Objective: To assess the long-term clinical impact of a broad-based ethics education program for medical houseofficers with specific emphasis on appropriate care for patients who have do-not-resuscitate (DNR) orders.
Design: Prospective, with an initial randomized phase.
Setting: The medical service of a university teaching hospital.
Participants: Medical houseofficers and their inpatients.
Interventions: A pilot program in 1988, and a full program with a two-year curricular cycle from 1989 to 1991.
Measurements and main results: The authors measured compliance with specific standards of care by reviewing charts of patients who had DNR orders at baseline (n = 39, 1988), after the pilot phase (n = 57, 1989), and at the end of the first curricular cycle (n = 56, 1991), noticing who wrote the DNR order, whether the reasons for the order and appropriate consent were documented, and whether there was documented attention to any of 11 concurrent care concerns (CCCs), such as spiritual needs, the appropriateness of tube feedings or pressors, and adjustment of analgesic dose. The percentage of DNR orders written by houseofficers increased from 26% in 1988 to 67% in 1991 (p < 0.01). The percentage of charts documenting the rationale and consent for the DNR order was consistently high. The percentage of charts documenting attention to any CCC increased from 68% in 1988 to 86% in 1991 (p < 0.01). The mean number of CCCs addressed per DNA order increased from 1.34 in 1988 to 2.14 in 1991. The mean number of CCCs addressed per DNR order for patients who had AIDS increased from 0.89 in 1988 to 2.25 in 1991 (p = 0.03).
Conclusions: The quality of care for patients who had DNR orders, both overall and for those who had AIDS, improved over long-term observation in the setting of an ethics education program for medical houseofficers. The results suggest that ethics education may alter physician practices and improve patient care.