The effect of transitions in insurance on emergency department utilization: A national perspective for the years 1996-2001

Abstract

Study objectives: There are millions of people in the United States who lack health insurance and even more who are intermittently insured. The effects of changes in insurance status may be particularly dramatic among individuals with chronic illnesses. Understanding the relationship between a change in insurance status and emergency department (ED) use is vital to delineate the role of the ED as a safety net provider and predict changes in ED utilization that are likely to result from changes in public insurance programs. The purpose of this study is to assess the changes in ED utilization that occur after an individual changes insurance status, especially among individuals who have chronic illnesses and may be at high risk. Methods: We conducted a secondary data analysis of the Medical Expenditure Panel Study (MEPS), a prospective cohort study of a nationally representative sample of the noninstitutionalized civilian United States population. MEPS is unique in its collection of data from the respondent, third-party payers, and medical providers. MEPS uses an overlapping panel design, with each panel followed up for 24 months. During 1996 to 2001, MEPS followed up 97,976 individuals, representing response rates ranging from 64% to 70%. We determined the average number of ED visits per month that the individual had before and after the month in which the insurance transition occurred. We categorized insurance for each month as private, Medicare, Medicaid/public, and uninsured. We then determined the months in which an individual changed from one category of insurance to another. We created an indicator of chronic illness according to the individual's having 1 or more conditions listed in the Charlson comorbidity index. We compared the ratio of the mean number of monthly ED visits after the insurance transition with the mean number of monthly ED visits before the insurance transition using an adjusted Wald statistic in Stata SE/8.0 that accounts for the complex survey design of MEPS. To ensure that the statistical significance was robust, we repeated the analyses on a 5% top-truncated sample. Results: During 1996 to 2001, MEPS revealed that 19,067 individuals had an insurance transition, representing an estimated 282,700,000 noninstitutionalized civilian Americans. Examining only the first transition during the 2-year study, 5,266 (91,546,514) went from private insurance to uninsured, 834 (11,674,911) individuals went from Medicaid/public insurance to private insurance, 2,711 (32,014,179) went from Medicaid/public insurance to uninsured, 5,344 (82,028,699) went from uninsured to private insurance, and 2,814 (31,750,476) went from uninsured to Medicaid. The ratio of mean number of ED visits after the transition compared with before the transition was 0.81 (95% confidence interval [CI] 0.64 to 0.99) for the private insurance to uninsured, 0.61 (95% CI 0.41 to 0.81) for the Medicaid/public insurance to private insurance, 0.58 (95% CI 0.41 to 0.75) for the Medicaid/public insurance to uninsured, 1.19 (95% CI 0.85 to 1.52) for the uninsured to private insurance, and 1.14 (95% CI 0.88 to 1.40) for the uninsured to Medicaid/public insurance. Among the subgroup with at least 1 disease from the Charlson comorbidity index, the ratio of mean number of ED visits after the transition compared with before the transition was 0.76 (95% CI 0.51 to 1.01) for the private insurance to uninsured, 0.44 (95% CI 0.19 to 0.70) for the Medicaid/public insurance to private insurance, 0.63 (95% CI 0.33 to 0.93) for the Medicaid/public insurance to uninsured, 0.76 (95% CI 0.39 to 1.13) for the uninsured to private insurance, and 1.01 (95% CI 0.64 to 1.38) for the uninsured to Medicaid/public insurance. Conclusion: A substantial number of Americans experienced insurance transitions during the 6 years of the study. A transition out of Medicaid/public insurance was associated with dramatic reductions in ED use regardless of whether the transition was into private insurance or uninsured. For the uninsured, gaining private insurance was associated with statistically insignificant increase in ED use for the general population and was associated with a statistically insignificant decrease in those with chronic illness. For the uninsured, acquiring Medicaid/public insurance had no noticeable effect on ED use. Medicaid/public insurance appears to be most highly associated with ED use, and loss of this coverage reduces ED use. More work is needed to determine the health consequences of this reduction in ED use.