Diagnostic Accuracy of Breath Tests to Detect Pulmonary Tuberculosis: A Systematic Review

Tuberculosis (TB) is a highly contagious infection and one of the world's leading causes of death from a single infectious agent. Currently, TB diagnosis can be established via mycobacterial cultures, Acid Fast Bacilli smear and molecular studies. In the ever-evolving landscape of medical advancements, breath tests have shown considerable promise. This systematic review aimed to evaluate the diagnostic accuracy of breath tests to detect pulmonary TB in various populations.

This systematic review was conducted according to the Preferred Reporting Items for Systematic Review and Meta-analyses (PRISMA) guidelines. We searched Embase and PubMed to identify observational studies published from database inception to May 2024. All observational studies evaluating the diagnostic accuracy of breath tests to detect pulmonary tuberculosis were included. Authors independently reviewed each article for eligibility and risk-of-bias. A senior reviewer was consulted for discrepancies.

The pooled sensitivity for the breath test in diagnosing TB was 0.85 (95% CI 0.78-0.90) whilst the pooled specificity was 0.83 (95% CI 0.72-0.90), although heterogeneity was high. Sub-group analysis by low/lower-middle World Bank income group status, high proportion of TB in test population, or use of a separate breath sampling kit did not reduce the heterogeneity. Publication bias was absent.

Our study found that pooled sensitivity and specificity of the breath tests in diagnosing pulmonary TB was high. Future research efforts can be directed towards investigating the diagnostic accuracy of electronic noses and gas chromatography combined with mass spectrometry, whilst improving standardisation and reproducibility of breath test techniques.