Does stopping smoking improve tuberculosis treatment outcomes?

01

TB relapse is not smoking relapse

Tuberculosis is a bacterial disease, usually affecting the lungs. In the older study below, ‘TB relapse’ meant a recurrent episode of TB by month 12 after being recorded as cured or having completed treatment at month 6. It did not mean returning to cigarettes. Likewise, ‘treatment success’ combined two register outcomes—cured and treatment completed—not a measurement that every lung change, symptom or risk of recurrence had disappeared.

[1][3][4]

02

Why the earlier 91% versus 80% is an association

Siddiqi and colleagues analysed quit status within a trial in Bangladesh and Pakistan, published in Thorax in 2022 and online in 2021. The parent trial randomly allocated a cessation medicine or placebo; this secondary comparison did not randomly assign people to quit or keep smoking. Its 2,273 analysed participants included 577 classified as continuously abstinent, verified at months 6 and 12. The other group also included missing or unverified abstinence, not only people known to keep smoking every day.

At month 12, recorded cure or treatment completion was 525/576 (91%) in the verified group versus 1,296/1,615 (80%) in the other classification. Among those with relevant month-6 completion/cure records and follow-up, TB recurrence by month 12 was 33/535 (6%) versus 208/1,450 (14%). Those denominators differ: the relapse percentage is not a recurrence rate among all 2,273 participants. Neither number is an individual's probability.

Adjustment included initial disease score and treatment adherence, but residual differences remain possible. The analysis excluded 68 people who died and others without required follow-up; a person had to survive to have abstinence verified at month 12. It cannot support a mortality claim. Participants mostly were men, daily smokers motivated to quit, with drug-sensitive pulmonary TB; extrapolation to other forms, treatment contexts or populations is limited.

[1]

03

What the newer randomized study did—and did not—establish

Zahid and colleagues' JAMA 2026 trial, online in December 2025, randomized 27 TB clinics in Bangladesh and Pakistan to TB-specific text-message support or usual written cessation information. All participants continued TB care. The primary six-month verified abstinence result was 41.7% versus 15.3%. This comparison tests that supported programme, not quitting itself randomly assigned, and not any generic app or collection of reminders. The primary outcome combined self-reported continuous abstinence with supporting breath-carbon-monoxide verification at month 6. That CO reading checks only recent abstinence; it cannot by itself prove no smoking throughout all six months.

Exploratory TB treatment success was 89.3% versus 85.6%, with a reported risk ratio (RR) of 1.2 and 95% confidence interval 0.9–1.6. RR compares the chance of this outcome between groups; 1 means no difference. The interval includes 1, so more verified quitting did not establish a clear treatment-success difference in this trial. Secondary outcomes were exploratory without adjustment for multiple comparisons; almost all participants were men, contact calls were part of the study context and follow-up ended with treatment. No personal cure or lasting abstinence guarantee follows.

[2]

04

A useful handoff question

Ask the TB care team: ‘How can cessation support be coordinated with my existing care, and which result are we discussing—stopping tobacco, treatment completion or recurrent TB?’ In England, NHS local stop-smoking services provide a separate support route. These research figures cannot determine medicines, treatment duration, whether someone is infectious, contact management or permission to return to work. Those questions belong to the treating team and local public-health service. No test result or treatment history needs to be entered here.

[3][4][5]

05

What to keep in mind

• Quit status was not randomized in the older outcome comparison.
• TB recurrence and returning to smoking are different events.
• A trial can improve quitting without clearly establishing improved TB treatment success.