STANDARDIZING AN ANNUAL REFERENCE COMPARATOR OF HEALTHCARE UTILIZATION OF PAINFUL DIABETIC NEUROPATHY (PDPN): A U.S. CLAIMS COHORT META-ANALYSIS FOR MODELING AND PAYER BENCHMARKING
Author(s)
Jesse Fishman, MSc, PharmD1, Lizandra Marcondes, MD2, Nathan Schuster, MD3, Hemant Kalia, MD4;
1Averitas, Director HEOR, Morristown, NJ, USA, 2Averitas, Morristown, NJ, USA, 3UCSD, San Diego, CA, USA, 4Savya Neuroscience Institute, Rochester, NY, USA
1Averitas, Director HEOR, Morristown, NJ, USA, 2Averitas, Morristown, NJ, USA, 3UCSD, San Diego, CA, USA, 4Savya Neuroscience Institute, Rochester, NY, USA
OBJECTIVES: To construct a standardized annual reference comparator from painful diabetic peripheral neuropathy (PDPN) claims studies to estimate background resource use, prevalence, and costs for United States (US) payer benchmarking; as variation in published claims studies with PDPN cohorts limits use in decision analytic models
METHODS: Targeted PubMed and Cochrane searches (1995-August 2025) identified U.S. cohort claims studies reporting healthcare use in diabetic neuropathy. Ten eligible studies contributed data on ~401,202 adults with PDPN and ~ 202,000 comparator adults with Type 2 Diabetes (T2DM) without PDPN. We extracted sample size, number with ≥1 encounter per resource category, follow‑up duration, and costs. Follow‑up was standardized to 12 months under a constant‑hazard assumption, converting encounters to events per patient‑year. One‑stage generalized linear mixed models with Poisson likelihood and log link (with log person‑time offset) estimated event rates (λ); binomial‑logit mixed models estimated 12‑month encounter probabilities (p) per resource type. Mean encounters for users were m = λ / p. Costs were CPI‑adjusted to 12‑month expenditures among users (2025 USD). Cost per encounter equaled cost per user divided by m. Sensitivity analyses used random‑effects meta‑analyses with Paule-Mandel τ² and Knapp-Hartung intervals.
RESULTS: Using PDPN claims studies, annual resource use rates per 1,000 patient-years were 500 emergency visits (ER), 533 inpatient admissions, 4,230 outpatient visits, and 10,990 physician visits. Respectively, the corresponding annual probabilities per resource were estimated to be p = 0.393, 0.411, 0.985, and 1.000. Annual costs per user were $4,557 (ER), $11,392 (inpatient), $21,329 (outpatient), and $10,688 (physician), yielding, respectively, encounter-level costs of ~ $3,582, $8,830, $4,966, and $972. T2DM comparators had 2-3-fold lower annual resource use; the sensitivity analysis showed consistency across results.
CONCLUSIONS: This reference event utilization provides standardized rates, probabilities, mean use, and costs for PDPN, enabling construction of economic models, and an annual cost benchmark for US payers.
METHODS: Targeted PubMed and Cochrane searches (1995-August 2025) identified U.S. cohort claims studies reporting healthcare use in diabetic neuropathy. Ten eligible studies contributed data on ~401,202 adults with PDPN and ~ 202,000 comparator adults with Type 2 Diabetes (T2DM) without PDPN. We extracted sample size, number with ≥1 encounter per resource category, follow‑up duration, and costs. Follow‑up was standardized to 12 months under a constant‑hazard assumption, converting encounters to events per patient‑year. One‑stage generalized linear mixed models with Poisson likelihood and log link (with log person‑time offset) estimated event rates (λ); binomial‑logit mixed models estimated 12‑month encounter probabilities (p) per resource type. Mean encounters for users were m = λ / p. Costs were CPI‑adjusted to 12‑month expenditures among users (2025 USD). Cost per encounter equaled cost per user divided by m. Sensitivity analyses used random‑effects meta‑analyses with Paule-Mandel τ² and Knapp-Hartung intervals.
RESULTS: Using PDPN claims studies, annual resource use rates per 1,000 patient-years were 500 emergency visits (ER), 533 inpatient admissions, 4,230 outpatient visits, and 10,990 physician visits. Respectively, the corresponding annual probabilities per resource were estimated to be p = 0.393, 0.411, 0.985, and 1.000. Annual costs per user were $4,557 (ER), $11,392 (inpatient), $21,329 (outpatient), and $10,688 (physician), yielding, respectively, encounter-level costs of ~ $3,582, $8,830, $4,966, and $972. T2DM comparators had 2-3-fold lower annual resource use; the sensitivity analysis showed consistency across results.
CONCLUSIONS: This reference event utilization provides standardized rates, probabilities, mean use, and costs for PDPN, enabling construction of economic models, and an annual cost benchmark for US payers.
Conference/Value in Health Info
2026-05, ISPOR 2026, Philadelphia, PA, USA
Value in Health, Volume 29, Issue S6
Code
EE223
Topic
Economic Evaluation
Topic Subcategory
Cost/Cost of Illness/Resource Use Studies
Disease
SDC: Diabetes/Endocrine/Metabolic Disorders (including obesity)