Dying for [no] coverage: Effects of Lack of Insurance on Mortality among Working-Age Adults

Introduction: Disparities in life expectancy between white and non-white populations in the US are marked. While infant mortality contributes to this difference, the majority of excess deaths occur in middle age. The purpose of our research is to ascertain whether health insurance helps ameliorate racial disparities in premature mortality in this age group. Method: Data were drawn from the 2004 National Health Interview Survey (NHIS) Linked Mortality File, developed by the National Center for Health Statistics. This file links respondent information from the NHIS to death records through 2002. Our analysis examines the years 1990-1991, restricted to adults between the ages of 45 and 64 (n=45,572). The dependent variable was mortality status as of 2002 (5,989 deaths). Our independent variables are race/ethnicity (white, non-Hispanic black, Hispanic, and other) and health insurance status when interviewed (privately insured, publicly insured, not insured). Demographic, resource and health status characteristics of the individual were held constant in multivariable analysis. Logistic regression was used to model the probability of death by 2002, with survey year included to account for differential follow time. All analyses are weighted to account for the complex sample design. Results: Overall, 81.5% of respondents reported private insurance, 6.7% public insurance, and 11.7%, uninsured. Minorities were more likely to be uninsured (19.1% black, 26.8% Hispanic, and 19.7% other adults, versus 9.2% of white respondents). In analysis holding only age, sex and year constant, uninsured persons had an elevated risk of death by 2002 (OR 1.50, CI 1.40-1.61), as did persons with public insurance when interviewed (OR 1.74, CI 1.64-1.85). In similar analysis, black adults had an elevated risk of death (OR 1.50, 95% (CI) 1.37-1.65) compared to white adults, while individuals of other race ethnicity did not differ. In fully adjusted analysis, uninsured white, black and “other” respondents had higher odds for death than white privately insured persons. Black and “other” privately insured persons did not differ from their white peers, while Hispanic privately insured adults had lower odds of death. Publicly insured persons were generally at higher risk of death. Discussion: Lack of health insurance markedly increases the risk of death among most racial/ethnic groups. Current regulations surrounding the provision of public health insurance generally limit coverage to persons already disabled by health problems. Mechanisms for expanding health coverage to persons not yet disabled by injury may forestall death during typical working years and yield long-term economic benefits.