Flexible parametric survival model for analysing censored time-to-event data among tuberculosis patients
No Thumbnail Available
Files
Date
2024-12-01
Journal Title
Journal ISSN
Volume Title
Publisher
University of Mosul
Abstract
Tuberculosis (TB) is a globally deadly infectious disease responsible for 10 million new cases and 1.5 million deaths annually. Shorter TB treatment regimens show promise in reducing this problem, but there is an improved treatment success rate in South Africa, while retreatment cases remain a concern. An important feature of time-to-event modelling is its ability to consider transition probabilities of heterogeneous subgroups with different risk profiles. Survival analysis is generally performed to accurately estimate the transition probabilities associated with the risk profiles. This study explored the application of a flexible parametric survival model for analysing censored time-to-event data among TB patients. The data were obtained from East London Central Clinic-TB unit, Eastern Cape, South Africa. In total, 174 patients were included in the analysis. The goodness of fit of the models was explored using Akaike information criterion (AIC). We estimated the hazard ratios (HR) and baseline cumulative hazards of our model, which are necessary to calculate individual transition probabilities, and compared the model with the Cox model and additive hazard model to determine the survival predictions of TB patients.The flexible parametric survival model produced hazard ratio and baseline cumulative hazard estimates that were similar to those obtained using the Cox proportional hazards model. The analysis revealed that sex (HR=0.49, 95% CI: 0.38, 0.62), antiretroviral therapy (ART), (HR=0.53, 95% CI: 0.34, 0.78), and diabetes (HR=0.58, 95% CI: 0.41, 0.78) were all statistically significant factors associated with improved treatment survival in tuberculosis patients. Flexible parametric survival models are a powerful tool for modelling time-toevent data and individual transition probabilities. It is of great importance to fit models by modelling the baseline, which makes it easier to make different types of predictions and allows for non-proportional hazards since it is an interaction.
Description
Keywords
Tuberculosis -- Patients, Survival analysis (Biometry), Proportional hazards models, Medical statistics
Citation
Azeez, A., Osuji, G., Alakija, T., Odeyemi, A., Mutambayi, R. and Madu, P. (2024) ‘Flexible parametric survival model for analysing censored time-to-event data among tuberculosis patients’, Iraqi Journal of Statistical Sciences, 21(2), pp. 162–177.