Combination of initial palliative prognostic index and score change provides a better prognostic value for terminally ill cancer patients: a six-year observational cohort study.

Abstract

The Palliative Prognostic Index (PPI) is among the most popular scores used to predict life expectancy in terminally ill patients worldwide. PPI assessed on the first day of palliative care might be inappropriate because the contribution from subsequent changes in a patient's condition are not taken into account. The aim of this study is to determine the utility of sequential PPI assessments as a better prognostic tool for patients with terminal cancer. In total, 2392 terminally ill cancer patients with initial and one-week PPI assessments under the palliative care consultation service between January 2006 and December 2011 at a single medical center in Taiwan were selected. Patients were categorized into initial PPI, Week 1 PPI, score change (initial PPI - Week 1 PPI; Δscore), and combined initial PPI and Δscore subgroups for survival analysis. Overall median survival was 32 days (range eight to 180 days), and 2183 patients (91.3%) died within 180 days of palliative care consultation service care. A significant difference in survival was observed among patient subgroups (P < 0.001). Subgroup survival analysis showed significant difference in patients with Δscores >0, 0, and <0 in each prognostic group categorized by initial PPI. The c-statistic for predicting life expectancy <30 days was significantly higher with the combined initial PPI and Δscore (c-statistic, 0.71; 95% CI, 0.694-0.731) than with the initial PPI (c-statistic, 0.63; 95% CI, 0.61-0.65), Week 1 PPI (c-statistic, 0.67; 95% CI, 0.652-0.690), or Δscore (c-statistic, 0.64; 95% CI, 0.62-0.66). Combination of initial PPI and score change is more useful than initial PPI for identifying patients with poor outcomes in good prognostic groups and patients with better outcomes in poor prognostic groups.