This study aimed to evaluate whether patients’ race influenced the diagnosis of those with symptoms of acute cardiac ischemia. Patients with a broad range of presenting symptoms suggestive of acute cardiac ischemia were included in a 10-site clinical trial that evaluated a predictive model to aid clinical decision-making for emergency department triage in 1993. 10,001 African American and white participants from this trial were included in the present analysis.

There were marked differences in presenting symptoms, symptom history, and final diagnoses between African American and white patients. However, among those who were diagnosed with acute myocardial infarction, there were no racial differences in severity as measured by Killip Class. Similarly, among those diagnosed with angina, there were no racial differences with respect to severity as measured by the Canadian Cardiovascular Society classification.

The authors then examined whether racial group predicted a diagnosis of acute myocardial infarction after adjusting for symptoms, insurance, and medical history. The odds ratio (OR) for African Americans versus whites was 0.54 (95% confidence interval =0l.41 to 0.68), indicating that African Americans were significantly less likely to receive a diagnosis of acute myocardial infarction than white patients with the same symptoms. A similar finding was reported for the diagnostic category "acute cardiac ischemia" (OR=0.59; 95% confidence interval=0.50 to 0.71). The diagnoses used for this study were assigned by a member of the research team, who made use of a structured set of symptom and test data and applied World Health Organization criteria. Thus, differences in diagnoses by racial group after accounting for presenting symptoms might
indicate that presenting symptoms are not equally predictive of specific diagnoses for African Americans and white patients.

It is noteworthy that there were several racial differences in characteristics of the cardiac patients in this study. African Americans were younger, predominantly female, more likely to be single, and more often without health insurance. African Americans had more previous systemic hypertension, were more often current smokers, and, among women, were more often diabetic than whites. On the other hand, white patients had more frequent histories of angina, myocardial infarction, and use of cardiac medications. Duration of symptoms prior to hospitalization also differed between African American and white patients, as the delay time was 1 hour longer for African American men and 18 minutes longer for African American women. The authors also conducted an analysis adjusted for demographic and insurance factors, which reduced these racial differences substantially, and eliminated the statistical significance of the difference for men.

This study points to a possible selection bias in research using administrative data. Because African Americans with symptoms of acute cardiac ischemia in this study are younger, they are less likely to be Medicare eligible and more likely, therefore, to not have health insurance. Thus, the research examining treatment differences between racial groups using Medicare population data may miss an important portion of African Americans lacking sufficient care.

It is important to note that an important patient group may have been omitted from this sample, specifically those who died due to cardiac arrest. This omission may have biased some of the findings. It is also uncertain whether racial biases existed in the process of assigning diagnoses, despite efforts to standardize this procedure.