A diagnostic model for impending death in cancer patients: Preliminary report

作者：Hui David^*; Hess Kenneth; dos Santos Renata; Chisholm Gary; Bruera Eduardo

来源：Cancer, 2015, 121(21): 3914-3921.

DOI：10.1002/cncr.29602

• death
• diagnosis
• mortality
• neoplasms
• palliative care
• physical examination

BACKGROUNDSeveral highly specific bedside physical signs associated with impending death within 3 days for patients with advanced cancer were recently identified. A diagnostic model for impending death based on these physical signs was developed and assessed.
METHODSSixty-two physical signs were systematically documented every 12 hours from admission to death or discharge for 357 patients with advanced cancer who were admitted to acute palliative care units (APCUs) at 2 tertiary care cancer centers. Recursive partitioning analysis was used to develop a prediction model for impending death within 3 days with admission data. The model was validated with 5 iterations of 10-fold cross-validation, and the model was also applied to APCU days 2 to 6.
RESULTSFor the 322 of 357 patients (90%) with complete data for all signs, the 3-day mortality rate was 24% on admission. The final model was based on 2 variables (Palliative Performance Scale [PPS] and drooping of nasolabial folds) and had 4 terminal leaves: PPS score 20% and drooping of nasolabial folds present, PPS score 20% and drooping of nasolabial folds absent, PPS score of 30% to 60%, and PPS score 70%. The 3-day mortality rates were 94%, 42%, 16%, and 3%, respectively. The diagnostic accuracy was 81% for the original tree, 80% for cross-validation, and 79% to 84% for subsequent APCU days.
CONCLUSIONSBased on 2 objective bedside physical signs, a diagnostic model was developed for impending death within 3 days. This model was applicable to both APCU admission and subsequent days. Upon further external validation, this model may help clinicians to formulate the diagnosis of impending death. Cancer 2015;121:3914-3921. (c) 2015 American Cancer Society.
In this prospective, observational study of hospitalized cancer patients, a diagnostic recursive partitioning model, based on bedside physical signs, has been developed for impending death within 3 days. This model is accurate, is applicable both at admission and during hospitalization, and may help clinicians to formulate the diagnosis of impending death.

• 出版日期2015-11-1