1Objectives: Physician recommendations for further medical treatment or palliative treatment only at the end of life may influence patient decisions. Little is known about the patient characteristics that affect physician-assessed quality of life or how such assessments are related to subsequent recommendations. Design, Setting, and Subjects: A 2010 mailed survey of practicing U.S. physicians (1,156/1,878 or 62% of eligible physicians responded).
Measurements and Main Results:Measures included an end of life vignette with five experimentally varied patient characteristics: setting, alimentation, pain, cognition, and communication. Physicians rated vignette patient quality of life on a scale from 0 to 100 and indicated whether they would recommend continuing full medical treatment or palliative treatment only. Cognitive deficits and alimentation had the greatest impacts on recommendations for further care, but pain and communication were also significant (all p < 0.001). Physicians who recommended continuing full medical treatment rated quality of life three times higher than those recommending palliative treatment only (40.41 vs 12.19; p < 0.01). Religious physicians were more likely to assess quality of life higher and to recommend full medical treatment. Conclusions: Physician judgments about quality of life are highly correlated with recommendations for further care. Patients and family members might consider these biases when negotiating medical decisions. (Crit Care Med 2016; XX:00-00) Key Words: critical care; end-of-life care; medical decision-making; physician recommendations; quality of Life; withdrawal of care P atient quality of life may play an important role in physician recommendations for further life sustaining treatments at the end of life. The limited efficacy of such treatments (1, 2) and an emphasis on quality of life together have led to greater use of palliative care resources and more frequent decisions to limit care (3-7). Physician recommendations often play an integral role in such choices, as patients require information about prognosis, treatment efficacy, and quality of life to make informed decisions (8, 9). How physicians perceive and make judgments about patient quality of life may impact the recommendations physicians offer, but few studies have investigated this at a national level.Prior studies have observed significant associations between patient quality of life and do-not-resuscitate (DNR) orders (10-13), but this may be due to patient-as opposed to physician-preference. Some studies of DNR orders neglected to consider patient quality of life (14, 15), whereas others were small (15-18) or in foreign countries (11,17). In a small Swiss study, physicians explicitly considered quality of life when implementing DNR orders in as many as 71% of cases (17). In the United States, single-center and regional studies have observed associations between physician assessment of quality of life and physician treatment preferences for patients (16,18). Yet no nationally representative study...