![[ Visit Client Website ]](images/banner.jpg)
BACKGROUND: Optimal implementation of the Affordable Care Act (ACA) requires methods to track change in insurance status and health coverage eligibility between three distinct uninsured groups: 1) uninsured and Medicaid-eligible, 2) uninsured and Exchange-eligible, and 3) uninsurable (undocumented, some legal immigrants). Monitoring change through baseline and ongoing assessment should identify and characterize the groups to inform geographically or demographically targeted outreach to improve health by expanding access to coverage and care.
METHODS: Our objectives were to identify neighborhoods and demographic groups with high rates of uninsured in 2013 (pre-ACA), and to track increases in coverage post-ACA. Groups with high rates of uninsured individuals were defined using data from two large weighted telephone surveys: the American Community Survey (ACS) and the Colorado Health Access Survey (CHAS). Pre-ACA coverage by neighborhood was estimated from five consecutive years of ACS data and developed a map of the uninsured. Sub-analyses combining health coverage and income projected Medicaid-eligible, Exchange-eligible, and uninsurable. After January 1, 2014, monthly Medicaid caseloads tracked overall increases in Medicaid coverage. Exchange enrollment data was used to estimate reductions in uninsured, using a national estimate for the proportion of enrollees who were previously uninsured.
RESULTS: Denver County is an urban county with of 640,000 persons at the center of the metropolitan area (2.9 million residents). Baseline estimates of Denver uninsured persons differed by 16% between CHAS (116,406) and ACS (100,653). Uninsured prevalence was highest among young adults (19-34 year olds), males, and Hispanics. Maps of uninsured rates showed large differences in lack of health coverage (1%-44%) by census tract. Sub-analysis of 2011 data suggested that 42,000 uninsured individuals were eligible for Medicaid under expanded ACA eligibility, 40,000 were eligible for exchange plans, and 20,000 were uninsurable. From onset (October 2013) through November 1, 2014, the Denver county Medicaid caseload increased by over 56,000 individuals (140% of estimate) with steep monthly increases in both adults and children. Over 15,000 individuals enrolled through the exchange, an estimated 9,000 of whom were previously uninsured. Denver county uninsured persons have decreased by more than half (55-65%) within the first year of implementation.
CONCLUSIONS: Accurate methods to monitor health insurance coverage (a key public health metric) merit careful epidemiological study. Identifying demographic and geographic characteristics of uninsured populations were useful. Enrollment data from Medicaid and the exchange suggest that telephone surveys underestimated uninsured populations and coverage-specific eligibility estimates. The ‘welcome mat’ affect contributed a large unanticipated overage of enrollment.