Predicting prognosis in patients with advanced cancer: A prospective study.

BACKGROUND: Prognosis is one of the most challenging questions with which physicians are confronted. Accuracy in the prediction of survival is necessary for clinical, ethical, and organizational reasons.

AIM: Evaluate young doctors' clinical prediction of survival and the aids they could get: expert opinion, Palliative Prognostic score, and Palliative Prognostic Index.

DESIGN: Prospective, observational study.

SETTING/PARTICIPANTS: Advanced cancer patients under observation of an inhospital palliative care team, from April to July 2014. A total of 38 patients were included, mostly male (65.8%), average age 68.5 years. Average survival time was 24 days. Follow-up concluded with death or after 90 days.

RESULTS: Young doctors' clinical prediction of survival was adequate at 10.5%, with 55.3% severe errors in an optimistic direction. Palliative care experts were more adequate (23.7%) and made less severe errors (42.1%). Palliative Prognostic score and Palliative Prognostic Index were even more adequate (47% and 55%, respectively) and made even less severe errors (0% and 11%, respectively). The best correlation with observed survival was achieved when palliative care experts used palliative prognostic score ( rs  = -0.629; p < 0.01).

CONCLUSION: Young doctors' clinical prediction of survival is often inadequate. Palliative Prognostic score, which includes clinical prediction of survival, calculated by palliative care experts had the best performance. Our results support the recommendation of using clinical prediction of survival together with prognostic scores.