Prognostic Factors in Cirrhosis Identified by Cox's Regression Model
- 21 September 2007
- journal article
- clinical trial
- Published by Ovid Technologies (Wolters Kluwer Health) in Hepatology
- Vol. 3 (6), 889-895
- https://doi.org/10.1002/hep.1840030601
Abstract
In a controlled clinical trial in 488 patients with chronic liver disease treated with prednisone or placebo, survival data were analyzed using Cox's proportional hazards model. A total of 162 variables were screened separately for prognostic and/or therapeutic effect by log‐rank analyses, whereby 46 clinical, biochemical, serological, and histological variables were isolated. Another five variables traditionally found to be important in patients with liver disease were included. After extensively checking the assumptions of the model, the 51 variables were, by a step‐wise procedure, reduced to a final model. It comprised, besides a treatment indicator, 12 variables with significant (p < 0.05) prognostic or therapeutic effect. The following eight variables had a significant prognostic effect: sex, age, prothrombin, acetylcholinesterase, eosinophil leucocytes in liver parenchyma, liver cell necrosis, inflammation in liver connective tissue, and efferent veins in parenchymal nodules. A prognostic index based on the final model is formed allowing calculation of 5 years survival probability. The usefulness of the prognostic index was tested by a cross‐validation method, and no statistical significant difference was found between the estimated and observed survivorship functions.This publication has 10 references indexed in Scilit:
- Azathioprine Versus Prednisone in Chronic Active Hepatitis and Non-Alcoholic CirrhosisScandinavian Journal of Gastroenterology, 1982
- Testing Goodness of Fit of Cox's Regression and Life ModelBiometrics, 1982
- Clinical relevance of restrictive morphological criteria for the diagnosis of cirrhosis in liver biopsiesLiver International, 1981
- An Application of Cox's Proportional Hazard Model to Multiple Infection DataJournal of the Royal Statistical Society Series C: Applied Statistics, 1979
- Covariance Analysis of Heart Transplant Survival DataJournal of the American Statistical Association, 1977
- Proportional Hazard Regression Models and the Analysis of Censored Survival DataJournal of the Royal Statistical Society Series C: Applied Statistics, 1977
- Analysis of Survival Data under the Proportional Hazards ModelInternational Statistical Review, 1975
- Sex, Ascites and Alcoholism in Survival of Patients with CirrhosisThe New England Journal of Medicine, 1974
- Linear Statistical Inference and its ApplicationsWiley Series in Probability and Statistics, 1973
- Asymptotically Efficient Rank Invariant Test ProceduresJournal of the Royal Statistical Society. Series A (General), 1972