Predictors of tuberculosis mortality among patients on treatment: A retrospective cohort study at a rural general hospital in Zambia.

BACKGROUND

Despite decades of global efforts and the availability of curative and preventive measures, tuberculosis (TB) continues to be the world's leading infectious disease cause of mortality, particularly in resource-limited settings. However, in developing countries, there is limited evidence on determinants of mortality during TB treatment, making it difficult to design targeted interventions. This study aimed to determine the incidence and factors associated with mortality among people diagnosed with TB at a rural health facility in Zambia.

METHODS

We conducted a retrospective cohort study among 316 individuals who started TB treatment between May 2019 and March 2024. Participants were followed from treatment initiation until a treatment outcome was recorded, with a maximum follow-up time of 12 months. Demographic and clinical information were collected from medical records using a data collection tool via KoboTool software. The primary outcome was mortality. We used a multivariable Cox regression model to identify factors associated with TB mortality using Stata version 15.

RESULTS

The participants had a median age of 41 years (IQR: 32-54) and a median follow-up of 174 days (IQR: 155-191). Overall, 19.0% (n = 60; 95% CI: 14.8-23.8) of participants died, corresponding to an incidence rate of 42.1 per 100 person-years based on 51,968 person-days of follow-up (95% CI: 31.6-54.0). Age (adjusted hazard ratio (aHR): 1.017, 95% CI: 1.000-1.033, P = 0.042), clinically diagnosed TB had a significantly higher risk of mortality compared to those with bacteriologically confirmed TB (aHR: 1.97, 95% CI: 1.14-3.41, p = 0.015), and People living with HIV (aHR: 1.84, 95% CI: 1.03-3.29, P = 0.039) were significantly associated with increased mortality. While being on family-based directly observed therapy (DOT) (aHR: 0.42, 95% CI: 0.24-0.74, P = 0.003) and every unit increase in baseline weight (aHR: 0.96, 95% CI: 0.92-0.99, p = 0.004) was associated with a reduced risk of mortality.

CONCLUSION

This study reveals a high incidence of TB-related mortality in a rural setting, which is significantly influenced by demographic and clinical determinants. The findings highlight the need for closer evaluation of patients with clinically diagnosed TB and strengthening family-based DOT as strategies to reduce mortality.