Quality of Care and Disenrollment in the New York State Children's Health Insurance Program (SCHIP)

[Objective] SCHIP retention is a focus of both state government and the research community. Despite the fact that most states use managed care plans to provide services to the low-income children in SCHIP, the effect of quality of care on disenrollment remains unclear. We examined whether higher quality measures in health plans were associated with lower disenrollment from SCHIP in New York. [Study Design] The study was based on a NY statewide cohort of children who were newly enrolled in SCHIP in 2001 and were followed up for about 15 months. A total of 1995 low-income children enrolled in 29 managed care plans were included. Quality of care was measured using seven Consumer Assessments of Health Plans Survey (CAHPS) scores and three Health Plan Employer Data and Information Set (HEDIS) scores, according to the 2002 NY Quality Assurance Reporting Requirements data. Based on the NY SCHIP universal billing files, we defined voluntary disenrollment as being disenrolled for at least two consecutive months. A theoretical model was derived from the satisfaction literature, and a discrete-time survival model was used to estimate the effect of quality of care on disenrollment. [Principle Findings] Nearly 36% of children experienced disenrollment during the study period. Younger children and children with unemployed parents were more likely to leave the program. Among children with better-educated parents, the probability of disenrollment was reduced by about 30% for those in plans with an average CAHPS score above the third quartile. Similarly, among children with special care needs, those in plans with an average HEDIS score above the third quartile were about 30% less likely to disenroll. The annual recertification generated 5-fold greater disenrollment compared to other time periods. During the recertification period, children with special care needs were more likely to disenroll, and plan outreach activity was associated with lower risk of disenrollment. [Conclusion] Higher quality of care measured by health plan CAHPS and HEDIS scores was associated with lower disenrollment, especially among children with special care needs and those with better-educated parents. [Policy Implication] Since prior studies have found that most children who disenroll become uninsured, the state government could support health plans to improve retention and reduce the number of uninsured by emphasizing delivery of high-quality care and through use of CAHPS/HEDIS measures. This would be particularly effective among vulnerable children with special healthcare needs. Increasing plan outreach activities could also help improve the retention in SCHIP.