This retrospective cohort study used data from 30 hospitals in northeast Ohio to compare observed, predicted, and risk-adjusted hospital mortality rates in white and African-American patients. The authors sought to assess whether predicted risks of death and risk-adjusted mortality would be similar or higher in African-Americans. The patient sample was comprised on 88,205 patients discharged with 6 diagnoses (acute myocardial infarction, congestive heart failure, obstructive airway disease, gastrointestinal hemorrhage, pneumonia, and stroke) between 1991 and 1993.

Results indicated that predicted risks of death were lower in African-Americans for 4 of the 6 diagnoses examined. (p<0.01) For all patients, the risk of death was lower among African-American patients compared to white patients. Examining the individual diagnoses, adjusted odds of hospital deaths were lower in African-Americans for 2 of the 6 diagnoses (p<0.01 for congestive heart failure and obstructive airways disease) and similar for the other diagnoses.

The authors commented that the finding of lower severity of illness at admission for African-Americans may reflect differences in access to care. Specifically, African-Americans may be more likely to be admitted due to difficulty in managing such medical problems as an outpatient or due to the uncertainty regarding subsequent follow-up care. An alternative explanation is that assessment of severity of illness—or predicted mortality—differs according to race. If African-Americans undergo diagnostic procedures less frequently, then assessed “severity” might be lower in African-Americans and unmeasured severity (i.e. detection bias) would be higher. By contrast, white patients might have higher “assessed” severity as a result of adverse effects of treatment during the initial period of hospitalization.