Smoking linked to lower success in drug-resistant TB treatment

Tuberculosis remains a major global health problem, and people who smoke are known to be at higher risk of worse TB treatment outcomes than those who do not smoke. Why smoking worsens outcomes has not been clear. To tackle that question, researchers led by corresponding author Matthew L. Romo analyzed data from the endTB Observational Study, a prospective multicountry cohort of people with multidrug-resistant and rifampicin-resistant tuberculosis (MDR/RR-TB). Participants in this study were treated with longer regimens that included bedaquiline and/or delamanid. The team set out to estimate how daily smoking (defined in the study as at least one cigarette per day at enrollment) affected the chance of having a successful end-of-treatment result, defined as cured or treatment completed. They also explored whether pathways related to patients leaving care — known as loss to follow-up — help explain any difference in outcomes between people who smoke and those who do not.

The analysis included 1,786 people from 12 countries, of whom 539 (30.2%) reported smoking at enrollment. To assess the impact of smoking on treatment success, the researchers used a statistical approach called marginal standardization to compare outcomes between smokers and non-smokers. They then simulated an intervention on loss to follow-up by censoring participants who were lost and applying inverse probability of censoring weights to create a pseudopopulation that represents what might happen if nobody was lost to follow-up. At treatment end, 73.5% of people who smoked achieved treatment success, compared with 80.3% of people who did not smoke, a crude risk difference of -6.8 percentage points (95% CI: -11.1, -2.6). After adjusting for baseline confounders, including demographics, social history, and comorbidities, the risk difference was -5.2 percentage points (95% CI: -14.1, 3.2). In the simulated pseudopopulation without loss to follow-up, the difference narrowed to -1.9 percentage points (95% CI: -10.2, 5.1).

The findings show that people who smoked had lower rates of successful MDR/RR-TB treatment in this large, multicountry cohort treated with regimens containing bedaquiline and/or delamanid. Importantly, when the researchers simulated eliminating loss to follow-up, the gap in success between smokers and non-smokers became much smaller. That pattern suggests that problems keeping people in care — rather than only a direct biological effect of smoking on the drugs or disease — played a meaningful role in the worse outcomes seen among people who smoke. For health programs, the results point to two clear targets: supporting smoking cessation and strengthening retention in care for people with MDR/RR-TB. While the study does not prove one single mechanism, it highlights that interventions to prevent patients from dropping out could reduce the treatment disadvantage seen in people who smoke and improve overall success with these longer, second-line TB regimens.