During the last decade, a consensus has emerged among physicians, legal experts, and ethicists that 'do not resuscitate' (DNR) orders (also called 'no code' orders) are appropriate under certain circumstances. Specifically, patients with serious illnesses should make shared decisions with their physicians regarding intensive care and cardiopulmonary resuscitation. Despite this consensus, however, our pilot study suggests that these principles are being applied inconsistently. Our present study will attempt to confirm or refute this impression by determining the frequency of DNR orders for patients with four common diseases with similar prognoses: 1) lung or colon cancer with distant metastates; 2) acquired immunodeficiency syndrome (AIDS); 3) cirrhosis of the liver with a history of esophageal varices; and 4) severe congestive heart failure with a history of ischemic coronary artery disease. We will review the charts of all patients discharged from the medical services of the three teaching hospitals of the University of California, San Francisco (UCSF) and a Health Maintenance Organization (HMO) hospital who belong to one of the four subgroups. Consecutive charts will be reviewed, and baseline admission data will be recorded and patient inclusion determined. Subsequently, the physician's orders will be reviewed to determine whether the patient received a DNR order during the hospitalization and, if so, whether there was an in-hospital deterioration which accounted for the order. A pilot study conducted retrospectively at the Fort Miley Veteran's Hospital showed the frequency of DNR orders without obvious in-hospital precipitants to be approximately: cancer - 40%, AIDS - 30%, cirrhosis - 5%, and congestive heart failure - 5%. This study will determine whether, despite similar prognoses, patients with AIDS and cancer are significantly more likely to receive DNR orders than patients with cirrhosis and heart failure. If this is true, and if physician attitudes or misconceptions about prognosis are responsible for this discrepancy, then principles of shared decision-making are not being applied uniformly across patient groups. Recognition of this discrepancy should encourage physicians to apply these principles more equitably, resulting in more appropriate utilization of resources, lower heath care costs, increased patient autonomy, and more humane terminal care for seriously ill patients.
| Lo, B (1990) Clinical ethics and HIV-related illnesses: issues in treatment and health services research. Med Care Rev 47:15-32 |
| Lo, B; Dornbrand, L (1989) Understanding the benefits and burdens of tube feedings. Arch Intern Med 149:1925-6 |
