also identifies factors that increased the likelihood that physicians would choose hospice care for themselves and examines how their preferences relate to the timing of end-of-life care discussions with patients.
"Having timely discussions with terminally-ill cancer patients to establish goals for end-of-life care is important to maximize the quality of patient care. But by and large we're not doing a good job at having these discussions early on," says lead author Garrett Chinn, MD, MS, of the Massachusetts General Hospital Division of General Medicine. "We know that patients facing terminal illness often wish to spend their remaining days at home, surrounded by loved ones. Since end-of-life care in the U.S. often stands in stark contrast to these preferences, it's important to identify factors that may facilitate cost-effective care that supports patient preferences."
The current study was conducted as part of the Cancer Care Outcomes Research and Surveillance (CanCORS) study, a consortium of centers across the country measuring the quality of care delivered to 10,000 patients with newly diagnosed lung or colorectal cancer. This report reflects the answers of close to 4,400 physicians caring for cancer patients - including primary care physicians, surgeons, oncologists, radiation oncologists and other specialists - to two survey questions.
Respondents were asked to indicate how strongly they agreed with the statement, "If I were terminally ill with cancer, I would enroll in hospice." They also were asked when they would discuss hospice care with an asymptomatic patient with terminal cancer who they estimated had 4 to 6 months to live - right away, when symptoms first develop, when no more options are available to treat the cancer, when the patient is admitted to the hospital, or when the patient or family asks about hospice care.
In their response to the question about personal hospice enrollment, 65 percent reported strong agreement with the statement, and 21 percent indicated they agreed "somewhat." Physicians who were female, who cared for more terminally ill patients or who worked in managed care settings were more likely to indicate strong agreement, while surgeons and radiation oncologists were less likely than primary care physicians or oncologists to do so.
Only 27 percent of respondents overall indicated they would discuss hospice care with the described patient "now." Waiting until symptoms appear was the choice of 16 percent; 49 percent would bring it up when no more therapeutic options were available; and upon hospital admission or when asked by a patient or family member each were chosen by 4 percent. Among physicians who strongly agreed that they would personally enroll in hospice care, almost 30 percent responded that they would discuss hospice care with the patient 'now,' while about 20 percent of all other respondents would do so.
"Our results suggest that most doctors would want hospice care for themselves, but we know that many terminally ill cancer patients do not enroll in hospice," says Nancy Keating, MD, MPH, of the Harvard Department of Health Care policy, senior author of the report. "In the overall CANCORS study, only about half of the patients who died of metastatic lung cancer had ever discussed hospice care with their physician. Our findings suggest that doctors with more negative personal preferences about hospice care may delay these discussions with patients, which indicates they may benefit from learning more about how hospice can help their patients."
Chinn adds, "Although a physician's personal care preferences may be quite important, we still do a poor overall job having timely end-of-life care discussions with our terminally-ill cancer patients. A lack of knowledge about guidelines for end-of-life care for such patients, cultural and societal norms, or the continuity and quality of communication with patients and family members are also factors that might act as barriers preventing physicians from 'practicing what they preach' in end of life care."