of 0.78 (95% CI 0.74-0.82) (Table 2). As shown in Fig. 2, both under- and overestimation of burn size was seen, estimation errors were larger in referrers. In burn experts, but not in referrers, a trend was seen between years of experience and the validity of burn size assessment (Fig. 3).

a. Experts

Figure 2. Bland–Altman plot of the difference in TBSA observer minus golden standard against the mean TBSA per photograph

a. Experts



b. Referrers

The horizontal lines indicate the mean difference (middle) and the upper and lower limits of agreement (±2 SD).

b. Referrers

Figure 2. Bland–Altman plot of the difference in TBSA observer minus golden standard against the mean TBSA per photographThe horizontal lines indicate the mean difference (middle) and the upper and lower limits of agreement (±2 SD).

Table 2. Burn size (%TBSA burned) and burn depth reliability and validity

Burn Size

Inter-Rater Reliability ICC (95% CI)

Validity ICC (95% CI)Gold standard: live assessment of burn experts at first presentation

Burn centre expert (n=7) 0.83 (0.76-0.89) 0.87 (0.84-0.89)Referring physicians (n=8) 0.68 (0.56-0.78) 0.78 (0.74-0.82)

Burn Depth Gold standard: clin-LDI Burn centre expert (n=7) 0.38 (0.25-0.53) 0.28 (0.18-0.38)Referring physicians (n=8) 0.26 (0.16-0.39) 0.13 (0.03-0.22)

Missing values: reliability burn size experts 2/50, referrers 2/50, validity burn size experts 2/350, referrers: 3/400, reliability burn depth experts 4/50 (in one expert 16 items missing, therefore excluded from reliability analysis), referrers 2/50 validity burn depth, experts 20/350, referrers 4/400

R1R2R3R4R5R6R7R8R9


130 | Chapter 7

Figure 3. Validity (ICCs) of photographic burn size assessment in experts

Results per period of experience in years, respectively 2, 2 and 3 observers per group

Figure 3. Validity (ICCs) of photographic burn size assessment in experts Results per period of experience in years, respectively 2, 2 and 3 observers per group

Burn depthThe photographic assessment of burn depth was not reliable, a low ICC between experts was found: 0.38 (95% CI; 0.25-0.53), and an even lower ICC between referrers: 0.24 (95% CI 0.14-0.37). The validity of burn depth assessment compared to clin-LDI was low, with an ICC of 0.28 (95% CI 0.18-0.38) for experts and 0.13 (95% CI 0.03-0.22) for referrers (Table 2 and 3). The positive predictive value of deep partial thickness and full thickness burns was 55.4% for experts and 51.9% for referrers. Sensitivity in burn experts of the photographic recognition of deep partial thickness and full thickness burns (when clin-LDI HP was >21 days) was 83.8% (95% CI 79.8-87.7), specificity (the photographic recognition of superficial dermal burns, when clin-LDI HP was ≤21 days) was 40.9% (95% CI 35.6-46.2). Sensitivity in referrers was 66.8% (95% CI 62.2-71.5) and the specificity 46.2% (95% CI 41.3-51.1). In other words, in both groups the most common mistake in photographic assessment was overestimation of burn depth. In referrers also many wounds were underestimated (Table 3). No relation between years of experience and validity was seen. Due to the possibility that the low ICCs found in burn depth validity were the result of unequal 3-point scales between photographic assessment and gold standard (respectively, superficial partial thickness, deep partial thickness, full thickness, vs. LDI: HP<14, HP 14-21, HP >21 days), we aimed to create equal scales by combining two gold standards: clin-LDI and biopsy results in one new gold standard. The validity of photographic assessment compared to the combined gold standard remained low, with an ICC of 0.37 (95%CI 0.25-0.54) in experts and an ICC of 0.19 (95% CI 0.09-0.29) in referrers.



Table 3. Photographic assessment of burn depth compared to clinical assessment + LDI

HP<14LDI resultsHP 14-21 HP>21 days

Total

Phot

ogra

phic

as

sess

men

t ex

pert

s (n

=7) Superficial partial

thickness69 3 25 97

Deep partial thickness

88 4 94 186

Full thickness 12 0 35 47

Total 169 7 154 330*

Phot

ogra

phic

as

sess

men

t re

ferr

ers

(n=8

)

Superficial partial thickness

93 5 61 159

Deep partial thickness

86 3 87 176

Full thickness 25 0 36 61

Total 204 8 184 396**

*In total 20 observations missing. ** In total 4 observations missing. HP, healing potential in days.

Treatment choiceThe sensitivity of photographic assessment of surgery indication in experts, compared to clin-LDI, was 49.7% (95% CI 44.2-55.2, n=73/147), specificity was 82.4% (95%CI 78.2-86.5. n=140/170). In referring physicians sensitivity and specificity were even lower, respectively 39.9% (95% CI 35.1-44.7, n=73/183) and 76.2% (95% CI 72.0-80.4, n=163/214). All wounds with a clin-LDI prediction of ≤21 days (n = 37), healed without surgery. Of the wounds (n=23) with a predicted healing potential of > 21 days, 20 were treated surgically, 3 healed without surgery. Sensitivity and specificity, using the actual treatment choice as a reference test, were 52.0% (95% CI 46.5-57.5) and 80.5% (95% CI 76.2-84.9) in burn experts and 42.1% (95% CI 37.3-47.0) and 76.1% (95% CI 71.0-80.2) in referrers, respectively.

Referral choiceBurn experts more often judged that referral was indicated than referring physicians, in 78.3% compared to 68.0% of cases (p<0.00). Within both experts and referrers a low agreement was found (ICC 0.25, 95% CI 0.15-0.38 and ICC 0.25, 95% CI 0.14-0.38, respectively).

Discussion

This was the first study into the clinimetrics of telemedicine in burn care with multiple observers and an accurate gold standard for the assessment of burn depth. We showed that photographic assessment by burn experts was a reliable and valid method to assess burn

R1R2R3R4R5R6R7R8R9


132 | Chapter 7

size. Contrary to our hypothesis, reliability and validity of photographic burn depth assessment was not acceptable. Subsequently, the need for surgery could not be assessed validly using photographs. Experts more often observed an indication for referral than referrers, but the agreement within the experts and referrer groups was low.Similar to a previous study on the photographic assessment of burn size11, in our study experts could make a valid and reliable assessment of burn size, which potentially improves referral and resuscitation choices. Whereas the previous study included only one observer, in the current study several burn experts (n=7) participated, which strengthens our conclusions. An extra observation made in our study was the trend between years of burn care experience and validity in burn experts. Jones et al. have reported similar findings on photographic assessment of burn depth by experts of different levels; poor validity was found in junior nurses and good validity in plastic surgeons22. It is important to realize that years of experience play a role in the use of telemedicine. For an reliable and valid assessment of burn depth more information is needed than the visual aspect obtained by a photograph. Our study results were not in accordance with some previously presented positive results of photographic assessment of burn depth20-22. Although the results for experts were better than for referrers, neither reliability nor validity was acceptable in any group. In accordance with Boccara et al. most errors in our study were due to overestimation of burn depth20. These results resemble studies on ‘live’ clinical assessment of burn depth in which the most common error was overestimation1,12,13. Since the optimal method of early burn depth assessment is not yet established, in our study several gold standards were used. Our study was the first telemedicine study that used clinical assessment in combination with LDI results. Possibly, the earlier presented positive results of photographic burn depth assessment were caused by the use of the gold standard live assessment by a burn expert at presentation, which is in timing almost equal to the photographic assessment (shortly after burn injury). However, live assessment in itself is accurate in only 50-75% of case1,12,13, therefore we chose the LDI prediction, with an accuracy of >95% between 2-5 days after injury, as gold standard. The addition of biopsy data resulted in (non-significant) improvement of sensitivity and specificity, but correct assessment of burn depth remained poor.Also the choice of conservative vs. surgical treatment could not be validly assessed using photographs. Whereas overestimation was the most common error in depth, surprisingly, the most common error in treatment choice was underestimation of the need for surgery. In Dutch burn care, because of heterogeneity of most burns, it is uncommon to perform early excision and grafting (within one week post-burn)28 preventing overzealous excision29. This could explain the reluctance to decide on surgery based on a photograph. Furthermore, there was only limited agreement between experts on the indication for referrals to specialized burn care, possibly partially caused by the mild severity of burn injuries in our study. In patients with



larger burn sizes and larger full thickness burn sizes, agreement for referral would probably be higher. Additionally, the referral criteria of the Emergency Management of Severe Burns (EMSB), do leave some room for referral choices in less severe burns (EMSB Manual, Dutch Version)30.Some limitations of our study must be mentioned. The burn sizes were rather small; range 0.5-10 % TBSA. We did not investigate whether our results apply to patients with a larger percentage TBSA burned. However, we expect that the difference in validity of burn size assessment between experts and referrers will be even higher in larger burns, as was described by Saffle et al.19. Furthermore, the burn size in our study population is representative of the majority of burn injuries in high-income countries like the USA in which 63% of the patients has a TBSA of <10%31. Therefore, our results are probably applicable to most burn injuries in high-income countries. Another point was the timing of the photograph (0-1 day post-burn). In daily practice, also photographs will be used made a few days post-injury, to obtain advice about wound treatment. From live assessment studies, it is known that validity of burn depth is higher when performed several days post-burn23. This probably also applies to photographic assessment of burn depth. But to mimic daily practice, we chose to study photographs of burns within 1 day post-burn. The last limitation was the (retrospective) design of our study; in practice only experts will photographically assess the burns and the referrers will perform a live assessment before considering telemedicine. Therefore, in practice, reliability and validity of assessment by referrers are possibly more favourable than in our study.The currently tested method is insufficient to completely assess a burn. A next step is to use live interactive video instead of photographs, this adds the opportunity of an extensive anamnesis and testing of capillary refill and sensibility8. The validity of live interactive video was only described for burn size assessment, until now19. There is no insight in the reliability and validity of burn depth assessment by live interactive video. Our research group is currently preparing a study into the clinimetrics of live interactive video assessment. Future research should also address the actual impact and cost-effectiveness of the introduction of telemedicine, e.g. the prevention of unnecessary referrals, in a clinical trial.

In conclusion, the use of photographs only enables a partial diagnosis of the burn injury, that of burn size, which can support referral and resuscitation choices. For burn depth, telemedicine by photographs is neither valid nor reliable, which makes it an inadequate diagnostic instrument to support surgical treatment choices. We recommend exploring other forms of telemedicine, like live interactive video, to find out if this leads to improved burn depth assessment.

R1R2R3R4R5R6R7R8R9


134 | Chapter 7

AcknowledgementsWe sincerely thank all Dutch burn experts and referring physicians from the Medical Centre Leeuwarden for their participation in this study. In addition, we thank Nicole Trommel, RN and Hans Eshuis, RN, for their contribution in the collection of the photographs and Bouke Boekema, PhD, and colleagues for assessing the biopsies. Finally, we thank Henrica de Vet, PhD and Wim Tuinebreijer, MD PhD, for their statistical advises.

Declaration of interestNone of the authors has a financial interest in any of the products, devices, or drugs mentioned in this manuscript.



References

1. Monstrey S, Hoeksema H, Verbelen J, et al. Assessment of burn depth and burn wound healing potential. Burns. 2008 Sep;34(6):761-9.

2. Berkebile BL, Goldfarb IW, Slater H. Comparison of burn size estimates between prehospital reports and burn centre evaluations. J Burn Care Rehabil. 1986 Sep-Oct;7(5):411-2.

3. Nagel TR, Schunk JE. Using the hand to estimate the surface area of a burn in children. Pediat Emerg Care. 1997; 13:254–5.

4. Lund CC, Browder NC. The estimation of areas of burns. Surg Gynecol Obst. 1944;79:352-8.5. Wallace AB. The exposure treatment of burns. The Lancet. 1951;257(6653):501–4.6. Baartmans MG, van Baar ME, Boxma H, et al. Accuracy of burn size assessment prior to arrival

in Dutch Burn centres and its consequences in children: a nationwide evaluation. Injury. 2012 Sep;43(9):1451-6.

7. Saffle JR. Telemedicine for acute burn treatment: the time has come. J Telemed Telecare. 2006; 12(1):1-3.

8. Wallace DL, Hussain A, Khan N, Wilson YT. A systematic review of the evidence for telemedicine in burn care: with a UK perspective. Burns. 2012 Jun;38(4):465-80.

9. Freiburg C, Igneri P, Sartorelli K, Rogers F. Effects of differences in percent total body surface area estimation on fluid resuscitation of transferred burn patients. J Burn Care Res. 2007;28:42– 48.

10. Hagstrom M, Wirth GA, Evans GR, Ikeda CJ. A review of emergency department fluid resuscitation of burn patients transferred to a regional, verified burn centre. Ann Plast Surg 2003;51:173–6.

11. Shokrollahi K, Sayed M, Dickson W and Potokar T. Mobile phones for the assessment of burns: we have the technology. Emerg Med J. 2007 November; 24(11): 753–755.

12. Niazi ZB, Essex TJ, Papini R, et al. New laser Doppler scanner, a valuable adjunct in burn depth assessment. Burns 1993; 19(6):485–9.

13. Jeng JC, Bridgeman A, Shivnan L, et al. Laser Doppler imaging determines need for excision and grafting in advance of clinical judgment: a prospective blinded trial. Burns 2003; 29(7):665–70.

14. Petrie N, Norbury W, Fogarty B, et al. The use of laser Doppler imaging to reduce operative intervention in the treatment of paediatric burns. In Abstract book, 12th congress of the International Society for Burn Injuries. Yokahama: ISBI; 2004.

15. Kim LH, Ward D, Lam L, Holland AJ. The impact of laser Doppler imaging on time to grafting decisions in pediatric burns. J Burn Care Res 2010; 31(2):328–332.

16. Turk E, Karagulle E, Aydogan C, et al. Use of telemedicine and telephone consultation in decision-making and follow-up ofburn patients: Initial experience from two burn units. Burns. 2011May;37(3):415-9.

17. Van den Bruel A, Cleemput I, Aertgeerts B, et al. The evaluation of diagnostic tests: evidence on technical and diagnostic accuracy, impact on patient outcome and cost effectiveness is needed. J Clin Epidemiol 2007, 60(11):1116–22.

18. Mokkink LB, Terwee CH, Patrick DL, et al. International consensus on taxonomy, terminology, and definitions of measurement properties for health-related patient- reported outcomes: results of the COSMIN study. J Clin Epid. 2010; 63:737-45.

19. Saffle JR, Edelman L, Theurer L, et al. Telemedicine evaluation of acute burns is accurate and cost-effective. J Trauma. 2009 Aug;67(2):358-65.

20. Boccara D, Chaouat M, Uzan C, et al. Retrospective analysis of photographic evaluation of burn depth. Burns. 2011 Feb;37(1):69-73.

21. Jones OC, Wilson DI, Andrews S. The reliability of digital images when used to assess burn wounds. J Telemed Telecare. 2003;9(1):S22-4.

22. Jones O. Measurements of the clinical competence of doctors and nurses to process telemedicine referrals for burns patients. J Telemed Telecare. 2005;11(1):89-90.

23. Hoeksema H, Van de Sijpe K, Tondu T, et al. Accuracy of early burn depth assessment by laser Doppler Imaging on different days post burn. Burns 2009, 35(1):36–45.

24. Vet de HCW, Terwee CB, Mokkink LB and Knol DL. Measurement in Medicine. Cambridge: Cambridge University Press; 2011.

R1R2R3R4R5R6R7R8R9


136 | Chapter 7

25. Hop MJ, Hiddingh J, Stekelenburg CM, et al. LDI study group. Cost-effectiveness of laser Doppler imaging in burn care in the Netherlands. BMC Surg. 2013 Feb 1;13:2.

26. Altman DG. Practical statistics for medical research. London: Chapman & Hall, 1991. 27. Latarjet J. A simple guide to burn treatment. International Society for Burn Injuries in collaboration

with the World Health Organization. Burns. 1995 May;21(3):221-5.28. Hoogewerf CJ, Hop MJ, Nieuwenhuis MK, et al. Early excision and grafting for burns (Protocol).

Cochrane Database of Systematic Reviews 2012, Issue 3. 29. Hop MJ, Hoogewerf CJ, van Baar ME, et al. A call for evidence: timing of surgery in burns. Burns.

2012 Jun;38(4):617-8. 30. The Education Committee of the Australian and New Zealand Burn Association. Emergency

management of severe burns (EMSB) course manual, Dutch version Dutch Burn Foundation (2009).

31. Latenser BA, Miller SF, Bessey PQ, et al. National Burn Repository 2006: a ten-year review. J Burn Care Res. 2007Sep-Oct;28(5):635-58.

Part III. Improving burn care efficiency:

therapeutics and costs

Chapter 8Early excision and grafting for burns

Cornelis J. HoogewerfM. Jenda Hop

Marianne K. NieuwenhuisEsther Middelkoop

Cornelis H. van der VliesMargriet E. van Baar

Cochrane Database of Systematic Reviews (amended form)In revision

R1R2R3R4R5R6R7R8R9


140 | Chapter 8

Abstract

Background: Burn injuries are an important health problem. The functional and cosmetic outcome of a burn depends on the size, depth and treatment of the burn. It is generally understood that superficial burns heal well with topical treatment alone, while deep partial-thickness and full-thickness burns often require surgical treatment. However, there is debate about the optimal timing of surgery.Objectives: To assess the effects of early excision and grafting on scar quality in people with burns of all depths.Search methods: We searched the Cochrane Wounds Group Specialised Register (searched 13 May 2015); the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library 2015, Issue 4); the Database of Abstracts of Reviews of Effects (DARE) (The Cochrane Library 2014, Issue 3); the NHS Economic Evaluation Database (The Cochrane Library 2014, Issue 8); Ovid MEDLINE (1946 to May 2015); Ovid MEDLINE (In-process & Other Non-Indexed Citations 12 May, 2015); Ovid EMBASE (1974 to 12 May, 2015); and EBSCO CINAHL (1982 to 13 May 2015) for relevant trials. We did not apply date or language restrictions.Selection criteria: Randomised controlled trials (RCTs), both published and unpublished, that evaluated the effects of early excision and grafting in people with burns were eligible for inclusion in this review.Data collection and analysis: Two review authors independently assessed and included the references identified by the search strategy. Included trials were assessed using a risk of bias form, and data were extracted using a standardised data extraction sheet. For dichotomous and continuous outcomes, we calculated risk ratios and mean differences, respectively, both with 95% confidence intervals.Main results: We included ten RCTs, comprising a total of 416 participants. All studies had relatively small sample sizes (13 to 85) and had some methodological limitations. Heterogeneity of interventions and outcomes prevented pooling of data for most results. Three studies addressed scar quality as an outcome but there was insufficient data to support any definite conclusions. Additional results indicated that early excision and grafting had a lower number of wound swabs positive for contamination and shorter length of hospital stay compared with conservative treatment. On the other hand, four studies addressed the proportion of participants requiring surgery for wound closure; 48% (34 out of 71) of the participants that received conservative treatment had wounds that healed without surgery. In addition, four studies showed participants who received delayed excision (conservative treatment) experienced less blood loss during surgery compared to participants in the early excision group.Authors’ conclusions: There is insufficient high quality research and evidence to enable definite conclusions to be drawn about the effects of early excision and grafting on scar quality in people with burns.


Early excision and grafting for burns | 141

Background

Burn injuries are an important health problem. Worldwide it is estimated that each year over 195,000 people die from fire-related burn injuries and millions more suffer from burn-related disabilities and disfigurement1. In the United States of America (USA), annually 40,000 people are admitted for burns, including 30,000 admissions to hospitals with specialised burn centres2. In the United Kingdom, approximately 13,000 people per year are admitted to hospital for treatment of burns3, while in the Netherlands the annual figure is about 1800 people4, 550 to 600 of whom are treated in one of the three Dutch burn centres5.Major improvements in burn care in the twentieth century mean that mortality rates from burn injuries have substantially decreased. This has resulted in a shift in attention from mortality towards morbidity, for example, the functional outcome after a burn injury6. It is generally understood that superficial burns heal well with few or no functional or aesthetic problems. However, the best treatment for deep partial or full-thickness burns remains controversial7. In the United States, there seems to be consensus on early excision and grafting8, while in Europe, usually, a more conservative approach is applied. In conservative treatment, surgery is postponed until clinical assessment indicates that surgery is inevitable. Eventually both United States and Europe excise and graft deep partial-thickness burns, but optimal timing of this intervention is unclear. Timing is important because early excision could reduce length of hospital stay but increases the risk of overzealous excision (removal of tissue that could have healed without surgery), while delayed excision reduces the risk of overzealous excision but increases the risk of wound infection and length of hospital stay. Overzealous excision could also increase risk of scarring. Scars usually appear when wound healing takes more than two weeks or after surgical treatment. Therefore, it might be beneficial to postpone surgery until it becomes evident whether a wound will heal within two weeks.

Description of the conditionA burn injury of the skin occurs when some or all the different layers of the skin are destroyed by physical energy such as hot liquid, flame, contact burns or ultraviolet/infrared radiation, radioactivity, electricity or chemicals9. In addition to the localisation of a burn and associated injuries, severity of burn wounds is characterised by size and depth. The size of a burn is assessed by the Total Body Surface Area percentage (TBSA%), which is the percentage of the skin surface area burned. The depth of a burn is determined by which layers of the skin are destroyed, only the epidermis or both the epidermis and the dermis. So far, no consensus has been reached on the exact classifications of burns, especially not in relation to the classification of depth10. In general, skin burns are classified by depth as either superficial partial-thickness burns, deep partial-thickness burns or full-thickness burns. In superficial partial-thickness burns only the epidermal layer and the superficial part of the dermis is destroyed. Healing

R1R2R3R4R5R6R7R8R9


142 | Chapter 8

generally occurs within two weeks due to the migration of epithelial cells from the stratum basale to the surface, with very little or no scarring. In deep partial-thickness burns, the epidermis and most of the dermis is destroyed with damage to deeper structures within the skin such as blood vessels, nerves, glands and hair follicles. The damage of glands and hair follicles hampers the migration of epithelial cells from the stratum basale surrounding these accessory structures. Healing generally takes more than two to three weeks and occurs due to the migration of epithelial cells from the wound edges and the sparse epithelial elements present in the bottom of the wound. Finally, full-thickness burns involve all layers of the skin and may involve structures underneath, such as muscle and bone, leaving little chance of healing from the epithelial elements in the bottom of the wound. In case of a very small full-thickness burn, healing might occur by contraction and epithelial cell growth from the wound edges8. Full-thickness burns will always result in scarring and often result in hypertrophic scarring. In addition, hypertrophic scarring may also occur when re-epithelialisation does not occur within two to three weeks11,12.Burn wound surfaces are sterile immediately following thermal injury, but they are rapidly colonised by a variety of micro-organisms13,14. Those micro-organisms not only originate from the patient’s own skin, respiratory and gastro-intestinal flora, but also from contact with contaminated external environmental surfaces, hands of healthcare workers and even air13-15. Burn wounds provide a favourable niche for microbial colonisation and proliferation because of their protein-rich environment and avascular necrotic tissue13,16. This avascularity of eschar (necrotic tissue) results in impaired migration of host immune cells and restricts delivery of systemically administered antimicrobial agents to that area. The most common burn wound pathogens are Staphyloccocus aureus and Pseudomonas aeruginosa17. Colonisation of burn wounds has been associated with delayed wound healing, increased need for surgical interventions and prolonged length of stay at burn centres18.In addition to local responses, severe burn injuries can also induce systemic responses like cardiovascular, respiratory, metabolic and immunological changes. The release of inflammatory mediators at the site of injury has a systemic effect once a burn reaches 30% of the total body surface area19. These systemic reactions, besides generating excessive oedema in burns as a result of increased capillary permeability, can further compromise wound healing. It is important to consider adequate local treatment as well as systemic management of a burn as this may influence the final outcome of the injury.Hypertrophic scarring occurs if the balance between collagen synthesis and breakdown is disrupted8. The post-burn hypertrophic scar might present itself as a pink to red, slightly thickened or as a red to purple inelastic mass of skin tissue. Functional impairment can occur, especially if a hypertrophic scar crosses a joint or surrounds openings like eyes. For instance, a contracture of the elbow may result in a limited range of motion and even cause restrictions in daily life activities. Another example are eyes that may not close due to the inelasticity and



contraction of the hypertrophic skin. Furthermore, scars can result in discomfort because of itching and sometimes cause neuropathic pain after burn injury20. The degree of hypertrophic scarring differs among individuals and depends on a variety of factors, one of which is time to wound healing, with hypertrophic scar formation being seen more often if wound healing takes more than 21 days11. In general, a deeper burn wound results in the formation of more hypertrophic tissue. Other factors involved with hypertrophic scarring are race, age, genetic factors, type of injury, anatomic region and mechanical tension on the wound21.

Description of the interventionThe focus of this review is early excision and grafting for burns. In this section we will elaborate on the different surgical excision and grafting techniques used in burns as well as issues of timing.The surgical excisions performed in burns can roughly be divided in escharotomy and surgical debridement. Escharotomy is a surgical procedure used to treat full-thickness, circumferential burns22. In those burns, the combination of a leathery-like, inelastic burned skin and the swelling of underlying tissue can obstruct circulation and cause cell death in healthy body parts due to ischaemia (e.g. a circumferential full-thickness burn around the forearm can cause necrosis of the hand with even amputation as a result). In order to prevent cell death and amputation, incisions (escharotomies) are made through the eschar until blood flow to the body part at risk is restored8. Furthermore, escharotomies to the chest are used to restore respiration and are indicated when burned skin prevents expansion of the chest necessary for respiration22. Surgical debridement on the other hand, is the excision with removal of all non-vital tissue of the burn wound and can be divided in two main approaches, namely: fascial and tangential excision23. Fascial excision (removal of all layers of eschar and underlying tissue and fat to the level of facis), or avulsion, is only practised in full-thickness burns and consists of tearing away the burned skin and underlying subcutaneous tissue. This technique will cause body deformities but reduces blood loss compared to tangential excision. Tangential excision is the most commonly used technique in burn surgery and often combined with grafting. With this technique all non-viable tissue is removed layer by layer, preserving as much viable tissue as possible8. Nowadays, this procedure can be even more precise with the development of hydro surgery which is especially used in critical functional and aesthetic areas23.Grafting is the transplantation of skin onto an excised burn wound in order to provide temporary or permanent wound closure. The timing of grafting can be immediate (i.e. in the same procedure as the initial excision) or delayed (i.e. in a subsequent procedure) and the skin transplant can originate from the same individual (autologous), another individual of the same species (allogeneic or homologous) or from a different species (xenogeneic or heterologous). There are two different kind of grafts, namely: full-thickness graft (FTG) and split skin graft (SSG). A FTG consists of the epidermis and all of the dermis, and is used more

R1R2R3R4R5R6R7R8R9


144 | Chapter 8

in reconstructive surgery than primary surgery. A SSG consists of the epidermis and some of the dermis, and is often used in primary surgery. A SSG can be applied as a meshed, unmeshed24 or Meek-Wall graft25. An untreated SSG is unmeshed, but when small incisions are made in the SSG, it can be expanded up to six times its original size and is called a meshed graft26. Another technique to increase the surface of a graft even more is the Meek-Wall technique. With this technique the SSG is divided into little squares, or skin islands, and transplanted on an excised burn wound25.Timing of excision can be divided into early and delayed or late excision, but definitions of these concepts are under debate. Choi et al. (2008) state that “all excision procedures done before the natural separation of the eschar are considered an early excision, and all done afterward are considered delayed excision”27, while Kirn et al. (1998) defined early excision as operative excision within seven days post-burn28. Herndon et al. (2007) advocate early excision and grafting in major burns without explicitly defining ‘early’8, and in the review by Ong et al. (2006) ‘early’ ranges from < 24 hours to < 144 hours post-burn29. In conclusion, there is no consensus on a cut-off point between early and delayed excision.

How the intervention might workEarly excision and immediate or delayed grafting for burns provides an early wound closure or coverage which might have beneficial effects on infection, survival and scarring. In particular, a reduced infection rate seems probable because early wound closure limits the time a wound is exposed to possible invading pathogens colonising the wound. Colonised burn wounds have been associated with delayed wound healing, increased need for surgical interventions and prolonged length of stay at burn centres18. The effects on survival might be a result of early excision of eschar. Removal of eschar is thought to decrease the release of inflammatory mediators which could cause systemic effects19,30. Systemic effects, like systemic inflammatory response syndrome (SIRS), sepsis and multi-organ failure, compromise wound healing and can cause death. Effects of early excision and grafting on scarring could be beneficial or harmful, depending on the depth of the burn wound. In full-thickness and deep partial-thickness burns there will be scarring regardless of intervention, but early excision and grafting could provide early wound closure and could reduce the severity of scarring11. Superficial partial-thickness burns, on the other hand, could heal without scarring when conservative treatment is applied, while surgical treatment will always leave scars. Often the depth of a burn is not either superficial partial-thickness or deep partial-thickness but a mix of those depths. In these cases, early excision and grafting could be overzealously executed and cause scarring in areas that would have been able to reepithelialise without surgical intervention. Conservative treatment could give the superficial partial-thickness areas time to reepithelialise with little or no scarring, but postponing excision of the deep partial-thickness areas too long will cause more extensive scars11.



Why it is important to do this reviewSurgical procedures are an important intervention in burns, but there is no consensus on the best timing of surgical intervention. One review states that “early excision of burns reduces mortality in patients without inhalational injury, increases blood transfusion requirements and reduces the length of hospital stay in patients” 29, but this review is non-systematic and not up-to-date. Several systematic reviews have been published in the field of wound care, but most of them focus on wound dressings31 and other topical treatments18,32 instead of surgical procedures. Despite these reviews, guidelines to support clinical decision-making in burn care are predominantly practice-based or are concerned with the general treatment of burns. For example, an evidence-based guideline was published, concerning the treatment of burns and scalds in primary care33. Systematic reviews are necessary to increase the body of underlying evidence for these guidelines. In conclusion, published reviews do not address the effectiveness of early excision and grafting in burns.

ObjectivesTo assess the effects of early excision and grafting on scar quality in people with burns of all depths.

Methods

Criteria for considering studies for this reviewTypes of studies: We considered all randomised controlled trials (RCTs), both published and unpublished, that evaluated the effects of early excision and grafting in people with burns. We decided to consider quasi-randomised controlled trials only in the absence of RCTs.

Types of participants: We considered studies that included people of any age with burns of any degree in any care setting. Any type of burn injury was eligible (flame, scald, chemical, etc.).

Types of interventions: Studies were considered for inclusion if early excision and grafting was applied and compared with any comparator intervention. Whilst there was no clear consensus on the definition of ‘early’ we defined early excision as excision within a week post-burn. We considered any kind of graft. Comparator interventions could include any other intervention (e.g. different excision technique or grafts, ointments), no intervention or a placebo intervention on condition that timing of excision was different compared to the intervention. The timing or type of grafting did not form part of the selection process.

R1R2R3R4R5R6R7R8R9


146 | Chapter 8

Types of outcome measures: Study outcome did not form part of the selection process. We divided outcomes into primary and secondary outcomes; these are listed below.Primary outcomes; three outcome measures were considered primary outcomes. These primary outcomes consisted of one positive and two negative outcomes. The positive primary outcome was the following:

• scar quality: observed and self reported (any definition of scar quality was accepted).The negative primary outcomes were the following:

• wound infection (any definition);• mortality.

Secondary outcomes; eight outcome measures were considered secondary outcomes. These outcomes were the following:

• proportion of burns requiring reconstructive surgery;• pain;• time to complete wound healing;• length of hospital stay;• adverse effects (e.g. need for re-grafting, reduced graft take, blood loss, complications

of surgery and surgery for wound closure in the conservative group);• patient satisfaction;• quality of life;• costs of treatment.

Search methods for identification of studiesElectronic searches: In May 2015 we searched the following electronic databases for reports of randomised controlled trials:

• Cochrane Wounds Group Specialised Register (searched 13 May 2015);• the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane

Library 2015, Issue 4);• the Database of Abstracts of Reviews of Effects (DARE) (The Cochrane Library

2014,Issue 3);• the NHS Economic Evaluation Database (The Cochrane Library 2014, Issue 8);• Ovid MEDLINE (1946 to 12 May 2015);• Ovid MEDLINE (In-process & Other Non-Indexed Citations 12 May, 2015);• Ovid EMBASE (1974 to 12 May 2015);• EBSCO CINAHL (1982 to 13 May 2015).

The search strategies for CENTRAL can be found in the published protocol of this review34. This search strategy was modified as appropriate for other databases. We combined the MEDLINE search with the Cochrane Highly Sensitive Search Strategy for identifying randomised trials in MEDLINE; sensitivity- and precision-maximising version (2008 revision);



Ovid format. This filter is published in the Cochrane Handbook for Systematic Reviews of Interventions (‘Cochrane Handbook’)35. The EMBASE and CINAHL searches were combined with the trial filters developed by the Scottish Intercollegiate Guidelines Network (SIGN)36. No date or language restrictions were applied.In addition we searched the International Clinical Trials Registry Platform Search Portal (http://apps.who.int/trialsearch/; searched 6 August 2013).

Searching other resources: We checked citation lists within all reports of included studies and major review articles in an effort to identify any additional relevant studies. We sent emails to all authors of included studies requesting information on unpublished data and ongoing studies.

Data collection and analysisSelection of studies: Without restrictions on language of publication or publication status, two review authors (CH and JH) independently assessed the titles and abstracts of studies identified from the search in terms of their relevance and design. We obtained full versions of articles if they matched the inclusion criteria from this initial assessment, or where this could not be determined. The same two review authors independently assessed full text articles and determined a final selection of trials eligible for this review. Another review author (MvB) evaluated any discrepancies and advised in case of disagreement.

Data extraction and management: Two review authors (CH and JH), working independently, extracted and summarised details of trials using a data extraction sheet. They extracted data on the following items:

• characteristics of the trial: method of randomisation, setting, location of care, country, source of funding;

• participants baseline information: number, age, gender, type of burn, percentage total body surface area (TBSA) burned, burn depth, inhalation injury, concurrent illness;

• intervention: time elapsed before treatment, percentage TBSA grafted, types of excision, types of grafting, concurrent interventions;

• comparator intervention (see above);• outcomes: types of outcomes measured, timing of outcomes;• results.

The authors resolved any discrepancies by discussion with a third review author (MvB), and contacted the trial authors when information was missing from published reports or clarification was needed. Data from trials published in duplicate were included only once, but were maximally data extracted.

R1R2R3R4R5R6R7R8R9


148 | Chapter 8

Assessment of risk of bias in included studies: Two review authors (CH and JH) made systematic and independent assessments of the risk of bias of each trial using the Cochrane ‘Risk of bias’ criteria37. The criteria relate to the following issues:

• sequence generation;• allocation concealment;• blinding of participants and care providers;• blinding of outcome assessors;• incomplete outcome data: assessment drop-out rate and intention-to-treat analysis;• selective outcome reporting;• other sources of bias: baseline similarity, co-interventions, compliance, similar timing

of outcome assessment and trial sponsorship.Risk of bias increases with each criterion that is judged to be negative. A detailed description of criteria for a judgement of ‘low risk’, ‘high risk’ or ‘unclear risk’ of bias can be found in the published protocol of this review34. Any discrepancies in judgement between the two review authors was resolved by discussion with a third review author (MvB). Final assessment of risk of bias was presented in a risk of bias summary figure, which presents all of the judgements in a cross-tabulation of study by entry. A plus (+), minus (-) or question mark (?) were used to indicate low, high or unclear risk of bias, respectively. This display of internal validity indicates the weight the reader may give to the results of each study.

Measures of treatment effect: Data analysis was performed according to the guidelines of the Cochrane Collaboration38. One review author (CH) entered quantitative data into RevMan, this was checked by another review author (JH) and analysed using the Cochrane Collaboration’s associated software (RevMan)39. For each outcome, summary estimates of treatment effect (with 95% confidence intervals (CI)) were calculated for every comparison. Dichotomous outcomes were presented as risk ratios (RR) (also called relative risks) (see Cochrane Handbook 9.2.238) with 95% CI, and continuous outcomes were presented as mean differences (MD) with 95% CI. We intended to use standardised mean differences (SMD) on occasions when studies assessed the same outcome (e.g. quality of life) but measured the outcome in different ways. Time to wound healing would be analysed as a survival (time-to-event) outcome if possible, using an appropriate analytical method (i.e. hazard ratio, Cochrane Handbook 9.2.638).

Unit of analysis issues: We addressed the level at which randomisation occurred in our analysis. In general, the unit of randomisation and measurement was likely to be the person. Any deviations were described and addressed in the analysis.



Dealing with missing data: We contacted the original investigators to request missing data whenever possible.

Assessment of heterogeneity: We planned to explore both clinical and statistical heterogeneity. Clinical heterogeneity was assessed using information on type of early excision, timing of early excision and body part burned. We planned to test statistical heterogeneity using the Chi2 test and estimate the amount of heterogeneity using the I2 statistic with 95% CI38,40, which examines the percentage of total variation across studies due to heterogeneity rather than to chance.

Assessment of reporting biases: We planned to measure publication bias by the Begg funnel plot41 and the Egger test42, if the included studies were homogeneous and sufficient in number.

Data synthesis: We planned to perform a meta-analysis for each primary outcome if clinical and statistical homogeneity indicated this would be appropriate40, and calculate summary estimates of treatment effect for every comparison. We planned to conduct a narrative overview, structured by the type of comparison, when statistical meta-analyses was inappropriate. We focused on direct comparisons between surgical interventions. No totals were calculated if trial heterogeneity was considerable (I2 statistic greater than 75%). If pooling was appropriate (I2 statistic less than 75%) we used both a fixed-effect (I2 statistic less than 50%) or a random-effects model (I2 statistic greater than 50%). The fixed-effect model ignores heterogeneity and gives an estimate of the intervention effect, assuming a single intervention effect. A random-effects model incorporates heterogeneity amongst studies38,43.

Subgroup analysis and investigation of heterogeneity: We planned to investigate heterogeneity through subgroup and sensitivity analyses38, when there was a sufficient number of studies in the meta-analysis (i.e. more than 10). We planned to conduct subgroup analysis for the following groups:

• % TBSA burned (<20% TBSA versus 20% or more TBSA);• age (<5 years versus 5 to 60 years versus 60 years and older).

Sensitivity analysis: If there was a sufficient number of studies in the meta-analyses, we planned to perform a sensitivity analysis showing how conclusions might be affected if studies at high risk of bias for sequence generation and allocation concealment were excluded from the analyses.

R1R2R3R4R5R6R7R8R9


150 | Chapter 8

Results

Description of studiesResults of the search: The search identified, after initial de-duplication, 606 articles. Two review authors (CH and JH) independently assessed the titles and abstracts of these articles and judged 28 citations to be potentially eligible for the review. Six citations appeared to be duplicates, decreasing the number of unique articles to 22. Full texts of the 22 eligible articles were obtained and assessed by the same two review authors, which resulted in the inclusion of 10 articles. They completed data extraction forms and the risk of bias table, and screened the references in the articles for additional eligible studies. No additional studies were identified with this “snowballing” method. An additional search in the International Clinical Trials Registry Platform Search Portal (http://apps.who.int/trialsearch/), the ClinicalTrials.gov (http://www.clinicaltrials.gov/) and the EU Clinical Trials Register (https://www.clinicaltrialsregister.eu/) did not result in additional, potentially eligible trials.

Included studies: Assessment of the 22 potentially eligible articles (28 citations) resulted in the inclusion of 10 studies (14 citations)44-57. The characteristics of these studies are described in Table 1 and are summarised below.

Health care settingsEight RCTs took place in burn centres; six in the USA44-46,54,55,57; one in England47 and one in Egypt53. One RCT took place in a Division of Plastic and Reconstructive Surgery in Lithuania48-52 and one in a General Surgery Department in India56. Four of the RCTs conducted in the USA were performed by the same research group44,46,54,57.

ParticipantsA total of 416 participants (224 wounds in intervention group, 226 wounds in control group) were recruited to the 10 included studies (range of sample size 13 to 85). It is possible that this number might be lower due to a possible overlap of participants between three studies46,54,57 which potentially would decrease the total to 366. Nine studies randomised the participants, whereas one study randomised 20 hands of 16 participants56. Two studies randomised the participants and, when applicable, analysed both hands in the allocated group48-53. Age and percentage TBSA burned of the included participants are summarised below. The mean age and standard error (SE) of the participants in the only paediatric study44 was 1.8 years (SE 0.4 years) in the intervention group and 1.9 years (SE 0.5 years) in the control group. The other nine studies included only adults48-52,54, participants between 17 and 55 years46, children and adults45,47,55,56 or were unclear whether children were excluded53,57. The mean age and standard deviation (SD) of the participants in those nine studies varied from 15.4 years (SD



14.6 years) to 46.7 years (SD 2.8 years). Mean percentage TBSA burned was reported in all 10 included studies and varied from 5.5% (SD 1.1%) in the intervention group in Engrav et al.(1983)45 to 66.7% (SEM 6.1%) in the intervention group in Rutan et al.(1986)54.

InterventionsAll studies compared early excision and grafting within a week post-burn with a kind of conservative treatment. Two studies only stated that conservative treatment consisted of delayed excision and grafting44 or grafting alone47. In four studies, conservative treatment consisted of daily application of silversulfadiazine45,54,55,57 and two studies performed delayed excision and grafting if no re-epithelialisation occurred after application of silversulfadiazine in 14 days48-52 or 21-24 days post-burn46. One study performed delayed excision and grafting when spontaneous separation of eschar occurred after vigorous irrigation by saline and application of antimicrobial ointments in the form of betadine or nitrofurazone53. In one study, conservative treatment consisted of application of honey dressings on alternate days56.

Primary outcomesThree studies included scar quality as an outcome of interest, but differed in their measurement. In Engrav et al.(1983), scar quality was presented as the proportion of scars that had blisters, abnormal contour, surface irregularity, hypertrophy or loss of motion45, whereas Maslauskas et al. (2005) used the Vancouver Scar Scale and a patient-rated cosmetic appearance of the scar on a four point scale (lower scores represent better scar quality for both scales)48-52. The third study56 used cosmetic wound appearance three months post discharge as scar quality measurement. Wound infection was reported in five studies and determined as clinically significant wound infection or sepsis44, number of days septic (4 out of 7 pre-specified criteria)46, positive wound swabs or blood cultures47 or positive wound swabs from wounds in which infection was suspected clinically48-52,56. Four studies reported mortality46,47,56,57.

Secondary outcomesSecondary outcomes reported in the ten studies included the proportion of burns requiring reconstructive surgery, time to complete wound healing, length of hospital stay, adverse effects and costs of treatment. One study reported the proportion of burns requiring reconstructive surgery45 and one study included wound healing as an outcome of interest, measuring it as time in days until 5% and 2% of whole skin loss was left47. Six studies44-46,53,56,57 reported length of hospital stay and one study reported costs of treatment45. Adverse effects were reported in six studies, four studies reported the proportion of burns requiring surgery for wound closure44,45,55,56, one study reported proportion of wounds with partial graft loss and need for re-grafting53 and one study reported percentage graft take five days post-operative56. Furthermore, one study reported complications of surgery45 and two studies reported blood

R1R2R3R4R5R6R7R8R9


152 | Chapter 8

loss44,46. None of the studies reported pain, patient satisfaction or quality of life as outcome measure.Sponsorship. None of the included studies provide information about sponsorship. Considering the nature of the interventions, the review authors made a judgement that sponsorship was unlikely.

Excluded studies: Ten studies (12 citations) were excluded for different reasons; four studies because they were not RCTs58-63, three studies because excision and grafting was not performed within a week post-burn64-66, one study because excision and grafting was not an intervention67 and two studies because there was no difference in timing of excision68,69.

Risk of bias in included studiesTwo review authors (CH and JH) independently assessed risk of bias in the ten included studies and initially disagreed on 14 judgements. All disagreements were resolved by discussion. Details of the risk of bias judgements for the ten studies are presented in a summary figure (Figure 1) and are described below.

Allocation (selection bias): For risk of bias assessment the term “allocation” included sequence generation and allocation concealment, which both had to be considered and are summarised below.

Sequence generationOf the ten included studies, three studies45,46,53 described the method of sequence generation adequately, that is by pulling cards45, using random number charts46 or random allocation software53. Additionally, one study56 described the method of sequence generation in personal communication, that is by chit method (chits representing the control or intervention group are drawn for a box generating a random allocation sequence). The other six studies stated only that participants were randomised, but did not describe the method of sequence generation44,47-52,54,55,57.

Allocation concealmentOf the ten included studies, two studies45,53 described the method of allocation concealment adequately, that is by pulling cards45 or using random allocation software53. Additionally, one study56 described the method of allocation concealment in personal communication, that is by chit method. The other seven studies did not describe the method of allocation concealment44,46-52,54,55,57.



Figure 1: Risk of bias summary: review authors' judgements about each risk of bias item for each

included study. A plus (+), minus (-) or question mark (?) represents a judgement of low, high or unclear

risk of bias, respectively.

Figure 1: Risk of bias summary: review authors’ judgements about each risk of bias item for each included study. A plus (+), minus (-) or question mark (?) represents a judgement of low, high or unclear risk of bias, respectively.

Blinding (performance bias and detection bias): Review authors had to judge the blinding of participants, care providers and outcome assessors. None of the ten studies reported blinding of participants or care providers, but the nature of treatments made it impossible to blind them. Nevertheless, reviewers made a judgement of low risk of bias because the outcomes were not likely to be influenced by the lack of blinding of participants and care providers. None of the ten studies reported blinding of outcome assessors. Nevertheless, reviewers made a judgement of low risk of bias for five studies44,46,54,55,57 because the outcomes were not likely to be influenced by the lack of blinding of outcome assessors. In contrast, the reviewers made a judgement of unclear risk of bias for three studies47-53 because some of the outcomes in

R1R2R3R4R5R6R7R8R9


154 | Chapter 8

these studies could have been influenced by the lack of blinding of outcome assessors. After personal communication with Engrav et al.(1983)45 and Subrahmanyam et al.(1999)56, the reviewers made a judgement of high risk of bias and low risk of bias, respectively. Engrav et al.(1983) performed no blinding which could have influenced outcome assessment45, whereas Subrahmanyam et al.(1999) described blinding as: “The assessor was the doctor mentioned in the acknowledgements and not aware of the alootment”56.

Incomplete outcome data (attrition bias): The item “incomplete outcome data” consisted of two topics: drop-out rate and intention-to-treat (ITT). The drop-out rate was described and acceptable (i.e. did not exceed 20% for short-term follow-up and 30% for long-term follow-up and does not lead to substantial bias) in seven studies44,47-54,56,57, whereas one study46 was judged “unclear” because there were discrepancies between the numbers in the tables and the numbers in the text. Drop-out rates in two studies45,55 were not acceptable and therefore judged to have high risk of bias. In Engrav et al.(1983), the drop-out rate exceeded 30% for long-term follow-up in the control group without description of reasons for drop-out45. In Salisbury et al.(1982), 28 participants started the study but only 16 completed the study without describing the reasons for drop-out55. None of the ten studies reported ITT-analysis explicitly, but it appeared likely for eight studies44,45,47-54,56,57, whereas one study46 was judged “unclear” because there were discrepancies between the numbers in the tables and the numbers in the text. Salisbury et al. (1982) clearly did not undertake ITT-analysis, as the control group was divided in a group that needed delayed excision and grafting and a group that healed with conservative treatment55. These groups were analysed separately while this was not pre-specified in the methods section, therefore, the reviewers made a judgement of high risk of bias.

Selective reporting (reporting bias): All ten studies were classified as free of suggestion of selective outcome reporting, including one study46 that did not report length of hospital stay and operating room procedures for a subgroup with inhalation injury. Because the high mortality in that subgroup prevented meaningful analyses for these outcomes in survivors, the reviewers made the judgement of low risk of bias.

Other potential sources of bias: Review authors considered four other potential sources of bias, that is, baseline characteristics, co-interventions, compliance and timing of outcome assessment. The baseline characteristics between intervention and control group were similar in seven studies44,45,47-53,55,56. In Herndon et al. (1989), the distribution of gender and etiology at baseline were not similar but unlikely to have influenced outcomes46. In Rutan et al. (1986), age and percentage TBSA burned were not similar between groups and could have influenced outcomes54. The baseline characteristics in Thompson et al (1987) were not similar



with regard to age, which could have influenced outcomes57. The co-interventions between intervention and control groups were similar in eight studies44-46,53-57, whereas no information was provided about co-interventions in two studies47-52. Compliance was acceptable in all ten studies and timing of outcome assessment between intervention and control group was similar in eight studies44-52,54,55,57. In Omar et al. (2011), follow-up was at two weeks and two months postoperatively rather than post-burn, which was on average 11 days later in the control group53. Because the outcomes of interest were related to the operation, this difference in timing of outcome assessment was judged “low risk of bias”. In contrast, timing of outcome assessment in Subrahmanyam et al. (1999) was judged “unclear” because the mean difference in timing between both groups was 25 days56. This difference in timing could have influenced outcome assessment, especially with regard to the outcome scar quality. The four potential sources of bias resulted in an overall judgement of “low risk of bias” for seven studies44-53,55, “high risk of bias” for two studies54,57; and one study56 was judged “unclear”.

Effects of interventionsHeterogeneity of studies with regard to interventions and outcomes prevented assessment of reporting biases and limited data synthesis to a narrative overview, structured by the type of comparison. Type of comparisons were delayed excision and grafting (comparison 1), application of antimicrobials and delayed excision and grafting if necessary (comparison 2), and honey dressings (comparison 3). The effects of interventions are presented in Table 1 and summarised below.

R1R2R3R4R5R6R7R8R9


156 | Chapter 8

18

Table 1. Summary of study characteristics of included studies

Study ID Inclusion criteria and main baseline characteristics

Interventions Outcomes

Desai et al. (1991) [44]

Scald injuries (which were not caused by grease) of clinically indeterminant depth. Mean Age (SE): I: 1.8(0.4); C: 1.9(0.5). %TBSA (Mean (SE)): I: 28(3); C: 24(3)

I (n = 12): Early excision and grafting (within 72 hours of admission). C (n = 12): Late excision and grafting (after at least two weeks of serial dressing changes and daily bathing).

Primary outcome Wound infection: I: 0; C: 0. Secondary outcomes Mean length of hospital stay (SE): I: 17 (2); C: 21 (3). Adverse effects: Proportion of burns requiring surgery: I: 12/12; C: 6/12. Adverse effects: Blood loss (Mean total body blood turnover(SE)): I: 1.2(0.3); C: 0.3(0.1).

Engrav et al. (1983) [45]

Indeterminant flame or scalds burns less than 20% TBSA burned. Mean Age (SD; range): I: 23.6 (3.4; 13 months to 56 years); C: 30.6 (4.3; 19 months to 78 years). %TBSA (Mean (SD)): I: 5.5(1.1); C: 7.3(0.9).

I (n=22): Early excision and grafting (within 7 days postburn). C (n = 25): Nonoperative treatment, twice daily hydrotherapy, debridement, and application of silversulfadiazine cream.

Primary outcomes Scar quality: Proportion hypertrophy: I: 3/17; C: 7/17 Proportion abnormal contour: I: 0/17; C: 2/17. Proportion surface irregularity: I: 8/17; C: 1/17. Proportion loss of motion (contracture): I: 0/17; C: 1/17. Proportion blisters: I: 2/17; C: 1/17. Secondary outcomes Number of patients requiring reconstructive surgery: I: 1/17; C:1/17. Mean length of hospital stay (SD): I: 16.4 (1.2); C: 25.0 (1.8); p<0.05. Adverse effects: Proportion of burns requiring surgery: I: 22/22; C: 12/25. Adverse effects: Complications of surgery: I: 1/22; C: 1/12. Mean total hospital costs in dollars (including physicians charges) (SD): I: 9,063 (1,144); C: 12,702 (1,270)..

Herndon et al. (1989) [46]

Burns >30% TBSA second-degree and >20% TBSA third-degree burns, admitted within 3 days postburn. Mean Age (SD) (calculated from 2 subgroups): I: 29 (11); C: 34 (11). %TBSA (Mean (SD)(calculated from 2 subgroups)): I: 60 (17); C: 63 (18).

I (n = 45): Early excision (within 72 hours of admission) and immediate wound coverage with meshed autografts 4:1. C (n = 40): Conservative wound-management with silver sulfadiazine and/or sulfamylon acetate antimicrobial cream; wounds were sharply debrided 21 to 24 days postburn and gradually grafted

Primary outcomes Wound infection: Mean (SD) days septic in subgroup without inhalation injury: I: 1.8 (2.5); C: 1.7 (1.7); Subgroup with inhalation injury: not reported. Mortality: I: 18; C: 28 Secondary outcomes Mean length of stay in days (SD): Subgroup without inhalation injury: I: 53 (38); C: 47 (25); Subgroup with inhalation injury: not reported. Adverse effects: Blood loss as mean (SD) total body blood turnover; calculated at 80 ml/kg body weight: Subgroup without inhalation injury: : I: 2.1 (2.2); C: 0.7 (0.6); Subgroup with inhalation injury: not reported.



19

Table 1 (continued)



Jackson et al. (1960) [47]

Burn patients Mean Age (SD; range): I: 18.4 (14.9; 2-52); C: 15.4 (14.6; 1-50) %TBSA (Mean (SD)): I: 35.4 (17.6); C: 35.6 (19.1)

I (n = 16): Early excision (on day of burning). C (n = 14): Delayed grafting (2-3 weeks).

Primary outcomes Wound infection: Not clearly reported, unable to report meaningful proportions due to mixed numbers of different studies. Mortality (n): I: 5 (16); C: 3 (14). Secondary outcome Time to complete wound healing in days: Mean days until 5% / 2% of whole skin loss (w.s.l.) left: Reported for subgroups only. Original data was reported and used for survival analyses.

Maslauskas et al. (2005) [48-52]

Hand burns Mean Age (SD): I: 39.75 (2.8); C: 46.68 (2.77) (p=0.085) %TBSA (Mean (SD)): I: 11.56 (2.57); C:15.32 (2.28) (p=0.23)

I (n = 24 (40 hands)): Early excision/necrotomy (within 7 days postburn) and grafting. C (n = 25 (39 hands)): Conservative treatment with silversulfadiazine until epithelialisation. Delayed necrotomy and grafting if no epithelialisation occurred during a 14 day period

Primary outcomes Scar quality: Mean Vancouver scar scale (SD; n): I: 3.65 (2.93; 38); C: 6.77 (2.96; 37), p<0.001. Mean cosmetic appearance on 4 point scale (SD): I: 1.9 (0.78); C:2.64 (0.84), p<0.001. Wound infection: Positive wound swab: I: 11; C: 33 Secondary outcomes None

Omar et al. (2011) [53]

Deep second and third degree hand burns. Mean Age (SD): I: 23 (6.86); C: 25 (8.3) %TBSA (Mean (SD)): I: 26 (3.47); C: 24 (3.68)

I (n = 20 (25 hands)): Early excision and grafting (before 6th day postburn). C (n = 20 (27 hands)): Wounds were subjected to vigorous irrigation by saline and application of antimicrobial ointments in the form of betadine or nitrofurazone. Dressing was carried out until spontaneous separation of eschar. Delayed excision and grafting.

Primary outcomes None Secondary outcome Wound healing: Complete graft take without re-grafting: I: 20 hands; C: 21 hands. Mean length of hospital stay (SD): I: 16 (2.5); C: 24 (3.4); p<0.05.

Rutan et al. (1986) [54]

Burns > 45% TBSA Mean Age (SEM): I: 25.7 (2); C: 37.5 (6.4) %TBSA (Mean (SEM)): I: 66.7 (6.1); C: 55.2 (2.5)

I (n = 7): Excision and grafting (within 72 hours of injury). C (n = 6): Conservative wound-management with daily hydrotherapy and twice a day applications of silver sulfadiazine and/or mafenide acetate (antimicrobial cream).

Primary outcomes None Secondary outcomes None

18

Table 1. Summary of study characteristics of included studies



Desai et al. (1991) [44]

Scald injuries (which were not caused by grease) of clinically indeterminant depth. Mean Age (SE): I: 1.8(0.4); C: 1.9(0.5). %TBSA (Mean (SE)): I: 28(3); C: 24(3)

I (n = 12): Early excision and grafting (within 72 hours of admission). C (n = 12): Late excision and grafting (after at least two weeks of serial dressing changes and daily bathing).

Primary outcome Wound infection: I: 0; C: 0. Secondary outcomes Mean length of hospital stay (SE): I: 17 (2); C: 21 (3). Adverse effects: Proportion of burns requiring surgery: I: 12/12; C: 6/12. Adverse effects: Blood loss (Mean total body blood turnover(SE)): I: 1.2(0.3); C: 0.3(0.1).

Engrav et al. (1983) [45]

Indeterminant flame or scalds burns less than 20% TBSA burned. Mean Age (SD; range): I: 23.6 (3.4; 13 months to 56 years); C: 30.6 (4.3; 19 months to 78 years). %TBSA (Mean (SD)): I: 5.5(1.1); C: 7.3(0.9).

I (n=22): Early excision and grafting (within 7 days postburn). C (n = 25): Nonoperative treatment, twice daily hydrotherapy, debridement, and application of silversulfadiazine cream.

Primary outcomes Scar quality: Proportion hypertrophy: I: 3/17; C: 7/17 Proportion abnormal contour: I: 0/17; C: 2/17. Proportion surface irregularity: I: 8/17; C: 1/17. Proportion loss of motion (contracture): I: 0/17; C: 1/17. Proportion blisters: I: 2/17; C: 1/17. Secondary outcomes Number of patients requiring reconstructive surgery: I: 1/17; C:1/17. Mean length of hospital stay (SD): I: 16.4 (1.2); C: 25.0 (1.8); p<0.05. Adverse effects: Proportion of burns requiring surgery: I: 22/22; C: 12/25. Adverse effects: Complications of surgery: I: 1/22; C: 1/12. Mean total hospital costs in dollars (including physicians charges) (SD): I: 9,063 (1,144); C: 12,702 (1,270)..

Herndon et al. (1989) [46]

Burns >30% TBSA second-degree and >20% TBSA third-degree burns, admitted within 3 days postburn. Mean Age (SD) (calculated from 2 subgroups): I: 29 (11); C: 34 (11). %TBSA (Mean (SD)(calculated from 2 subgroups)): I: 60 (17); C: 63 (18).

I (n = 45): Early excision (within 72 hours of admission) and immediate wound coverage with meshed autografts 4:1. C (n = 40): Conservative wound-management with silver sulfadiazine and/or sulfamylon acetate antimicrobial cream; wounds were sharply debrided 21 to 24 days postburn and gradually grafted

Primary outcomes Wound infection: Mean (SD) days septic in subgroup without inhalation injury: I: 1.8 (2.5); C: 1.7 (1.7); Subgroup with inhalation injury: not reported. Mortality: I: 18; C: 28 Secondary outcomes Mean length of stay in days (SD): Subgroup without inhalation injury: I: 53 (38); C: 47 (25); Subgroup with inhalation injury: not reported. Adverse effects: Blood loss as mean (SD) total body blood turnover; calculated at 80 ml/kg body weight: Subgroup without inhalation injury: : I: 2.1 (2.2); C: 0.7 (0.6); Subgroup with inhalation injury: not reported.

R1R2R3R4R5R6R7R8R9


158 | Chapter 8

Comparison 1: early excision and grafting compared with delayed excision and grafting (2 RCTs, 54 participants)Two studies compared early excision and grafting with late excision and grafting after serial dressing changes for at least two weeks44 or with delayed grafting two to three weeks post-burn47. Both studies did not describe the method of sequence generation an allocation concealment (selection bias) adequately and therefore the reviewers made a judgement of

20

Table 1 (continued) Study ID Inclusion criteria

and main baseline characteristics


Salisbury et al. (1982) [55]

Hand burns Mean Age (Range): I: 23.8 (7-40); C(healed): 19.3 (13-30); C(operated): 23.6 (7-53) %TBSA (Mean (Range)): I: 34.8 (26-58); C(healed): 33.3 (26-39); C(operated): 22.4 (7-38)

I (n = 8): Early excision and grafting (within 5 days postburn). C (n = 12): Conservative treatment with daily hydrotherapy, eschar debridement, topical application of silversulfadiazine cream and biological dressings (porcine xenograft or cadaver allograft).

Primary outcomes None Secondary outcome Adverse effects: Proportion of burns requiring surgery: I: 8/8; C: 8/12.

Subrahmanyam et al. (1999) [56]

Burns < 30% TBSA burned, haemodynamically stable and between 10 and 40 years of age. Mean Age (SD): I: 23 (1); C: 22 (2) %TBSA (Mean (SD)): I: 23 (4); C: 24 (4)

I (n = 25): Early excision and grafting (before 6th day postburn). C (n = 25): Honey dressings. On alternate days with topically applied unprocessed honey, after the wounds had been washed with normal saline.

Primary outcomes Scar quality: Excellent/good/fair: I: 8/14/2; C:12/10(Excellent or good/fair) Wound infection: Positive bacterial swab when suspected clinically: I:7/71; C:42/123, p<0.05. Mortality: I: 1; C: 3. Secondary outcomes Mean length of hospital stay (likely SD): I: 21 (4); C: 46 (19); p<0.001. Adverse effects: Proportion of burns requiring surgery: I: 25/25; C: 11/22. Adverse effects: Mean (SD) blood replacement units: I: 35 (12); C: 21 (15). % graft take 5 days post-operative (n): I: 100% (19), 95% (5 or 6); C: 100% (2) 40%-84% (9), p<0.05.

Thompson et al. (1987) [57]

Burns >30% TBSA burned. Mean Age (SD)(calculated from 4 subgroups): I: 28 (8); C: 42 (14). %TBSA (Mean (SD)(calculated from 4 subgroups)): I: 59 (10); C: 54 (12).

I (n = 24): Early excision and grafting (within 72 hours of admission; which was max 5 days postburn). C (n = 26): Conservative treatment. Daily tubbings with debridement of the burn wound and application of silver sulfadiazine or sulfamylon acetate prophylactic topical antimicrobial cream three times a day

Primary outcome Mortality: I: 9; C: 17. Secondary outcomes Mean length of hospital stay in days for survivors (n)(calculated from 4 subgroups): I: 46 (15); C: 62 (9). Adverse effects: Blood loss: I: 124.7; C: 52.4

Abbreviations: I = intervention group; C = control group; TBSA = Total Body Surface Area; SE = Standard Error; SD = Standard Deviation; p = P value, SEM = Standard Error of the Mean.



unclear risk for selection bias. The mean percentage TBSA burned ranged from 24% (SE3) in the control group in Desai et al. (1991) to 35.6% (SD19.1) in the control group in Jackson et al. (1960).

Primary outcomesScar quality: Scar quality was not reported in these studies.Wound infection: In Desai et al. (1991) none of the 24 participants showed clinically significant signs of wound infection or sepsis44. Jackson et al. (1960) reported wound infections in a RCT, a pilot trial and an experimental trial together, preventing data extraction for the RCT only47.Mortality: Jackson et al. (1960) reported non-significant higher proportions of deaths in the early excision group (5 out of 16) compared to the delayed grafting group (3 out of 14); risk ratio (RR) 1.46 (95% CI 0.42 to 5.03; P = 0.55)47.

Secondary outcomesTime to complete wound healing: In Jackson et al. (1960), time to complete wound healing was reported as mean days until 5% and 2% of whole skin loss was left (i.e. 95% and 98% of full thickness area healed)47. Jackson et al. (1960) analysed time to complete wound healing (a time to event outcome) as a continuous variable, which is inappropriate and potentially misleading (since it cannot take account of people who did not heal). However, Jackson et al. (1960) reported original data allowing re-analyses with survival analyses (Kaplan-Meier). Median days until 95% of the wound was healed was 34 days (interquartile range, 23 to 61) for the early excision group (n = 16) and 41 days (interquartile range, 27 to 48; log-rank test, P = 0.77) for the delayed excision group (n = 14). Median days until 98% of the wound was healed was 57 days (interquartile range, 23 to 81) for the early excision group and 52 days (interquartile range, 27 to 67; log-rank test, P = 0.51) for the delayed excision group47.Length of hospital stay: Desai et al. (1991) reported mean length of hospital stay with standard error; SDs were calculated for our analysis. The mean length of hospital stay was 17 days (SD 6.9) in the early excision group and 21 days (SD 10.4) in the delayed excision group (mean difference -4.00; 95% CI -11.07 to 3.07; P = 0.27)44.Adverse effects: Desai et al. (1991) compared the proportion of participants in the intervention and control group that required surgery, which is an adverse effect in the control group. The nature of the intervention prescribed that all participants in the early excision group required surgery (12 out of 12), whereas six out of 12 participants in the delayed excision group required surgery; RR 1.92 (95% CI 1.10 to 3.35; P = 0.02)44. Furthermore, Desai et al. (1991) reported blood loss as a result of excision; the mean total body blood turnovers (TBBT) was significantly higher in the early excision group (1.2; SD 1.0) compared to the delayed excision group (0.3; SD 0.2) (mean difference 0.90; 95% CI 0.28 to 1.52; P = 0.004)44.

20

Table 1 (continued) Study ID Inclusion criteria

and main baseline characteristics


Salisbury et al. (1982) [55]

Hand burns Mean Age (Range): I: 23.8 (7-40); C(healed): 19.3 (13-30); C(operated): 23.6 (7-53) %TBSA (Mean (Range)): I: 34.8 (26-58); C(healed): 33.3 (26-39); C(operated): 22.4 (7-38)

I (n = 8): Early excision and grafting (within 5 days postburn). C (n = 12): Conservative treatment with daily hydrotherapy, eschar debridement, topical application of silversulfadiazine cream and biological dressings (porcine xenograft or cadaver allograft).

Primary outcomes None Secondary outcome Adverse effects: Proportion of burns requiring surgery: I: 8/8; C: 8/12.

Subrahmanyam et al. (1999) [56]

Burns < 30% TBSA burned, haemodynamically stable and between 10 and 40 years of age. Mean Age (SD): I: 23 (1); C: 22 (2) %TBSA (Mean (SD)): I: 23 (4); C: 24 (4)

I (n = 25): Early excision and grafting (before 6th day postburn). C (n = 25): Honey dressings. On alternate days with topically applied unprocessed honey, after the wounds had been washed with normal saline.

Primary outcomes Scar quality: Excellent/good/fair: I: 8/14/2; C:12/10(Excellent or good/fair) Wound infection: Positive bacterial swab when suspected clinically: I:7/71; C:42/123, p<0.05. Mortality: I: 1; C: 3. Secondary outcomes Mean length of hospital stay (likely SD): I: 21 (4); C: 46 (19); p<0.001. Adverse effects: Proportion of burns requiring surgery: I: 25/25; C: 11/22. Adverse effects: Mean (SD) blood replacement units: I: 35 (12); C: 21 (15). % graft take 5 days post-operative (n): I: 100% (19), 95% (5 or 6); C: 100% (2) 40%-84% (9), p<0.05.

Thompson et al. (1987) [57]

Burns >30% TBSA burned. Mean Age (SD)(calculated from 4 subgroups): I: 28 (8); C: 42 (14). %TBSA (Mean (SD)(calculated from 4 subgroups)): I: 59 (10); C: 54 (12).

I (n = 24): Early excision and grafting (within 72 hours of admission; which was max 5 days postburn). C (n = 26): Conservative treatment. Daily tubbings with debridement of the burn wound and application of silver sulfadiazine or sulfamylon acetate prophylactic topical antimicrobial cream three times a day

Primary outcome Mortality: I: 9; C: 17. Secondary outcomes Mean length of hospital stay in days for survivors (n)(calculated from 4 subgroups): I: 46 (15); C: 62 (9). Adverse effects: Blood loss: I: 124.7; C: 52.4

Abbreviations: I = intervention group; C = control group; TBSA = Total Body Surface Area; SE = Standard Error; SD = Standard Deviation; p = P value, SEM = Standard Error of the Mean.

R1R2R3R4R5R6R7R8R9


160 | Chapter 8

Other secondary outcomes: Proportion of burns requiring reconstructive surgery, pain, patient satisfaction, quality of life and costs of treatment were not reported in these studies.

Comparison 2: early excision and grafting compared with antimicrobial agents and delayed excision and grafting if necessary (7 RCTs, 312 participants/ 346 wounds)Seven studies compared early excision and grafting with an antimicrobial agent and delayed excision and grafting if necessary. Six studies45,46,48-52,54,55,57 used silversulfadiazine (SSD) and one study53 used betadine or nitrofurazone in the control group. The reviewers made a judgement of low risk of bias for selection bias for two studies45,53 and one study46 received a judgement of low risk of bias for sequence generation and unclear risk of bias for allocation concealment. The other four studies did not describe the method of sequence generation and allocation concealment (selection bias) adequately and therefore the reviewers made a judgment of unclear risk of bias for selection bias. In two studies45,48 the mean percentage TBSA burned was less than 20% and ranged from 5.5% (SD1.1) in the intervention group in Engrav et al. (1983) tot 15.2 (SD2.28) in the control group in Maslauskas et al. (2005). In the other five studies, the mean percentage TBSA burned was 20% or more and ranged from 24% (SD3.68) in the control group in Omar et al (2011) to 66.7% (SEM 6.1) in the intervention group in Rutan et al. (1986).

Primary outcomesScar quality: Two studies reported scar quality as an outcome. Engrav et al.(1983) reported scar quality as the proportion of participants who had hypertrophy, abnormal scar contour, scar surface irregularity, loss of motion or blisters45. Hypertrophy occurred in three out of 17 participants in the early excision group and seven out of 17 in the control group that received SSD; risk ratio (RR) 0.43 (95% CI 0.13 to 1.39; P = 0.16). Abnormal scar contour occurred in none of the participants in the early excision group and two out of 17 in the control group; RR 0.20 (95% CI 0.01 to 3.88; P = 0.29). Scar surface irregularity occurred in eight out of 17 participants in the early excision group and one out of 17 in the control group; RR 8.00 (95% CI 1.12 to 57.20; P = 0.04). Loss of motion due to contractures occurred in none of the participants in the early excision group and one out of 17 in the control group; RR 0.33 (95% CI 0.01 to 7.65; P = 0.49). Blisters occurred in two out of 17 participants in the early excision group and one out of 17 in the control group; RR 2.00 (95% CI 0.20 to 20.04; P = 0.56)45. Maslauskas et al. (2005) reported statistically significant lower mean Vancouver Scar Scale score in the early excision group (3.65; SD 2.93) compared to the control group that received SSD (6.77; SD 2.96) (mean difference -3.12; 95% CI -4.45 to -1.79; P < 0.0001)48-

52. Furthermore, Maslauskas et al. (2005) reported statistically significant lower mean cosmetic appearance on a 4-point scale (1 represents normal appearance and 4 represents unsatisfactory appearance) in the early excision group (1.9; SD 0.78) compared to the control



group (2.64; SD 0.84) (mean difference -0.74; 95% CI -1.10 to -0.38; P = 0.0001)48-52.Wound infection: Two studies reported wound infection as an outcome. Herndon et al. (1989) reported mean days septic in a subgroup of survivors without inhalation injury, which was 1.8 (SD 2.5) for the 22 participants in the early excision group and 1.7 (SD 1.7) for the 10 participants in the delayed excision group (mean difference 0.10; 95% CI -1.38 to 1.58; P = 0.91)46. Maslauskas et al. (2005) reported statistically significant less positive wound swabs in the early excision group (11 out of 40) compared to the control group that received SSD (33 out of 39); RR 0.33 (95% CI 0.19 to 0.55; P < 0.0001)48-52.Mortality: Two studies reported mortality as an outcome. Herndon et al. (1989) reported significant lower proportions of deaths in the early excision group (18 out of 44) compared to the delayed excision group (28 out of 40); RR 0.58 (95% CI 0.39 to 0.88; P = 0.01)46. Thompson et al. (1987) reported non-significant lower proportions of deaths in the early excision group (9 out of 24) compared to the delayed excision group (17 out of 26); RR 0.57 (95% CI 0.32 to 1.03; P = 0.06)57. Both studies were considered sufficiently similar to pool (in the absence of significant heterogeneity (P = 0.96; I2 = 0%; Figure 2), a fixed effect model was used). Pooling both studies, 27 out of 68 participants in the early excision group and 45 out of 66 in the delayed excision group died; RR 0.58 (95% CI 0.41 to 0.81; P = 0.002; Figure 2).

Figure 2: Forest plot of comparison: 2 Early excision and grafting vs conservative treatment with SSD

and subsequent excision and grafting, outcome: Mortality.

Figure 2: Forest plot of comparison: 2 Early excision and grafting vs conservative treatment with SSD and subsequent excision and grafting, outcome: Mortality.

Secondary outcomesProportion of burns requiring reconstructive surgery: One study45 reported the proportion of burns requiring reconstructive surgery in approximately one year follow-up. One out of 17 participants required reconstructive surgery in both the early excision group and the control group.Length of hospital stay: Four studies45,46,53,57 reported length of hospital stay. Pooling was not possible due to missing data57 and clinical heterogeneity (hand burns in participants with relatively small45 or medium53 mean percentage TBSA burned; or participants with major burns46). Engrav et al. (1983) reported significant shorter mean hospital stay in the early excision group (16.4 days; SD 1.2) compared to the delayed excision group (25.0 days; SD 1.8) (mean difference -8.60; 95% CI -9.47 to -7.73; P <0.001)45. Omar et al. (2011) reported

R1R2R3R4R5R6R7R8R9


162 | Chapter 8

significant lower mean hospital stay in the early excision group (16 days; SD 2.5) compared to the delayed excision group (24 days; SD 3.4) (mean difference -8.00; 95% CI -9.85 to -6.15; P <0.001)53. Herndon et al. (1989) reported the mean length of hospital stay in a subgroup of survivors without inhalation injury only, which was 53 days (SD 38) for the 22 participants in the early excision group and 47 days (SD 25) for the 10 participants in the delayed excision group (mean difference 6.00; 95% CI -16.19 to 28.19; P = 0.65)46. Mean length of hospital stay in Thompson et al. (1987) was reported without variance data, and only for survivors in different groups independently. Summarised mean length of hospital stay for all groups was 46 days for the 15 participants in the early excision group and 62 days for the 9 participants in the delayed excision group57.Adverse effects: Two studies reported proportion of burns requiring surgery for wound closure as an outcome45,55. The nature of the intervention prescribed that all participants in the early excision group required surgery. Both Engrav et al. (1983)45 and Salisbury et al. (1982)55 studied hand burns and were considered sufficiently similar to pool (in the absence of significant heterogeneity (P = 0.23; I2 = 29%; Figure 3), a fixed effect model was used). Pooling both studies, all of the 30 participants in the early excision group and 20 out of 37 in the delayed excision group required surgery; RR 1.82 (95% CI 1.34 to 2.46; P < 0.001; Figure 3). Three studies reported adverse effects related to the operative treatment45,46,57. Engrav et al. (1983) reported complications of surgery, which occurred in both the early excision group (upper airway obstruction; 1 out of 22) and the delayed excision group (donor site infection; 1 out of 12); RR 0.55 (95% CI 0.04 to 7.96; P = 0.66)45. Herndon et al. (1989) reported blood loss (mean total body blood turnovers (TBBT)) as a result of excision in survivors without inhalation injury, which was statistically significant more in the early excision group (2.1; SD 2.2; n = 22) compared to the delayed excision group (0.7; SD 0.6; n = 10) (mean difference 1.40; 95% CI 0.41 to 2.39; P = 0.006)46. Thompson et al. (1987) reported blood loss (mean units of blood infused) without variance data and only for different groups independently. Summarised mean units of blood infused for all groups was 124.7 units in the early excision group and 52.4 units in the delayed excision group57. Omar et al. (2011) reported the proportion of wounds with partial graft loss and need for re-grafting, which occurred in 5 out of 25 hands in the early excision group and 4 out of 27 hands in the delayed excision group; RR 1.35 (95% CI 0.41 to 4.47; P = 0.62)53.Costs of treatment: One study, Engrav et al. (1983) reported mean total hospital cost including physicians charges in United States dollars, which were statistically significant lower in the early excision group (9,063; SD 1,144) compared to the delayed excision group (12,702; SD 1,270) (mean difference -3,639; 95% CI -4,329.19 to -2,948.81; P < 0.001)45.Other secondary outcomes: Time to complete wound healing, pain, patient satisfaction and quality of life were not reported in these studies.



Figure 3: Forest plot of comparison: 2 Early excision and grafting vs conservative treatment with SSD

and subsequent excision and grafting, outcome: Proportion of participants requiring surgery.

Figure 3: Forest plot of comparison: 2 Early excision and grafting vs conservative treatment with SSD and subsequent excision and grafting, outcome: Proportion of participants requiring surgery.

Comparison 3: early excision and grafting compared with honey dressings (1 RCT, 50 participants)One study compared early excision with honey dressings in 50 people with less than 30% TBSA burned56. The reviewers made a judgement of low risk of bias for selection, performance, detection, attrition and reporting bias. The mean percentage TBSA burned was 23$ (SD 4) in the intervention group and 24% (SD3) in the control group.

Primary outcomesScar quality: Subrahmanyam et al. (1999) assessed wound appearance three months post discharge but did not describe the measurement instrument sufficiently56. Scars were analysed as the proportion of wounds that had “excellent or good results”, while “fair result” was the only other category stated. Excellent or good results were reported in 22 out of 24 participants in the early excision group and 12 out of 22 in the control group that received honey dressings; RR 1.68 (95% CI 1.13 to 2.51; P = 0.01). Additional data provided in personal communication conflicted the published data.Wound infection: Wound infection in Subrahmanyam et al. (1999) was determined with positive bacterial swabs when wound infection was suspected clinically. Subrahmanyam et al. (1999) reported statistically significant fewer positive wound swabs in the early excision group (7 out of 71) compared to the control group that received honey dressings (42 out of 123); RR 0.29 (95% CI 0.14 to 0.61; P = 0.001)56.Mortality: Subrahmanyam et al. (1999) reported non-significantly lower proportions of deaths in the early excision group (1 out of 25) compared to the honey treated group (3 out of 25); RR 0.33 (95% CI 0.04 to 2.99; P = 0.33)56.

Secondary outcomesLength of hospital stay: The mean length of hospital stay in Subrahmanyam et al. (1999) was significantly shorter in the early excision group (21 days; SD 4) compared to the honey treated group (46 days; SD 19) (mean difference -25.00 95% CI -33.10 to -16.90; P < 0.001)56.Adverse effects: Subrahmanyam et al. (1999) compared the proportion of participants in the intervention and control group that required surgery56. The nature of the intervention prescribed

R1R2R3R4R5R6R7R8R9


164 | Chapter 8

that all participants in the early excision group required surgery (25 out of 25), whereas 11 out of 22 participants in the honey treated group required surgery; RR 1.96 (95% CI 1.30 to 2.96; P = 0.001). However, graft take five days post-operatively was less successful in the honey treated group. Nineteen participants in the early excision group had 100% graft take, whereas the remainder (five or six participants, unclear if deceased participant is included) had 95% graft take. Only two of the 11 excised participants in the honey treated group had 100% graft take five days post-operative, whereas the remainder nine ranged between 40% to 84% graft take. Subrahmanyam et al. (1999) reported a significantly better (P < 0.05) graft take in favour of the early excision group, however, the method of analysis was unclear and original data were not reported, therefore re-analyses was not possible56. Furthermore, Subrahmanyam et al. (1999) reported blood loss as mean (SD) blood replacement units. The mean blood replacement units was significantly higher in the early excision group (35 units; SD 12) compared to the honey treated group (21 units; SD 15) (mean difference 14.00; 95% CI 6.47 to 21.53; P = 0.0003)56.Other secondary outcomes: Proportion of burns requiring reconstructive surgery, time to complete wound healing, pain, patient satisfaction, quality of life and costs of treatment were not reported in this study.

Discussion

Summary of main resultsWe included ten randomised controlled trials in this review that evaluated early excision and grafting in burns. Studies compared early excision and grafting with either:

• delayed excision and grafting (comparison 1; two studies),• application of antimicrobial agents and delayed excision and grafting if necessary

(comparison 2; seven studies), or• application of honey dressings (comparison 3; one study).

The variety in control interventions and differences in outcome measures between studies made pooling of data inappropriate for most results, therefore, the results have been presented in a narrative overview by comparison. This summary of main results is divided in three primary outcomes (scar quality, wound infection and mortality) and all secondary outcomes have been combined.

Scar qualityOverall, there was insufficient data to support any definite conclusions that early excision and grafting improved scar quality compared to delayed excision. One study showed statistically significant better scar quality in one out of five reported scar outcomes45 and favoured delayed



excision. Two studies showed statistically significant better scar quality for three outcomes measures after early excision48-52,56, although one of those studies56 provided additional data in personal communication that conflicted with the published data and therefore these results should be interpreted with caution. In summary, three outcome measures favoured early excision, one outcome measure favoured delayed excision and four outcome measures showed no statistically significant differences with regard to scar quality.

Wound infectionFive studies addressed wound infection as an outcome but differed in outcome measurement. Two studies reported statistically significant less positive wound swabs in the early excision group48-52,56, whereas two studies found no statistically significant difference in clinical signs of infection between early excision and delayed excision44,46. One study reported wound infections in a RCT, a non-randomised pilot trial and an non-randomised experimental trial together, preventing meaningful data extraction for the RCT only47. A cautious conclusion might be that although early excision reduces the number of positive wound swabs, there appears to be no clinically significant reduction in wound infection.

MortalityFour studies addressed mortality as an outcome. Herndon et al. (1989)46 and Thompson et al. (1987)57 were considered sufficiently similar to pool and analyses showed significant lower proportions of deaths in the early excision group compared to the delayed excision group. These results should be interpreted with caution as these studies might have an overlap in participants. The other two studies reported non-significant higher47 or lower56 proportions of deaths in the early excision group. Overall, there was insufficient data to support any definite conclusions.

Secondary outcomesExcept for pain, patient satisfaction and quality of life, all other secondary outcomes (proportion of burns requiring reconstructive surgery, time to complete wound healing, length of hospital stay, adverse effects and costs of treatment) were addressed by at least one study. No statistically significant differences were found for the proportion of burns requiring reconstructive surgery45 and time to wound healing47. Six studies addressed length of hospital stay, three studies significantly favoured early excision45,53,56, two studies reported no significant difference44,46 and one study favoured early excision but reported mean length of hospital stay without variance data57. A cautious conclusion might be that length of hospital stay is shorter with early excision and grafting compared to conservative treatment.Seven studies addressed one or more adverse effects, that is proportion of participants requiring surgery for wound closure, need for re-grafting, reduced graft take, blood loss and

R1R2R3R4R5R6R7R8R9


166 | Chapter 8

complications of surgery. Four studies addressed the proportion of participants requiring surgery for wound closure44,45,55,56, which can be regarded as an adverse effect in both the intervention and control group. Ideally, burn wounds in the control group heal completely with conservative treatment only, therefore delayed surgery is an adverse effect in the control group. On the other hand, an adequate randomisation procedure presumes the creation of comparable groups and therefore this outcome indicates that a proportion of participants in the early excision group received unnecessary surgery from a wound healing perspective, which can be regarded as an adverse effect in the early excision group. The nature of the intervention prescribed that all participants in the early excision group received surgery. All four studies reported that significantly less participants in the control group required surgery compared to the early excision group. Overall, 52% (37 out of 71) of the participants in the control group required surgery. Consequently, 48% of the participants had wounds that healed without surgery. In addition, Omar et al. (2011) reported no significant difference in the proportion of wounds with partial graft loss and need for re-grafting between early and delayed excision53, whereas Subrahmanyam et al. (1999) reported a significantly better graft take five days post-operative in favour of the early excision group56. However, the method of analysis in Subrahmanyam et al. (1999) was unclear and original data were not reported, therefore re-analyses was not possible. Four studies addressed blood loss as a result of excision as an adverse effect of treatment. Desai et al. (1991), Herndon et al. (1989) and Subrahmanyam et al. (1999) reported significantly more blood loss in the early excision group44,46,56. The fourth study reported similar findings without variance data, therefore re-analyses was not possible57. Complications of surgery were reported in Engrav et al. (1983) and consisted of one case of upper airway obstruction in the early excision group and one case of donor site infection in the delayed excision group45.The final secondary outcome addressed was costs of treatment. Engrav et al. (1983) reported significantly lower mean total hospital costs in the early excision group compared to the delayed excision group45. A cautious overall conclusion might be that early excision reduces the length of hospital stay, whereas conservative treatment reduces the proportion of participants that receive a surgical intervention. In addition, participants who receive delayed excision (conservative treatment) experience less blood loss during surgery compared to participants in the early excision group. A reason for our caution is that heterogeneity of studies prevented pooling for most outcomes and therefore none of the results were sufficient to support any definite conclusions.

Overall completeness and applicability of evidenceThe objective of this review was to assess the effectiveness of early excision and grafting on scar quality in people with burns of any depth. Unfortunately, only three out of ten studies addressed scar quality as an outcome and differed in outcome measurement. With regard to



mortality, major improvements in burn care in the twentieth century have decreased mortality rates. One of those major improvements is early excision in major burns, but other more recent improvements in newly developed topical antibiotics and critical care might diminish the effect of early excision on mortality. Since three out of four studies that reported mortality were conducted more than 25 years ago, this raises the question whether the results of those older studies are still applicable to present-day burn care. In addition, none of the included studies addressed the patient reported outcomes of pain, satisfaction or quality of life. Therefore, overall completeness has clearly not been achieved. The included studies were heterogeneous, so we could not assess publication bias with a Begg funnel plot or an Egger test. In addition, the cautious conclusion that there appears to be no clinically significant reduction in wound infection between early excision and conservative treatment might be applicable to specialised burn centres in developed countries only. In contrast to most health care facilities in developing countries, those burn centres have the facilities and resources that are necessary to prevent wound infection.

Quality of the evidenceThe evidence combined in this review was of insufficient quality to allow definite conclusions to be drawn. Most of the ten included studies had relatively small sample sizes, ranging from 13 to 85, and the total of 416 participants might be lower due to a possible overlap of participants between three studies46,54,57. While pooling data from small trials could increase statistical power and give a more precise overall estimate of effect size, most of the studies in this review did not compare similar interventions or differed in outcome measures, which generally prevented pooling. Most of the included studies had methodological limitations regarding sequence generation, allocation concealment and blinding of outcome assessment. Only two studies described sequence generation and allocation concealment adequately45,53 and one study provided this information in personal communication56. One study described sequence generation adequately but did not provide information about the allocation concealment46 and the other six studies only stated that participants were randomised. The nature of the interventions made blinding of participants and care providers not possible, but outcome assessors could have been blinded. This was done in six out of ten studies, whereas three studies did not describe blinding of outcome assessors47-53 and one study did not blind the outcome assessor45. Furthermore, drop-out rate in two studies was unacceptable high without reasons given45,55, one study did not perform intention-to-treat (ITT) analysis55 and one study had a minor discrepancy between the number of participants in the text and the number of participants in the tables46. As a result of all these deficiencies, evidence from the included studies should be interpreted with caution.

R1R2R3R4R5R6R7R8R9


168 | Chapter 8

Potential biases in the review processPotential bias in the review process might have occurred due to a possible participant overlap between three studies46,54,57 that were conducted in the same hospital with an overlapping study period. We were unable to confirm whether there was a participant overlap, or determine how this would alter our conclusions, should it indeed be the case. Another potential bias might have arisen as a result of the minimal response to our queries from authors of the eligible studies. The review authors tried to contact study authors by email in an attempt to retrieve all possible data to assess the studies thoroughly. No contact details were available for Jackson et al. (1960) and Salisbury et al.(1982)47,55. Despite issuing a reminder, we received replies only from Engrav et al. (1983), Maslauskas et al. (2005) and Subrahmanyam et al. (1999)45,48-52,56. As a result of those answers, we judged the sequence generation and allocation concealment of Subrahmanyam et al. (1999) to be of “low risk of bias” instead of “unclear risk of bias” and we judged the blinding of outcome assessment of Engrav et al. (1983) to be of “high risk of bias” instead of “unclear risk of bias”. Maslauskas et al. (2005) provided information on citations only.

Agreements and disagreements with other studies or reviewsThe results of this review are largely in accordance with the results of Ong et al. (2006), but some conclusions in that non-systematic and not up-to-date review were slightly premature 29. Ong et al. (2006) states that “early excision of burns reduces mortality in patients without inhalational injury, increases blood transfusion requirements and reduces the length of hospital stay in patients”. These conclusions were based on six studies, all with methodological limitations and including one study that was excluded in this review because it was not an RCT61-63. Since none of the included studies provided firm evidence, conclusions by Ong et al. (2006) should have been more cautious. Furthermore, the small number of included studies in this review is in accordance with two Cochrane reviews that focused on wound dressings31 and topical silver32 that included 26 and 20 RCTs, respectively. Those reviews included more studies, which was expected considering the broader search with regard to the intervention. Wasiak et al. (2008) included all RCTs that assessed burn wound dressings31 and Storm-Versloot et al. (2010) included all RCTs that assessed silver containing wound dressings and topical agents32. Another review investigated the methodological quality of randomised controlled trials in burn care70. Danilla et al. (2009) included 257 eligible studies (from OVID Medline 1950 to January 2008) and concluded that “the reporting standards of RCTs are highly variable and less than optimal in most cases”. Furthermore, their results showed an increase in RCTs over time without a significant improvement in methodological quality70. These findings are only partly in line with our results; methodological quality in the ten included RCTs is similar over time, but our review showed no increase in published RCTs over time as only two out of the ten included RCTs were performed in the 21st century.



Authors’ conclusionsImplications for practiceThere is insufficient high quality research and evidence to enable definite conclusions to be drawn about the effects of early excision and grafting on scar quality in people with burns. Nonetheless, results indicate that early excision and grafting reduces positive wound swabs and length of hospital stay, whereas conservative treatment reduces the proportion of participants that receive a surgical intervention. In addition, participants who receive delayed excision (conservative treatment) experience less blood loss during surgery compared to participants in the early excision group.

Implications for researchThere is a need for large, well-designed trials that compare early excision and grafting with conservative treatment in burns. In order to improve methodological quality, future studies should be designed in conjunction with a trials expert and a statistician and should follow the CONSORT guidelines on reporting71. Appropriate sequence generation and allocation concealment methods should be used in order to reduce the risk of selection bias, and blinding should be attempted to avoid performance and detection biases. Although it is difficult to blind participants and care providers (performance bias), to some extent it is possible to blind outcome assessors (detection bias). A sample size calculation should be used in order to increase statistical power and give a more precise overall estimate of effect size. Furthermore, future trialists might add pain, patient satisfaction and quality of life to their outcomes of interest, as these outcomes are especially important for patients. In addition, future trialists should include scar quality as an outcome of interest, preferably measured with validated scar assessment tools that incorporate both the patient’s and professional’s perspective, like the Patient and Observer Scar Assessment Scale (POSAS)72 (http://www.posas.org) that has been found reliable and valid73.

AcknowledgementsThe review authors would like to thank Johannes van der Wouden for his advice on methodology and the Dutch Association of Burn Survivors for their comments on outcome measures. Furthermore, the review authors would like to thank the peer referees who refereed the protocol and/or review: Wounds Group Editors: Joan Webster, Andrew Jull, Sonya Osborne, Trial Search Co-ordinator: Ruth Foxlee, Amande Briant, Methodologist: Emma Maund, Anne-Marie Bagnall, Statistican: Evan Kontopantelis and referees Jeffrey Shupp, Heather Cleland, Stephanie Kondos, Pramod Kumar and Ruth Ropper for their comments. The authors would also like to acknowledge the contribution of the copy editor Jenny Bellorini.

R1R2R3R4R5R6R7R8R9


170 | Chapter 8

Declarations of interestNo conflict of interest.

Sources of support Internal source: The Association of Dutch Burn Centres, NetherlandsExternal source: NIHR/Department of Health (England), (Cochrane Wounds Group), UK



References

1. World Health Organization. WHO fact sheet on burns. www.who.int/mediacentre/factsheets/fs365/en/index.html 2012 (accessed 27 August 2013).

2. American Burn Association. Burn incidence and treatment in the United States: 2012 Fact Sheet. www.ameriburn.org/resources_factsheet.php (accessed 26 August 2013).

3. Hettiaratchy S, Dziewulski P. ABC of burns: Introduction. BMJ 2004;328(7452):1366-8.4. Ormel W. Brandwonden. Amsterdam, The Netherlands: Consumer Safety Institute, 2010 February.5. Hoogewerf CJ, van Baar ME, Hop MJ, Bloemen MCT, Middelkoop E, Nieuwenhuis MK. Burns to

the head and neck: epidemiology and surgery of facial burns. Burns 2013;39(6):1184-92.6. Van Baar ME, Essink-Bot ML, Oen IMMH, Dokter J, Boxma H, Van Beeck EF. Functional outcome

after burns: a review. Burns 2006;32(1):1-9.7. Hop MJ, Hoogewerf CJ, van Baar ME, van der Vlies CH, Middeloop E. A call for evidence: timing

of surgery in burns. Burns 2012;38(4):617-8.8. Herndon 2007. Total Burn Care. 3rd edition. Philadelphia: Saunders Elsevier, 2007.9. World Health Organization. Facts about injuries. Burns. http://www.who.int/violence_injury_

prevention/publications/other_injury/en/burns_factsheet.pdf 2004 (accessed 27 August 2013).10. Klasen HJ. History of burns. Rotterdam, The Netherlands: Erasmus Publishing, 2004.11. Cubison TCS, Pape SA, Parkhouse N. Evidence for the link between healing time and the

development of hypertrophic scars (HTS) in paediatric burns due to scald injury. Burns 2006;32(8):992-9.

12. Deitch EA, Wheelahan TM, Rose MP, Clothier J, Cotter J. Hypertrophic burn scars: analysis of variables. Journal of Trauma 1983;23(10):895-8.

13. Erol S, Altoparlak U, Akcay MN, Celebi F, Parlak M. Changes of microbial flora and wound colonization in burned patients. Burns 2004;30(4):357-61.

14. Wysocki AB. Evaluating and managing open skin wounds: colonization versus infection. AACN Clinical Issues 2002;13(3):382-97.

15. Weber J, McManus A. Nursing Committee of the International Society for Burn Injuries. Infection control in burn patients. Burns 2004;30(8):A16-24.

16. Barret JP, Herndon DN. Effects of burn wound excision on bacterial colonization and invasion. Plastic and Reconstructive Surgery February 2003;111(2):744-50.

17. Nagoba BS, Gandhi RC, Hartalkar AR, Wadher BJ, Selkar SP. Simple, effective and affordable approach for the treatment of burns infections. Burns 2010;36(8):1242-7.

18. Vermeulen H, van Hattem JM, Storm-Versloot MN, Ubbink DT, Westerbos SJ. Topical silver for treating infected wounds. Cochrane Database of Systematic Reviews 2007, Issue 1. Art. No.: CD005486. DOI: 10.1002/14651858.CD005486.pub2.

19. Hettiaratchy S, Dziewulski P. ABC of burns: Pathophysiology and types of burns. BMJ 2004;328(7453):1427-9.

20. Van Loey NE, Bremer M, Faber AW, Middelkoop E, Nieuwenhuis MK. Itching following burns: epidemiology and predictors. British Journal of Dermatology 2008;158(1):95-100.

21. Bloemen MCT, Van der veer WM, Ulrich MMW, Van Zuijlen PPM, Niessen FB, Middelkoop E. Prevention and curative management of hypertrophic scar formation. Burns 2009;35(4):463-75.

22. Orgill DP, Piccolo N. Escharotomy and decompressive therapies in burns. Journal of Burn Care and Research 2009;30(5):759-68.

23. Klein MB, Hunter S, Heimbach DM, Engrav LH, Honari S, Gallery E, et al. The Versajet water dissector: a new tool for tangential excision. Journal of Burn Care and Rehabilitation 2005;26(6):483-7.

24. Papini R. ABC of Burns: Management of burn injuries of various depths. BMJ 2004;329(7458):158-60.

25. Meek CP. Successful microdermagrafting using the Meek-Wall microdermatome.. American Journal of Surgery 1958;96(4):557-8.

26. Orgill DP. Excision and skin grafting of thermal burns. New England Journal of Medicine 2009;360(9):893-901.

R1R2R3R4R5R6R7R8R9


172 | Chapter 8

27. Choi M, Panthaki ZJ. Tangential excision of burn wounds. Journal of Craniofacial Surgery 2008;19(4):1056-60.

28. Kirn DS, Luce EA. Early excision and grafting versus conservative management of burns in the elderly. Plastic and Reconstructive Surgery 1998;102(4):1013-7.

29. Ong YS, Samuel M, Song C. Meta-analysis of early excision of burns. Burns 2006;32(2):145-50.30. Barret JP, Herndon DN. Modulation of inflammatory and catabolic responses in severely burned

children by early burn wound excision in the first 24 hours. Archives of Surgery 2003;138(2):127-32.

31. Wasiak J, Cleland H, Campbell F. Dressings for superficial and partial thickness burns. Cochrane Database of Systematic Reviews 2008, Issue 4. Art. No.: CD002106. DOI: 10.1002/14651858.CD002106.pub3.

32. Storm-Versloot MN, Vos CG, Ubbink DT, Vermeulen H. Topical silver for preventing wound infection. Cochrane Database of Systematic Reviews 2010, Issue 3. Art. No.: CD006478. DOI: 10.1002/14651858.CD006478.pub2.

33. New Zealand Guidelines Group. Management of burns and scalds in primary care. Wellington, New Zealand: Accident Compensation Corporation, 2007.

34. Hoogewerf CJ, Hop MJ, Nieuwenhuis MK, Middelkoop E, Van Baar ME. Early excision and grafting for burns. Cochrane Database of Systematic Reviews 2012, Issue 3. Art. No.: CD009715. DOI: 10.1002/14651858.CD009715.

35. Lefebvre C, Manheimer E, Glanville J, on behalf of the Cochrane Information Retrieval Methods Group. Chapter 6: Searching for studies. In: Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration, 2011. Available from www.cochrane-handbook.org.

36. Scottish Intercollegiate Guidelines Network (SIGN). Search filters. http://www.sign.ac.uk/methodology.filters.html#random (accessed 27 September 2010).

37. Higgins JPT, Altman DG (editors). Chapter 8: Assessing risk of bias in included studies. In: Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration, 2011. Available from www.cochrane-handbook.org.

38. Deeks JJ, Higgins JPT, Altman DG, on behalf of the Cochrane Statistical Methods Group and the Cochrane Bias Methods Group (Editors). Chapter 9: Analysing data and undertaking meta-analyses. In: Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration, 2011. Available from www.cochrane-handbook.org.

39. Review Manager (RevMan) [Computer program]. Version 5.2. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2012.

40. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ 2003;327(7414):557-60.

41. Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics 1994;50(4):1088-101.

42. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997;315(7109):629-34.

43. Ioannidis JP, Patsopoulos NA, Evangelou E. Uncertainty in heterogeneity estimates in meta-analyses. BMJ 2007;335(7626):914-6.

44. Desai MH, Rutan RL, Herndon DN. Conservative treatment of scald burns is superior to early excision. Journal of Burn Care and Rehabilitation September/Oktober 1991;12(5):482-4.

45. Engrav LH, Heimbach DM, Reus JL, Harnar TJ, Marvin JA. Early excision and grafting vs. nonoperative treatment of burns of indeterminant depth: a randomized prospective study. The Journal of Trauma 1983;23(11):1001-4.

46. Herndon DN, Barrow RE, Rutan RL, Rutan TC, Desai MH, Abston S. A comparison of conservative versus early excision. Therapies in severely burned patients. Annals of Surgery 1989;209(5):547-52.

47. Jackson D, Topley E, Cason JS, Lowbury EJL. Primary excision and grafting of large burns. Annals of Surgery August 1960;152(2):167-89.



48. Maslauskas K, Rimdeika R, Kaikaris V. The advantages of early surgical treatment of deep dermal hand burns. In: 16th Conference of the European Wound Management Association; 2006, 18-20 May; Prague, Czech Republic. 2006:126.

49. Maslauskas K, Rimdeika R, Rapolienė J, Ramanauskas T. Analysis of burned hand function (early versus delayed treatment). Medicina (Kaunas) 2005;41(10):846-51.

50. Maslauskas K, Rimdeika R, Rapolienė J, Saladžinskas Ž, Samsanavičius D, Kaikaris V. The comparison of two methods of treatment evaluating complications and deficiency of functions of hands after deep partial skin thickness hand burns. Medicina (Kaunas) 2009;45(1):37-45.

51. Maslauskas K, Rimdeika R. Surgical treatment of deep dermal and full thickness hand burn injuries. In: 2nd World Union of Wound Healing Societies Meeting; 2004, 8-13 July; Paris, France. 2004:106.

52. Rimdeika R, Maslauskas K. The efficiency of aggressive surgical versus conservative treatment of deep dermal hand burns. In: European Wound Management Association Conference; 2005, 15-17 September; Stuttgart, Germany. 2005:292.

53. Omar MT, Hassan AA. Evaluation of hand function after early excision and skin grafting of burns versus delayed skin grafting: a Randomized clinical trial. Burns 2011;37(4):707-13.

54. Rutan TC, Herndon DN, Van Osten T, Abston S. Metabolic rate alterations in early excision and grafting versus conservative treatment. The Journal of Trauma 1986;26(2):140-2.

55. Salisbury RE, Wright P. Evaluation of early excision of dorsal burns of the hand. Plastic and Reconstructive Surgery April 1982;69(4):670-5.

56. Subrahmanyam M. Early tangential excision and skin grafting of moderate burns is superior to honey dressing: a prospective randomised trial. Burns 1999;25(8):729-31.

57. Thompson P, Herndon DN, Abston S, Rutan T. Effect of early excision on patients with major thermal injury. The Journal of Trauma 1987;27(2):205-7.

58. Barret JP, Herndon DN. Effects of burn wound excision on bacterial colonization and invasion. Plastic and Reconstructive Surgery February 2003;111(2):744-50.

59. Baux S, Mimoun M, Kirsch JM, Guero S, Faivre JM Spagnoli AM. Recent burns of the hand early excision-graft versus conventional treatment: a retrospective study during two years [Brûlures récentes des mains excision-greffe précoce contre traitement conventionnel: étude rétrospective sur deux années]. Annales de Chirurgie de la Main 1987;6(4):276-81.

60. El Hadidy M, Cavallini M, Colonna M, Rizzo F, Signorini M. Contraction and growth of deep burn wounds covered by non-meshes and meshed split thickness skin grafts in humans. Burns 1994;20(3):226-8.

61. Sørensen B, Fisker NP, Steensen JP, Kalaja E. Acute excision or exposure treatment? Final results of a three-year randomized controlled clinical trial. Scandinavian Journal of Plastic and Reconstructive Surgery 1984;18(1):87-93.

62. Sørensen B, Pedersen EK, Fisker NP Steensen JP. Acute excision or exposure treatment? In: Rivista Italiana di Chirurgia Plastica, Padova, 25-27 September. Vol. 13. 1981.

63. Sørensen B. Acute excision or exposure treatment? Results of a randomized controlled clinical trial. Scandinavian Journal of Plastic and Reconstructive Surgery 1979;13(1):115-8.

64. Edstrom LE, Robson MC, Macchiaverna JR, Scala AD. Prospective randomized treatments for burned hands: nonoperative vs. operative. Preliminary report. Scandinavian Journal of Plastic and Reconstructive Surgery 1979;13(1):131-5.

65. Ganai AA, Darzi MA, Chowdri NA, Iqbal S, Zaroo MI. Prospective randomized comparative study of early surgical excision and conservative treatment for deep burns of hand. JK-Practitioner 2002;9(1):24-7.

66. Mohammadi AA, Bakhshaeekia AR, Marzban S, Abbasi S, Ashraf AR, Mohammadi MK, Toulide-ie HR, Ravakkolian AR. Early excision and skin grafting versus delayed skin grafting in deep hand burns (a randomised clinical controlled trial). Burns 2011;37(1):36-41.

67. Krieger Y, Berezovski B, Silverstein E, Gurfinkel R, Arnon O, Vigoda D, Shoham Y, Rosenberg N, Levi A, Sagi A, Rosenberg L. Enzymatic debridement of deeply burned hand: an update. In: Burns; European Burns Association 14th Annual Congress; 2011, 14-17 September; The Hague, The Netherlands. Vol. 37 (Supplement 1). 2011:S19.

R1R2R3R4R5R6R7R8R9


174 | Chapter 8

68. Anniboletti T, Palombo M, Fasciani L, Delli Santi G, Palomba P. The use of Versajet hydrosurgery: 5 years experience. In: Burns; European Burns Association 14th Annual Congress; 2011, 14-17 September; The Hague, the Netherlands. Vol. 37 (Supplement 1). 2011:S19.

69. Munster AM, Smith-Meek M, Shalom A. Acellular allograft dermal matrix: immediate or delayed epidermal coverage? Burns 2001;27(2):150-3.

70. Danilla S, Waisak J, Searle S, Arriagada C, Pedreros C, Cleland H, et al. Methodological quality of randomised controlled trials in burns care. A systematic review. Burns 2009;35(7):956-61.

71. Moher D, Hopewell S, Schulz KF, Montori V, Gøtzsche PC, Devereaux PJ, Elbourne D, Egger M, Altman DG. CONSORT 2010 Explanation and Elaboration: Updated guidelines for reporting parallel group randomised trials. BMJ 2010;340:c869.

72. Draaijers LJ, Tempelman FR, Botman YA, Tuinebreijer WE, Middelkoop E, Kreis RW, van Zuijlen PP. The patient and observer scar assessment scale: a reliable and feasible tool for scar evaluation. Plastic and Reconstructive Surgery 2004;113(7):1960-5; discussion 1966-7.

73. Van der Wal MB, Tuinebreijer WE, Bloemen MC, Verhaegen PD, Middeloop E, van Zuijlen PP. Rasch analysis of the Patient and Observer Scar Assessment Scale (POSAS) in burn scars. Quality of Life Research 2012;21(1):13-23.

Chapter 9Cost study of dermal substitutes and topical negative

pressure in the surgical treatment of burns

M. Jenda Hop, Monica C.T. Bloemen, Margriet E. van Baar,

Marianne K. NieuwenhuisPaul P.M. van Zuijlen,

Suzanne Polinder, Esther Middelkoop

TOPSKIN study group

Burns. May;40(3):388-96

R1R2R3R4R5R6R7R8R9


176 | Chapter 9

Abstract

Background: A recently performed randomised controlled trial investigated the clinical effectiveness of dermal substitutes (DS) and split skin grafts (SSG) in combination with topical negative pressure (TNP) in the surgical treatment of burn wounds. In the current study, medical and non-medical costs were investigated, to comprehensively assess the benefits of this new treatment.Methods: The primary outcome was mean total costs of the four treatment strategies: SSG with or without DS, and with or without TNP. Costs were studied from a societal perspective. Findings were evaluated in light of the clinical effects on scar elasticity. Results: Eighty-six patients were included. Twelve months post-operatively, highest elasticity was measured in scars treated with DS and TNP (p=0.027). The initial cost price of treatment with DS and TNP was €2912 compared to treatment with SSG alone €1703 (p< 0.001). However, mean total costs per patient did not differ significantly between groups (range €29 097- €43 774). Discussion: Costs of the interventional treatment contributed maximal 7% to the total costs and total costs varied widely within and between groups, but were not significantly different. Therefore, in the selection of the most optimal type of surgical intervention, cost considerations should not play an important role.


Cost study of dermal substitutes and topical negative pressure in the surgical treatment of burns | 177

Background

In the past decades, due to a significant reduction in mortality, the focus in burn care has shifted from improving survival to quality of life and scar quality. Full thickness burns usually heal with considerable scars, leading to both esthetical and functional problems. One of the factors causing scars is a lack of dermal tissue in the healing wound. In recent years, the application of dermal substitutes (DS) in combination with skin grafts to improve scar quality has been investigated in both burn and other wounds1-7. In reconstructive burn wounds, with this technique, an improvement in scar quality was found4. The use of dermal substitutes in acute burns, however, did not lead to improvement of scar quality in previous studies of our research group4,8-10. Thus, to improve the effectiveness of dermal substitutes in burns, additional strategies need to be investigated.

Our research group recently performed a randomised controlled trial (TOPSKIN study) to test if the combined application of a dermal substitute and a split-skin graft under topical negative pressure (TNP) would improve graft take and lead to an improved scar quality. It was postulated that the previously reported limited effects of dermal substitutes in burns were due to a lower initial graft take in burns treated with dermal substitutes and split skin grafts, versus burn wounds treated with split skin grafts only. Previous research showed an improved graft take after treatment with TNP therapy, compared with standard dressings in skin grafted wounds11-13. Therefore, in our RCT, the combined application of dermal substitution, split-skin grafting and TNP therapy was investigated. The clinical outcomes of this study have already been reported. To summarise, graft take and wound epithelialisation did not reveal significant differences between standard treatment with split skin graft alone and treatment with dermal substitution, split-skin grafting and TNP therapy combined. Twelve months postoperatively, higher scar elasticity was measured in scars treated with dermal substitution, split skin grafting and TNP therapy compared to scars treated with dermal substitution and split skin grafting14. Until now, however, there was no insight in the costs involved.

Next to clinical effectiveness, costs need to be evaluated in order to comprehensively assess the benefits of a new treatment. Healthcare costs associated with treatment of patients with burns are significant as these patients often need intensive treatment including time and material consuming surgical and non-surgical wound care, intensive care and long rehabilitation15-20. Treatment of patients with burns is known to be more expensive than, for example, the treatment of patients with HIV/AIDS or stroke survivors17. With healthcare getting increasingly expensive and pressure on budgets rising, cost and outcome studies are important to demonstrate economic viability of new technologies. Cost studies provide valuable insight into the distribution of costs and allow the development of specific cost reducing measures.

R1R2R3R4R5R6R7R8R9


178 | Chapter 9

Up till now, there are only few studies available into the costs of burn care15-17,19-24. Costs were, in most of these studies, only ‘roughly’ calculated, using charges and/or with little information about the differentiation of these costs15,17,21,22,24. To date, no cost studies have been performed on the use of dermal substitutes in combination with TNP-therapy in burns. The aim of this study was to comprehensively analyse the medical and non-medical costs of patients with burns that were surgically treated with dermal substitutes and split skin grafts in combination with topical negative pressure, in a multicentre randomised controlled trial.

Patients and methods

Study design and participantsThe TOPSKIN study was a multicentre randomised controlled trial. Details of the study methods have been reported previously14.Patients were recruited in the clinic or outpatient clinic of the three Dutch burn centres in the Netherlands (Beverwijk, Groningen, Rotterdam), from October 2007 until February 2010.Inclusion criteria were:

1. deep dermal or full thickness burn wounds requiring skin transplantation2. age ≥ 18 yrs3. TBSA full thickness burns ≤ 15%4. study wound surface area min. 10 cm² and max. 300 cm²5. informed consent

Exclusion criteria were:1. wounds without adequate possibility to apply TNP2. infected wounds3. severe cognitive dysfunction or psychiatric disorders4. immunocompromized patients5. pregnancy

InterventionPatients were randomly assigned to one of the following intervention groups:1. Dermal substitute (Matriderm®), 1:1.5 meshed split skin graft (SSG) and TNP therapy

(DS-TNP)2. Dermal substitute and SSG (DS)3. SSG and TNP therapy (TNP)4. Standard treatment: SSG alone (ST)In surgery, after avulsion, tangential excision or scrubbing, a dermal substitute was applied in the patients of group 1 and 2, and on top of that a meshed split skin graft was positioned.



In group 3 and 4, a meshed split skin graft was immediately applied after avulsion/tangential excision or scrubbing. TNP therapy, with a Vacuum-Assisted Closure (VAC) system, was subsequently applied in group 1 and 3 and left in place for 3-5 days.

Outcome measuresCosts: The primary endpoint for this study was defined as mean total costs of the four treatment strategies. To assess the true economic impact of the introduction of a new intervention, not only intervention costs but also other possible cost changes need to be incorporated in a cost analysis. Therefore, total costs included healthcare and non-healthcare costs from the day of admittance until 12 months after randomisation.The economic evaluation was performed in accordance with the Dutch guidelines for such analyses25. Costs were studied from a societal perspective, including direct healthcare costs (inpatient and outpatient costs), direct non-healthcare costs (travel costs) and indirect non-healthcare costs (productivity loss). Real medical costs were calculated by multiplying the volumes of healthcare use with the corresponding unit prices. The cost calculation was performed according to the bottom up approach (following the micro-costing method of Gold et al.26), based on a detailed inventory of all cost items. Resource use was inventoried in all burn centres. Cost prices, on the other hand, were inventoried in one burn centre and used for all centres to prevent measuring cost price differences between burn centres, instead of cost differences between the four intervention groups. The costs applied to the financial year 2010.Hospital costs included intervention costs, hospital days, and other variable costs:

• The surgical costs of interventional treatments were defined as all costs associated with the procedures performed within that strategy. The costs per unit of the surgical procedures were determined with cost-accounting taking into consideration the initial investment of equipment (including TNP equipment), investments during use, maintenance, number of years of use and discounting the number of procedures per year, material costs (including Matriderm and TNP material), personnel costs and a 35.5% raise for housing and overhead in accordance with the guidelines.

• Hospital day costs, both Intensive care Unit (ICU) and non-ICU, were calculated by multiplying the length of hospital stay with the cost price. Hospital day costs consisted of personnel costs (including burn physician costs), material, equipment, food and laundry, and a 35.5% raise for housing and overhead costs.

• Variable costs included: transport to the hospital, other surgical procedures, dressings, medication, diagnostics, allied health professionals/ non-burn physicians, pressure garments, splints, and outpatient visits.

R1R2R3R4R5R6R7R8R9


180 | Chapter 9

Data regarding patients’ baseline characteristics and healthcare use were obtained from patient records.Other healthcare costs and non-healthcare costs were calculated based on charges as a proxy of real costs (based on Hakkaart et al.25). Patients received a questionnaire after 3 and 12 months with questions regarding non-hospital costs, including nursing-home and rehabilitation care, visits to general practitioners and allied health professionals outside the hospital. Indirect non-healthcare costs included loss of economic productivity due to absence from work (human capital method). Direct non-healthcare costs included patient travel costs.

Clinical effectiveness: The clinical effectiveness of this study was presented in detail elsewhere14. In this paper we present the primary clinical outcome: scar elasticity, measured with the cutometer. The parameter maximal skin extension (Uf) (in mm) was used, as this was demonstrated to be the most reliable parameter27. To eliminate influence of different anatomic locations, elasticity was analysed using the ratio of the scar and normal skin.

Statistical analysisDifferences in scar elasticity were analysed using Factorial Anova. Post hoc analysis with Bonferroni correction was performed, to correct for multiple testing in the four treatment groups.Missing cost items were imputed by randomisation group. Since cost data per patient (but not per day of care) are typically highly skewed, nonparametric bootstrap techniques (bootstrapping) were used to derive p-values of the costs and 95% confidence interval for the differences in costs. We planned to perform a cost-effectiveness analysis between group DS-TNP and group ST, however, in absence of significant differences in effects (elasticity) between these groups, we performed a cost-minimisation analysis in which only costs were compared.Data analysis was performed using SPSS Statistics 20 and RStata.

Results

Patient characteristics and clinical effectiveness In total 86 patients were included. The four intervention groups were comparable with respect to patient and burn characteristics. The total group had a mean age of 47 years, with predominantly flame burns, frequently on the arms (Table 1).



Table 1. Characteristics of 86 patients randomised to four different interventionsCharacteristics DS-TNP DS TNP ST

N= 21 N= 23 N= 22 N= 20Male/female 11/10 12/11 12/10 14/6Age at admission (years) 44 (18-70) 48 (21-84) 49 (22-76) 53 (20-84)TBSA burned (%) 8.0 (0.5-21.0) 9.6 (0.5-28.0) 10.0 (0.2-45.0) 7.7 (0.8-30.5)Aetiology Flame 13 14 9 14 Scald 0 3 5 0 Other 8 6 8 6Location Arm 10 12 11 10 Leg 6 4 8 8 Trunk 5 7 3 2Mean length of hospital stayIncluding mean ICU days

21 (2-43)1 (0-7)

24 (0-58)3 (0-36)

22 (0-71)2 (0-21)

21 (0-46)1 (0-14)

TBSA: Total Body Surface AreaLength of hospital stay: hospital day, ICU days and day care

There was a significant difference in elasticity ratio (scar/normal skin) twelve months post-operatively (ratio Uf 0.80, p=0.027). After post hoc testing, it was demonstrated that elasticity of the DS-TNP group was significantly higher compared to the DS group (p=0.012) [14]. There were no significant differences between the experimental groups (DS-TNP, DS, TNP) compared to the standard treatment: SSG alone (ST) (Table 2).

Table 2. Clinical effectiveness: scar elasticityDS-TNP DS TNP ST P-valueN= 21 N= 23 N= 22 N= 20

Scar elasticity 12 months PO, Ratio Uf[95% CI]

0.80 [0.62-0.98]

0.51 [0.33-0.68]

0.66 [0.49-0.84]

0.69 [0.53-0.86]

P = 0.027

PO: post operatively, Ratio: scar/normal skin, Uf: maximal skin extension (mm). Statistical analysis was performed using Factorial Anova with post hoc testing with Bonferroni correction. Post hoc test with Bonferroni correction shows a significant difference (p= 0.012) between group DS-TNP and DS.

Costs The initial cost prices of the interventional treatment, including all personnel, equipment, material, housing and overhead costs, were €2912, €2218, €2180, €1703, for DS-TNP, DS, TNP and ST respectively (Table 3). The highest intervention costs were found in the DS-TNP group (p<0.001), due to a combination of relatively high personnel, equipment and material costs. In groups DS-TNP and TNP the use of topical negative pressure therapy led to high equipment costs: the costs of 6 days use of TNP were €254. In the dermal substitution groups (DS and DS-TNP) the high material costs were caused by the use of Matriderm® (€ 333/sheet, one or two sheets per patient).

R1R2R3R4R5R6R7R8R9


182 | Chapter 9

Table 3. Full cost prices (€, 2010) of the four surgical interventional treatments

Cost categoryDS-TNP DS TNP ST P values

***N= 21 N= 23 N= 22 N= 20

Personnel 883 720 904 681Equipment * 402 146 408 151Materials ** 864 771 297 425Housing/overhead (35.5%) 763 581 571 446Total costs 2 912 2 218 2 180 1 703 P<0.001

*Anaesthesia equipment, smoke extraction, diathermia, surgical instruments and VAC system (costs for the use of 6 days were €254)**All materials, including Matriderm®: the costs of one Matriderm® sheet were € 333, one or two sheets/patient were used (one sheet for approximately 1 % TBSA)***Statistical analysis was performed using the independent samples Kruskal-Wallis test

A detailed overview of the mean costs of specialised burn care per patient per surgical treatment group is shown in Table 4. Specialised burn care represented on average 66% of all costs and was €24 277, €29 787, €24 871, and € 23 302 for DS-TNP, DS, TNP, and ST group respectively; costs did not significantly differ between the four groups (p=0.768). Two significant differences were found in specialised burn care: high diagnostic costs were found in the DS-group (p<0.001), due to one patient with a long ICU stay (36 days), and low clinical consultation costs were found in the DS-TNP group (p=0.003). In specialised burn care, costs of intramural care, consisting of ICU, non-ICU hospital days and day care, were by far the highest cost category in all treatment groups (up to 75% of the specialised burn care costs and up to 59% of the total costs per patient). The majority of included patients was clinically admitted (94.1%), a minority (5.9%) was treated as outpatients, undergoing surgery in day care. The mean length of hospital stay was respectively 21 days (including 1 ICU day), 24 days (3 ICU days), 22 day (2 ICU days), 21 days (1 ICU day) in the DS-TNP, DS, TNP, and ST group. In the DS group, high ICU costs were observed, this was due to one patient with a long ICU stay (36 days).Another high cost category within the specialised burn care were the overall treatment costs, including surgery, wound dressing etc. (DS-TNP €5340; DS €4849; TNP €4575; ST €4058; p=0.126). The interventional treatment was the highest cost item within the treatment costs. However, these intervention costs represented maximally 12% of total specialised burn care costs and 7% of the total costs per patient. The other (non specialised burn care) healthcare cost were relatively low, except for the TNP group (DS-TNP €2242; DS €2525; TNP €7765; ST € 699; p=0.111). The high costs in the TNP group were due to one patient with a rehabilitation centre stay of 210 days.



Patient (work absence and travel) costs were relatively high (with a mean of 25% of the total costs per patient) and varied significantly between the four treatment groups (DS-TNP €14 584; DS € 9 704; TNP €11 138; ST €5 096; p=0.045). For all intervention groups this could largely be attributed to costs of work absence. Most patients were in their working age and the mean productivity loss was respectively 478, 314, 355 and 155 hours per patient. Mean total costs per patient in the four arms did not differ significantly (DS-TNP €41 103; DS €42 017; TNP €43 774; ST €29 097; p=0.105). We observed a trend for higher costs in the experimental groups (DS-TNP, DS and TNP) compared to the standard care (SSG).To study the influence of TBSA burned on the total costs, a sub-analysis of costs by TBSA groups was performed. Higher costs were seen in more severe burned patients (p<0.001). Within these TBSA groups still differences in costs, although non-significant, were found between the intervention groups (Figure 1).

1

Figure 1. Mean total costs per patient by intervention group, in different TBSA groups (€, 2010)

Figure 1. Mean total costs per patient by intervention group, in different TBSA groups (€, 2010)

R1R2R3R4R5R6R7R8R9


184 | Chapter 9

Table 4. Mean costs of healthcare use, work absence and travel costs (€, 2010) per patient

Cost categoryDS-TNP DS TNP ST

P-valuesN= 21 N= 23 N= 22 N= 20 N= 21

Transport hospital 258 226 219 203

Intramural burn careHospital days 13 973 14 672 13 373 13 778ICU days 2 521 5 985 3 369 2 435Readmittance days 588 1 712 1 825 894Day care 135 74 154 157Total intramural [95% CI]

17 217[13 660-22 020]

22 443 [14 660-33 090]

18 722 [12 760- 26 674]

17 264 [12 600-

23 360]P=0.753

*****Diagnostic proceduresSwabs 339 575 264 452Lab tests 209 491 117 262Others * 56 314 39 160Total diagnostic procedures[95% CI]

604 [460-640]

1 380 [960-1680]

421 [310-480]

874 [640-970] P<0.001

TreatmentInterventional treatment study wound 2 912 2 218 2 180 1 703Surgical treatment other wounds ** 712 615 702 351Wound care *** 890 897 674 1 077Medication, intramural 133 279 172 262Blood products (erythrocytes) 189 56 - 13Pressure garments 248 716 461 612Reconstructive surgery study wound 0 0 79 40Reconstructive surgery other wounds 256 69 308 0Total treatment[95% CI]

5 340 [4 620-6 470]

4 849 [4 170-

5 770]

4 575 [3 690-5 650]

4 058 [3 520-

4 860] P=0.126

Clinical consultations Physiotherapist 96 135 87 140Occupational therapist 35 133 46 84Psychologist 33 29 49 37Skin therapist - 2 3 3Psychiatrist - - 81 85Total clinical consultations[95% CI]

163 [120-210]

299 [200-400]

266 [140-420]

350 [240-470] P=0.003

Outpatient burn careOutpatient burn care **** 405 413 441 392Occupational therapy 103 105 87 96Plastic surgeon 115 55 122 38Physiotherapist 62 6 0 14Rehabilitation physician 10 0 15 4



Others(skin therapist, after care nurse, nurse practitioner)

0 12 4 8

Total outpatient burn care[95% CI]

695 [480-960]

590 [490-700]

669 [550-830]

552 [420-720] P=0.615

Total costs specialised burn care[95% CI]

24 277[20 020-30 140]

29 787 [20 900-

41 770]

24 871 [17 770-34 130]

23 302 [17 990-30 460]

P=0.768

Other healthcare costsRehabilitation centre 971 1 774 3 616 0Nursing home 0 0 1 947 0General practitioner 38 36 62 10Home (nursing) care 688 585 1 651 588Extramural physiotherapy 544 131 489 101Total other healthcare costs[95% CI]

2 242 [720-

4 890]

2 525 [390-6 580]

7 765 [1 690-16 320]

699 [260-1 350]

P=0.111

Patient costsWork absence 14 358 9 425 10 711 4 657Travel costs 226 279 428 439Total patient costs[95% CI]

14 584 [7 630-22 840]

9 704 [5 620-14 740]

11 138 [6 010-17 310]

5 096 [2 420-

8 290]P=0.045

Total costs per patient[95% CI]

41 103

[29 910-47 300]

42 017 [28 540-

38 640]

43 774 [28 260-57 200]

29 097

[20 640-35 080]

P=0.105

* other diagnostics: X-ray, CT scan, bronchoscopy**some patients needed also surgery for other wounds than the study wound *** wound care costs consist of material costs only, both in- and outpatient setting**** without material (dressing) costs *****p-values and 95% confidence intervals for the differences in distribution of the cost categories were derived from 2000 bootstrap samples drawn with replacement

Discussion

In this study, the costs of four surgical interventions in patients with burns were assessed in detail in order to provide an integrated evaluation of the total medical and non-medical costs. We aimed for a cost-effectiveness analysis on the introduction of dermal substitutes in combination with TNP therapy in burn care. However, due to subtle differences in scar elasticity, a cost-minimisation analysis was performed, in which only costs between different intervention groups were compared. Differences in costs were not related to effects (a so-called incremental cost effectiveness ratio), because effects between the experimental and standard treatment group were statistically the same.

R1R2R3R4R5R6R7R8R9


186 | Chapter 9

This is the first detailed cost inventory in a clinical trial on dermal substitution in patients with burns. Total costs varied between €29 097 and €43 774. No significant differences were found in total costs between the four treatment groups. The costs of the interventional treatment with dermal substitution, SSG and TNP therapy were significantly the highest, but these costs contributed up to 12% of the specialised burn care costs and up to 7% of the total costs. The major cost item was intramural burn care. The second important cost item was patient costs, including work absence and travel costs. Total cost levels varied strongly between and within treatment groups. However, no significant differences between treatment groups were found, because of heterogeneity of patients within the separate treatment groups, highly skewed cost data per patient (which is normal for cost data), and the small sample size.

Only few complete cost inventories and cost-analyses into the costs of burns were published before. Therefore, we have minimal insight into the distribution of costs in burn care, which impedes the possibility of providing more cost-effective burn care. We found one study on the costs and the effects of TNP therapy and conventional dressing methods on split skin grafts in patients with burns in India11. An improved graft take was shown with the use of TNP (from 87.5% to 96.7% at day 9). The cost inventory was limited, and confined to material costs only (costs of foam, tube and wound dressing) of an average sized burn wound: US$ 9.95 (€ 7.77). The more costly items like VAC equipment and hospital days, were not taken into account. To our knowledge, no cost studies on the use of dermal substitutes in burn care were presented before. It was only assumed that the use of dermal substitutes is expensive. In this study, we have shown that the use of dermal substitutes and TNP therapy indeed involved higher surgical costs than standard surgical burn care, but that the surgical therapy represented only a small part of the total costs.In this study, hospital days represented the highest cost category, which was also shown in other studies on burn care costs16,17,19,28. The second important cost item of our study, patients costs (mainly work absence costs), was recently also studied by Sanchez et al.. In their study work absence costs represented even the highest cost item17.

This study had some limitations. First, the study had enough statistical power to detect clinical effectiveness, but groups were relatively small for a cost analysis. In future, we would like to perform cost analyses in larger patient groups.Secondly, the interventional treatment was only applied on a small study wound (1 or 2 % TBSA). Mean total costs were the result of all patient and injury (e.g. TBSA) characteristics. Therefore non-significant differences in mean total costs (with a trend for higher costs in the ‘experimental groups’ versus ‘standard care’) cannot be contributed to the study intervention only, but are the result of the total burn care, sometimes including multiple wounds in one patient. However, it is impossible to distinguish all costs related to the study intervention from



all other burn care costs. Where possible, costs were directly linked to the intervention: costs of interventional therapy and costs of reconstructive surgery of the study wound. Furthermore, higher costs in more severe burns were found. However, (non-significant) differences between intervention groups still seem to exist. Another limitation was the limited follow up time; 12 months post-operatively. Consequently, the costs of possible reconstructive surgery or other rehabilitation activities after this period were not considered. Possibly, in the long term, DS and SSG in combination with TNP therapy will lead to less reconstructive surgery and subsequently lower costs.Furthermore, the calculated mean costs per patient cannot be generalised to all patients with burns, because we studied a selection of surgically treated adult patients with burns with a maximum of 15% TBSA full thickness burns.Finally, the cost calculations were performed in Dutch burn centres only. Although the multicentre analysis of resource use will improve generalizability compared to a single-centre analysis, cost prices and resource use in other countries are probably different29 due to other healthcare systems and different burn care practices. However, since our standards of burn care are comparable to other high-income countries, our cost overview will be indicative for these countries as well.

To conclude, no differences in total costs between the four intervention groups were shown and the intervention costs represented only a small part of the total costs per patient. Therefore, cost considerations should not play an important role in the selection of the most optimal treatment, out of the four described in this study. As a consequence, the clinical effectiveness of dermal substitution and split skin grafting in combination with topical negative pressure must be decisive. This trial already showed an improvement in scar elasticity after applying topical negative pressure in wounds treated with dermal substitutes. Nevertheless, no statistical improvement could be demonstrated versus the standard split skin graft treatment. In our opinion, this was mainly a consequence of the multicentre setup of the study, which increased the variation in outcome parameters, in combination with relatively small treatment groups. In addition, the effects were probably limited because of the use of split skin grafts meshed in a 1:1.5 ratio, which was demonstrated before to have less effect on scar improvement than in larger mesh expansions8,10. Before implementation, further evaluation of the surgical treatment with dermal substitutes in larger patient groups with burns is therefore necessary to accomplish an improved scar quality compared to standard treatment with split skin graft alone.We recommend including full cost comparisons in all randomised controlled trials, in order to comprehensively assess the cost consequences of a treatment change.

R1R2R3R4R5R6R7R8R9


188 | Chapter 9

Funding This research was financially supported by grants of the Dutch Burns Foundation (11.102 and 07.109).

Conflict of interest statementThe authors declare that they have no conflict of interest.

AcknowledgementsThe authors would like to thank O.J. van de Breevaart, from the financial department of the Maasstad Hospital Rotterdam, for his input in the cost inventory. We also thank the TOPSKIN study group: G.I.J.M. Beerthuizen MD, PhD, H. Boxma MD, PhD, J. Dokter MD, S.M.H.J. Scholten MD, PhD, F.R.H. Tempelman MD, P.D.H.M. Verhaegen, MD, PhD, A.F.P.M. Vloemans MD, and M.BA. van der Wal, MD.In addition, we thank J. Eshuis RN, J. Hiddingh MSc, RN, and H. Hofland MScN, for collecting part of the data.



References

1. Wainwright DJ. Use of an acellular allograft dermal matrix (AlloDerm) in the management of full-thickness burns. Burns1995; 21: 243–8.

2. Dantzer E, Queruel P, Salinier L, Palmier B, Quinot JF. Dermal regeneration template for deep hand burns: clinical utility for both early grafting and reconstructive surgery. Br J Plast Surg 2003; 56: 764–74.

3. Marston WA, Hanft J, Norwood P, Pollak R. The efficacy and safety of Dermagraft in improving the healing of chronic diabetic foot ulcers: results of a prospective randomized trial. Diabetes Care 2003; 26: 1701–5.

4. Van Zuijlen PP, Van Trier AJ, Vloemans JF, Groenevelt F, Kreis RW, Middelkoop E. Graft survival and effectiveness of dermal substitution in burns and reconstructive surgery in a one-stage grafting model. Plast Reconstr Surg 2000; 106: 615–23.

5. Canonico S, Campitiello F, Della Corte A, Fattopace A. The use of a dermal substitute and thin skin grafts in the cure of “complex” leg ulcers. Dermatol Surg 2009; 35: 195–200.

6. Branski LK, Herndon DN, Pereira C, Mlcak RP, Celis MM, Lee JO, et al.. Longitudinal assessment of Integra in primary burn management: a randomized pediatric clinical trial. Crit Care Med 2007; 35: 2615–23.

7. Donohue KG, Carson P, Iriondo M, Zhou L, Saap L, Gibson K, Falanga V. Safety and efficacy of a bilayered skin construct in full-thickness surgical wounds. J Dermatol 2005; 32: 626–31.

8. Van Zuijlen PP, Vloemans JF, van Trier AJ, Suijker MH, van Unen E, Groenevelt F, et al.. Dermal substitution in acute burns and reconstructive surgery: a subjective and objective long-term follow-up. Plast Reconstr Surg 2001;108: 1938–46.

9. Van Zuijlen PP, Lamme EN, van Galen MJ, van Marle J, Kreis RW, Middelkoop E. Long-term results of a clinical trial on dermal substitution. A light microscopy and Fourier analysis based evaluation. Burns. 2002 Mar;28(2):151-60.

10. Bloemen MC, van Leeuwen MC, van Vucht NE, van Zuijlen PP, Middelkoop E. Dermal substitution in acute burns and reconstructive surgery: a 12-year follow-up. Plast Reconstr Surg 2010; 125: 1450–9.

11. Petkar KS, Dhanraj P, Kingsly PM, Sreekar H, Lakshmanarao A, Lamba S, et al.. A prospective randomized controlled trial comparing negative pressure dressing and conventional dressing methods on split-thickness skin grafts in burned patients. Burns. 2011 Sep;37(6):925-9.

12. Moisidis E, Heath T, Boorer C, Ho K, Deva AK. A prospective,blinded, randomized, controlled clinical trial of topical negative pressure use in skin grafting. Plast Reconstr Surg 2004; 114:917–22.

13. Scherer LA, Shiver S, Chang M, Meredith JW, Owings JT. The vacuum assisted closure device: a method of securing skin grafts and improving graft survival. Arch Surg 2002; 137: 930–3; discussion 933–4.

14. Bloemen MC, van der Wal MB, Verhaegen PD, Nieuwenhuis MK, van Baar ME, van Zuijlen PP, Middelkoop E. Clinical effectiveness of dermal substitution in burns by topical negative pressure: A multicentre randomized controlled trial. WoundRepair Regen. 2012 Nov;20(6):797-805.

15. Sahin I, Ozturk S, Alhan D, Açikel C, Isik S. Cost analysis of acute burn patients treated in a burn centre: the Gulhane experience. Ann Burns FireDisasters. 2011 Mar 31;24(1):9-13.

16. Ahn CS, Maitz PK. The true cost of burn. Burns. 2012 Nov;38(7):967-74.17. Sanchez JL, Bastida JL, Martínez MM, Moreno JM, Chamorro JJ. Socio-economic cost and health-

related quality of life of burn victims in Spain. Burns. 2008 Nov;34(7):975-81.18. Pellatt RA, Williams A, Wright H, Young AE. The cost of a major paediatric burn. Burns. 2010

Dec;36(8):1208-14. Epub 2010 May 23.19. Hemington-Gorse SJ, Potokar TS, Drew PJ, Dickson WA. Burn care costing: the Welsh experience.

Burns. 2009 May;35(3):378-82.20. Takayanagi K, Kawai S, Aoki R. The cost of burn care and implications for efficient care. Clin

Perform Qual Health Care. 1999 Apr-Jun;7(2):70-3.

R1R2R3R4R5R6R7R8R9


190 | Chapter 9

21. Jeng JC. Growth rings of a tree: progression of burn care charges abstracted from a decade of the national Burn Repository. J Burn Care Res. 2007 Sep-Oct;28(5):659-60.

22. Klein MB, Hollingworth W, Rivara FP, Kramer CB, Askay SW, Heimbach DM, Gibran NS. Hospital costs associated with pediatric burn injury. J Burn Care Res. 2008 Jul-Aug;29(4):632-7.

23. Sánchez JL, Perepérez SB, Bastida JL, Martínez MM. Cost-utility analysis applied to the treatment of burn patients in a specialized center. Arch Surg.2007 Jan;142(1):50-7; discussion 57.

24. Kai-Yang L, Shi-Hui Z, Hong-Tai T, Yi-Tao J, Zhao-Fan X, Dao-Feng B, et al.. The direct hospitalisation costs of paediatric scalds: 2-year results of a prospective case series. Burns. 2009 Aug;35(5):738-45.

25. Hakkaart-van Roijen L, Tan SS,Bouwmans CAM. Handleiding voor kostenonderzoek; methoden en standaard kostprijzen voor economische evaluaties in de gezondheidszorg. Diemen: College voor zorgverzekeringen; 2010.


27. Draaijers LJ, Botman YA, Tempelman FR, Kreis RW, Middelkoop E, van Zuijlen PP. Skin elasticity meter or subjective evaluation in scars: a reliability assessment. Burns. 2004 Mar;30(2):109-14.

28. Griffiths HR, Thornton KL, Clements CM, Burge TS, Kay AR, Young AE. The cost of a hot drink scald. Burns. 2006 May;32(3):372-4.

29. Wheeler JR, Harrison RV, Wolfe RA, Payne BC. The effects of burn severity and institutional differences on the costs of care. Med Care. 1983 Dec;21(12):1192-203.

Chapter 10General Discussion

R1R2R3R4R5R6R7R8R9


192 | Chapter 10


General Discussion | 193

General discussion

The main goal of this thesis was to search for methods to optimise efficiency in burn care. To be able to implement new interventions in burn care, it is necessary to consider adequate economic evaluation studies to compare both effects and costs of different interventions. In this thesis we searched for methods to optimise efficiency in burn care, by identifying the healthcare and non-healthcare costs of burn care, important costs items and predictors for high costs and evaluating the effectiveness and cost-effectiveness of upcoming diagnostics and treatment options in burn care.In the discussion, the three parts of the thesis will be discussed separately, respectively: ‘overview of burn care costs’, ‘improving burn care efficiency: diagnostics and costs’, and ‘improving burn care efficiency: therapeutics and costs’. Finally, the concluding remarks will be presented.


Outcomes and implicationsThis thesis demonstrated that burn care is associated with high costs. Both healthcare and non-healthcare costs, like productivity loss, are substantial in the short-term and long-term.We aimed to identify both healthcare and non-healthcare costs of burn care, and to find important cost items and predictors for high costs. Therefore we conducted a systematic review, a retrospective and a prospective cohort study, which are described in this thesis. In our review, we presented that the mean total healthcare costs per burn patient in high-income countries were €56,010 and that burn care was in general more expensive than other injuries. In our prospective study on burn care costs within 3 months post burn, mean costs per patient were €26,540 per patient, varying from €740 to €235,560, in a population with a mean burn size of 8% total body surface area. Next to the wide methodological variation between the included studies in the review, one of the reasons for the differences in mean total costs per burn patient between the review and our prospective study was the difference in mean TBSA (total body surface area burned).Burn centre stay accounted for 81% of the total specialised burn care costs, high costs were mainly caused by high burn centre day prices. Often reference prices are used for cost calculations, for example, the standard hospital day price (€462, year 2012). presented by Hakkaart et al.1. In burn care, no clear costs prices for burn centre days were described until this thesis. In our review we did reveal 34 studies that described day prices, but often methodology of day price calculation was poorly described and the prices varied widely. In our prospective study, we calculated burn centre day prices in detail, with the help of annual

R1R2R3R4R5R6R7R8R9


194 | Chapter 10

accounts, which consisted of personnel, material, equipment, nutrition, medication, housing and overhead. We found that the costs of a non-ICU burn centre day were €948 and of a ICU burn centre day €2,966. The calculation of cost prices is important in cost studies (Table 1). We recommend to use our day prices for future economic evaluations in high-income countries.Other important cost items were the costs of surgery in the acute phase (before wound healing) and of reconstructive surgery. In our prospective study, the costs of surgery were the second highest cost item of specialised burn care, although mean costs per patients were far lower than mean costs of total burn centre stay (Table 1). Although often studied before4-6, the mean total costs of wound care were only as high as one burn centre day (mean of €906 per patient) In our retrospective study on reconstructive surgery post-burn we showed that one out of eight patients required reconstructive surgery post-burn, the mean medical costs of reconstructive surgery per patient over a 10 year period were € 8342 (range 863 – 80 217). As 13% of all admitted burn patients underwent reconstructive surgery, the mean costs of reconstructive surgery per admitted patient were: €1084 (0.13*8342) (Table 1).In this thesis, we aimed to complement existing knowledge gaps on burn care costs. We demonstrated that next to healthcare costs in the acute phase of the burn injury also non-healthcare costs, like productivity loss, and long-term healthcare costs, like reconstructive surgery were significant (Table 1). The only authors who presented costs of productivity loss/work absence before were Sanchez et al.2,3; they presented even higher work absence costs compared to our results, probably because of their methods of cost calculations (human capital approach). With our study we could confirm that work absence results in high costs after burn injuries, which makes it an important cost item to include in future cost studies.We showed that a larger burn size (%TBSA burned) was strongly associated with higher burn care costs. For the interpretation and comparison of cost studies, it is important to realise that there is a strong association between total costs and mean TBSA. In our study population the mean TBSA was 8%, which is relatively low compared to most previous studies; in about 50% (43/91) of the studies in our review that reported TBSA, mean TBSA burned was over 20%. One of the reasons for the low TBSA in our population was a limited availability of our burn centre due to unexpected renovation activities in a part of the study period (admission was restricted to adults ≤40% TBSA burned and children <10% TBSA burned only). Burn centre related costs per 1% TBSA were €4,245 in our prospective study. This was comparable to the mean cost per %TBSA established in our cost review (€3768), and appeared to be a stable predictor of total burn care costs in different TBSA categories. Therefore, the one percent TBSA price can be used to roughly estimate total burn care costs of a patient. In our literature review we found a total of only three cost-effectiveness analyses. These three CEA’s discussed different dressings and compared their costs. As shown in our prospective cost study, dressings represent only a small part of total burn care costs. Thus, in economic



evaluations of dressings it is more interesting to analyse the influence of a new dressing on total burn care costs, including important items like LOS, than only comparing dressing costs. Since our review, several new economic evaluations in burn care were published, again mainly on topical treatment strategies4-6. Since burn care is expensive care and there is an on-going need for improvement of burn diagnostics and treatment, there is a growing interest in economic evaluation studies in burn care.

Table 1. Summary of most important cost resultsMean costs per patient**

Source * Financial year

One burn centre day €948 Prospective cost study 2012One ICU burn centre day €2,966 Prospective cost study 2012Total burn centre stay €16,575 Prospective cost study 2012Surgery in acute phase*** €1,298 Prospective cost study 2012Reconstructive surgery €1,084 Retrospective study 2011Work absence €5,255 Prospective cost study 2012Costs per % TBSA**** €4,245 Prospective cost study 2012

€3,768 Literature review 2012Total costs per patient €26,540 Prospective cost study 2012

€56,009 Literature review 2012

* chapter 2,3,4 of this thesis, **mean costs per patient admitted to a Dutch burn centre, *** defined as surgery before wound closure, **** specialised burn care costs.

Recommendations for future studiesWith the results of our cost studies we can conclude that to improve efficiency of burn care from a cost perspective, we should search for ways to improve care with a limited length of hospital stay, an early return to work and a low requirement for (reconstructive) surgery. Attempts like this should, of course, always be balanced against quality of care. We recommend to include these important cost items in future economic evaluations in burn care.

In this thesis, we calculated the costs of productivity loss with a follow-up of 3 months post injury, but we also found that 31% of all burn patients with a paid job is still absent from work 3 months post burn. Therefore, costs of productivity loss will be even higher when analysed in studies with a longer follow-up. Furthermore, 45% of the yearly admitted burn patients is admitted outside the burn centres/specialised burn care, to general hospitals, and we do not have insight in the costs of these patients yet7,8. Although, we expect that the inventory of non-specialised burn care cost and long-term costs will be difficult, because it is a labour intensive trajectory in terms of data collection, we recommend to further complete the insight in burn injury costs, by designing cost-studies in burn centres with a long follow-up time (e.g. 5 years) and designing cost studies outside the burn centres.

R1R2R3R4R5R6R7R8R9


196 | Chapter 10

The design and conduct of a cost study can be a challenge9. In our literature review, methods of cost calculations were often unclear or poorly described, which impeded us to perform a meta-analysis. For future cost studies and economic evaluations we recommend the use of a standard approach to improve the quality and harmonisation of economic evaluation studies, optimise comparability and improve insight into burn care costs and efficiency, e.g. by using the recently published CHEERS statement9. Economic studies in burn care should preferably be performed from a societal perspective, including productivity loss1, and adapt an adequate time horizon. We expect that a time frame of several (e.g. 5) years post-burn is necessary, but this hypothesis should be explored further by cost studies with a longer follow-up. In this thesis we carefully calculated burn centre day (both ICU and non-ICU) prices, existing of personnel, material (without dressing costs), equipment, food and laundry, medication, and overhead. Other, non-fixed costs, like dressings and surgery should preferably be calculated separately. We suggest to use our results as a reference price for future studies. In case of significant differences in clinical effects between interventions, a cost-effectiveness analysis should be performed9. Cost-utility analysis (CUA’s) is the preferred analyses by policy makers and for this purpose EuroQol-5D should be included in economic evaluations. In the studies from this thesis we eventually refrained from a CEA/CUA because of small differences in primary clinical effects.

Part II. Improving burn care efficiency: diagnostics and costs

Outcomes and implicationsThe diagnostic instrument mainly focussed on in this thesis was laser Doppler imaging (LDI). First, we conducted a RCT on the cost-effectiveness of LDI. Secondly, we used LDI as a gold standard in another attractive tool in burn care: telemedicine.In part II of this thesis we showed that the introduction of LDI improved the timing of therapeutic decisions and has the potential for cost-savings of €875 per scanned patient by reducing LOS in surgical treated patients. Next to that, we searched for methods to efficiently improve burn diagnosis outside the burn centres and showed that telemedicine by using photographs is only suitable for the assessment of burn size, not depth.

We showed that laser Doppler imaging is effective in Dutch burn care by improving therapeutic decision making, as early as 2-5 days post-burn. In a subgroup of admitted patients requiring surgery, an earlier decision for surgery and a shorter wound healing time in the LDI group (16.0 versus 19.9 days, p=0.022) was found. With our study results we were able to show significant improvements in burn care treatment, however, we were not able to present differences in costs between the LDI and control group. We expect that therapeutic decision-making could



improve even further when clinicians will have fully incorporated the LDI in daily practice, and this will eventually lead to cost savings by earlier surgery and a subsequent decreased length of stay (LOS). We also can conclude that LDI is especially useful in admitted patients, only a few outpatients met the inclusion criteria of our trial and in these patients no differences were seen between randomisation groups. All three centres that participated in this study bought the LDI afterwards and use it regularly, which can be considered as a great result of our study. The LDI was bought even before the study results were available. Apparently, clinicians considered the LDI to improve quality of care even before final study results were presented.

Improving burn care efficiency can be pursued by changing specialised burn care, but also by changing the organisation of burn care. In the Netherlands, patients with burn injuries can be treated by their general practitioner, in general hospitals, or in one of the three burn centres. In the last decades, the number of patients admitted to the burn centres increased, while the mean TBSA of the admitted patients decreased7,10. This is probably the result of changes in referral criteria in the late nineties10. The introduction of telemedicine is a potential way to improve burn care efficiency and prevent unnecessary referrals. As mentioned before, the initial assessment of burn size and depth is important, but difficult and depends on experience of the clinician. To improve burn depth assessment, we introduced LDI in the burn centres. Outside the burn centres, in e.g. in general hospitals, only a few patients with burns are seen monthly. It is not efficient to introduce LDI in these hospital because the necessary experience with scanning and interpretation cannot be generated with a low frequency of burn injuries. Next to that, the price of an LDI scan per patient becomes far higher, when this instrument is used only a few times a year. So, to improve burn diagnosis in general hospitals, on accident sites and by general practitioners we should search for other possibilities.

In this thesis we investigated telemedicine as an instrument to improve burn diagnosis by using photographs judged by experts. With our study, we were able to give a more nuanced view on the diagnostic abilities of the most simple and least costly form of telemedicine. Photographs can be used accurately to assess burn size, but not burn depth by burn experts, in contrast to several previous studies11-13. Therefore, the search for an ideal form of telemedicine continues. In the Netherlands, recently ‘Teleburn’ was introduced: a smartphone with direct connection to burn experts, used in trauma helicopters and several emergency departments14. This live interactive video method gives the opportunity of an extensive anamnesis and testing of capillary refill and sensibility15. A this point, we do not have insight in the effectiveness of teleburn yet.

R1R2R3R4R5R6R7R8R9


198 | Chapter 10

Recommendations for future studiesIt would be interesting to evaluate diagnostic and therapeutic decisions in our burn centres again in a few years. The learning curve of LDI introduction will be definitely not be a problem anymore by then, and possibly, OR logistic will have changed, enabling even more efficient surgical treatment. Next to that, a long-term follow up of scar quality of the patients included in our LDI trial should be obtained, in order to gain more insight in the influence of different treatment strategies on scar quality of wounds of which we know the exact burn depth. The effects of the introduction of LDI depend on local treatment preferences and protocols. It would be useful to perform an RCT in burn centres with a treatment preference of early excision and grafting, to analyse if the introduction of LDI also improves therapeutic choices in this treatment setting, and if it will prevent unnecessary surgery, as described in the retrospective cohort study of Petrie et al.16.

We recommend the use of LDI in future intervention studies as a gold standard for burn depth. In studies on, for example, new topical treatment options it is important to know the precise burn depth before applying new interventions and draw conclusions on effectiveness. LDI has obvious advantages over the current gold standard: biopsy17, because it is not invasive and is used for the whole burn wound, instead of a small part. Next to LDI, it is good to keep interest in other new techniques developed for the diagnosis of burn depth. LDI also has disadvantages, the device is large and difficult to position, the scanning takes several minutes, and it can be accurately used between 48 hours and 5 days only18. Recently, a new Laser Doppler Line Scanner (LDLS) was brought to the market, which can rapidly scan small wounds as accurate as the ‘conventional’ LDI, which seems to be more ideal for younger patients19-20. Maybe it is possible to reach even better results with other techniques that measure dermal vasculature, e.g. by near infrared spectroscopy, and laser speckle imaging18. These techniques have to be investigated further and validated before we can use them in practice and research. At this point, we recommend LDI to be used in practice.

To further improve efficiency of burn care outside the walls of the burn centres, it is important to further explore the possibilities of telemedicine in burn care. We recommend to conduct a study in which the clinimetrics and effectiveness of Teleburn is evaluated, followed by an economic evaluation. We expect that a live interactive video connection, like Teleburn, can improve initial burn care and be cost-effective as well by the prevention of unnecessary referrals, and an unnecessary delay of optimal treatment.



Part III. Improving burn care efficiency: therapeutics and costs

Part III of this thesis showed that the optimal timing of burn surgery is still not clear and that in burn surgery dermal substitutes can be used effectively without significant impact on total costs.

Outcomes and implicationsIn our Cochrane review we concluded that there was no single definition of early excision and grafting, nor sufficient evidence on the effectiveness of early excision and grafting on scar quality. Unclear definitions complicate the debate on early vs. delayed excision: one study considered early excision to be all procedures before the natural separation of eschar, which can take weeks21, another study defined early excision as surgery within 24 hours post-burn22, and others described several time points between these two extremes22,23. It is possible that treatment preferences are more alike than realised. Current performers of delayed excision and grafting probably hardly ever wait until natural separation of eschar, while performers of early excision and grafting do not operate all burns within 24 hours. In severe, clearly full thickness burns, probably a similar timing of surgery, within the first days post burn, is applied. Does earlier excision and grafting lead to an improved scar quality and functional outcome as described in Engrav et al. and Maslaukas et al.24,25? This probably depends on the depth of burn wounds in which early excision and grafting is performed. In full thickness wounds, most burn experts agree that early excision and grafting is preferred to decrease mortality and morbidity, as proposed by Dr. Zora Janzekovic in the 1970s26. In partial thickness wounds, however, it is more difficult to determine which wounds will benefit most from early excision and grafting. With the help of LDI, we are better equipped to assess burn depth, and choose which wounds should be operated surgically.

In our search for improved healing quality, we also performed a cost-effectiveness analysis on the use of dermal substitutes and skin grafts under topical negative pressure. Because of subtle differences in clinical effectiveness (scar elasticity), a cost-minimisation analysis was performed: total costs between intervention groups did not differ significantly, and the intervention costs contributed only to a small part (7%) of total burn care costs. Thus, the assumption that the use of dermal substitutes leads to cost raises could not be confirmed. In the implementation of dermal substitutes in burn care clinical effectiveness should therefore be decisive instead of costs. The effectiveness, in the form of improved scar quality, of dermal substitutes was not clearly shown in the trial described in this thesis, but was demonstrated before in both acute and reconstructive burn wounds27-29. Therefore, we believe that dermal substitutes should get a role in daily practice of specific acute burns (for example in functional areas) and reconstructive wounds.

R1R2R3R4R5R6R7R8R9


200 | Chapter 10

Recommendations for future studiesIn order to gain more insight in the actual effects of early excision and grafting on scar quality, the need for reconstructive surgery, and costs, we recommend to perform new prospective randomised controlled studies. First, for such a trial early excision and grafting should be clearly defined. We recommend to define early excision and grafting as excision within one week, in agreement with our Cochrane review. To accurately assess the burn depth of wounds evaluated in such a trial, we recommend the use of LDI. We hypothesize that early excision and grafting leads to an earlier wound healing, decreased LOS, improved scar quality and decreased costs. One of the current arguments against early excision and grafting is the fear for overzealous excision30. We expect that the introduction of LDI can be part of the solution; by accurately assessing burn depth within 2-5 days post burn, it seriously decreases the risk of unnecessary surgery17.

Therapeutic decisions in heterogeneous burn wounds remain problematic, for example, in wounds with a predicted healing time varying from 14-21 days to >21 days. To reach an optimal scar quality, should surgery be performed early on those wounds, or should be waited for spontaneous healing of some parts of the wounds and perform (early or delayed) surgery on the non-healing parts? We recommend to design new RCT’s to gain more insight in this subject. In these RCT’s scar quality should be the primary outcome, the population should consist of patients with mixed wounds with healing times varying from 14-21 day to >21 days and the intervention early surgery (within one week) vs. a delayed approach.

In our trial we could not convincible prove the effectiveness of dermal substitutes. But in earlier studies improvement of scar quality by dermal substitutes was shown27,28. We recommend to continue the search for the optimal dermal substitute, as currently is done in, for example the Euroskingraft project by Reichmann et al.31, and by the group of Steven Boyce et al.32. Probably, for the optimal skin substitute, patients’ own cells are necessary, e.g. by using undifferentiated proliferating cells on carriers. Promising results are shown with cell-based burn wound treatments in experimental studies and animal models33, but clinical effectiveness has to be further explored.

Concluding remarks of this thesis

In burn care, shown to be expensive care in this thesis, economic evaluations are still scarce. In this thesis we extended our insight in burn care costs and performed several economic evaluations. To further improve burn care efficiency we should focus on both quality of care and costs. The results of our cost studies suggest that to improve efficiency of burn care from



a cost perspective, we should search for ways to improve care with a limited length of hospital stay, an early return to work and a low requirement for (reconstructive) surgery. Attempts like this should, of course, always be balanced against quality of care. Carefully performed economic evaluations provide essential insights for the implementation of new interventions. With carefully performed economic evaluations we mean economic evaluations with a broad perspective, including the most important costs items of burn care, in order to generate thorough recommendations on the cost-effectiveness of new diagnostic and treatment options in burn care. This is exactly what we did in our first economic evaluation on dermal substitutes in combination with topical negative pressure therapy. Given the increasing healthcare expenditures and the limited healthcare budgets, taking into account cost-effectiveness data in guideline development seems appropriate. Therefore, more studies should be performed investigating the cost-effectiveness of burn care.

Next to costs, quality of burn care should be improved to optimise burn care efficiency. By optimising diagnostics and therapeutics in burn care, we aim to eventually reach an optimal scar quality and quality of life in patients after burn injuries. We successfully introduced LDI in Dutch burn care and showed its’ benefits on burn depth assessment and therapeutic decision-making. An early, accurate burn depth diagnosis plays a key role in efficient burn care. It facilitates an early decision on operative or non-operative treatment, which can improve treatment of both centres with a preference for early excision and grafting or for delayed excision and grafting. Next to that, LDI can be used as a gold standard for different kind of studies in which burn depth is measured, such as shown in our study on telemedicine. We expect that telemedicine will be a cost-effective manner to improve burn diagnosis and care outside the burn centre walls and strongly recommend to further analyse the effects of the use of telemedicine in burn care. Finally, in this thesis we made steps in the search for optimal healing quality after burn injuries. We showed that one out of eight patients requires reconstructive surgery in the 10-years after initial burn centre admittance. We wish to decrease the need for reconstructions in future by accomplishing better healing quality after burn injuries. The use of dermal substitutes can contribute to this goal. Hopefully, in the near future, we will find more and even better techniques to reach the most optimal scar quality in patients who have to live with the consequences of burn injuries.In conclusion, to improve burn care efficiency both evaluation of effects and costs of interventions are required. In burn diagnostics and treatment on-going advancements are achieved, but economic evaluation studies remain scarce. In this thesis we were able to extend our cost insights of patients with burns and to produce usable reference prices for future cost studies. Furthermore, we evaluated several upcoming and promising diagnostics and therapeutics in burn care and were able to assess their cost-effectiveness, which is inevitable to introduce new interventions in the current economic climate with limited budgets

R1R2R3R4R5R6R7R8R9


202 | Chapter 10

and increasing expenditures. Therefore, we recommend to always consider the inclusion of cost-effectiveness analyses in future intervention studies in burn care.



References

1. Hakkaart-van Roijen L, Tan S, Bouwmans C. Handleiding voor kostenonderzoek. Methoden en standaard kostprijzen voor economische evaluaties in de gezondheidszorg. Geactualiseerde versie 2010.


3. Sánchez JL, Perepérez SB, Bastida JL, Martínez MM. Cost-utility analysis applied to the treatment of burn patients in a specialized center. Arch Surg. 2007 Jan;142(1):50-7.

4. Sheckter CC, Van Vliet MM, Krishnan NM, Garner WL. Cost-effectiveness comparison between topical silver sulfadiazine and enclosed silver dressing for partial-thickness burn treatment. J Burn Care Res. 2014 Jul-Aug;35(4):284-90.

5. Verbelen J, Hoeksema H, Heyneman A, Pirayesh A, Monstrey S. Aquacel(®) Ag dressing versus Acticoat™ dressing in partial thickness burns: a prospective, randomized, controlled study in 100 patients. Part 1: burn wound healing. Burns.

6. Liu H, Xiao S, He Q, Zhou P, Liu F, Peng H. [Cost-effect analysis of 2 therapeutic regimens for deep II burn patients]. Zhong Nan Da Xue Xue Bao Yi Xue Ban. 2012 Nov;37(11):1112-6.

7. Dokter J, Vloemans AF, Beerthuizen GI, van der Vlies CH, Boxma H, Breederveld R, Tuinebreijer WE, Middelkoop E, van Baar ME; Dutch Burn Repository Group. Epidemiology and trends in severe burns in the Netherlands. Burns. 2014 Nov;40(7):1406-14.

8. Draisma JA Brandwonden Ongevalscijfers. Amsterdam Veiligheid NL, May, 2014.9. Husereau D, Drummond M, Petrou S, Carswell C, Moher D, Greenberg D, Augustovski F, Briggs

AH, Mauskopf J, Loder E; CHEERS Task Force. Consolidated Health Economic Evaluation Reporting Standards (CHEERS) statement BMJ. 2013 Mar 25;346:f1049.

10. Vloemans AF, Dokter J, van Baar ME, Nijhuis I, Beerthuizen GI, Nieuwenhuis MK,Kuijper EC, Middelkoop EM. Epidemiology of children admitted to the Dutch burn centres. Changes in referral influence admittance rates in burn centres. Burns.2011 Nov;37(7):1161-7.

11. Boccara D, Chaouat M, Uzan C, et al. Retrospective analysis of photographic evaluation of burn depth. Burns. 2011 Feb;37(1):69-73.

12. Jones OC, Wilson DI, Andrews S. The reliability of digital images when used to assess burn wounds. J Telemed Telecare. 2003;9(1):S22-4.

13. Jones O. Measurements of the clinical competence of doctors and nurses to process telemedicine referrals for burns patients. J Telemed Telecare. 2005;11(1):89-90.

14. www.teleburn.org15. Wallace DL, Hussain A, Khan N, Wilson YT. A systematic review of the evidence for telemedicine in

burn care: with a UK perspective. Burns. 2012 Jun;38(4):465-80.16. Petrie N, Norbury W, Fogarty B, Philip B, Baraat J, Dziewulski P. The use of the laser Doppler

imaging to reduce operative intervention in the treatment of paediatric burns. Abstract Book, International Society for Burn Injuries 2004;(Abstract Book, International Society for Burn Injuries).

17. Monstrey S, Hoeksema H, Verbelen J, Pirayesh A, Blondeel P. Assessment of burn depth and burn wound healing potential. Burns. 2008 Sep;34(6):761-9.

18. Kaiser M, Yafi A, Cinat M, Choi B, Durkin AJ. Noninvasive assessment of burn wound severity using optical technology: a review of current and future modalities. Burns. 2011 May;37(3):377-86.

19. Hoeksema H, Baker RD, Holland AJ, Perry T, Jeffery SL, Verbelen J, Monstrey S. A new, fast LDI for assessment of burns: a multi-centre clinical evaluation. Burns. 2014 Nov;40(7):1274-82.

20. Holland AJ, Ward D, La Hei ER, Harvey JG. Laser doppler line scan burn imager (LDLS-BI): sideways move or a step ahead? Burns. 2014 Feb;40(1):113-9.

21. Choi M, Panthaki ZJ. Tangential excision of burn wounds. J Craniofac Surg. 2008 Jul;19(4):1056-60.

22. Ong YS, Samuel M, Song C. Meta-analysis of early excision of burns. Burns. 2006 Mar;32(2):145-50. Epub 2006 Jan 18. Review.

23. Kirn DS, Luce EA. Early excision and grafting versus conservative management of burns in the elderly. Plast Reconstr Surg. 1998 Sep;102(4):1013-7.

R1R2R3R4R5R6R7R8R9


204 | Chapter 10

24. Engrav LH, Heimbach DM, Reus JL, Harnar TJ, Marvin JA. Early excision and grafting vs. nonoperative treatment of burns of indeterminant depth: a randomized prospective study. J Trauma. 1983 Nov;23(11):1001-4.

25. Maslauskas K, Rimdeika R, Rapoliene J, Ramanauskas T. Analysis of burned hand function (early versus delayed treatment). Medicina. 2005;41(10):846-51.

26. Janzekovic Z. A new concept in the early excision and immediate grafting of burns. J Trauma. 1970 Dec;10(12):1103-8.

27. van Zuijlen PP, Van Trier AJ, Vloemans JF, Groenevelt F, 468 Kreis RW, Middelkoop E. Graft survival and effectiveness of 469 dermal substitution in burns and reconstructive surgery in 470 a one-stage grafting model. Plast Reconstr Surg 471 2000;106:615–23.

28. Bloemen MC, van Leeuwen MC, van Vucht NE, van Zuijlen, Middelkoop E. Dermal substitution in acute burns and 494 reconstructive surgery: a 12-year follow-up. Plast Reconstr 495 Surg 2010;125:1450–9.

29. Haslik W, Kamolz LP, Manna F, Hladik M, Rath T, Frey M.J. Management of full-thickness skin defects in the hand and wrist region: first long-term experiences with the dermal matrix Matriderm. Plast Reconstr Aesthet Surg. 2010 Feb;63(2):360-4.

30. Hop MJ, Hoogewerf CJ, van Baar ME, van der Vlies CH, Middelkoop E. A call for evidence: timing of surgery in burns. Burns. 2012 Jun;38(4):617-8.

31. www.euroskingraft.eu32. Sander EA, Lynch KA, Boyce ST. Development of the mechanical properties of engineered skin

substitutes after grafting to full-thickness wounds. J Biomech Eng. 2014 May;136(5):051008.33. Gardien KL, Middelkoop E, Ulrich MM. Progress towards cell-based burn wound treatments. Regen

Med. 2014 Mar;9(2):201-18.

AddendumSummary

About the author

Bibliography

Dankwoord

R1R2R3R4R5R6R7R8R9


206 | Addendum


Summary | 207

Summary

In this thesis we searched for methods to optimise efficiency in burn care. Therefore, both healthcare and non-healthcare costs of burn care were studied in detail, and important cost items and predictors for high costs were identified. Furthermore, the effectiveness and cost-effectiveness of upcoming diagnostics and treatment options in burn care were evaluated. In this summary the main findings of the thesis will be presented.


In Part I of the thesis we aimed to identify both healthcare and non-healthcare costs of burn care and to find important cost items and predictors for high costs. Therefore we conducted a systematic review, a retrospective and a prospective cohort study. We demonstrated that burn care is associated with high costs. Both healthcare costs, especially hospital stay, and non-healthcare costs, like work absence, are substantial in the short- and long-term.In the systematic literature review on burn care costs (Chapter 2) we presented that burn care was in general more expensive than other injuries. The mean total healthcare cost per burn patient in high-income countries was €56,010. Significant cost categories in the cost review were hospital stay, followed by surgery. Also productivity loss and informal care seemed to be major cost categories, but were presented by one author only (in two articles of Sanchez et al.). In our prospective cohort study on burn care costs within 3 months post-burn (Chapter 3), conducted in a Dutch burn centre, mean costs per patient were €26,540 per patient, varying from €740 to €235,560, in a population with a mean burn size of 8% total body surface area. The major cost categories found in our literature review could be confirmed in the prospective cohort study; burn centre days and ICU days were the most important burn care cost categories, followed by work absence and surgical interventions. Burn centre stay accounted for 81% of the total specialised burn care costs, mainly caused by high burn centre day prices. In burn care, no clear costs prices for burn centre days were described until this thesis. In our prospective study, we calculated burn centre day prices in detail, with the help of annual accounts, which consisted of personnel, material, equipment, nutrition, medication, housing and overhead. We found that the costs of a non-ICU burn centre day were €948 and of a ICU burn centre day €2,966. We recommend using these day prices for future economic evaluations in high-income countries. In non-healthcare costs, work absence represented the main cost category. The mean costs per patient of work absence were even substantially higher than of surgery (€5,255 versus €1,298). Patient and injury characteristics associated with high costs were a higher age, flame burns and a higher %TBSA burned (max. 80%).

R1R2R3R4R5R6R7R8R9


208 | Addendum

Our retrospective study on reconstructive surgery after burn injuries (Chapter 4) showed that one out of eight patients needs reconstructive surgery after a burn injury. Mean number of reconstructive procedures per patient was 3.6 (range 1-25). Frequently reconstructed locations were hands and head/neck. The most important indication was scar contracture and the most applied technique was release plus random flaps/ skin grafting. Reconstructive surgery represents a significant cost category in burn care, with a mean of €8342 (range 928-70,067) per patient that undergoes reconstructive surgery in the first decade post burn. With the cost insights gained in the first part of this thesis, a baseline was established for the design of new cost-effectiveness studies in Dutch burn care, which are highly desirable because burn care is expensive care and there is an on-going need for improvement of burn diagnostics and treatment. We also recommend to further extend our insight in burn care costs in future, by designing cost studies in and outside burn centres with a long follow-up time (e.g. 5 years).

Part II. Improving burn care efficiency: diagnostics and costs

In Part II of this thesis different applications of the diagnostic instrument LDI were described: 1) LDI as an add-on test in burn depth diagnosis, and 2) LDI + clinical assessment as a gold standard to investigate the validity of photographic assessment of burn depth. In our RCT on the cost-effectiveness of LDI (Chapter 5&6) we showed that laser Doppler imaging is effective in Dutch burn care by improving therapeutic decision making, as early as 2-5 days post-burn: on the day of randomisation clinicians decided significantly more often on operative or non-operative treatment in the LDI group (p<0.001) instead of postponing their treatment choice. In a subgroup of admitted patients requiring surgery, an earlier decision for surgery and a shorter wound healing time in the LDI group (16.0 versus 19.9 days, p=0.022) was found. In our study we were able to show significant improvements in burn care treatment, however, we were not able to present differences in costs between the LDI and control group. We expect that therapeutic decision-making can even further improve when clinicians will have fully incorporated the LDI in daily practice, and this will eventually lead to cost savings by earlier surgery and a subsequent decreased length of stay (LOS), with the potential to save € 875 per patient. All three centres that participated in this study bought the LDI afterwards and use it regularly, which can be considered as a great result of our study.

In this thesis we also investigated the validity and reliability of telemedicine (by using photographs judged by experts) as an instrument to improve burn size and depth diagnosis (Chapter 7). With our study, we were able to give a more nuanced view on the diagnostic abilities of the most simple and least costly form of telemedicine. Photographs can be used


Summary | 209

accurately by burn experts to assess burn size, but not burn depth. This is in contrast to several previous studies. Because photographic assessment showed to have limitations, the search for an ideal form of telemedicine continues. In the Netherlands, recently ‘Teleburn’ was introduced: a smartphone with direct connection to burn experts, used in trauma helicopters and several emergency departments. This live interactive video method gives the opportunity of an extensive anamnesis and testing of capillary refill and sensibility. A this point, we do not have insight in the effectiveness of Teleburn yet, but we recommend to research its’ effectiveness and cost-effectiveness.

Part III. Improving burn care efficiency: therapeutics and costs

Part III of this thesis showed that the optimal timing of burn surgery still is not clear and that in burn surgery dermal substitutes can be used effectively without significant impact on total costs.

Early excision and grafting vs. delayed excision and grafting; it remains a hot topic in burn care. An important finding of our Cochrane review (Chapter 8) was a very basic one: the definition of early excision and grafting in literature is still unclear. Unclear definitions complicate the debate on early vs. delayed excision. It is possible that treatment preferences are more alike than realised. Current performers of delayed excision and grafting probably hardly ever wait until natural separation of eschar, while performers of early excision and grafting do not operate all burns within 24 hours. In severe, clearly full thickness burns, probably a similar timing of surgery, within the first days post burn, is applied. Outcome measures of the studies included in our review differed widely, preventing us of pooling data. All together: no clear recommendations were derived from this review. There was insufficient data to support any definite conclusions on the effects of early excision and grafting on scar quality. Does earlier excision and grafting lead to an improved scar quality and functional outcome as described in Engrav et al. and Maslaukas et al.? This probably depends on the depth of burn wounds in which early excision and grafting is performed. In full thickness wounds, most burn experts agree that early excision and grafting is preferred to decrease mortality and morbidity, as proposed by Dr. Zora Janzekovic in the 1970s. In partial thickness wounds, however, it is more difficult to determine which wounds will benefit most from early excision and grafting. With the help of LDI, we are better equipped to assess burn depth, and choose which wounds should be treated surgically. We strongly recommend to all burn specialists to publish their experience with early vs. delayed grafting or even conduct a randomised controlled trial on this topic and use the LDI as an gold standard for burn depth assessment.

R1R2R3R4R5R6R7R8R9


210 | Addendum

In our search for an improvement of burn care efficiency, we elaborated on optimal burn depth diagnosis, the optimal selection of burn wounds requiring surgery, and the optimal timing of surgery in this thesis. Another possible way to improve healing quality is optimisation of surgical techniques. In this thesis we performed a cost-effectiveness analysis on the use of dermal substitutes and skin grafts under topical negative pressure (Chapter 9). Because of subtle differences in clinical effectiveness (scar elasticity), a cost-minimisation analysis was performed: total costs between intervention groups did not differ significantly, and the intervention costs contributed only to a small part (7%) of total burn care costs (€2912 in the intervention group and €1703 in the control group: split skin graft alone). Thus, the assumption that the use of dermal substitutes leads to cost raises could not be confirmed. In the implementation of dermal substitutes in burn care clinical effectiveness should therefore be decisive instead of costs. The effectiveness, in the form of improved scar quality, of dermal substitutes was not clearly shown in the trial described in this thesis, but was demonstrated before in both acute and reconstructive burn wounds. Therefore, we believe that dermal substitutes should get a role in daily practice of specific acute burns (for example in functional areas) and reconstructive wounds.

ConclusionTo improve burn care efficiency both evaluation of effects and costs of interventions are required. In burn diagnostics and treatment on-going advancements are achieved, but economic evaluation studies remain scarce. In this thesis we were able to extend our cost insights of patients with burns and to produce usable reference prices for future cost studies. Furthermore, we evaluated several upcoming and promising diagnostics and therapeutics in burn care and were able to assess their cost-effectiveness, which is inevitable to introduce new interventions in the current economic climate with limited budgets and increasing expenditures. Therefore, we recommend to always consider the inclusion of cost-effectiveness analyses in future intervention studies in burn care.


About the author | 211

About the author

Martine Jenda Hop werd geboren op 6 december 1983 in Zwolle. Ze werd geboren in het ziekenhuis, waarmee haar eerste kennismaking met de medische wetenschap een feit was. Jenda groeide op in het Gelderse Wezep met haar ouders, zusje Erika en spaniël Buck. Ondanks een jarenlange bijbaan bij de Wezepse Boni, koos ze na haar eindexamen gymnasium op het Zwolse Agnieten College (locatie Carolus Clusius) voor een studie Geneeskunde aan de Rijksuniversiteit Groningen.

Tegen het einde van haar studie koos Jenda ervoor haar scriptie te schrijven bij het Brandwondencentrum van Groningen. Reconstructieve chirurgie leek haar bijzonder, en ze hoopte er op deze manier iets over te leren. Hier werd ze zo enthousiast, dat ze niet alleen besloot haar laatste coschap op de plastische chirurgie in Leeuwarden te doen, maar bovendien tot haar eigen verbazing een interesse in onderzoek had ontwikkeld. Dit kwam niet in de laatste plaats door haar scriptiebegeleider Marianne Nieuwenhuis. Van haar leerde Jenda dat als je echt iets wil weten of bereiken, daar ook een pad van onderzoek bij hoort. En dat het oké is dat dat soms taai is. Ze begon voorzichtig te flirten met de mogelijkheid tot promoveren, maar besloot zich eerst op een opleidingsplek tot plastisch chirurg te richten.

Na haar afstuderen ging ze aan de slag als ANIOS, eerst bij de algemene chirurgie in het Martini Ziekenhuis in Groningen, later bij de plastische chirurgie in Doetinchem. In 2010 kon ze opnieuw aan de slag op de plastische chirurgie in Leeuwarden, nadat de ene na de andere arts er in zwangerschap was uitgebarsten. Daar waren ze zo over haar te spreken dat ze haar een opleidingsplek aanboden. Die plek kwam alleen pas in 2013 beschikbaar. Door een gelukkige combinatie van timing, talent en heel hard werken verwierf ze voor de tussentijd een promotieplek in het Brandwondencentrum van het Maasstad Ziekenhuis Rotterdam, onder promotor Esther Middelkoop. Na het promotietraject dat naast bijzonder interessant en leerzaam, inderdaad soms ook best taai was, werkt Jenda sinds december 2013 met veel enthousiasme als AIOS plastische chirurgie in Leeuwarden, waar ze samenwoont met Marius.

R1R2R3R4R5R6R7R8R9


212 | Addendum


Bibliography | 213

Bibliography

Hop MJ, Polinder S, van der Vlies CH, Middelkoop E, van Baar ME. Costs of burn care: a systematic review. Wound Repair Regen. 2014 Jul-Aug;22(4):436-50.

Hop MJ, Langenberg LC, Hiddingh J, Stekelenburg CM, van der Wal MB, Hoogewerf CJ, van Koppen ML, Polinder S, van Zuijlen PP, van Baar ME, Middelkoop E. Reconstructive surgery after burns: a 10-year follow-up study. Burns. 2014 Dec;40(8):1544-51.

Hop MJ, Moues CM, Bogomolova K, Nieuwenhuis MK, Oen IM, Middelkoop E, Breederveld RS, van Baar ME. Photographic assessment of burn size and depth: reliability and validity. J Wound Care. 2014 Mar;23(3):144-5, 148-52.

Hop MJ, Bloemen MC, van Baar ME, Nieuwenhuis MK, van Zuijlen PP, Polinder S, Middelkoop E; TOPSKIN Study Group. Cost study of dermal substitutes and topical negative pressure in the surgical treatment of burns. May;40(3):388-96.

Hoogewerf CJ, van Baar ME, Hop MJ, Bloemen MC, Middelkoop E, Nieuwenhuis MK. Burns to the head and neck: Epidemiology and predictors of surgery. Burns. 2013 Sep;39(6):1184-92.

Hop MJ, van Baar ME, Nieuwenhuis MK, Dokter J, Middelkoop E, van der Vlies CH. Bepaling van brandwondendiepte Klinische inschatting en laser-doppler-imaging. WCS Nieuws Tijdschrift. 2013 Mar; 29(1):38-41.

Hop MJ, Hiddingh J, Stekelenburg C, Kuipers HC, Middelkoop E, Nieuwenhuis MK, Polinder S, van Baar ME; LDI study group. Cost-effectiveness of laser Doppler imaging in burn care in the Netherlands. BMC Surg. 2013 Feb 1;13:2.

Hoogewerf CJ, Van Baar ME, Hop MJ, Nieuwenhuis MK, Oen IM, Middelkoop E. Topical treatment for facial burns. Cochrane Database Syst Rev. 2013 Jan 31;1:CD008058.

Hop MJ, van Baar ME, Nieuwenhuis MK, Dokter J, Middelkoop E, van der Vlies CH. Determining burn depth: clinical assessment and laser Doppler imaging. NedTijdschr Geneeskd. 2012;156(31):A4810.

Hop MJ, Hoogewerf CJ, van Baar ME, van der Vlies CH, Middelkoop E. A call for evidence: timing of surgery in burns. Burns. 2012 Jun;38(4):617-8.

R1R2R3R4R5R6R7R8R9


214 | Addendum

Hop MJ, van der Biezen JJ. Ray reduction of the foot in the treatment of macrodactyly and review of the literature. J Foot Ankle Surg. 2011 Jul-Aug;50(4):434-8.

Hop MJ, Nieuwenhuis MK, Beerthuizen GIJM. Epidemiologie van gelaatsverbrandingen in het brandwondencentrum in Groningen. Nederlands Tijdschrift voor Plastische Chirurgie 2011;1:9-13.


Dankwoord | 215

Dankwoord

Het is moeilijk niet in clichés te vervallen bij het schrijven van een dankwoord. Toch wil ik graag meerdere mensen bedanken, dus er zit niet anders op dan de clichés te omarmen.

Allereerst wil ik graag alle patiënten bedanken die deel hebben genomen aan de diverse studies. Ik was blij verrast door de bereidwilligheid van veel patiënten (en/of hun ouders). Zonder jullie hulp was dit proefschrift niet tot stand gekomen!

Er zijn veel collega’s met wie ik met plezier heb samengewerkt en die ik graag wil bedanken. Als eerste uiteraard, Margriet, copromotor. Dank voor je geweldige begeleiding: ik kreeg veel vrijheid van je, maar je was ook altijd bereikbaar voor hulp, een heel prettige combinatie. Ik heb met veel plezier met je gewerkt en van je geleerd.Esther, net als Kees was ik een promovendus van je ‘op afstand’, onder de hoede van Margriet. Ik heb veel bewondering voor de manier waarop jij je leerstoel invult en ging na onze afspraken in Beverwijk altijd met nieuwe inzichten weg. Dank hiervoor.Suzanne, net als Margriet, mijn copromotor. Jij tilt ‘efficiency’ naar een hoger level met super snelle reacties, die altijd inhoudelijk scherp zijn. Ik vind je een inspirerende wetenschapper.

Kees H., mijn partner in crime. Jij was al even bezig met je promotie onderzoek, toen ik startte in Rotterdam, en daardoor toch wel een beetje mijn voorbeeld (dat kan ik nu wel toegeven). Ik waardeer je persoonlijkheid en humor zeer en denk dat we elkaar op vakgebied goed aanvullen. Dank voor een mooie tijd!Jan, Irma en Kees v.d. V., dank dat ik met jullie een kamer mocht delen. Naast gezelligheid en goede koffie, heb ik veel over het reilen en zeilen in de kliniek van jullie mogen leren. Helma, ik vond het fijn met je te mogen werken, behalve kennis over brandwonden en onderzoek, ook veel andere zaken van je geleerd! Nicole, tijdens mijn promotietraject maakte jij de switch van IC- naar onderzoeksverpleegkundige en dat deed je enthousiast en grondig, je hebt veel werk voor de onderzoeken in dit proefschrift verricht. Dank hiervoor!Jakob, ik leerde je tijdens mijn scriptie kennen en was toen al fan. Wat een feest dat ik nu vanuit R’dam weer met je mocht samenwerken. Bedankt!Onderzoekcollega’s uit Beverwijk en Groningen, dank voor al jullie inspanningen rondom o.a. de LDI studie, dankzij al jullie is het een prachtige multicenter studie geworden.Alle andere collega’s uit het brandwondencentrum van Rotterdam. Jullie zijn een prachtig, hardwerkend team en hebben mij vanaf het begin een welkom gevoel gegeven. Dank jullie wel!

R1R2R3R4R5R6R7R8R9


216 | Addendum

Beste paranimfen, als eerste: papa. Sinds ik kan praten, vraag ik jou mij uit te leggen hoe de wereld in elkaar zit, dit tot grote irritatie van mijn kleine zusje (toch papa?). Het is voor mij dan ook niet meer dan logisch om jou tijdens de verdediging van mijn proefschrift aan mijn zijde te hebben. Beste Marianne, andere paranimf, naast mijn vader kan ik geen meer geschikte persoon bedenken dan jij. Jij was destijds mijn scriptiebegeleider en tijdens deze periode ontdekte ik tot mijn eigen, en volgens mij ook tot jouw, verbazing dat wetenschappelijk onderzoek eigenlijk heel leuk is. In deze periode is het zaadje geplant: ooit zou ik gaan promoveren. Nu dat ook echt gaat lukken, moet jij natuurlijk naast me staan!

Tenslotte, lieve Marius, samen met jou kan ik alles. Dank voor je geduld, aanmoediging en liefde de afgelopen jaren. Op onze toekomst samen!


M. Jenda Hop


M.J. H

op

Uitnodiging

voor het bijwonen van de openbare verdediging

van het proefschrift

‘Improving burn care efficiency’

door M. Jenda Hop

donderdag 12 november

11:45

in de aula Vrije Universiteit Amsterdam

De Boelelaan 1105

na afloop bent u van harte welkom

op de receptie ter plaatse

paranimfen: Rende Hop

[email protected]

en

Marianne [email protected]

Ook wil ik u graag uitnodigen voor het promotiefeest vanaf 20:30 in Pont 13

Haparandadam 501013 AK Amsterdam

Proefschrift Hop

Documents

Transcript of Proefschrift Hop