How To Calculate The Cost Of Diabetic Foot Ulcers

Size: px

Start display at page:

Download "How To Calculate The Cost Of Diabetic Foot Ulcers"

Prudence Campbell
3 years ago
Views:

1 Medical, Drug, and Work-Loss Costs of Diabetic Foot Ulcers Brad Rice, PhD; 1 Urvi Desai, PhD; 1 Alice Kate Cummings, BA; 1 Michelle Skornicki, MPH; 2 Nathan Parsons, RN BSN; 2 and Howard Birnbaum, PhD 1 1. Analysis Group, Inc., Boston, MA 2. Organogenesis Inc., Canton, MA Prepared for: ISPOR 18 th Annual International Meeting May 21, 2013 New Orleans, LA This research was supported by Organogenesis, Inc.

1. Analysis Group, Inc., Boston, MA 2. Organogenesis Inc.

2 Background and Objective Background: According to the American Diabetes Association, the annual cost of diabetes, which affects 25.8 million people in the U.S., was $245 billion in 2012 $176 billion in direct medical costs and $69 billion in reduced workforce productivity 1,2 One common complication of diabetes is the development of foot ulcers Foot ulcers have been estimated to affect 1% to 6% of patients with diabetes annually and up to 25% of diabetes patients over their lifetime 3,4 While diabetes patients with foot ulcers can require substantial amounts of resource use, little is known about the burden of these ulcers imposed on the U.S. healthcare system and payers Objective: Estimate annual per-patient medical, prescription drug, and work-loss costs of diabetic foot ulcers (DFUs) using de-identified administrative claims data For this purpose, DFU patients and non-dfu patients with diabetes (controls) were identified using two databases: ages 65+ from a 5% random sample of Medicare beneficiaries (Standard Analytical Files) and ages from a privately-insured population (OptumInsight) PAGE 2

have been estimated to affect 1% to 6% of patients with diabetes annually and up to 25% of diabetes patients over their lifetime 3,4 While diabetes patients with foot ulcers can require substantial

3 Periods of Analysis Study period Index date 12 mos 12 mos * Diagnosis period Patients with a DFU diagnosis during * were identified, with the date of each patient s most recent episode in this timeframe defined as the index date (the index date for the control group was randomly assigned to ensure similar timing distribution) Baseline period Patient characteristics in the 12 months prior to treatment were assessed to create treatment and control groups with comparable characteristics Follow-up period Resource use and costs of treatment and control groups were compared for 12 months post-index to determine burden of illness * Medicare analysis through 2010 PAGE 3

distribution) Baseline period Patient characteristics in the 12 months prior to treatment were assessed to create treatment and control groups with comparable

4 Sample Selection and Patient Counts 1 Identify all patients with medical claims (Medicare, ; private ins., ) Patients with at least one medical claim Private ins.: N = 8,398,397 Medicare: N = 2,285,018 Treatment Control 2 Identify relevant patient population > 1 foot ulcer diagnoses in relevant period following > 2 diabetes diagnoses > 2 diabetes diagnoses in medical history, but no foot ulcer diagnoses in the relevant time period 3 Ensure complete visibility of medical and pharmacy utilization Continuous (non HMO) coverage, and age restrictions in the study period Continuous (non HMO) coverage, and age restrictions in the study period 4 Identify the first date of new ulcer episodes No foot ulcer claims in the 12 months preceding a recent foot ulcer claim 5 Identify index date Select most recent episode Select a random medical claim meeting above criteria Patient characteristics are analyzed among treatment and control groups Analytical sample Private ins.: n = 5,681 Medicare: n = 29,681 Potential controls Private ins.: n = 113,337 Medicare: n = 201,757 PAGE 4

medical history, but no foot ulcer diagnoses in the relevant time period 3 Ensure complete visibility of medical and pharmacy utilization Continuous (non HMO) coverage, and age restrictions in the

5 Propensity Score Matching and Outcomes Propensity score matching: To provide an unbiased estimation of the incremental costs due to DFU, DFU patients were matched to patients in the non-dfu diabetic control population using a greedy matching method based on: Gender Year of index date Baseline healthcare costs (+/- 10%) Propensity score (within ¼ std. dev.) variables include age, comorbidities, medical resource use, and prescription drug use (private insurance only) Outcomes: Total and incremental direct healthcare as well as work-loss costs in the 12 months post-index were estimated for DFU and matched control patients Healthcare costs were estimated overall and for selected services containing a diagnosis or procedure code for ulcers, ulcer-related infections, or amputation recorded on the claim Work-loss costs were estimated for the subgroup of privately-insured patients with disability and wage information available following Birnbaum et al. (2000) PAGE 5

) variables include age, comorbidities, medical resource use, and prescription drug use (private insurance only) Outcomes: Total and incremental direct healthcare as well as work-loss costs in the 12

6 Pre- and Post-Match Baseline Characteristics Although the DFU patients differed from the control population on nearly all baseline characteristics, these differences were largely eliminated after matching Selected baseline characteristics among DFU patients and non-dfu diabetic controls Private insurance Pre-match Post-match Selected characteristics DFU n = 5,681 Non-DFU controls n = 113,337 DFU n = 4,536 Non-DFU controls n = 4,536 Male, % 59.7% 55.0%* 59.0% 59.0% Age, mean * * Comorbid conditions, % Diabetes w/ complications 43.8% 14.2%* 36.3% 36.5% Peripheral vascular disease 15.2% 3.5%* 9.6% 9.3% Renal disease 14.1% 3.1%* 8.6% 8.2% Depression 10.7% 7.7%* 9.6% 10.0% Cerebrovascular disease 10.5% 4.4%* 7.6% 6.9% COPD 7.9% 3.4%* 5.9% 5.5% Malignancies (incl. leukemia) 6.6% 4.9%* 6.1% 6.0% Myocardial infarction 3.5% 1.7%* 2.1% 2.1% Healthcare costs, mean $30,718 $12,338* $14,239 $14,244 Note: 8 of 17 comorbidities included in the propensity score shown. * Statistically different from the DFU cohort at p<0.05. PAGE 6

DFU n = 4,536 Non-DFU controls n = 4,536 Male, % 59.7% 55.0%* 59.0% 59.0% Age, mean 55.3 52.8* 55.0 54.5* Comorbid conditions, % Diabetes w/ complications 43.8% 14.2%* 36.3% 36.

7 Results: Per-Patient Annual All-Cause Healthcare Cost Differential DFU patients were over twice as costly as diabetic patients for private insurers and almost twice as costly for Medicare Per patient costs DFU patients Matched non-dfu diabetic controls $35,000 $30,000 $30,309 Private insurance Medicare $25,000 $27,040 $20,000 $16,286 $11,296 $15,000 $14,022 $15,743 $10,000 $9,316 $5,000 $0 $5,922 $3,053 $2,869 $5,098 Not applicable Overall Selected services Workloss costs Overall Selected services Workloss costs Medical and prescription drug costs Medical costs Note: Selected services include those with a diagnosis or procedure code for ulcers, ulcer-related infections, or amputation recorded on the claim. PAGE 7

$15,743 $10,000 $9,316 $5,000 $0 $5,922 $3,053 $2,869 $5,098 Not applicable Overall Selected services Workloss costs Overall Selected services Workloss costs Medical and

8 Results: Components of Cost Differential Inpatient and outpatient services comprised almost 75% of the cost differential for private insurance, with a more even distribution across places of service in Medicare Per-patient annual cost $18,000 $12,000 $16, % 8.6% 3.2% 7.5% 12.3% Medical costs Prescription drug Other medical Emergency department Home health care Outpatient/physician office Inpatient $11, % 20.7% $6, % 18.0% 20.3% $0 Private insurance 20.7% Medicare * Includes use of nursing home care, skilled nursing facilities, rehabilitation centers, hospice, durable medical equipment, and some specialist services (e.g., chiropractor). PAGE 8

3% Medical costs Prescription drug Other medical Emergency department Home health care Outpatient/physician office Inpatient $11,296 20.3% 20.7% $6,000 62.5% 18.0% 20.

9 Results: Annual Incremental Payer Burden of DFU Using publicly-available incidence estimates, the estimated annual U.S. burden of illness is in the range of $10 $15 billion DFU per-patient healthcare cost differential $11,296 $16,286 X 0.9 million patients / year 1 Estimated DFU annual payer burden = $10 $15 billion + Work-loss and other indirect costs 1. Estimated based on Type 2 diabetes population of 25.8 million and the midpoint (3.5%) of annual DFU incidence estimates. 1,3,4 PAGE 9

9 million patients / year 1 Estimated DFU annual payer burden = $10 $15 billion + Work-loss and other indirect costs 1.

10 Limitations and Conclusions Limitations: As with any claims data analysis, this analysis is based on ICD-9 and CPT codes to identify diagnoses and procedures as opposed to actual observance of medical conditions and resource use Results may not generalize to other patient populations (e.g., Medicaid) Results potentially underestimate incremental burden due to factors such as: Costs not covered through Medicare (supplemental insurance) Excludes additional indirect costs (e.g., quality of life) Disproportionate removal of relatively high-cost DFU patients during the matching process Conclusions: After accounting for differences in baseline characteristics, DFU patients had significantly higher medical resource use, resulting in them being approximately twice as expensive as the matched controls The incremental annual cost of DFU is in the range of $10 $15 billion, suggesting that presence of DFU imposes substantial burden on payers beyond that of care for diabetes alone The study highlights the need for improved preventive measures and optimized treatment for DFUs to help avoid some severe and costly outcomes such as amputation PAGE 10

neralize to other patient populations (e.g.

11 References 1. Centers for Disease Control and Prevention. National diabetes fact sheet: national estimates and general information on diabetes and prediabetes in the United States, Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, American Diabetes Association. Economic costs of diabetes in the U.S. in 2012, Diabetes Care 2013;36(4): Singh N, DG Armstrong, and BA Lipsky, Preventing foot ulcers in patients with diabetes, JAMA 2005;293(2): Margolis D, DS Malay, OJ Hoffstad, et al., Incidence of diabetic foot ulcer and lower extremity amputation among Medicare beneficiaries, 2006 to Data Points #2 (prepared by the University of Pennsylvania DEcIDE Center, under Contract No. HHSA I). Rockville, MD: Agency for Healthcare Research and Quality. January AHRQ Publication No. 10(11)-EHC009-1-EF, January Birnbaum HG, M Barton, PE Greenberg, et al., Direct and indirect costs of rheumatoid arthritis to an employer, Journal of Occupational and Environmental Management 2000;42(6): PAGE 11

Singh N, DG Armstrong, and BA Lipsky, Preventing foot ulcers in patients with diabetes, JAMA 2005;293(2):217-228 4. Margolis D, DS Malay, OJ Hoffstad, et al.

There may be up to 5 to 6 million cases of. Treatment Costs of Community- Acquired Pneumonia in an Employed Population*

There may be up to 5 to 6 million cases of. Treatment Costs of Community- Acquired Pneumonia in an Employed Population* Treatment Costs of Community- Acquired Pneumonia in an Employed Population* Gene L. Colice, MD, FCCP; Melissa A. Morley, MA; Carl Asche, PhD; and Howard G. Birnbaum, PhD Background: Community-acquired