First participant recruited into the landmark RADIO-TB trial

13 Oct 2026

Ground-breaking trial aims to transform the management of people with early-stage tuberculosis (TB) identified through community chest x-ray screening

The RADIO-TB trial has recruited its first participant. The participant was enrolled by the team at the Biomedical Research and Training Institute (BRTI) in Bulawayo, Zimbabwe, marking the start of this international trial that will recruit 784 participants across Zimbabwe, South Africa and Pakistan. Recruitment in South Africa and Pakistan is expected to open shortly.

Sponsored by University College London (UCL), RADIO-TB is funded by the Wellcome Trust and brings together a network of investigators. It is being carried out across three high TB burden countries.

RADIO-TB is the first randomised clinical trial in more than 35 years to focus specifically on adults and adolescents with radiographically apparent, bacteriologically unconfirmed tuberculosis identified through chest x-ray based active case finding.

A growing challenge in TB screening

Tuberculosis remains one of the world's leading causes of death from an infectious disease. However, millions of people with TB continue to go undiagnosed in communities, especially when they have mild symptoms and therefore don’t seek care. Increasingly, national TB programmes are turning to community-based chest X-ray screening, supported by artificial intelligence and computer-aided detection software, to actively find people with TB disease in its early stages. Getting these people onto treatment helps to interrupt TB transmission.

This approach is helping uncover a large group of individuals who have chest X-ray abnormalities strongly suggestive of TB but who either have negative sputum molecular tests or are unable to produce sputum. Despite the absence of bacteriological confirmation, clinicians may still judge TB to be the most likely diagnosis.

These individuals are classified as having bacteriologically unconfirmed TB. Evidence suggests they have a substantial risk of progressing to microbiologically confirmed disease if left untreated. Those who do receive treatment are typically offered the same six-month regimen used for people with more advanced disease, despite the possibility that substantially shorter treatment may be sufficient. Yet at the same time many are not referred for further care and lost to follow-up.

Addressing a major evidence gap

The RADIO-TB trial has been designed to address the question: how should people with early, radiographically apparent TB be managed?

The study will recruit 784 adults and adolescents identified through symptom-agnostic chest X-ray screening programmes in communities and health facilities. Participants will be randomised to one of six management strategies, including standard six-month treatment, shorter treatment durations ranging from two to five months, or close follow-up with deferred treatment only if disease progression occurs.

Using an innovative Response Over Continuous Intervention (ROCI) design developed by methodologists at UCL, the trial aims to identify the shortest effective duration of treatment while also quantifying the benefits of treatment compared with observation alone. Importantly, the study will provide the first robust current estimates of the number needed to treat to prevent progression of disease in this population. This information is critical for policymakers, clinicians and national TB programmes as chest X-ray based screening expands globally.

International collaboration

The trial is being delivered through a partnership between investigators in Zimbabwe (BRTI), South Africa (Aurum Institute, Wits Health Consortium Clinical HIV Research Unit), Pakistan (National University of Medical Sciences), the UK (UCL and London School of Hygiene & Tropical Medicine), Sweden (KTH Royal Institute of Technology) and Switzerland (FIND) with funding from the Wellcome Trust.

Image shows an output from CAD4TB localising an abnormality on a screening chest x-ray suggestive of TB.

Further information: