Doctors' Recognition of Approaching Death in Terminal Patients Seen Unreliable

By Will Boggs, MD
NEW YORK (Reuters Health) Aug 17 - Physician estimates of survival in terminally ill cancer patients are quite imprecise, especially for patients near death, according to findings published in the August 1st issue of the Journal of Clinical Oncology.
"We spend a lot of time discussing the prognoses of patients in order to tailor palliative therapy to the individual life expectancy," Dr. Stephan Gripp told Reuters Health. "I was especially astonished that we were obviously unable to recognize those persons who died very soon. I previously believed that prognostication would be more difficult and imprecise with longer survival."
Dr. Gripp from the University Hospital Duesseldorf, Germany, and colleagues investigated how survival of palliative cancer patients relates to subjective prediction of survival, objective prognostic factors, and individual psychological coping.
Survival (for less than 1 month, for 1-6 months, and more than 6 months) was appraised correctly by 61.3% of a pool of physicians, 54.9% of experienced radiation oncologists, and 62.7% of tumor board physicians, the authors report.
Errors in prediction were skewed toward optimistic prognosis by pooled physicians and tumor board physicians, but not by experienced radiation oncologists.
Correct assessments of patients who died within 1 month were rare, the investigators say. From 48% to 78% (depending on physician group) of these patients were predicted to survive 1 to 6 months, and 17% to 23% were predicted to survive more than 6 months.
Based on clinical variables, favorable prognostic factors included primary colorectal cancer and breast cancer (relative to other primary tumors), the researchers note, whereas metastatic spread to the brain, Karnofsky performance score below 50%, use of morphine, dyspnea, and elevated white blood cell count conveyed a worse prognosis.
"Do not offer any specific date of death," Dr. Gripp advised. "Considering the potentially devastating consequences, honest communication on prognosis must be vague, emphasizing the enormous uncertainty. Doctors are not necessarily better than other persons in prognostication."
J Clin Oncol 2007;25:3313-3320.