Intermittent insurance is a risk factor for emergency department use: Implications for the Affordable Care Act roll-out

Objective: With Affordable Care Act (ACA) implementation, the formerly uninsured are at increased risk of emergency department (ED) use because of nationwide primary care provider shortages. To address gaps in the literature as to whether intermittent insurance coverage affects ED use, we evaluated the association between socioeconomic factors (e.g., age, sex, insurance status) and ED use. While the Medical Expenditure Panel Survey (MEPS) is known to underreport ED use overall, it is a rich, longitudinal source of nationally representative data on all ages and insurance groups, and a unique national database for exploring these questions.

Methods: Data on 14,788 respondents were obtained from the 2006-07 MEPS (prior to the ACA and the recession that began in 2008). We included respondents to all 5 survey rounds, obtaining information on their socioeconomic characteristics and reported ED use. We used weights to create nationally representative estimates within 5 insurance subgroups and multivariable logistic regression to identify correlates of ED use.

Results: The data represent 272.9 million civilian, noninstitutionalized individuals (51.4% female, baseline mean age=36 years). The insurance groups were: continuous private, 48.6%; continuous Medicare, 13.2%; continuous Medicaid, 8.2%; continuously uninsured, 10.4%; and intermittent insurance (any type), 18.2%. About 48% of individuals enrolled in Medicaid at any time experienced gaps in coverage, as did 25% of privately insured and 2% of Medicare enrollees. The private and uninsured groups had the lowest reported ED use rates per 100 (10.1 and 10.4); intermittent, 15.3; Medicaid, 17.0; Medicare, 19.9. Risk factors for ED use included: age, female sex, black race, less education, lower income, intermittent insurance, Medicaid insurance, prior ED use, inability to access care, and any cognitive or physical limitations. Despite having the highest reported (unadjusted) ED use rate, Medicare enrollees were not significantly more likely to use the ED after controlling for other risk factors. Adjusted odds ratios for ED use were similar for intermittent coverage (1.38) and Medicaid (1.48), relative to the uninsured (P<.001).

Discussion: Gaps in insurance coverage and Medicaid are significantly associated with increases in reported ED use compared to lack of insurance, whereas continuous private insurance and Medicare coverage are not. The ACA will increase Medicaid and private enrollment, yet maintaining continuous coverage may be difficult for many of the newly insured. Combined with current and predicted shortages of primary care providers, these factors could contribute to continued unsustainable growth in national ED use.