Identifying Some Risk Factors of Time to Recurrent Relapses in Bipolar I Disorder Patients using Frailty Model of Survival Analysis

Abstract

Objective: Bipolar I disorder patients often experience relapse once and even more with no limit on number of relapses. The time to relapses of these patients are rarely studied particularly considering heterogeneity across individuals. The aim of this study was to identify some risk factors of time to recurrent relapses in bipolar I disorder patients with a recurrent event model in survival analysis. Methods: In a retrospective longitudinal study, data of medical records of 526 bipolar I disorder patients who had referred to Razi Psychiatric hospital in Tehran from 1993 to 2011 with at least one relapse, without relapse in other hospitals or home in this duration, were investigated and time to recurrent relapses were collected in months. Semi-parametric penalized frailty model which consider whit-in subject correlation and heterogeneity across individuals, was applied to identify the risk factors of recurrent time to relapses. Results: Significant frailty parameter (p<0.001) prove presence of heterogeneity among data. In frailty model the effects of substance abuse (p=0.041), regular fluctuation (p=0.002) and marital status (p=0.009) were significant on the hazard of recurrent times to relapses but other variables showed no significant effect. Conclusions: Substance abuse, marital status and RF are important risk factors in order to plan for postpone the time to next relapses. More studies are required to clear out the effect of other covariates with this model.