Background
Methods
-
A COI study was developed following an ingredients-based approach where unit costs for resources are multiplied by the quantity used. A decision tree model (see Fig. 1) was constructed to show cases of patients at high-risk of ulceration, diabetic foot and related outcomes. A population prevalence-based approach was performed to illustrate the costs and benefits that can be achieved in the first year of each strategy. All sources for key parameters are listed in Table 1.Table 1Epidemiology and cost inputsParameterValueLower estimateUpper estimateReferenceEpidemiologya. Total population with Type 2 diabetes mellitus942,000[11]b. % diabetics at high risk for ulcer21.90 %c. Prevalence of ulcer in those at high-risk22.18 %15.53 %28.84 %c1. Effectiveness (reduced prevalence) with standard care45.00 %30.00 %60.00 %c2. Effectiveness (reduced prevalence) with standard care plus temperature monitoring78.81 %65.00 %85.00 %d. % with ulcer receiving hospital care30.00 %21.00 %39.00 %[16] Own assumption: values for lower and upper estimate +/−30 %e. % with ulcer receiving outpatient wound management70.00 %61.00 %79.00 %[16] Own assumption: values for lower and upper estimate.f. % at hospital with amputation and heal47.75 %33.43 %62.08 %[16] Own assumption: values for lower and upper estimate +/−30 %f1. % with major amputation55.80 %[16]f2. % with minor amputation44.20 %[16]g. % at hospital with debridement and heal39.45 %25.12 %53.77 %[16] Own assumption: values for lower and upper estimate +/−30 %h. % at hospital that die of diabetic foot after an amputation12.80 %[16]Direct costs (in 2012 US dollars)PreventionSub-optimal care654584Standard care185129240Standard care plus temperature monitoring406285528TreatmentWound management without hospitalization79Debridement1,022Minor amputation5,153Major amputation7,3605,1529,568Indirect costs (in 2012 US dollars)Productivity loss from premature death6,719Peru's basic salary in year 2012; 3 % discount rate.×
-
The sub-optimal care or baseline situation was assumed to include an annual visit to a physician and podiatrist where routine tests are performed once, but neither education nor appropriate footwear is provided.
-
The second strategy, called standard care, was based on IDF guidelines [6], which emphasizes attention by a cadre of medical professionals, bi-monthly consultations, a higher frequency of tests, podiatrist consultation with foot evaluation, education and use of insoles (see Table 2).Table 2Major assumptions in cost-of-illness estimationA. Wound management without hospitalizationOutpatient: 1 first visit, 2 visits of control, 3 minor healing procedures in a hospital. Test: HbA1C, lipid profile, X-ray. Medication: Clindamycin 300 mg qid for 2 weeks.B. DebridementInpatient: Emergency consultation, 6 days of hospitalization, evaluation by anesthesiologist and cardiologist, anesthesiology medication and surgical materials, debridement procedure, intermediate care unit and 6 wound healing procedures. Test: Pre-surgery tests, antibiogram, HbA1C, lipid profile and X-ray. Medication: Intravenous antibiotic (Ampicillin/Sulbactam 1.5 g qid for 3 days), oral antibiotics for 11 days and peripheral line. Outpatient: Consultations with physician until healed at the hospital and materials for dressing changes.C. AmputationInpatient: Emergency consultation, 10 days (minor amputation) or 19 days (major amputation) of hospitalization, evaluation by anesthesiologist and cardiologist, anesthesiology medication and surgical materials, amputation procedure, intermediate care unit and blood transfusion. Test: Pre-surgery tests, bacteriology study, HbA1C, lipid profile, white cells count, X-ray, Doppler echography, arteriography, MRI, tissue biopsy. Medication: Intravenous antibiotic (Ampicillin/Sulbactam 1.5 g qid for 3 days in minor amputation and 5 days in major amputation), oral antibiotics (11 days in minor and 16 days in major amputation) and peripheral line. Outpatient: Consultations with physician and podiatrist until healed, materials for dressing changes (assuming that a nurse or a trained person at home is in charge of this procedure). Others: Rehabilitation sessions (40 for minor amputation and 50 for major amputation), orthopedic supplies for foot amputation (crutches and orthopedic foot) or for leg amputation (crutches, orthopedic leg, wheelchair), caregiver at home (conservative assumption of 6 months at Peru´s basic salary or 12 months working partial time).D. Premature deathWe assumed that 2 years (retirement age of 65) of paid productive work were lost due to the death and discounted at an annual rate of 3 %. Minimum wage rate in Peru amounts to PEN 750 in year 2012 (equivalent to US$284). We assumed a monthly income equal to minimum wage. The estimated indirect cost was US$6,719, which is the net value of the lost earnings for the next 2 years.E. Sub-optimal careOutpatient: 1 annual consultation with physician and podiatrist. Test: 1 annual testing of HbA1C, lipid profile, creatinine, electrocardiogram, X-ray.F. Standard careOutpatient: 6 consultations with physicians, 1 consultation with the podiatrist and 1 education session with a nurse. Test: 3 annual evaluations of HbA1C, 1 annual testing of lipid profile, 2 creatinine tests, 2 electrodiagrams and 1 X-ray. Others: protective footwear (a pair).G. Standard care plus temperature monitoringSimilar to standard care, but in addition: thermometer and daily phone calls assisted by a nurse or a trained person (about 5 minutes per patient everyday).
-
Thirdly, a standard care plus temperature monitoring strategy was included, which added daily self-monitoring of foot temperature to the standard care. This involves the use of a handheld thermometer for recording and monitoring fluctuations in foot temperature of individuals. If patients find a high degree of fluctuation in temperature between both feet, higher or equal to 2.2 °C, they need to call a nurse and schedule a consultation. This strategy has been applied in a number of clinical trials yielding consistent positive results [8‐10].
-
Our calculations considered the Peruvian adult population (18–79 years old) with Type 2 diabetes mellitus who were at high-risk of developing diabetic foot ulcer due to severe neuropathy with foot deformity or a history of ulceration. We used the IDF national data for the prevalence of Type 2 diabetes in Peru, estimated at 5.38 % in year 2011 [11]. This resulted in nearly 942,000 people having the disease. The prevalence of being at high-risk for diabetic foot was defined as the prevalence rate of severe neuropathy among diabetic patients, assumed to be 21.9 % based on previous studies for developing countries [12, 13]. It was calculated that the population in Peru with severe neuropathy was 206,298 persons. We have excluded patients with peripheral ischemia from our approach given that this population was also excluded in clinical trial studies exploring temperature monitoring as a preventative strategy for diabetic foot [8‐10].
-
The three different prevention strategies affected a reduction in the one-year ulceration rate for those that were at high-risk for diabetic foot. To illustrate the achievable health benefits, it was assumed that each strategy was introduced at scale. For the sub-optimal care strategy as baseline, the ulceration prevalence rate was assumed to be 22.18 %. This figure came from considering the ulceration prevalence rate for the standard care in a clinical trial study reported as 12.2 % [9], together with an effectiveness estimations of 45 % derived from the United States Center for Disease Control and Prevention (CDC) [14, 15] when moving from a scenario of sub-optimal care to standard care. An even higher reduction of 79 % in the baseline rate of ulceration is achieved with the standard care plus temperature monitoring, achieving a 4.7 % of ulceration rate [9].
-
We determined the likely clinical outcomes for different scenarios. The data was for a one-year period related to diabetic foot treatment and complications such as ulcer development, wound management, debridement procedure, amputation and death.
-
We considered the parameters from a Brazilian study [16] that provides rates for hospitalization and outcomes attributed to diabetic foot ulceration in patients attended by the public health system. This kind of study has not been implemented in Peru and there is a lack of reliable epidemiological data from the health statistics at hospitals. In the Brazilian COI study [16], 70 % of ulcerated patients heal without hospitalization, and therefore, we assumed that those patients only receive outpatient wound management. The other 30 % of patients require hospitalization, with 39.5 % healing with primary care (assumed as debridement procedure), 47.8 % healing by a minor or major amputation, and 12.8 % dying after amputation surgery. Our assumption here is that all patients who die had an amputation in the same year. Among those patients who undertook amputation surgery in the Brazilian study, 44.2 % received a minor amputation and 55.8 % a major amputation. Table 1 shows the values used and assumptions made for this study.
-
Direct costs of resources used to prevent or manage the disease were registered under a societal perspective. We identified the main medical procedures that can be carried out in Peru related to each stage of diabetic foot. We followed an ingredients-based approach for our costing analysis (see Additional file 1). We estimated the number of specific procedures, personnel, medical supplies, examinations and medications for each disease stage according to the IDF guidelines. The unit costs of treatment were taken from a list of tariffs in a public hospital of the Ministry of Health in Peru [17]. We considered generic drug prices [18].
-
Costs were registered at the local currency Peruvian Nuevos Soles (PEN) and converted to US dollars (US$), considering an average exchange rate of 2.64 PEN per dollar in year 2012 [19].
-
Additionally, the indirect costs due to premature death were measured using a human capital approach. Minimum wage rate in Peru amounts to US$284 in year 2012. We assumed a monthly income equal to minimum wage and a mean age of 63 years for premature mortality, according to the primary data of patients with neuropathy in a public hospital in Lima [12]. We considered that two years, with a retirement age of 65, of paid productive work were lost due to the death and discounted at an annual rate of 3 %.
-
Cost-effectiveness is examined in terms of cost per deaths averted. The analysis consisted in estimating the incremental cost-effectiveness ratio (ICER) by estimating the additional costs and deaths averted for each of the preventive strategies compared to the sub-optimal strategy. Since the analysis only looked at costs and effects for one-year, there was no discounting.
-
We conducted multiple one-way sensitivity analysis to assess the uncertainty of the key input parameters. The variables of interest were independently varied according to their plausible range from Table 1. Most of the variables were varied using a range of +/− 30 %, given that uncertainty ranges were not available in the literature. The resulting change in cost-effectiveness was then taken and the variables were ordered in a tornado diagram from the most to the least sensitive parameter.
-
The model assumed that healthcare utilization is 100 % and constant for the entire population. Finally, our approach does not discriminate by rural–urban setting, assuming that all patients attend urban hospitals.
Ethics
Results
Description | All strategies |
Sub-optimal care
|
Standard care
|
Standard care plus temperature monitoring
| |||
---|---|---|---|---|---|---|---|
Prevention strategy
| |||||||
Cost per person | US$65 | US$185 | US$406 | ||||
Total people | 206,298 | 206,298 | 206,298 | ||||
Total cost of prevention | US$13,350,763 | US$38,129,966 | US$83,849,832 | ||||
Complications
| Cost/person | No. of people | Total cost | No. of people | Total cost | No. of people | Total cost |
Hospital | |||||||
Healing with debridement | US$1,022 | 5416 | US$5,535,166 | 2979 | US$3,044,341 | 1148 | US$1,172,820 |
Healing with major amputation | US$7,360 | 3658 | US$26,921,264 | 2012 | US$14,806,695 | 775 | US$5,704,219 |
Healing with minor amputation | US$5,153 | 2897 | US$14,930,136 | 1594 | US$8,211,575 | 614 | US$3,163,476 |
Death with major amputation | US$7,360 | 981 | US$7,216,590 | 539 | US$3,969,125 | 208 | US$1,529,089 |
Death with minor amputation | US$5,153 | 777 | US$4,002,215 | 427 | US$2,201,218 | 165 | US$848,010 |
Subtotal
|
US$26,048
|
13728
|
US$58,605,371
|
7551
|
US$32,232,954
|
2909
|
US$12,417,613
|
Outpatient | |||||||
Healing with outpatient visit | US$79 | 32032 | US$2,514,669 | 17618 | US$1,383,068 | 6787 | US$532,821 |
Total No. of people/Total cost of treatment |
45761
| US$61,120,040 |
25168
| US$33,616,022 |
9696
| US$12,950,435 | |
Total cost | US$74,470,803 | US$71,745,988 | US$96,800,267 |
Comparison | Standard care vs. sub-optimal care | Standard care plus temperature monitoring vs. sub-optimal care | Standard care plus temperature monitoring vs. standard care |
---|---|---|---|
Differences in costs (∆ costs) | |||
Direct costs | -US$2,724,815 | US$22,329,464 | US$25,054,279 |
Direct costs + Indirect costs | -US$8,037,824 | US$13,024,440 | US$21,062,264 |
Differences in deaths (∆ deaths) | 791 | 1,385 | 594 |
Differences in major amputations | 2,087 | 3,656 | 1,568 |
Incremental cost-effectiveness ratio (∆ costs/∆ deaths) | |||
Direct costs | Cost-saving | ICER = 16,124 | ICER = 42,169 |
Direct costs + Indirect costs | Cost-saving | ICER = 9,405 | ICER = 35,450 |