Analysis of Time to Relapse among Schizophrenia Patients in St. Amanuel Mental Specialized Hospital, Addis Ababa, Ethiopia

Abstract

Objective: To investigate the associated factors that affect time to relapse in patients with schizophrenia at St. Amanuel Mental Specialized Hospital, with recurrent events model. Material and Methods: A hospital-based retrospective review of the medical records of 332 patient with schizophrenia, covering a two-year period, were examined. Parametric frailty models were used to determine the correlation between discharge times and relapse, and to identify risk factors using R-software. Results: The distribution number of relapses was 162, and the median survival time of patients with schizophrenia was 665 days. The unobserved heterogeneity in patients and correlation between relapses, as estimated by the Weibull-gamma frailty model, was p-value≤0.001 and Kendall’s Tau (Γ)=0.498. This indicated that there was heterogeneity among participants and a correlation between relapses. The final model showed that the effect of psychiatric comorbidity (hazard ratio (HR)=6.522, p-value≤0.001), employment status (HR=5.334, p-value=0.001), history of suicide attempt (HR=2.167, p-value=0.003) and history of traditional treatment (HR=1.973262, p-value=0.021) had a significant effect on the hazard time to relapse. Onset Age of schizophrenia and drug adherence were not predictive. Subsequent relapses are likely dependent on both the first and previous relapses. Conclusion: Comorbidity, employment status, history of suicide attempt and history of traditional treatment are imperative risk factors influencing the hazard of time to relapse, which increases the risk of relapse.

Keywords

• frailty models
• proportional hazard model
• recurrent events
• relapse
• schizophrenia patients