Decide-TB

The challenge

Missed cases are a well-known challenge in tuberculosis, particularly so for childhood TB. More than a million children and adolescents develop TB each year. Of these, an estimated 250,000 die, 95% of them without a diagnosis. When TB is diagnosed and treated, nearly all children survive.

Tuberculosis diagnosis is particularly difficult in children. Symptoms are non-specific, bacterial load is typically low, and sputum samples may be hard to collect. Diagnosis is therefore generally based on assessment of signs and symptoms and whatever test results are available. WHO has developed two treatment decision algorithms to guide clinician decision-making (one with and one without chest X-ray results), but there is very little data to support their use.

The project

The Decide-TB project is integrating a wide range of data sources to generate two refined treatment decision algorithms, backed up by practical tools to support clinical decision-making.

These tools will include additional refinements, including modules specifically designed for TB diagnosis in children living with HIV and for children with severe acute malnutrition, where different criteria need to be applied. In addition, while the standard course of treatment for TB is six months of antibiotics, four months’ treatment has been shown to be effective in children with non-severe TB. The new tools will therefore point clinicians towards this option when appropriate. 

To evaluate these tools, the project is organising a cluster randomised trial in Mozambique and Zambia. As well as assessing the effectiveness of the approach at TB case detection, this study incorporates an implementation research aspect, and will also collect evidence on the feasibility and practicalities of introduction. National TB Control Programmes in Mozambique and Zambia are project partners, so the results of the study will feed directly into policymaking.

Existing health data platforms (DHIS2) will be adapted to support data collection, facilitating data use for action by national disease control programmes. The project will also collate individual patient-level data to provide a comprehensive evidence synthesis of the diagnostic accuracy of new treatment decision algorithms.

Impact

The Decide-TB project will provide critical evidence on the effectiveness and feasibility of new tools to detect TB in children. It will:

• Reveal how effective the treatment decision algorithms and clinical decision support systems are for detecting TB in children.
• Generate a package of evidence for the WHO, to inform global guidance and development of materials to support implementation.
• Directly facilitate implementation in countries participating in the project.

If shown to be effective, the new tools would deliver multiple benefits, reducing the number of missed cases of TB in children, enhancing detection in vulnerable children such as those living with HIV or affected by severe acute malnutrition, and leading to greater use of shortened treatment regimens for those with less severe disease. This would ensure that more children receive TB treatment and reduce the number of children dying unnecessarily from the disease.