New research sponsored by the Agency for Healthcare Research and Quality shows black Medicare patients hospitalized for heart failure or pneumonia in three large States received poorer overall quality of care than other Medicare patients treated for the same illnesses. Racial disparities occurred even in basic hospital services, such as physical exams, simple diagnostic tests, standard drug therapies, and patient history-taking.
The researchers also found that quality of care differences for the black heart failure patients tended to be more pronounced in community hospitals than in teaching facilities (mostly large, urban medical centers associated with medical schools).
The quality disparities may have caused one additional death among every 200 patients treated, according to the study's lead author, Harvard Medical School researcher John Z. Ayanian, M.D. This represents a one-half percent increase in the 30-day death rate of the black patients.
In this study, the researchers found, for example, that only 32 percent of the black pneumonia patients were given antibiotics within 6 hours of admission, compared with 53 percent of the other Medicare patients being treated for the same condition. The black patients also were less likely to have had their blood cultures collected on the first or second day of hospitalization. Prompt administration of antibiotics and collection of blood cultures have been associated with lower mortality rates in prior research.
The researchers also looked at the quality of care being provided to male and female Medicare patients, regardless of race. Overall, the quality was roughly equivalent. However, the men received better care than women from doctors, while the women tended to receive better nursing care than men. The findings are based on reviews by separate panels of physicians and nurses of the medical records of nearly 2,200 Medicare patients 65 years of age and older treated for heart failure or pneumonia in 1991 and 1992 in 501 hospitals in Illinois, New York, and Pennsylvania.
The physicians provided implicit review, meaning they studied each patient's medical record for aspects of care such as prognosis, tests and treatments given, treatment goals, and the discharge plan, following which they rated the patient's care according to a scale ranging from excellent to very poor. The registered nurse panel reviewed patients' records for adherence to explicit process criteria, such as how thoroughly doctors examined patients' hearts and lungs or how closely nurses monitored their blood pressure. All the reviewers were blinded to the researchers' hypothesis that the quality of care would differ by the patients' race and sex. Roughly 35 percent of the heart failure patients and 23 percent of the pneumonia patients were black. The researchers controlled for the patients' age, sex, race, household income, severity of illness on admission, and the characteristics of the hospitals where they were treated.
The authors said that although the data used for the analysis were derived from 1991 and 1992 medical records, clinical practice for heart failure and pneumonia has not changed since that time in systematic ways that would likely alter their findings.
The study was conducted as part of an AHRQ-funded research project (HS06331) to use patient outcomes to assess quality of health care. The project was directed by Harvard School of Public Health researcher Arnold M. Epstein, M.D.
AHRQ is allocating $13.5 million this fiscal year for studies that will help speed the pace of translating research into practice to help reduce or eliminate differences in quality of care. About half this amount is earmarked for projects that specifically address racial and ethnic disparities by identifying and implementing quality improvement strategies focused on minority populations.
For more information, see "Quality of care by race and gender for congestive heart failure and pneumonia," by Dr. Ayanian, Joel S. Weissman, Ph.D., Scott Chasen-Taber, Ph.D., and Dr. Epstein, in the December 1999 issue of Medical Care 37(12), pp. 1260-1269.