Venue: The Fuqua School of Business, Duke University, 1 Towerview Drive, Durham, NC 27708-0120
Presentation
The Effects of Race and Insurance on the Likelihood of Potentially Avoidable Hospitalizations in Tennessee
Purpose: The objective of this study is to determine quantitatively the influence of race on the likelihood of potentially avoidable hospitalizations (PAHs). Another objective is to identify other key drivers of PAHs, specifically the role of different insurance plans such as private commercial insurance, Medicare, and TennCare (Tennessee's Medicaid managed care program), and of the absence of insurance.
Background: Research suggests that ambulatory-care sensitive conditions (ACSCs) are potentially avoidable. These hospitalizations can be avoided, that is, they represent PAHs, when clinicians deliver timely and effective outpatient treatment to individuals who actively participate in their own care, follow a healthy life style, and engage in responsible personal behavior (AHRQ 2004). Research has also suggested that high incidence of PAHs may be indicative of underlying problems with access to primary care or deficiencies in outpatient care management and follow-up. The Agency for Healthcare Research and Quality (AHRQ) released in 2004 an analytical tool called Prevention Quality Indicators or PQIs for state and local health agencies to apply the ACSC concept for evaluating access to primary care. In 2006, this analytical tool was updated and we apply this updated version to Tennessee hospital inpatient discharge records to identify PAHs in this study.
Data and Method: The data used are derived form the Tennessee Hospital Discharge Data System or HDDS (http://health.state.tn.us/statistics/), which contains clinical and administrative information recorded in the federal UB-92 claims forms filed for each discharge occurring at a licensed Tennessee hospital. The analytical file based on the HDDS data for 2002-2004 was then linked to two additional databases: first to the 2004 Area Resource File (http://www.arfsys.com/) for relevant county-level socio-demographic and other information based on the admitting hospital's county of location, and then to the Tennessee Joint Annual Report of Hospitals (TN-JAR) for 2002-2004 for information on the characteristics of the hospital where the patient was admitted (http://health.state.tn.us/statistics/). We use the logistic regression technique to estimate the likelihood of PAHs using patient characteristics and hospital characteristics as explanatory variables.
Results: The study finds that black patients are 35.1% more likely than white patients of similar characteristics to experience PAHs. However, the extra risk is limited to chronic conditions. Black patients' risk for PAHs actually drops below the white level for acute conditions after controlling for covariates. We also find that the risk of PAH declined slightly from 2002 to 2004 for patients admitted for acute ACSCs but not for chronic ACSCs. Another interesting finding is that the black-and-white gap in PAH risk is confounded by insurance, with black Medicare patients having a 19.4% higher risk than white Medicare patients, and with the increase in risk rising to 37.3% for black TennCare patients, 68.5% for black commercial patients, and to 90.8% for black uninsured patients. The confounding effects of race and insurance on PAHs, together with the extra risks of chronic conditions, deserve greater attention.