keywords: Cox Regression Analysis, dfbeta values, Kaplan Meier plots, Risk factors, Schoenfeld residuals, Survival, Tuberculosis
Tuberculosis is a chronic infectious disease that has been one of the major causes of mortality in Nigeria and Benue State in particular. The objective of this study is to identify the risk factors associated with Tuberculosis, estimate the survival time probabilities of the disease covariates, compare the survival time probabilities with respect to the risk factors and evaluate simultaneously the effect of these factors on survival using Cox Proportional Hazard model. This longitudinal epidemiological study selected two renowned Health Centers in Makurdi Metropolis. Kaplan Meier estimates and plots were used to assess the survival pattern for covariates under study. Multivariate analysis to test the association of independent variables with Tuberculosis mortality was performed using Cox Regression Analysis. Out of the registered patients, some died during the study while others were censored. The Wald, likelihood and Logrank Test were conducted to test for the significance of the model. The Schoenfeld residual and DFBeta plot were used to test the proportional hazard model assumptions and outliers respectively. Analysis of Cox regression parameters and hazard ratio show that age, HIV status, Initial weight and type of Tuberculosis diagnosed had statistically significant difference in survival experience of tuberculosis patients whereas patients’ gender, Tuberculosis category and smear test result show no effect on patients’ survival experience. It was recommended that special concern be given to Tuberculosis patients with HIV Positive status and middle age persons. Also, there is need to strengthen the follow up of patients with Tuberculosis treatment from the day of anti-TB treatment introduction to completion days.