Cost of treating diabetic foot ulcers in five different countries

Transcription

1 DIABETES/METABOLISM RESEARCH AND REVIEWS Diabetes Metab Res Rev 2012; 28(Suppl 1): Published online in Wiley Online Library (wileyonlinelibrary.com).2245 REVIEW ARTICLE Cost of treating diabetic foot ulcers in five different countries Peter Cavanagh 1 * Christopher Attinger 2 Zulfiqarali Abbas 3,4 Arun Bal 5 Nina Rojas 6 Zhang-Rong Xu 7 1 Department of Orthopaedics and Sports Medicine, The University of Washington, Seattle, USA 2 Georgetown University, Washington, DC, USA 3 Muhimbili University of Health and Allied Sciences, Dar es Salaam, Tanzania 4 Abbas Medical Centre, Dar es Salaam, Tanzania 5 All India Institute of Diabetes Mumbai, Mumbai, India 6 Hospital Regional of Concepción, Concepción, Chile 7 306th Hospital Beijing, Beijing, China *Correspondence to: Peter R. Cavanagh, Department of Orthopaedics and Sports Medicine, The University of Washington Medical Center, Box , 1959 NE Pacific Street, Seattle, WA 98195, USA. cavanagh@uw.edu Abstract Most estimates in the literature for the economic cost of treating a diabetic foot ulcer (DFU) are from industrialized countries. There is also marked heterogeneity between the complexity of cases considered in the different studies. The goal of the present article was to estimate treatment costs and costs to patients in five different countries (Chile, China, India, Tanzania, and the United States) for two hypothetical, but well-defined, DFUs at the extreme ends of the complexity spectrum. A co-author, who is a treating physician in the relevant country, was asked to choose treatment plans that represented the typical application of local resources to the DFU. The outcomes were pre-defined as complete healing in case 1 and trans-tibial amputation in case 2, but the time course of treatment was determined by each investigator in a manner that would be typical for their clinic. The costs, in local currencies, for each course of treatment were estimated with the assistance of local hospital administrators. Typical reimbursement scenarios in each country were used to estimate the cost burden to the patient, which was then expressed as a percentage of the annual per capita purchasing power parity-adjusted gross domestic product. There were marked differences in the treatment plans between countries based on the availability of resources and the realities of local conditions. The costs of treatment for case 1 ranged from Int$102 to Int$3959 in Tanzania and in the United States, respectively. The cost for case 2 ranged from Int$3060 to Int$ in Tanzania and in the United States, respectively. The cost burden to the patient varied from the equivalent of 6 days of average income in the United States for case 1 to 5.7 years of average annual income for case 2 in India. Although these findings do not take cost-effectiveness into account, they highlight the dramatic economic burden of a DFU for patients in some countries. Copyright 2012 John Wiley & Sons, Ltd. Keywords cost; foot ulcer; health economics; comparative medicine; reimbursement Introduction Received: 12 September 2011 Revised: 8 October 2011 Accepted: 11 October 2011 The present knowledge of the costs of treating a diabetic foot ulcer (DFU) is drawn primarily from industrialized countries, notably the United States, Sweden, and The Netherlands [1 9] (Table 1). Direct cost estimates (in 2010-adjusted US dollars) range from to US$3096 [4] for a Wagner grade 1lesiontoUS$107900foranulcerresultinginamputation[1].Thecosts reported are usually those billed (or sometimes fees collected) by the treating institution, but the cost burden to the patient is rarely reported. An examination of the cost of treating the same condition in different countries can be instructive as it provides insight into the effect of different health-care and reimbursement models on the patient s economic responsibility for treatment. Thus, the goal of the current study is to estimate the Copyright 2012 John Wiley & Sons, Ltd.

2 108 P. Cavanagh et al. Table 1. Summary of the costs of healing diabetic foot ulcers from the literature. All costs are in US$ adjusted for inflation as of 2010 Author Year Country Setting Costs (US$) Apelqvist et al. [9] 1990 Sweden University hospital without critical ischaemia with critical ischaemia with minor amputation with major amputation Holzer et al. [8] United States Private employer-sponsored insurance plans 3096 Wagner Grade 1 or Wagner Grade 3 or Wagner Grade 5 Ramsey et al. [7] United States Large HMO in attributable direct costs for 2-year post diagnosis Harrington et al. [6] United States US Medicare population in attributable direct costs Horswell et al. [5] 2003 United States Public hospital (lower costs in staged management program versus standard care) Ragnarson Tennvall Sweden Comprehensive review article direct costs and Apelqvist [4] van Houtum et al. [3] 1992 The Netherlands National database total direct costs for a single amputation and average length of hospital stay Prompers et al. [2] EU 14 European diabetic foot centres (Eurodiale) in total costs (direct and indirect) for a foot ulcer Hunt et al. [1] 2011 United States Identified surgical intervention, length of stay in the hospital, and intensive care unit time as drivers of the treatment cost HMO, Health Maintenance Organization. cost to the patient and to the institution of treating simple and complex DFUs in five different countries with widely varying health-care practices, reimbursement policies, and gross domestic products (GDPs). We also plan to apply econometric adjustments so that the costs can be meaningfully compared across countries. Methods Two well-defined hypothetical DFUs at the extreme ends of the complexity spectrum were developed (Table 2). Case 1 is an uncomplicated 4-cm 2 plantar ulcer of 6 weeks duration under the first metatarsal head in a well-controlled diabetic, neuropathic, non-smoking, 55-year-old woman with normal renal function and a body mass index of 28. Her ankle-brachial index (ABI) is between 0.8 and 1.0; there are no soft-tissue defects, clinical signs of infection, prior history of amputation, or ankle involvement. The ulcer is classified as PEDIS ([1/4 cm 2 /1/1/2], Wagner grade 1, Texas grade 1, stage A) [10 12], and full healing occurred in 8 weeks. Case 2 is a 20-cm 2 plantar ulcer of 6 months duration under the heel of a poorly controlled 75-year-old man who smokes and has a body mass index of 40. He has an ABI <0.7 and stage 3 chronic kidney disease with a creatinine >220 mm. The wound shows signs of deep infection, and there is a soft-tissue defect at the ankle that is believed to be infected. The patient has a contralateral trans-tibial amputation and an ipsilateral first-ray amputation. The ulcer is classified as PEDIS ([3, 20 cm 2, 3, 4, 2], Wagner grade 3, Texas stage D). The ipsilateral limb was eventually amputated. Each investigator was asked to choose locally relevant treatment options for each patient from the following palate of possible interventions: CT scan, dermal replacement, external fixation, hospitalization, hyperbaric oxygen, intravenous antibiotics, MRI scan, post-healing off-loading footwear, operating-room debridement, oral antibiotics, orthopaedic consult, outpatient clinic visits, outpatient debridement, plain X-ray, prosthesis, rehabilitation visits, renal consult, re-vascularization, soft-tissue flap, total contact cast/walker, trans-tibial amputation, and vascular consult. The time course and staging of the treatment plan was determined by each investigator in a manner that would be typical for their clinic. With the help of their administrators, each investigator also provided actual direct costs to the institution in local currencies for each treatment component. It shouldbenotedthateachinstitutionwasinanurban area, and treatment practices and costs vary considerably in rural areas of the same country. Investigators also supplied information on the most typical local reimbursement policy, and this was applied to calculate patient cost burden. The assumptions made in each country were as follows: In Chile, where approximately 70% of the population is covered by a government insurance plan, a 0 40% co-pay above the fixed fee for each condition paid by insurance is typical. We assumed the worst case of 40%. In urban China, we assumed that health insurance will typically reimburse 90% of the treatment cost with 10% to be paid by the patient. Because only approximately 5% of people in urban India have health insurance, we assumed 100% of the cost was borne by the patient.

3 Cost of Treating Diabetic Foot Ulcers in Five Countries 109 Table 2. Characteristics of the two hypothetical patients and their wounds Wound 1 Wound 2 Age Body Mass Index Sex Female Male Smoker No Yes Location MTH1 Heel HbA 1c 8.5% 11% Lesion size 4 cm 2 20 cm 2 Wagner grade Grade 1 (partial/full thickness ulcer) Grade 3 (deep with osteitis) Texas grade Grade 1 (superficial wound not involving tendon, capsule, or bone) stage A (clean wounds) Stage D (ischaemic infected wounds) PEDIS grade Perfusion 1 3 Extent/size 4 cm 2 20 cm 2 Depth/tissue loss 1 3 Infection 1 4 Sensation 2 2 Neuropathy LOPS LOPS Ischaemia ABI >0.8 ABI <0.7 Infection No Yes Soft-tissue defect No Yes Duration of wound 6 weeks 6 months Prior ipsilateral amputation No Yes loss of first ray Prior contralateral amputation No Yes trans-tibial amputation Renal status No symptoms: microalbuminuria (UAE <20 mg/min) Ankle involvement None Suspected infection ABI, ankle-brachial index; LOPS, Loss of Protective Sensation; UAE, urinary albumin excretion. Chronic kidney disease stage 3 creatinine >220 mm The patients in Tanzania were assumed to be treated at government-run urban clinics where they are responsible for 100% of the cost. In the United States, a hospital typically collects only 65% of costs billed, and we therefore used collections rather than the amounts billed to estimate the actual cost. Medicare patients (older than 65 years) with supplemental insurance typically have their costs covered (with some limitations), but we assumed a 20% co-pay to exemplify patients with no supplemental insurance. The aforementioned assumptions represent an oversimplification of five complex medical economies but were necessary for the purposes of the study. We made two cascading adjustments to generate costs to the patient that would be comparable across the five countries that represent an extremely wide range of national wealth and personal income. First, rather than simply using exchange rates for conversion of treatment costs, an adjustment for the difference in purchasing power of the local currency for citizens of each country was made using 2010 estimates of purchasing power parity (PPP) published by the World Bank [13]. PPP between two countries is defined as the rate at which the currency of one country needs to be converted into that of a second country to ensure that a given amount of the first country s currency will purchase the same volume of goods and services in the second country as it does in the first [14]. This is sometimes called the Market Basket Index, the Big Mac Index, or the Starbucks Latte Index to exemplify the different costs for the same items even after exchange-rate adjustment of local currencies. This results in a value in international dollars that compares the purchasing power in each country to that of the United States. Data for 2010 show these values to be Int$0.71, Int$0.53, Int$0.37, Int$0.37, and Int$1, for Chile, China, India, Tanzania, and the United States, respectively. Thus, Int$0.37 in India or in Tanzania would buy the same amount of goods and services as Int$1 in the United States. The second step was to convert the GDP of the five countries to international dollars, again to adjust for PPP[15]. On this scale, the five countries, the United States, Chile, China, India, and Tanzania, were ranked 9th, 49th, 94th, 138th, and 167th in the world, respectively. The division of this PPP-adjusted GDP by the population of each country results in an annual per capita value in international dollars that can be thought of as an index of the average annual earning and hence the ability to pay for treatment (Table 3). These values for Chile, China, India, Tanzania, and the United States are Int$16 659, Int$8268, Int$3419, Int$1481, and Int $47 248, respectively. Results The treatment plan for the same ulcer was not the same in any of the five countries. Notable in case 1 were the use of oral antibiotics in India and in Tanzania, 14 days of inpatient treatment in China, and the absence of off-loading footwear after healing in China and in Tanzania. The costs

4 110 P. Cavanagh et al. Table 3. Various indices and costs for cases 1 and 2 in the different countries Units Chile China India Tanzania United States Purchasing power parity Ratio Per capita PPP-adjusted GDP Int$ Case 1 Total treatment cost Int$ Patient co-pay % Patient co-pay Int$ Patient co-pay a Months of income Case 2 Total treatment cost Int$ Patient co-pay % Patient co-pay Int$ Patient co-pay a Month of income PPP, purchasing power parity; GDP, gross domestic product. a Calculated as [12* (patient co-pay in Int$/per capita PPP-adjusted GDP)]. for treatment of case 1 varied from Int$102 in Tanzania to Int$3959 in the United States, while the patient s contribution to the cost varied from Int$102 (Tanzania) to Int$1192 in India (Table 3). When expressed as a quotient of per capita PPP-adjusted GDP, the range represents between 1.7% (United States) and 34.9% (India) so that 6 days of average income in the United States and 127 days of average annual income in India would be required to pay the patient s share of the cost of treatment for this uncomplicated ulcer. The treatment plan for case 2 in the United States included a long and progressive attempt at limb salvage using all the available options mentioned earlier with the exception of outpatient debridement, a CT scan, dermal replacement, a total contact cast, and oral antibiotics. In Tanzania, the proposed treatment progressed quickly to trans-tibial amputation after inpatient debridement failed to heal the wound. The costs of treatment for case 2 varied dramatically from Int$3060 in Tanzania to Int$ in the United States. The range of patient costs implies that between 146 days (Chile) and 5.7 years (India) of average annual income would be required to pay for the patient s portion of the treatment cost. Discussion This hypothetical study of five different international clinics has generated both the lowest cost (Int$102: case 1 in Tanzania) and the highest cost (Int$ : case 2 in the United States) ever reported for treating a DFU. The maximum cost resulted from a prolonged, and ultimately failed, multifaceted attempt at limb salvage in a complex patient with a large infected wound and several comorbidities by an internationally known team from a major US academic medical centre. The cost to treat the same ulcer, with the same endpoint, in Tanzania was estimated as Int$3060, which is only 1.6% of the US cost. However, the burden for the patient is primarily not represented by the adjusted absolute cost but by the patient s responsibility for payment, which is a function of both insurance practices and annual per capita PPP-adjusted GDP. This leads to the finding that the most expensive country for a patient with a complex DFU is India, where 5.7 years (68.8 months) of income are required to pay for treatment compared with only approximately 3 months of income in Chile and in China. In Tanzania, more than 2 years of averageincomeisrequiredtopaythepatient s portion of the cost of the complex ulcer while the amount for the United States is equivalent to 9.6 months of average income. In addition to the assumptions described earlier regarding reimbursement practices, there are a number of limitations to the current study that should be considered. The results are entirely derived from the investigators opinions of the treatment plans for each case, and no actual patient data were collected. We did not use techniques such as Markov modelling [16] to predict the potentially different outcomes at each location, which might have resulted from the different treatment plans. Consequently, we cannot comment on cost-effectiveness, which is arguably the most important factor in the economics of treatment [17]. Quality-adjusted life years gained [18] were not calculated for the treatments. In addition, no judgements were made concerning the merits of treatment options and the reimbursement practices at the different sites. It could also be argued that the percentage of annual per capita PPP-adjusted GDP is a flawed estimator of actual income because it does not take into account income distribution or susceptibility to ulceration by income category. Also, several of the investigators reported marked differences between the costs in urban versus ruralsettings, but we have focused here on the urban clinics. Despite the aforementioned limitations, the authors believe that the data presented here provide a useful perspective on the cost burden of DFUs. It is counterintuitive that the large expenditure from a high-technology limb salvage campaign in the United States is only the median-adjusted cost to the patient from the five countries considered. In countries where the burden of cost to the patient is so high, it is likely that many patients will decline treatment while some who chose treatment will face financial ruin. These results illuminate why

5 Cost of Treating Diabetic Foot Ulcers in Five Countries 111 medical bankruptcies are commonplace in India and why the percentage of GDP spent on medical care in the United States (17.4% in 2009 [19]) is so high. It is also notable that in some countries, such as Chile, where the cost of ulcer care is comparatively reasonable because of federal programs, there is no mandate for post-healing footwear, which may prevent recurrence. Acknowledgements Helpful conversations with Drs Tony Berendt and Andrew Boulton during the genesis of this manuscript were appreciated. We are grateful to Shabneez Gangji, Kulsum Ramadhani, and other staff in Dar es Salaam for their assistance. Conflict of interest The authors declare that there is no conflict of interest associated with this manuscript with the exception that Cavanagh holds equity in DIApedia LLC; is an inventor on US patents , , and , which elucidate a load relieving dressing and a method of insole manufacture for off-loading; has received honoraria from Merck, Eli Lilly, and ConvaTec; and is a consultant to Langer (UK). References 1. Hunt NA, Liu GT, Lavery LA. The economics of limb salvage in diabetes. Plast Reconstr Surg 2011; 127(Suppl 1): 289S 295S. 2. Prompers L, Huijberts M, Schaper N, et al. Resource utilisation and costs associated with the treatment of diabetic foot ulcers. Prospective data from the Eurodiale study. Diabetologia 2008; 51: van Houtum WH, Lavery LA, Harkless LB. The costs of diabetes-related lower extremity amputations in the Netherlands. Diabet Med 1995; 12: Ragnarson Tennvall G, Apelqvist J. Healtheconomic consequences of diabetic foot lesions. Clin Infect Dis 2004; 39(Suppl 2): S132 S Horswell RL, Birke JA, Patout CA. A staged management diabetes foot program versus standard care: a 1-year cost and utilization comparison in a state public hospital system. Arch Phys Med Rehabil 2003; 84: Harrington C, Zagari MJ, Corea J, Klitenic J. A cost analysis of diabetic lower-extremity ulcers. Diabetes Care 2000; 23: Ramsey SD, Newton K, Blough D, et al. Incidence, outcomes, and cost of foot ulcers in patients with diabetes. Diabetes Care 1999; 22: Holzer SE, Camerota A, Martens L, Cuerdon T, Crystal-Peters J, Zagari M. Costs and duration of care for lower extremity ulcers in patients with diabetes. Clin Ther 1998; 20: Apelqvist J, Ragnarson-Tennvall G, Larsson J, Persson U. Long-term costs for foot ulcers in diabetic patients in a multidisciplinary setting. Foot Ankle Int 1995; 16: OyiboSO,JudeEB,TarawnehI,NguyenHC, Harkless LB, Boulton AJ. A comparison of two diabetic foot ulcer classification systems: the Wagner and the University of Texas wound classification systems. Diabetes Care 2001; 24: Schaper NC. Diabetic foot ulcer classification system for research purposes: a progress report on criteria for including patients in research studies. Diabetes Metab Res Rev 2004; 20(Suppl 1): S90 S Abbas ZG, Lutale JK, Game FL, Jeffcoate WJ. Comparison of four systems of classification of diabetic foot ulcers in Tanzania. Diabet Med 2008; 25: World Bank PPP conversion factor. ppp&language=en [11 September 2011]. 14. International Monetary Fund WEO country groups and purchasing power parity (PPP). org/external/pubs/ft/weo/faq.htm#q4d [10 September 2011]. 15. The World Bank Data: countries and economies. org/ org/ display=graph [9 Setpember 2011]. 16. Wolf CF, Gu NY, Doctor JN, Manner PA, Leopold SS. Comparison of one and twostage revision of total hip arthroplasty complicated by infection: a Markov expected-utility decision analysis. J Bone Joint Surg Am 2011; 93: Driver VR, Fabbi M, Lavery LA, Gibbons G. The costs of diabetic foot: the economic case for the limb salvage team. J Vasc Surg 2010; 52: 17S 22S. 18. Cleemput I, Neyt M, Thiry N, De Laet C, Leys M. Using threshold values for cost per quality-adjusted lifeyear gained in healthcare decisions. Int J Technol Assess Health Care 2011; 27: Organisation for Economic Co-operation and Development Health expenditure and financing. org/index.aspx [11 September 2011].