Mortality after acute myocardial infarction in hospitals that disproportionately treat black patients

Abstract

Background: African Americans are more likely to be seen by physicians with less clinical training or to be treated at hospitals with longer average times to acute reperfusion therapies. Less is known about differences in health outcomes. This report compares risk-adjusted mortality after acute myocardial infarction (AMI) between US hospitals with high and low fractions of elderly black AMI patients.

Methods and results: A prospective cohort study was performed for fee-for-service Medicare patients hospitalized for AMI during 1997 to 2001 (n=1,136,736). Hospitals (n=4289) were classified into approximate deciles depending on the extent to which the hospital served the black population. Decile 1 (12.5% of AMI patients) included hospitals without any black AMI admissions during 1997 to 2001. Decile 10 (10% of AMI patients) included hospitals with the highest fraction of black AMI patients (33.6%). The main outcome measures were 90-day and 30-day mortality after AMI. Patients admitted to hospitals disproportionately serving blacks experienced no greater level of morbidities or severity of the infarction, yet hospitals in decile 10 experienced a risk-adjusted 90-day mortality rate of 23.7% (95% CI 23.2% to 24.2%) compared with 20.1% (95% CI 19.7% to 20.4%) in decile 1 hospitals. Differences in outcomes between hospitals were not explained by income, hospital ownership status, hospital volume, census region, urban status, or hospital surgical treatment intensity.

Conclusions: Risk-adjusted mortality after AMI is significantly higher in US hospitals that disproportionately serve blacks. A reduction in overall mortality at these hospitals could dramatically reduce black-white disparities in healthcare outcomes.

Figure 1. Index of Comorbidity and AMI Severity by the Average Percentage of African-American AMI Patients Admitted to the Hospital

Legend: The graph reports the average index of Comorbidity and AMI Severity by hospital decile according to the average percentage of African-American AMI patients admitted to that hospital. Multiple indicators for severity are used: the presence of vascular disease, pulmonary disease, dementia, diabetes, renal failure, cancer, and the location of the infarct, anterior, inferior, sub-endocardial, or other. These indicators are combined into one index using as weights the coefficients from a prediction model for 90-day mortality. Thus the index predicts 90-day mortality based on comorbidities and severity of the AMI, after adjusting for age, gender, and race. This index is intended to test the hypothesis that AMI patients are sicker in hospitals that disproportionately admit African-Americans. The graph indicates that this hypothesis is rejected; indeed those patients admitted to the lowest decile (no African-American admissions) experience elevated risk factors.

Figure 2. Risk-Adjusted 90-Day Mortality after AMI, by the Average Percentage of African-American AMI Patients Admitted to the Hospital

Legend: The graph reports 90 day mortality after adjusting for age, gender, race, comorbidities, and location of the infarct (anterior, inferior, sub-endocardial, other). Comorbidities include presence of vascular disease, pulmonary disease, dementia, diabetes, renal failure, and cancer. Hospital ownership and treatment characteristics are listed in Table 2, and include teaching hospital, government non-federal ownership, non-government not for profit, investor owned (for profit), hospital PTCA and CABG rates and annual AMI volume. Income refers to beneficiary’s zip-code income. Region refers to the 4 Census regions. A joint test of the importance of hospitals deciles is significant at the p < .001 level.

Figure 3. Risk-Adjusted 90-Day Mortality after AMI, by Race and Average Percentage of African-American AMI Patients Admitted to the Hospital

Legend: This regression includes all covariates described in the Legend for Figure 2, but with black and white hospital decile effects allowed to differ. To improve statistical power, Deciles 2–6, which together comprise 11 percent of the black AMI sample, are combined.

Figure 4. Distribution of African-American and White Patients, by the Average Percentage of African-American AMI Patients Admitted to the Hospital

Legend: The graph reports the share of each racial group (relative to all African-American or white AMI patients in the Medicare fee-for-service population) treated in hospitals within each decile category. A joint test of the importance of hospitals deciles is significant at the p < .001 level.