A Spatial Temporal Analysis Of Survival Among Tb And Hiv Co–Infected Patients In Erongo Region, Namibi

Abstract

Namibia is among the Sub-Sahara African countries with a high burden of TB and HIV coinfection in the world. The interaction between TB and HIV/AIDS makes the diagnosis and management of the co-infection difficult. The purpose of this study is to identify risk factors, assess spatial and space-time clusters mortality among TB and HIV in Erongo Region. A retrospective cohort study from 2003 to 2017 was carried out using data consisting of 3145 subjects from all the 16 health facilities providing ART and TB management in Erongo Region, Namibia. Firstly, the study used the Kaplan Meier to estimate the survival function. Cox PH model was used to model survival and identify risk factors among TB and HIV co-infected patients. Global Moran’s I and LISA statistics were used to help analyse the spatial distribution and clusters of the disease across settings. Finally, the study used Bayesian Inference in STAR models to assess the spatial and space-time clusters of TB and HIV Mortality. Of the 3145 patients, 424 (13.5%) defaulted treatment, 42 (1.3%) stopped treatment, 542 (17.2%) were transferred out and 1673 (53.2%) were alive at the end of the study. About 464 (14.8%) patients experienced event of interest (death), during the period. Only marital status and sex were not associated with death in the univariate analysis with a p-value of 0.30 and 0.07 respectively. Mortality was higher in Usakos district as compared to the other 3 health district in the region. Mortality worsened with increasing WHO clinical stage. In multivariate analysis, the covariate Function violated the PH assumptions with a p-value of 0.01. Global Moran’s I shows that the EAs tend clusters, 0.17 (p-value