The purpose of this study was to determine whether African Americans come to intensive care units (ICUs) sicker than their white counterparts, and whether, after adjustment for patient characteristics, there are racial differences in ICU length of stay, therapy, and hospital mortality rates.

Data were collected from a nationwide sample of hospitalized adult ICU patients; hospitals with more than 200 beds were selected. Data were collected for the period from May 1988 to February 1990.

There were three main findings:

1. There were differences by race in the characteristics of patients treated in ICUs. Compared with whites, African Americans were younger, more likely to be admitted from an emergency department rather than operating/recovery rooms, and more severely ill (according to Acute Physiology Score, receipt of a high APACHE III score (>90), severely compromised activities of daily living, and prior chronic renal disease, dialysis treatment, diabetes mellitus, and injection drug use). Insurance status also differed significantly by race.

2. Both the reason for use of the ICU and the type of treatment received in the ICU differed by race. Compared with whites, non-operative ICU admissions were more frequent and elective surgery was less frequent among African Americans. It is noteworthy that 10% of the white patients in the ICU were admitted for major vascular surgery, while only 2.6% of the African Americans were admitted for major vascular surgery. Since African Americans have higher risk factors for diseases that might lead to major vascular surgery, we must consider how access to treatment and access to appropriate treatment influence the results in this study. However, 11 of the 14 most common reasons for admission were the same for both racial groups.

Evaluation of monitoring and therapeutic interventions while in the ICU found several racial patterns (the largest difference was in nurse monitoring of hourly vital signs: 77.0% of African Americans versus 86.4% of whites), and mean “therapeutic intervention scores” were significantly different both on the first day and over the first week in the ICU. However, the authors stated that these differences in resource use were not clinically significant. Length of stay was longer for African Americans, but there was only a 0.34-day difference. The difference was small even after adjusting for patient and hospital characteristics.

3. There were racial differences in outcome as measured by actual survival rate and difference between actual and predicted survival rates at discharge. Insurance status had an effect on predicted mortality.

This study was limited to patients who were admitted to an ICU, and the authors conclude that “once a patient is admitted to an ICU, race appears to have little effect on therapy and no effect on patient outcomes.”