Competing Risk Model for Time to Development of Tuberculosis among Adults on Combination Antiretroviral Treatment

Abstract

The purpose of this study was to identify factors affecting the time to development of tuberculosis in the presence of competing risks. In this case death before developing tuberculosis was deemed a competing risk because it altered the occurrence of the outcome of interest being time to development of tuberculosis from baseline. We used data from a randomized longitudinal clinical trial study called the “Tshepo” study. The “Tshepo” study was a 3-year randomized clinical study following 650 ART-naïve adults (69.4% female) from Botswana who initiated first-line NNRTI-based ART. Participants were assigned in equal proportions (in an open-label, unblinded fashion) to one of 6 initial treatment arms and one of two adherence arms using permuted block randomization. Randomization was stratified by CD4+ cell count (less than 200 cells/mm3, 201 - 350 cells/mm3) and by whether the participants had an adherence assistant. Classical methods such as the Kaplan-Meier method and standard Cox proportional hazards regression were used to analyze survival data ignoring the competing event(s) which may have been inappropriate in the presence of competing risks. The idea was to use competing risk models to investigate how different treatment regimens affect the time to the development of TB and compare the results to those obtained using the classical survival analysis model which does not account for competing risks. Amongst 38 patients who died 15.8% of them developed tuberculosis whilst 84.2% of those who died did not develop the outcome of interest. The hazard ratio of treatment C was 1.069 implying that the risk of developing TB in patients taking treatment C is about 6.9% higher compared to those taking treatment A having adjusted for baseline age, baseline BMI, baseline CD4, Hemoglobin and gender. Similarly, after accounting for competing risks the hazard ratio for treatment C was about 1.89 implying that the risk of developing TB amongst those taking treatment C was about 89% higher as compared to those taking treatment A. From the obtained results it was thus concluded that the standard Cox model of time to event data in the presence of competing risks underestimated the hazard ratios hence when dealing with data with multiple failure events it is important to account for competing events.