An Evaluation of the Impact of Continuity of Care on Cost of Services in New and Renewing CHIP Enrollees

Objective. To assess the impact of continuity of care as measured by enrollment history on the cost of services in Colorado’s state insurance program for children in low-income families, the Child Health Plan Plus. Design. A retrospective cohort study using data extracted from the eligibility/enrollment system and the claims payment system for inpatient and specialty services. All children enrolled in the program for at least 90 days and managed under the state managed-care network from its inception to April 1st of 2000 were included in the analysis (n-11841). Enrollees were separated into three enrollment groups: new members, renewing members, and members transition from a prior outpatient services group. The odds by enrollment groups: new members, renewing members, and members transition from a prior outpatient services program. The odds by enrollment group an average monthly claims greater than the 75th percentile amongst those with claims (n-2661) and the odds of presence of any claim in the full data set were compared means of a logistic regression model. In order to permit inferences regarding the total cost of care, all models included age group as a cofactor. Presence of chronic or congenital disease, ethnicity, primary language, urban residence, and the presence of multiple parents or caretakers were tested using univariate chi-square analysis and significant factors (p [less than] 0.1) were included as effects in the models. Results. All cofactors with the exception of ethnicity were significant in both models. Chronic disease status was associated with an increased risk of all claims. (OR=4.961, p [less than] 0.0001) and high claims (OR=9.639, p [less than] 0.0001). Renewal status was associated with a decreased risk of any claim (OR=0.921, p [less than] 0.0936) while rollover status was associated with both an increased risk of any claim (OR=1.157, p=0.0034) and a decreased risk of high claims (OR=0.731, p=0.0033). Residence in rural counties increased the risk of any claim (OR=0.0428, p [less than] 0.0001) and an increased risk of high claims (OR=1.472, p=0.0376). Conclusions. Enrollment history had a modest, though significant, effect on the presence of any claim and the specific presence of high claims. Rollover status was protective increasing utilization of the system overall but reducing the risk of high claims whereas renewal status only reduced the risk of utilization overall. The absence of multiple adults in the household showed a pattern consistent with episodic use of the system. Relative to urban residents, residents of rural counties were more likely to have used any service and less likely to have incurred high claims. This effect may be attributable to the transition of patients to HMO networks in urban counties. Linguistic or cultural barriers as reflected in primary language strongly decreased utilization, suggesting that these may be important barriers to address in outreach and member education.