Abstract
OBJECTIVES: TB disease during pregnancy is associated with poor maternal and neonatal outcomes, and is a leading non-obstetric cause of maternal death. However, optimal detection strategies remain uncertain. We aimed to identify the optimal screening approach for TB disease in pregnant women. METHODS: We searched Ovid MEDLINE, Embase + Embase Classic, Web of Science, and CENTRAL to identify antenatal screening studies for TB disease. The yield, number needed to screen (NNS), and positive predictive value (PPV) were calculated for each method. Pooled estimates were generated using random-effects meta-analyses. Narrative synthesis was conducted to summarise secondary outcomes. RESULTS: We included 33 studies. Pooled yield for symptom screening (SS) was 7.26 [95% CI: 0.70, 19.25] cases per 1,000 versus 5.12 [95% CI: 0.79, 12.39] for TST/IGRA. NNS was 138 [95% CI: 51.95, 1,428.57] for SS versus 1,667 [95% CI: 537.63, 1,000,000] for TST/IGRA. SS pooled PPV was 3.85% [95% CI: 1.23-7.57%], and <0.01% [95% CI: <0.01-0.05%] for TST/IGRA. Narrative synthesis indicated antenatal SS is low-cost, feasible, and acceptable but poorly implemented. CONCLUSION: In pregnancy, symptom screening demonstrates highest yield and lowest NNS, is low-cost, feasible and acceptable. While currently optimal, the low PPV underscores the need for TB screening tools tailored to pregnant populations.