Semiparametric Bayesian models for clustering and classification in presence of unbalanced in-hospital survival

Keywords

Statistical learning
Health Analytics
Code:
21/2012
Title:
Semiparametric Bayesian models for clustering and classification in presence of unbalanced in-hospital survival
Date:
Tuesday 1st May 2012
Author(s):
Guglielmi, A.; Ieva, F.; Paganoni, A.M.; Ruggeri, F.; Soriano, J.
Download link:
Abstract:
In this work, Bayesian semiparametric logit models are fitted to grouped data related to in-hospital survival outcome of patients hospitalised with ST-segment Elevation Myocardial Infarction diagnosis. Dependent Dirichlet Process priors are considered for modelling the random-effects distribution of the grouping factor (hospital of admission), in order to provide a cluster analysis of the hospitals. The clustering structure is highlighted through the optimal random partition that minimises the posterior expected value of a suitable loss function. Two are the main goals of the work: to provide model-based clustering and ranking of the providers according to the similarity of their effect on patients’ outcome, and to make reliable predictions on the survival outcome at patient’s level, even when the survival rate itself is strongly unbalanced. The study is within a project, named Strategic Program of Regione Lombardia, and is aimed at supporting decisions in healthcare policies.
This report, or a modified version of it, has been also submitted to, or published on
Journal of the Royal Statistical Society(2014) - Series C, 63 (1): 25-46