The effect of a pediatric rare disease on subscriber retention rates for commercial health insurers in the United States

Abstract

BACKGROUND: The advent of personalized medicine creates opportunities for regenerative therapies to deliver extended clinical benefit from a single administration. Policymakers and health insurers in the United States are evaluating coverage and payment arrangements for these therapies. One challenge involved in these evaluations is the perception that subscribers change insurers relatively often. However, the effect of pediatric rare diseases on retention rates for commercial health insurers has not been well studied. OBJECTIVE: To develop estimates for subscriber retention by a commercial insurer for up to 10 years. METHODS: Three population cohorts were examined: (1) all subscribers, (2) subscribers with any dependent aged 16 years or younger, and (3) subscribers with any dependent aged 16 years or younger diagnosed with a chronic rare disease that typically results in a debilitating state or high mortality, usually associated with high health care costs regardless of whether a treatment is available. The analysis was conducted for a sample of fully insured and self-insured group business within the commercial health insurance market for these study cohorts. The MarketScan Commercial Claims and Encounters research database covering the time period from January 1, 2007, through December 31, 2016, was used as the basis for the analysis. Subscribers were included in the family-based cohorts beginning with the first observed month with a dependent aged 16 years or younger and were retained in the analysis until the subscriber or insurer withdrew from the dataset (whichever came first). Subscribers were included in the family-based rare disease cohorts if their qualifying dependent was reported with at least 2 occurrences for any of the rare diseases studied. A Kaplan-Meier estimator was used to produce retention rates for all populations for up to 120 months. An adjustment for interval censoring was applied to the family-based cohorts. A log rank test with chi-square statistic was used to determine statistical significance. RESULTS: The analysis found that the subscriber retention rate within the self-insured groups was higher than within the fully insured groups (P<0.0001). In addition, the probability of retaining subscribers with a dependent aged 16 years or younger compared with all subscribers was significantly greater (P<0.0001). The analysis also found the probability of retaining subscribers with a qualifying dependent with a rare disease compared with subscribers with any dependent aged 16 years or younger was significantly greater (P<0.0001). CONCLUSIONS: This study demonstrated that families with a child with a rare disease remained with their commercial health insurer longer than families who did not have a child with a rare disease. The analysis will be a useful resource when evaluating alternative payment arrangements and cost/benefit analyses of regenerative therapies that offer an extended duration of clinical benefit.