Setting: Active tuberculosis (TB) case finding (ACF) in Phnom Penh, Cambodia using light-emitting diode fluorescence microscopy (FM).
Objective: To evaluate the smear-positive yield of frontloaded (same-day) smear microscopy in ACF.
Design: All presumptive TB cases screened through ACF were asked to provide three sputum specimens: two spot specimens on Day 1 and a morning specimen on Day 2 (spot-spot-morning, SSM). Laboratory technicians blinded to previous results read the smears using FM. We considered only SSM series with at least one positive smear to calculate the proportion of TB cases missed and to determine the difference between the spot-spot (SS) and spot-morning (SM) approach.
Results: Of 4616 presumptive TB patients enrolled, 3306 provided three sputum samples. Of 2957 (89.4%) who followed the SSM approach, 188 (6.4%) were smear-positive: 177 on SM and 160 on SS. The incremental yield of the second sputum sample was 18.1% for SM vs. 9.4% for SS. Relative to any smear-positive case detected by SSM, 28/188 (14.9%, 95%CI 10.1-20.8) TB cases would be missed by SS vs. 11/188 (5.9%, 95%CI 3.0-10.2) by SM. The difference in the proportion of missed TB patients was 9.0% (P = 0.006).
Conclusion: ACF frontloaded sputum microscopy is inferior in terms of smear-positive yield: the SS approach would have missed a significant proportion of smear-positive TB.