Validation of Scoring System for Survival Prediction in Terminal Cancer Patients in Korea

Lee, In Chul; Kim, Chan Kyong; Suh, Sang Yeon; Kim, Young Sung; Cho, Kyung Hee; Kang, Hee Cheol; Youn, Bang Bu

Abstract

BACKGROUND: Accurate prediction of survival in terminal cancer patients is important for planning effective palliative care. But, the prediction of survival most often relies on the physicians' prediction. Recently, simple prognostic scores such as Palliative Prognostic Index and Palliative Prognostic Score have been developed to estimate duration of survival. The aim of this study was to validate these prognostic scores and physicians' prediction for terminally ill cancer patients in Korea to determine its value in clinical practice.
METHODS
The subjects of this study were 40 terminal cancer inpatients of one hospital who died between March to May 2005. All patients' Palliative Prognostic Index, Palliative Prognostic Score, and physicians' prediction were recorded on admission by a physician.
RESULTS
When a Palliative Prognostic Index of more than 6 was adopted as a cut-off point 3 weeks' survival was predicted with a sensitivity of 90% and a specificity of 60%. When the three groups were grouped by Palliative Prognostic Scores (group A: < or =5.5, group B: 5.6~11, group C: >11), the 30 day survival probability were 60% for group A, 16.7% for group B, and 4.3% for group C, respectively. Physicians' prediction showed moderate correlation (correlation coefficient=0.604) with actual survival and had a tendency to overestimate survival.
CONCLUSION
Palliative Prognostic Index was proved to be a reliable survival prediction tool in Korea. Palliative Prognostic Score had a tendency to overestimate survival as compared with other studies. Physicians' prediction had a correlation with actual survival, and it was presumed to be more accurate when combined with other prognostic score.