To Which Extent Do Physician Orders for Life-Sustaining Treatment (Polst) Reflect Patients' Preferences of Care at the End of Life?

Abstract

To assess to which extent Physician Orders for Life-Sustaining Treatment (POLST) reflect patients' preferences of care at the end of life. Two independent researchers interviewed hospitalized adult patients for whom one of their physicians would not be surprised if the patient died within one year. One interviewer followed an adapted version of the “POLST conversation” script produced by the Coalition for Compassionate Care of California and the other followed the approach proposed by Sudore & Fried for advance care planning conversations. Patients were classified as belonging to one of four groups according to their preferences of care at the end of life as assessed during each interview: 1) comfort care only, 2) limited treatment, 3) full treatment for a limited period of time, 4) full treatment for an unlimited period of time. We tested the hypothesis that the agreement between those interviews would yield a weighted kappa coefficient of at least 0.75. Sixty-two patients were included in the study. Mean age was 62 years (SD: 14y) and 21 (34%) patients were women. Median number of years of schooling was 5 years (IQR 4-11). The second interview happened within 1 to 3 days after the first interview. Median score at the Palliative Performance Scale was 70% (IQR 50-80%). The raw proportion of agreement regarding the classification of patients concerning their treatment preferences at the end of life was 95.2%, with a kappa coefficient of 0.89 (95% CI: 0.77-1.00). We provide evidence supporting the POLST paradigm as a means to translate patients' preferences of care into medical orders consistent with those preferences.