A competing risk analysis of predictors of time to lost to follow-up among adults with TB/HIV coinfection in Bahir Dar

Lost to follow-up (LTFU), defined as interrupting anti-TB treatment for ≥ 8 consecutive weeks or missing anti-retroviral therapy (ART) appointments for > 90 days, is a barrier to TB/HIV coinfection management. Poor treatment adherence, a driver of multidrug resistance in TB/HIV, poses critical challenges to case management. Overestimating effect sizes when considering mutually exclusive events, like LTFU from ART and anti-TB treatment, can occur if sources of error are not properly accounted for in competing events. However, studies estimating the effect sizes of predictors of time to LTFU using competing risk analysis are scarce. Hence, this study aimed to investigate the predictors of time to LTFU among adults with TB/HIV coinfection. We conducted a multicenter facility-based retrospective follow-up study. We randomly selected 471 TB/HIV coinfected adults. Data were extracted using standardised checklists. The LTFUs from ART and anti-TB treatment were events of interest and competing events, respectively, and others were censored. The data were entered into Epi data and then exported to Stata and Rstudio. Statistical differences were tested by Gray's test, and the cumulative incidence of each event was estimated by a cumulative incidence function. Bivariable and multivariable competing risk regression models were fitted, and variables with p values < 0.05 were considered significant predictors. Incidence rates of LTFU for ART and TB treatment were 3.90 and 19.17 per 1000 person-months of observation (PMOs), respectively. The predictors of ART LTFU included rural residence (adjusted subdistribution hazard ratio (SDHR): 3.39), WHO stage IV (SDHR: 2.88), haemoglobin < 11 g/dl (SDHR: 3.56), and opportunistic infections (OIs) (SDHR: 3.65). For TB treatment LTFU, significant predictors were rural residence (SDHR: 0.11), divorced (SDHR: 2.81), widowed (SDHR: 5.92), BMI < 18.5 (SDHR: 0.41), ambulatory functional status (SDHR: 2.59), adverse drug effects (SDHR: 2.87), and poor ART adherence (SDHR: 5.72). Considering errors in competing events, ART LTFU was higher among rural dwellers, individuals with advanced disease, nutritional deficits, or adverse drug effects requiring prioritised, multifaceted interventions. Targeted strategies such as intensified monitoring, adherence counselling, nutritional support, proactive management of drug-related side effects, marital instability, OIs and poor ART adherence should be integrated into the existing ART/TB program to mitigate LTFU.