PURPOSE

The use of the palliative prognostic index (PPI) when used only at an initial assessment might be inappropriate as a prognostic tool because it does not reflect the patient's clinical course. The purpose of this study was to assess the utility of PPI score change (∆score) between two assessments as a prognostic tool in terminally ill cancer patients categorized as having a poor prognosis.

METHODS

A total of 1,035 terminally ill cancer patients categorized as having a poor prognosis (initial PPI score >6) under palliative care between January 2006 and December 2011 at a single medical center in Taiwan were selected. Patients were categorized by magnitude of ∆score between the initial PPI and week 1 PPI assessments into five groups (<-20, -20 to 0, 0, 0 to 20, and >20 %) for survival analysis.

RESULTS

The median survival was 22 days (range, 8-180 days) in all patients. Median survival duration was 78, 32, 23, 17, and 14 days, and the death rate at the study end was 78.9, 87.1, 96.2, 100, and 100 % in each group, respectively. The c-statistic value for predicting life expectancy less than 30, 60, and 90 days was significantly higher with magnitude of ∆score than with the initial PPI score (p < 0.05).

CONCLUSIONS

Magnitude of PPI score change within 1-week interval provides a significant difference in survival prediction and is more reliable than initial PPI alone to identify terminally ill cancer patients with better outcome potential in those patients considered to have a poor prognosis.