Duke 2008 | ASHE | Equity and Efficiency in Health and Healthcare

Purpose: This study examines the levels of non-urgent Emergency Department (ED) use in Tennessee and quantifies the effects of race and insurance on such. Background: Non-urgent ED use is a nationwide problem, crowding Emergency Rooms and delaying care which has the potential for negatively effecting health care outcomes. Non-urgent visits to the ED cost more; the same care provided in a primary care provider setting can usually be received at a much lower cost. Further, continuity of care suffers when patients seek primary care in the ED. Continuity of care is necessary to effectively manage chronic illness such as diabetes and asthma to prevent flare-ups which result in ?preventable? hospitalizations. Since there is a link between primary care access and quality barriers and non-urgent ED visits, this study will examine the rate at which minority and low income patients/Medicaid patients, who are known to experience barriers to primary care, use the ED for non-urgent care. Hypotheses: This study will examine two hypotheses. Firstly that patient race has no effect on the rate of non-urgent ED use. Secondly, that the type of insurance has no effect on the rate of non-urgent ED use. Patients who report TennCare (Tennessee's Medicaid) as their primary insurance are by definition low income patients. Thus looking at the rate at which TennCare patients over/under use the ED for non-urgent care is synonymous with examining the rate at which low income patients over/under use the ED for non-urgent care. Data Sources: The primary date source is the Tennessee Hospital Discharge Data System (HDDS) which is maintained by the Tennessee Department of Health. All Tennessee hospitals submit patient claims data on the standardized UB-92 billing form for each calendar year to the Department of Health. Secondary data comes from the Joint Annual Report on Hospitals, also distributed by the Tennessee Department of Health. An algorithm designed by the Center for Health and Public Service Research at New York University will be used to determine the probability that an ED visit is non-urgent. Methodology: Each patient visit in the HDDS outpatient data will be analyzed using the NYU algorithm to determine the probability of the visit being non-urgent. This probability (NU) will be analyzed using ordered logistic regression to determine the individual contributions of race and insurance to the probability of a visit being non-urgent while controlling for other factors known to influence the rate of non-urgent ED visits such as region of the State and ED Trauma Center status.

Potential Contribution: Non-urgent ED visits are a common problem with no easy solution. Determining which insurance providers and racial groups are overrepresented among non-urgent ED users enables efforts at reducing non-urgent ED visits to be focused on those populations where they will have most effect.