May 24, 1993

Discussion of Preferences for Life-Sustaining Care by Persons With AIDSPredictors of Failure in Patient-Physician Communication

Author Affiliations

From the Section on Health Services and Policy Research, Division of General Medicine, Department of Medicine, Brigham and Women's Hospital (Drs Haas, Weissman, and Epstein), the Department of Health Care Policy, Harvard Medical School (Drs Haas, Weissman, Cleary, Gatsonis, and Epstein), Harvard Community Health Plan (Dr Goldberg), Boston Department of Health and Hospitals and the Department of Epidemiology and Biostatistics, Boston University School of Public Health (Dr Seage), the Center for Survey Research, University of Massachusetts (Drs Fowler and Massagli), and the Department of Medicine, Beth Israel Hospital (Dr Makadon), Boston, Mass.

Arch Intern Med. 1993;153(10):1241-1248. doi:10.1001/archinte.1993.00410100067010

Objectives:  To assess the determinants of communication about resuscitation between persons with acquired immunodeficiency syndrome (AIDS) and their physician.

Design and Setting:  Structured patient interview at a staff-model health maintenance organization (HMO), an internal medicine group practice at a private teaching hospital, and an AIDS clinic at a public hospital.

Patients:  289 persons with AIDS.

Main Results:  Only 38% of patients had discussed their preferences for resuscitation with their physician. Using logistic regression, we found that patients were less likely to have discussed resuscitation with their physician if they were nonwhite (odds ratio [OR], 0.49; 95% confidence interval [CI], 0.24 to 0.99), had never been hospitalized (OR, 0.52; 95% CI, 0.27 to 0.99), or were cared for in the HMO (OR, 0.44 relative to the private teaching hospital; 95% CI, 0.23 to 0.82). Patients were more likely to have discussed their preferences if they were not currently taking zidovudine (OR, 1.76; 95% CI, 1.02 to 3.03) and if they had decided to defer life-sustaining therapy (OR, 2.30; 95% CI, 1.35 to 3.91). Among nonwhites, those with a nonwhite physician were more likely to have discussed resuscitation (OR, 4.38; 95% CI, 1.13 to 16.93). Of patients who had not discussed their preferences for life-sustaining care, 72% wanted to do so. Patient desire for discussion of this issue did not vary by race, severity of illness, hospitalization status, use of zidovudine, or site of care.

Conclusions:  A majority of persons with AIDS in this study had not discussed their preferences for life-sustaining care with their physician, despite the desire to do so. Interventions to improve patient-physician communication about resuscitation for nonwhites and other groups at risk of inadequate discussion might lead to clinical decisions that are more consistent with patient preferences.(Arch Intern Med. 1993;153:1241-1248)