Are There Racial Disparities in Emergency Medical Services? Evidence from Mississippi
33 Pages Posted: 17 Oct 2006
Date Written: September 2006
Abstract
This paper analyzes the existence and scope of possible racial disparities in Emergency Medical Services using cardiac related incidence-level data for the state of Mississippi during 1995-2004.
We use incidence-level EMS data through the Mississippi Emergency Medical Services Information System (MEMSIS) collected by the Mississippi Department of Health. Demographic data from the U.S. Census Bureau, the Bureau of Labor Statistics, and the U.S. Department of Commerce. Heart disease rates from the Center for Disease Control and Prevention.
We estimate the relationships between race and two key measures of the quality of EMS - response time and respondent training - controlling for a variety of demographic and EMS organizational features that plausibly affect these performance measures. We also estimate models of the probability that the victim has already died when EMS arrives. Our study focuses on EMS incidents from 1995 to 2004 for which the initial call was related to cardiac problems for individuals aged 90 or below, including all calls made from a residence or public place.
We select all cardiac-related EMS incidences between 1995 and 2004 from the Mississippi Emergency Medical Services Information System (MEMSIS). We exclude Inter-facility transfers (e.g. between extended care facilities and doctors' offices and clinics). We merge the MEMSIS data with county level demographics based on the caller's county.
Our analysis provides little or no evidence of race-related disparities in EMS response time or quality of the responders. The only exception is for the population over 80 years old, where the evidence is relatively weak and plausibly related to imperfect controls for population density. In the case of response time, the evidence indicates faster response times for African Americans when we control for population density at the county level (population per square mile) only. However, when we include a county-level measure of population density that incorporates differences in African-American and white population density by census tract, the "negative" disparity in response times disappears. We find no economically or statistically significant difference in mean response time or the frequency of responses of eight minutes or below for African-Americans versus whites. Our analysis of possible differences in the likelihood of death-on-arrival at the scene provides some evidence that African-Americans on average are more likely to have died than whites.
This paper provides little evidence of race-related disparities in EMS and highlights the importance of properly controlling for own-race density measures. Evidence of race-related disparities in EMS response time for population over 80 years old is weak and plausibly related to imperfect controls for population density and disparities in living arrangements. Evidence that African-Americans on average are more likely to have died when EMS arrives at the scene may reflect differences in episode severity or possibly longer durations between the onset of symptoms and initiation of calls.
Keywords: Racial disparities, Emergency medical services, Cardiac care, Density measures
JEL Classification: I1
Suggested Citation: Suggested Citation