Decisions about Resuscitation: Inequities among Patients with Different Diseases but Similar Prognoses

Abstract

STUDY OBJECTIVE: To assess whether decisions about "do-not-resuscitate" (DNR) orders are made equitably in patients with different diseases but similar prognoses. DESIGN: Retrospective cohort study. SETTING: Three teaching hospitals: a university referral center, a county hospital serving a largely indigent population, and a Veterans Administration hospital. PATIENTS: Consecutive patients with any of the four following discharge diagnoses: the acquired immunodeficiency syndrome (AIDS) (100 patients); unresectable non-small-cell lung cancer (51 patients); cirrhosis with esophageal varices (51 patients); and severe congestive heart failure with coronary artery disease (115 patients). MEASUREMENTS AND MAIN RESULTS: Do-not-resuscitate orders were written for 52% of patients with AIDS and 47% of patients with cancer but for only 16% of patients with cirrhosis and 5% of patients with congestive heart failure (P less than 0.0001). Although DNR orders were associated with functional and mental status, reason for admission, and severity of illness, the strong association between DNR orders and disease category persisted after adjustment for these potential confounders by multiple logistic regression. A survey of housestaff showed that DNR orders were discussed more frequently with patients who had AIDS or lung cancer than with patients who had cirrhosis or heart failure, despite an accurate understanding of the generally similar prognoses among the four groups. CONCLUSIONS: Despite relatively similar prognoses, patients with AIDS or lung cancer are much more likely to receive DNR orders than patients with cirrhosis or severe congestive heart failure. This discrepancy cannot be explained by differences in severity of illness among patients or by misunderstandings of prognosis by clinicians. From our data, we cannot determine if patients with cirrhosis or heart failure receive too few DNR orders or if patients with AIDS or lung cancer receive too many. Our findings should encourage physicians to determine the preferences of patients about life-sustaining treatments more equitably.