(28.2 percent). For leukemia, pediatrics and oncology-hematology helpful scores were similar (21.2 percent and 20.5 percent, respectively), and the lowest score was in family and primary care medicine (10.3 percent).
The analysis of key end-of-life index words showed an overall paucity of references, consistent with the content domain analyses.
A physician consulting a textbook on the treatment of a potentially fatal condition is most likely to find no specific information that will help care for the patient who does, indeed, die. In a minority of cases, useful information may be found. In both studies (Carron et al., 1999; Rabow et al., 2000), the scoring was generous, erring on the side of giving higher rather than lower scores, so even these scores may overestimate the useful content. The investigators also did not rate how useful or complete the information was. However, Carron and colleagues found that more often than not, when information about prognostic factors and disease progression was present, it was vague and would not be helpful in caring for a patient (e.g., the admonition that “supportive care is all that can be offered at this point”).
Knowing that many physicians have little experience with dying and little training to help them, Carron and colleagues commented that “standard reference textbooks should provide at least the essentials of good practice.” Yet, in fact, physicians cannot rely on these texts for much-needed information: on advance care planning, decisionmaking, the effect of death and dying on a patient’s family, or symptom management. Most texts do not describe the way that people with a disease generally die.
The findings from these textbook reviews are so stark that they cannot, and in fact have not, been ignored. Partly in response to these studies, some textbook publishers have commissioned updates for particular chapters. In addition, the Robert Wood Johnson Foundation has begun a Textbook Revisions Project with the goal of working with publishers and editors to ensure that end-of-life chapters are added to textbooks and that end-of-life information is added to other chapters as appropriate (Gibson, 2000).
In 1998, American Society of Clinical Oncology conducted the first and only large-scale survey of U.S. oncologists about their experiences in providing care to dying patients. The questionnaire consisted of 118 questions about end-of-life care under eight headings, one of which was education and training (Hilden et al., 2001). All U.S. oncologists who reported that