Before and After the ACA Exchanges-Comparing the Individual and Employer-Sponsored Insured Populations

Thesis (Master's)--University of Washington, 2018Introduction: To properly assess risk, health insurers need information on their insured population. Sociodemographic characteristics and previous utilization patterns are elements of this assessment. With the implementation of the ACA and the establishment of health insurance exchanges, little of this type of information was available to health insurers about new entrants into the individual market through these exchanges. This report compares insurance enrollees through three mechanisms and in two time periods to provide data on how the exchange market population may differ from those with employer-sponsored insurance or individual and family directly purchased insurance. Methods: Using enrollment and claims data, we describe sociodemographics, clinical risk factors and utilization patterns for enrollees of Kaiser Permanente Washington in 2008-2009 and 2015-2016. Enrollees were divided into those with employer-sponsored insurance, individual and family directly purchased insurance, and individual and family exchange purchased insurance. The exchange enrollees were further examined based on receipt of ACA provided subsidies. We conducted chi-square tests to compare the insurance types between the time periods. Results: In the 2015-2016 time peiod, the subsidized exchange population was older (35.8% 56-64 years old) compared to the employer-sponsored insurance population (21.8%), and the unsubsidized exchange population was younger (28.2% 27-35 years old) compared to the employer-sponsored insurance population (18.3%). 70.1% of the unsubsidized exchange population chose a bronze or catastrophic plan. We also observed unadjusted utilization differences. Those with employer-sponsored insurance had the highest rates of utilization in all categories in both time periods, and those with unsubsidized exchange insurance had the lowest rates of utilization in most categories. As an example, 44.5% of those with unsubsidized exchange insurance had a primary care visit compared with 66.2% with employer-sponsored insurance. The subsidized exchange population had high emergency room use (98.7 visits per 1,000 person-years) compared to the unsubsidized exchange population (48.7 visits). Conclusions: We found the subsidized exchange population was older than the other groups, but we did not find the overall high utilization rates in this population that were expected with the initiation of the exchanges. The unsubsidized exchange population was younger than the other groups and more likely to choose lower cost plans with high deductibles, which may explain their lower utilization patterns. The employer-sponsored insurance and the direct purchase markets showed little impact of the ACA implementation in their sociodemographics or utilization rates