See One, Do One, Teach One?

Abstract

Background: Medical residents commonly discuss resuscitation decisions with hospitalized patients. Previous studies suggest that the quality of these discussions is poor. Objective: To learn about residents' experience with donot-resuscitate (DNR) discussions and their attitudes toward them. Methods: Medical house officers on the wards of three teaching hospitals were eligible to participate. A subset had previously audiotaped actual DNR discussions as part of a study that described the quality of discussions. In a self-administered questionnaire, house officers rated their performance conducting a recent DNR discussion, stated their attitudes, and described their experience learning to talk to patients about these issues. Results: One hundred one (88%) of 115 residents responded to the survey. Eighty-six (90%) of 96 stated they had done a good job with the discussion and 78 (77%) of 101 reported feeling comfortable discussing the topic with patients. Ninety-four (94%) of 100 residents said they discuss code status with all seriously ill patients and while on the medical wards they conduct a median of one DNR discussion per week. On average, they had observed four discussions conducted by more senior clinicians. One third of the residents had never been observed talking to patients about DNR decisions and 71% had been observed two or fewer times. Conclusions: These findings help explain the observations about the quality of DNR discussions. House staff "see" and "do" these discussions, but are not taught through observation and feedback. We recommend that communication about end-of-life treatment decisions be treated as a medical skill to be taught with the same rigor as other clinical procedures. (Arch Intern Med. 1996;156:1285-1289)N