Predicting survival in patients with advanced cancer in the last weeks of life: How accurate are prognostic models compared to clinicians’ estimates?

作者：David Hui, Jeremy Aaron Ross, Minjeong Park, Rony Dev, Marieberta Vidal, Diane Liu, Carlos Eduardo Paiva, Eduardo D Bruera · 发表于：Palliative Medicine · 年份：2019 · DOI：10.1177/0269216319873261 · 被引用次数：53 · 研究领域：Palliative Care and End-of-Life Issues、Management of metastatic bone disease、Pain Management and Opioid Use

BACKGROUND: It is unclear if validated prognostic scores such as the Palliative Performance Scale, Palliative Prognostic Index, and Palliative Prognostic Score are more accurate than clinician prediction of survival in patients admitted to an acute palliative care unit with only days of survival. AIM: We compared the prognostic accuracy of Palliative Performance Scale, Palliative Prognostic Index, Palliative Prognostic Score, and clinician prediction of survival in this setting. DESIGN: This is a pre-planned secondary analysis of a prospective study. SETTING/PARTICIPANTS: We assessed Palliative Performance Scale, Palliative Prognostic Index, Palliative Prognostic Score, and clinician prediction of survival at baseline. We computed their prognostic accuracy using the Concordance index and area under the receiver operating characteristics curve for 7-, 14-, and 30-day survival. RESULTS: A total of 204 patients were included with a median overall survival of 10 days (95% confidence interval: 8-11 days). The Concordance index for Palliative Performance Scale, Palliative Prognostic Index, Palliative Prognostic Score, and clinician prediction of survival were 0.74, 0.71, 0.70, and 0.75, respectively. The areas under the curve for these approaches were 0.82-0.87 for 30-day survival, 0.75-0.80 for 14-day survival, and 0.74-0.81 for 7-day survival. The four prognostic approaches had similar accuracies, with the exception of 7-day survival in which clinician prediction of survival was ...