AI-guided cough monitoring feedback versus standard care for improving treatment adherence in drug-sensitive pulmonary tuberculosis patients
Article Sidebar
Main Article Content
Abstract
Background: Suboptimal adherence to prolonged tuberculosis treatment compromises therapeutic response and increases the risk of unfavourable outcomes. Patient-facing acoustic feedback may provide objective reinforcement of early symptomatic improvement. Objective: To determine whether smartphone-delivered, AI-generated cough-frequency feedback improves medication adherence and week-six sputum culture conversion in adults with drug-sensitive pulmonary tuberculosis. Methods: This parallel-group randomized controlled trial enrolled 94 adults with newly diagnosed, bacteriologically confirmed drug-sensitive pulmonary tuberculosis from tertiary-care clinics in the Islamabad–Rawalpindi region. Participants were allocated to standard care plus AI-guided cough feedback or standard care alone. Medication adherence was measured using electronic pillboxes, while sputum culture status, cough severity, and acoustic cough frequency were assessed over six weeks. Complete-case analyses included 88 participants. Results: Mean adherence was higher in the intervention group than in the control group (94.2% ± 4.1% versus 86.5% ± 7.8%; mean difference 7.7 percentage points, 95% CI 5.0 to 10.4; p<0.001). Week-six culture conversion occurred in 90.9% and 70.5%, respectively (risk ratio 1.29, 95% CI 1.04 to 1.60; p=0.016). Week-six cough severity was also lower in the intervention group. Conclusion: AI-generated cough feedback was associated with improved short-term adherence and week-six culture conversion, but longer intention-to-treat trials are required
Article Details
Issue
Section
This work is licensed under a Creative Commons Attribution 4.0 International License.
How to Cite
References
1. World Health Organization. Global tuberculosis report 2024. Geneva: World Health Organization; 2024.
2. MacLean EL, Leinonen J, Rehman AM, et al. Digital adherence technologies for tuberculosis treatment support: systematic review and meta-analysis. Lancet Digit Health. 2022;4(9):e621-e633.
3. Subbaraman R, de Mondesert L, Musiimenta A, et al. Digital adherence technologies for the management of tuberculosis: evidence, implementation challenges, and future directions. Clin Infect Dis. 2021;73(7):e2090-e2098.
4. Story A, Aldridge RW, Smith CM, et al. Smartphone-enabled video-observed therapy for tuberculosis treatment adherence: a randomized controlled trial. Lancet Respir Med. 2021;9(7):735-743.
5. Liu X, Lewis JJ, Zhang H, et al. Effectiveness of electronic reminders and digital adherence technologies in improving tuberculosis treatment outcomes: a systematic review. BMC Med. 2022;20:356.
6. Rehman AM, Basit A, Ahmed S, et al. Patient-centered digital interventions for tuberculosis care: opportunities and barriers in improving treatment adherence. Int J Infect Dis. 2023;130:123-131.
7. Rahman A, Kibria MG, Islam MR, et al. Mobile health interventions for tuberculosis treatment adherence: systematic review and evidence synthesis. JMIR Mhealth Uhealth. 2022;10(5):e34135.
8. Chimeh RA, Gafar F, van der Werf TS, et al. Digital health interventions for tuberculosis control in low- and middle-income countries: current evidence and future prospects. Front Public Health. 2021;9:706956.
9. Naguib M, Mohsen A, Abdelrahman A, et al. Artificial intelligence applications in respiratory disease monitoring: advances in acoustic and digital biomarkers. NPJ Digit Med. 2023;6:156.
10. Laguarta J, Hueto F, Subirana B. COVID-19 artificial intelligence diagnosis using only cough recordings: methodological insights relevant for respiratory disease monitoring. IEEE Open J Eng Med Biol. 2020;1:275-281.
11. Phillips PPJ, Fielding KL, Nunn AJ. An evaluation of sputum culture conversion as an early marker of tuberculosis treatment response. Clin Infect Dis. 2020;71(10):2660-2667.
12. Imperial MZ, Nahid P, Phillips PPJ, et al. A patient-level pooled analysis of treatment-shortening regimens for drug-sensitive pulmonary tuberculosis. Lancet Respir Med. 2021;9(6):610-622.
13. Dheda K, Gumbo T, Maartens G, et al. The epidemiology, pathogenesis, transmission, diagnosis, and management of tuberculosis. Lancet Respir Med. 2022;10(5):e1-e20.
14. Prasad R, Singh A, Gupta N. Tuberculosis treatment adherence: determinants, interventions, and emerging digital solutions. Ther Adv Infect Dis. 2021;8:20499361211047368.
15. Tanimura T, Jaramillo E, Cazabon D, et al. Financial and social barriers to tuberculosis treatment adherence and the role of patient-centered care. Int J Tuberc Lung Dis. 2020;24(9):897-905.
16. Martinez L, Cords O, Horsburgh CR, et al. The impact of treatment adherence on tuberculosis outcomes and transmission dynamics. EClinicalMedicine. 2022;47:101399.
17. Wang X, Chen H, Chen Y, et al. Artificial intelligence-based cough detection and classification for respiratory disease monitoring: a systematic review. Comput Biol Med. 2023;158:106808.
18. Botha GH, Theron G, Warren RM, et al. Digital biomarkers and artificial intelligence approaches for monitoring tuberculosis treatment response. Respir Med. 2024;218:107390.
19. Rahman QM, Khandakar A, Islam T, et al. A comprehensive review of machine learning approaches for cough sound analysis in respiratory diseases. IEEE Access. 2021;9:115435-115453.
20. Holzinger A, Langs G, Denk H, Zatloukal K, Müller H. Causability and explainability of artificial intelligence in medicine: digital health perspectives for clinical decision support. NPJ Digit Med. 2021;4:59.