Evaluating Smartphone-Based Cough Monitoring for Tuberculosis Screening and Triage in Uganda: A Mixed-Methods Evaluation

Symptom-based tuberculosis (TB) screening often relies on cough reporting, which may be unreliable. Passive cough monitoring via smartphone apps offers a less subjective alternative. This study evaluated the diagnostic accuracy and implementation feasibility of smartphone-based cough frequency monitoring for identifying individuals likely to have TB in Uganda. We enrolled adults ([&ge;]15 years) screened for TB in community settings or tested at health facilities in Kampala. Participants underwent microbiological testing and 48-hour cough monitoring using the Hyfe Research app. TB status was determined using Xpert MTB/RIF Ultra and culture, and inverse probability weighting was applied to adjust for differential enrollment. We compared cough frequency between individuals with and without TB and assessed diagnostic accuracy using weighted ROC curves. We also conducted staff interviews, analyzed thematically, to explore implementation challenges. Of 884 enrolled participants, 197 had valid cough recordings and TB status (101 from community screening, 96 from health facilities). Individuals with TB had higher median cough frequency than those without: 2.2 (interquartile range 0.8-6.1) vs 0.9 (0.4-2.0) coughs per hour in the community and 6.7 (2.6-27.5) vs 2.4 (0.9-4.8) in facilities; Wilcoxon P < 0.0001 for both. AUCs for smartphone-based cough recording were 0.69 (95% CI: 0.58-0.79) in the community and 0.76 (95% CI: 0.6-0.88) in health facilities. At 90% sensitivity, specificities were 20% (95% CI: 0-22%) and 46% (95% CI: 1-32%), respectively. Smartphone-recorded cough frequency was moderately correlated with self-reported cough severity, respiratory-related quality of life scores (Saint Georges Respiratory Questionnaire), and staff-observed coughs. Staff cited device visibility, stigma, and security concerns as barriers to implementation. While smartphone-recorded cough frequency was associated with TB status, it did not meet diagnostic accuracy thresholds for stand-alone screening or triage. Implementation challenges also limited data collection. Addressing operational barriers will be critical to future development and deployment of cough monitoring tools for TB screening.