management (Table 9-3). The responses are subjective, but again, they suggest strongly that students are not prepared to care for dying patients as well as they could be during their undergraduate medical education.

The 1997–1998 American Medical Association (AMA) Annual Survey of Graduate Medical Education was sent to 7,861 residency programs (all of those accredited by the Accreditation Council on Graduate Medical Education), of which 96.5 percent responded. The survey asked whether each program had a structured curriculum in end-of-life care. (No more specific definitions of what might be included in an end-of-life curriculum were provided, so the term may have been interpreted differently by different respondents.)

Overall, 60 percent of programs reported that they did have a structured curriculum, but there was tremendous variability among programs in different specialties. Of the types of physicians most likely to care for dying patients

• 92 percent of programs in family practice and internal medicine and 98 percent in critical care medicine reported positively, and

• between 60 percent and 70 percent of programs in obstetrics-gynecology, pediatrics, psychiatry, and surgery reported positively.

The results of these recent surveys suggest that undergraduate medical and residency training lacks adequate content in end-of-life care, but without much detail. One would like to know what topics are covered in end-of-life education, the format (i.e., lectures, discussions, clinical experience), how much time is devoted to each subject, and how well students are prepared by the extent and types of training they receive. This information has not been assembled in a comprehensive way, but pieces of it are ex-