The purpose of this study was to determine if racial disparities in the diagnosis and treatment of AIDS patients, specifically the treatment of Pneumocystis carinii Pneumonia (PCP), existed in a health care system with minimal financial barriers. Those included in the study were over 18 years of age and had 1.) received medical care at the study hospital during the period from January 1, 1987 through December 31, 1990 and 2.) microbiologic confirmation of PCP or physician notes that indicated that PCP most likely accounted for the pulmonary process. Patients were characterized by age and sex, health insurance status, risk group and disease characteristics, and the characteristics of the hospital to which they were admitted. For the assessment of treatment, the measures were the use and timing of PCP medications. Because “virtually all patients received PCP therapy,” the investigators focused on the day that these medications were started following admission. In-hospital mortality rates were used to describe outcomes of care.

The findings indicated that “there were significant racial variations in the diagnostic patterns of care at non-VA hospitals, but not in VA hospitals.” In VA hospitals, use of bronchoscopy within two days of admission was 9.3% for blacks, 10.0% for Hispanics and 12.3% for whites; the respective rates in non-VA Hospitals were 14.9%, 20.0%, and 25.6%. Thus, “black patients at the non-VA hospitals were significantly less likely to undergo a bronchoscopy during their hospital stay; if they received it, these black patients underwent this procedure later in their hospital stay.” The racial differences in treatment were also evident in the use of anti-PCP medications within two days of admission. In VA hospitals, 69.8% of blacks, 75.0% of Hispanics and 76.6% of whites received the medication; in non-VA hospitals the respective rates were 77.0%, 74.0% and 76.2%.

Not surprisingly, “the unadjusted relative odds of dying during the hospitalization for black patients at non-VA hospitals were approximately 70% higher than those for white patients at non-VA hospitals (relative odds, 1.71; 95% CI, 1.23 to 2.45; P<.001).” Though disease severity accounted for this higher mortality rate in most cases, “for Hispanic patients at non-VA hospitals, adjustment for patient, disease, and hospital characteristics accounted for most but not all, of their higher relative odds of in-hospital death.”