Body mass index and insulin use as identifiers of high-cost patients with type 2 diabetes: A retrospective analysis of electronic health records linked to insurance claims data.
Academic Article
Overview
abstract
AIMS: To study the association of body mass index (BMI) and insulin use with type 2 diabetes-related healthcare expenditures (T2D-HE). MATERIALS AND METHODS: Retrospective study using de-identified electronic health records linked to insurance claims data. Study included a prevalence-based sample of overweight or obese patients with antihyperglycaemic-treated T2D. Patients had ≥1 A1c measurement in 2014 (last observed = index A1c), ≥1 BMI measurement within ±90 days of index (average BMI = baseline BMI), and continuous enrolment for 180 days before (baseline) through 395 days after index (day 30-395 = follow-up). BMI was categorized as: 25 to 29.9 kg/m2 = overweight; 30 to 34.9 kg/m2 = obese class I (OCI); 35 to 39.9 kg/m2 = OCII; ≥40 kg/m2 = OCIII. Multivariable regressions were used to examine one-year follow-up T2D-HE as a function of BMI, insulin use, an interaction term between BMI and insulin use, and patient demographics. RESULTS: Study included 13 026 patients (mean age = 63.6 years; 48.1% female; 29.5% overweight, 31.6% OCI, 20.3% OCII, 18.6% OCIII; 25.3% insulin users). Baseline insulin use rates monotonically ranged from 19.7% in overweight patients to 33.0% in OCIII patients (P < 0.001). Together, BMI and insulin use were jointly associated with one-year follow-up T2D-HE, which monotonically ranged from $5842 in overweight patients with no insulin to $17 700 OCIII insulin users, P < 0.001. Within each BMI category, insulin users' one-year T2D-HE was at least double that of non-users. Additional analyses of all-cause healthcare expenditures yielded consistent results. CONCLUSIONS: BMI and insulin use represent simple stratifiers for identifying high-cost patients. OCIII insulin users incurred the greatest annual healthcare expenditures; these patients may be an ideal group for targeted interventions.
