Emergency room use and access to primary care: evidence from Houston, Texas.

To cope with the rising number of uninsured, communities around the country are pursuing a variety of strategies to expand local health care safety nets. One measure that has been suggested to evaluate what is working is primary care-related emergency department (ED) visits. In this paper, we evaluate the applicability of this measure as an access indicator by examining its correlation with other indicators of medical under-service. We obtained ED visit data from safety net hospitals in Houston, Texas and applied the New York University ED Algorithm to estimate the rate of visits that were primary care-related. We then examined at the ZIP code level the correlation of primary care-related ED visits per 1,000 population with the federal government's Index of Medical Underservice (IMU), the poverty rate, and the uninsurance rate. Primary care-related ED visits were found to be weakly correlated with the IMU and strongly correlated with the rate of uninsurance and poverty. These findings suggest that the combination of this indicator with other measures of access could be used to monitor and evaluate local initiatives designed to expand care and coverage to the medically underserved.