In a study of decisions not to treat febrile patients, we reviewed the medical records of 1256 people admitted to nine extended-care facilities in Seattle during 1973. Fever, defined as two temperatures of 38.33°C to 38.83°C (101 to 101.9°F) within 24 hours or one temperature ≥38.88°C (102°F), developed in 190 patients before two years of stay. Active treatment, defined as antibiotics or hospitalization (or both), was ordered for fever in 109 patients, of whom 10 (9 per cent) died. Active treatment was not ordered for 81 patients, of whom 48 (59 per cent) died. The pre-decision factors that showed a significant relation (P<0.05) to such nontreatment were: diagnosis, mental status, mobility, pain, narcotics prescribed, size of the facility, relation of the physician to the patient and medical-record statements documenting the patient's deterioration or plans for nontreatment in general. This pattern of nontreatment suggests that physicians and nurses did not intend to treat these patients actively and that high mortality was expected. (N Engl J Med 300:1246–1250, 1979) PHYSICIANS have been accused of prolonging life at any cost. However, surveys of health professionals have found that many (50 to 70 per cent) are disposed to withdraw or withhold life-prolonging treatment; some (30 per cent) would favor euthanasia in certain settings.1,2 The experience of one of us (N.K.B.) and conversations with other primary-care physicians confirmed many anecdotal examples of physicians’ willingness not to treat life-threatening illness actively under certain circumstances. Attempting to obtain more useful information, we reviewed the medical records of patients dying in extended-care facilities and found that a frequent cause of death was untreated infection (i.e.,. © 1979, Massachusetts Medical Society. All rights reserved.
