Using Joint Models To Study The Association Between CD4 Count And The Risk Of Death In TB/HIV Data

Résumé

Abstract Background: Joint modeling is an active area of research which has seen an increasing interest in medical research for various dynamic medical scenarios for studying possible relationships between longitudinal biomarkers and survival outcomes. The association structure that links the survival and the longitudinal sub-models is of great importance within the joint modeling framework, nevertheless, rationale for selecting it has received relatively little attention in the literature. In light of this, our main aim is to explore alternative functional forms of the association structure linking the CD4 count and the risk of death and provide rationale for selecting the optimal association structure for our data. Methods: We used data from the study conducted by the Centre for the AIDS Programme of Research inSouth Africa (CAPRISA) AIDS Treatment programme, the Starting Antiretroviral Therapy at Three Points in Tuberculosis (SAPiT) study, which was an open-label, three armed randomised, controlled trial conducted between June 2005 and July 2010 (N=642). We considered five association structures and to select the final model we utilized the Deviance Information Criterion (DIC), where smaller values indicate better model adjustments to the data. Results: We observed similar characteristics of the study participants across the three study arms. The median CD4 count was 154.5, 149.0 and 140.0 in the early integrated, late integrated and sequential arms, respectively. The random effects association structure was found to be the best functional form for our data, in particular, the baseline levels as well as the longitudinal evolution of the underlying CD4 count were found tobe strongly associated with the risk of death. Conclusions: In this paper we have shown that the "current value" association structure is not always the bests tructure that expresses the correct relationship between the outcomes in all settings, which is why it is crucial to explore alternative clinically meaningful association structures that links the longitudinal and survival processes. Keywords: Time-to-event data; longitudinal data; joint models; CD4 count; mortality; Association structures

Citer ce document

Accès au document

Ce document est hébergé sur son dépôt institutionnel d'origine.

Auteur(s)

Statistiques

Consultations : 1

Téléchargements : 0