Stepped-wedge randomised trial of AI-guided site selection for tuberculosis case-finding in Pakistan (SPOT-TB)
Across 3,936 screening camps and 269,254 people in all four provinces of Pakistan, mobile x-ray teams switched in random order from conventional site selection to locations recommended by the MATCH-AI software; tuberculosis yield per camp was 7% higher but not significantly so in the intention-to-treat analysis (adjusted risk ratio 1.07, 95% CI 0.94-1.22), and 32% higher in a per-protocol analysis of camps held within 5 km of a recommended site.
- Model
- MATCH-AI (Mapping and Analysis for Tailored disease Control and Health system strengthening) geographic targeting software
- Field
- Medicine
- Date
- 2026-11-12
- Human collaborators
- Amna Mahfooz, Faran Emmanuel
What was found
The trial was embedded in community active case-finding run by Pakistan's National TB Control Program and private partners with Global Fund support. Thirty mobile x-ray van teams working in 68 districts were randomised to the time at which they moved from choosing camp sites by field knowledge, experience and historical data to sites recommended by MATCH-AI. Between 1 August 2023 and 28 September 2024 there were 2,046 control and 1,890 intervention camps. The primary outcome was the number of people diagnosed with bacteriologically confirmed tuberculosis per camp, analysed by intention to treat, with a validated per-protocol analysis restricted to camps within 5 km of an AI-recommended location. The trial is registered as NCT06017843 and was funded by the Gates Foundation and the Global Fund through Mercy Corps.
Novelty check
Europe PMC, 2026-11-16, tuberculosis in the title with artificial intelligence or machine learning, case-finding and a randomised design: 67 records. The top results were this trial, an earlier version of it from May 2026, and reviews and implementation reports on AI reading of chest x-rays, which is a different use of AI from choosing where to screen. Nothing in this registry covered tuberculosis case-finding.
Caveats and known objections
The prespecified primary analysis is null: the 7% difference by intention to treat has a confidence interval running from a 6% decrease to a 22% increase. The positive finding is per-protocol, and restricting to camps that teams actually held near recommended sites conditions on behaviour after randomisation and may select favourable settings. The authors note that routine programme data did not capture implementation fidelity or process quality. The outcome is yield per camp rather than per person screened or a population effect on transmission. The abstract does not say who built MATCH-AI or whether the investigators have a stake in it. Autonomy is ai-assisted: the software recommends where to screen, and teams decided where camps were actually held.
Trial registration
That a study was registered, with the outcomes it declared and the date it declared them, so a published endpoint can be checked against the one planned. It does not certify the results, their analysis, or any clinical benefit; registry records are submitted by the sponsor and are not peer reviewed.
No independent check is recorded on this entry yet.
Reading the primary source closely enough to say whether it supports the claim counts as a check, and you are credited on the entry.
Or on GitHub: submit a check challenge the grade send a correction or send a pull request
Flag this for triage
Signals order the review queue and nothing else. They are never published, and they never move a grade: that takes a citation.
Entry history (1 event)
- AddedEntered the registry graded Peer reviewed and AI-assisted.
Entries are never deleted. A grade that does not hold up is downgraded on the record, with the reason beside it.
Graded peer reviewed for verification and ai-assisted for autonomy. What these mean.
Cite this entry
whataifound.org. (2026). Stepped-wedge randomised trial of AI-guided site selection for tuberculosis case-finding in Pakistan (SPOT-TB). whataifound.org: A Registry of AI Scientific and Mathematical Discoveries. https://whataifound.org/finding/2026-11-12-spot-tb-ai-case-finding-rct
BibTeX
@misc{whataifound-centreforglobalpublichealthpakistan-2026-rct,
title = {Stepped-wedge randomised trial of AI-guided site selection for tuberculosis case-finding in Pakistan (SPOT-TB)},
author = {{whataifound.org}},
year = {2026},
howpublished = {whataifound.org: A Registry of AI Scientific and Mathematical Discoveries},
note = {Result by Centre for Global Public Health, Pakistan. Verification: Peer reviewed. Autonomy: AI-assisted.},
url = {https://whataifound.org/finding/2026-11-12-spot-tb-ai-case-finding-rct}
}