• Jean-Philippe Durand, Olivier Mir, Romain Coriat, Anatole Cessot, Sylvain Pourchet, and François Goldwasser.
• Department of Medical Oncology, University Hospital Cochin, Université Paris Descartes, Assistance Publique–Hôpitaux de Paris, 27, Rue du Faubourg Saint Jacques, 75014 Paris, France. jean-philippe.durand@cch.aphp.fr
• Support Care Cancer. 2012 Apr 1;20(4):857-64.

BackgroundOncologists often overestimate survival of advanced cancer patients. This study aimed to validate a score for survival prediction in terminally ill cancer patients.MethodsBetween 2004 and 2008, a prospective study was performed in 500 consecutive advanced cancer patients referred to a palliative care unit. Evaluation at admission included physical examination and routine blood tests. On a randomly selected training set, independent factors assessable at inclusion predicting 2-week survival by a multiple logistic regression were assigned integer-rounded weights to develop a risk index score, which was tested on a validation set.ResultsOn the training set (334 patients), predictive factors were: urea >12 mmol/L (weight = 5, odds ratio (OR) = 3.72, 95% confidence interval (95%CI) = [1.59; 8.71], p = 0.002), Karnofsky Performance Status ≤30% (weight = 4, OR = 3.28, 95%CI = [1.80; 6.01], p < 0.001), leucocytes >15 g/L (weight = 3, OR = 2.49, 95%CI = [1.18; 5.25], p = 0.017), transthyretin ≤0.05 g/L (weight = 3, OR = 2.42, 95%CI = [1.16; 5.04], p = 0.019) and male gender (weight = 2, OR = 2.25, 95%CI = [1.28; 3.97], p = 0.005). On the validation set (166 patients), the Cochin Risk Index Score (CRIS) ≥ 7 identified high-risk patients, with a positive predictive value of 78%.ConclusionWe validated the CRIS for survival prediction in terminally ill cancer patients.

21 keywords...

hide…