(C-147) Does Health Plan Enrollment Data Support Long-term Follow-up of Patients with Chronic Disease? Assessment of the Healthcare Integrated Research Database (HIRD®)

Background: Duration of health plan enrollment is an important consideration in the design of pharmacoepidemiologic studies using administrative health care payer (claims) databases.

Objectives: To understand enrollment duration within a large US health plan, we conducted a descriptive analysis of commercial and managed Medicare members. We compared enrollment duration by member characteristics and five chronic diseases: type 2 diabetes mellitus (T2DM), breast cancer (BC), heart failure (HF), multiple sclerosis (MS), or Alzheimer’s disease (ALZD).

Methods: We conducted a retrospective analysis using payer claims from the Healthcare Integrated Research Database (HIRD®). Study members had an outpatient office visit for any reason or were newly diagnosed with any of the five conditions between 10/01/2016 and 09/30/2020. Index date was date of first office visit claim or first claim with chronic disease. We required enrollment for at least a year before index date. Individuals with a diagnosis for the same chronic condition in the prior year were excluded. We calculated enrollment duration from index date through 09/30/2024. Enrollment duration was calculated in months by age categories (≤ 17, 18-44, 45-64, ≥65 years) and race/ethnicity, and reported as median (interquartile range) months.

Results: A total of 11.7 million members had an office visit during the study period. Mean age varied by condition, ranging from 48 years (MS) to 83 years (ALZD); overall mean age was 39 years. Post-index enrollment duration was a median of 38 months (14 – 80) among all members with an outpatient visit, with 36% enrolled for ≥5 years. Enrollment duration was similar for members with T2DM, BC, and MS, median (IQR): 44 (18-68), 48 (19-69) and 40 (16-67) months, respectively. Enrollment was shorter for those with HF, median (IQR) of 35 (11-62) and ALZD, 30 (12-54) months. Enrollment duration varied by age, ≤ 17: median (IQR) of: 47 (15-85), 18-44: 31 (12-68), 45-64: 39 (14-81) and ≥65: 53 (19-93), and was longest for White, non-Hispanics, 41 (15-83) and shortest for Asians, 35 (13-74) months.

Conclusions: Results show that large payer databases can support pharmacoepidemiologic studies requiring longer follow-up periods.