Molecular Detection of Airborne <i>Mycobacterium tuberculosis</i> in South African High Schools

Abstract

<b>Rationale:</b> South African adolescents carry a high tuberculosis disease burden. It is not known if schools are high-risk settings for <i>Mycobacterium tuberculosis</i> (MTB) transmission. <b>Objectives:</b> To detect airborne MTB genomic DNA in classrooms. <b>Methods:</b> We studied 72 classrooms occupied by 2,262 students in two South African schools. High-volume air filtration was performed for median 40 (interquartile range [IQR], 35-54) minutes and assayed by droplet digital PCR (ddPCR)-targeting MTB region of difference 9 (RD9), with concurrent CO₂ concentration measurement. Classroom data were benchmarked against public health clinics. Students who consented to individual tuberculosis screening completed a questionnaire and sputum collection (Xpert MTB/RIF Ultra) if symptom positive. Poisson statistics were used for MTB RD9 copy quantification. <b>Measurements and Main Results:</b> ddPCR assays were positive in 13/72 (18.1%) classrooms and 4/39 (10.3%) clinic measurements (<i>P</i> = 0.276). Median ambient CO₂ concentration was 886 (IQR, 747-1223) ppm in classrooms versus 490 (IQR, 405-587) ppm in clinics (<i>P</i> < 0.001). Average airborne concentration of MTB RD9 was 3.61 copies per 180,000 liters in classrooms versus 1.74 copies per 180,000 liters in clinics (<i>P</i> = 0.280). Across all classrooms, the average risk of an occupant inhaling one MTB RD9 copy was estimated as 0.71% during one standard lesson of 35 minutes. Among 1,836/2,262 (81.2%) students who consented to screening, 21/90 (23.3%) symptomatic students produced a sputum sample, of which one was Xpert MTB/RIF Ultra positive. <b>Conclusions:</b> Airborne MTB genomic DNA was detected frequently in high school classrooms. Instantaneous risk of classroom exposure was similar to the risk in public health clinics.