Key points are not available for this paper at this time.
OBJECTIVE: To evaluate variations in patient characteristics, hospital mortality, intensive care unit (ICU) length of stay, and treatment among African-American and white patients admitted to the ICU. DESIGN: Prospective, inception cohort study. SETTING: Forty-two ICUs at 40 U.S. hospitals, including 26 hospitals that were randomly selected and 14 volunteer institutions, primarily large university or tertiary care centers. PATIENTS: A consecutive sample of 17,440 ICU admissions. MEASUREMENTS AND MAIN RESULTS: Selected demographic, physiologic, and treatment information for an average of 415 admissions at each ICU, and payor information at 36 of 40 hospitals. Outcomes were compared using the ratio of observed to risk-adjusted predicted hospital mortality rate, ICU length of stay, and resource use during ICU day 1 and the first seven ICU days. Compared with 14,006 white patients admitted to the ICU, 2,450 African-American patient admissions were significantly (p < .0001) younger, had a higher mean severity of disease, and a greater proportion of nonoperative and emergency department admissions. African-Americans had fewer life-threatening Acute Physiology and Chronic Health Evaluation III (APACHE III) comorbidities, but a higher prevalence of severe compromise in activities of daily living, diabetes mellitus, chronic renal disease, and intravenous drug abuse. There was no significant racial difference in risk-adjusted hospital mortality rate. For African-Americans, adjusted ICU length of stay was significantly (p < .0003) shorter, and the first 7 days of resource use was significantly (p < .0004) lower, but the differences were small (3% to 4%). CONCLUSIONS: After adjusting for variations in patient characteristics at ICU admission, race has no significant effect on hospital survival. The small but statistically significant differences in adjusted ICU length of stay and resource use could indicate undertreatment for African-Americans or overtreatment for whites.
Williams et al. (Sat,) studied this question.