Survival analysis shows tuberculosis patients with silicosis experience earlier mortality and need employer-led care models in occupational settings in India

Abstract

Abstract India’s high tuberculosis (TB) burden is exacerbated by concurrent silicosis, which increases TB susceptibility and worsens treatment outcomes. Limited studies on TB patients with silicosis highlight the need to address this vulnerable group’s specific challenges, particularly to improve diagnosis and management. This retrospective cohort study analyzed survival data from 137 silico-tuberculosis and 2,605 TB-only patients in Khambhat, India, using Kaplan-Meier curves, log-rank tests, and comparisons between Cox proportional hazards and accelerated failure time (AFT) models. The lognormal AFT model, selected for its lowest Akaike Information Criterion (AIC), estimated survival times based on age, gender, HIV status, and prior TB treatment. Among the 2,742 patients, 309 (11%) died within 27 months. Median time from diagnosis to outcome was shorter for deceased patients (1.7 months) than for censored patients (5.6 months, p < 0.001, median test). Kaplan-Meier analysis showed a significantly steeper survival decline for silico-tuberculosis patients (p < 0.001, log-rank test). Silico-tuberculosis was associated with a two-fold increased mortality risk (HR = 2.0, 95% CI: 1.4-3.0, p < 0.001, Cox-proportional hazards regression). The lognormal AFT model indicated silico-tuberculosis patients had 36% of the median survival time compared to TB-only patients (16 vs. 44 months). These findings highlight significantly earlier mortality in silico-tuberculosis patients, underscoring the need for targeted, employer-led care models and TB-silicosis collaborative screening within India’s TB program for high-risk occupational groups.

Keywords

• Silicosis
• Tuberculosis
• Silico-tuberculosis
• Survival analysis
• Cox regression
• Accelerated failure time model