A prospective, multicenter cohort study to validate a simple performance status-based survival prediction system for...

Area of Science:

• Oncology
• Palliative Care
• Prognostic Indices

Background:

• Survival prediction systems like the Palliative Prognostic Index (PPI) are crucial for terminally ill patients.
• Oncologists are more familiar with the Eastern Cooperative Oncology Group (ECOG) performance status (PS) than the Palliative Performance Scale (PPS).
• The Performance Status-Based Palliative Prognostic Index (PS-PPI) was developed as a modified PPI utilizing ECOG PS for easier oncologist adoption.

Purpose of the Study:

• To validate the PS-PPI, a simplified survival prediction tool for oncologists.
• To compare the accuracy and utility of PS-PPI against the established PPI.

Main Methods:

• A multicenter, prospective cohort study involving 2346 patients across 58 palliative care services in Japan.
• Physicians calculated both PS-PPI and PPI for all enrolled patients.
• Comparative analysis of survival predictions, sensitivity, and specificity between PS-PPI and PPI, including subgroup analyses for different care settings and chemotherapy status.

Main Results:

• Survival predictions significantly differed between the PPI and PS-PPI risk groups (P < .001).
• The PS-PPI demonstrated higher sensitivity, while the PPI showed higher specificity.
• Both indices exhibited strong predictive performance with areas under the receiver operating characteristic curves >0.78 for survival prediction from 3 weeks to 180 days.

Conclusions:

• The PS-PPI is as accurate as the PPI in predicting prognosis for advanced cancer patients.
• PS-PPI proves effective for both short- and long-term survival predictions.
• The PS-PPI is a valuable tool for predicting survival, including for patients undergoing chemotherapy.