Introduction
TB disease screening programmes need to ensure programme effectiveness and understanding the test properties of individual and combined screening tests is key. There are knowledge gaps about the effect of adding TB infection (TBI) tests to screening migrant screening algorithms (as in Italy and Sweden). We will perform an RoR to summarise the evidence of test properties for all TB screening and confirmatory tests to detect prevalent TB disease to inform parametrisation of a mathematical modelling study estimating combined test properties.

Methods
Two independent reviewers will include all relevant systematic reviews from PubMed, Web of Science, Embase, Cochrane Library without any language or geographical restriction. Sensitivity and specificity of tests from primary papers included in the SRs will be pooled in random-effects meta-analysis. Sensitivity analysis will limit of pooled estimates to high quality papers. The RoR is registered with PROSPERO (CRD42023393129)

Results
Preliminary evidence shows that test sensitivities range between 42.1% (95% confidence interval, CI 36.6-47.7%) for prolonged cough to 94.7%, CI 92.2-96.4%) for any CXR abnormality and specificity between 65.1% (CI 53.3-75.4%) for any TB symptom and 95.6% (CI 92.6-97.4%) for TB-specific CXR abnormalities1.  

Conclusions
Current migrant screening algorithms for migrants are suboptimal and adding high sensitivity TB infection tests may potentially increase TB yields. Results of this RoR will inform mathematical modelling to optimise migrant screening algorithms.
 
1Van?t Hoog A., et al (2022). Cochrane Database of Systematic Reviews