Characterizing change in cough frequency and associated factors among patients with pulmonary tuberculosis using HYFE App in Kampala- Uganda

Abstract

Background: The most important means of pulmonary tuberculosis transmission is cough, which invariably results from an inflammatory reaction to pulmonary mycobacterial infection. A noticeable decrease in coughing is presumptively assumed as an indicative sign to adequate response to TB therapy and this lowers risks of increased transmission. Determining cough frequency is important in assessing severity of disease, evaluating infectiousness, and predicting patient outcomes. The study aimed to characterize change in cough frequency and associated factors among patients with confirmed pulmonary tuberculosis using Hyfe application which identifies the explosive sounds that might be coughs. Then these sounds are analyzed and, once confirmed as coughs, they are recorded on dashboard. Methods: A longitudinal retrospective study was conducted with both descriptive and analytical components among 100 participants’ from R2D2-Study that recruited over 120 newly diagnosed pulmonary tuberculosis (PTB) patients whose coughs were recorded using the Hyfe Application between Sept 2021 and Dec 2022 at Mulago national referral hospital and Kisenyi health center IV study sites. An abstraction tool was used in data collection. Time effect analysis was done and observed changes with effect of time on cough frequency were identified. Using R ‘glmmTMB’ package analysis, a random coefficient poisson mixed-effects model with robust standard errors was used to determine independent predictors of cough frequency.Results: The majority of the participants were aged 18-39 (71%, n=71) years with more than half being male (55%, n=51). There was a 90% reduction in the rate of coughs at day 14 compared to day1. BMI-underweight (IRR=0.32, 95% CI 0.13-0.92), BMI-overweight (IRR=0.07, 95% CI 0.01-0.39), interaction between time and diabetes (IRR=0.66, 95% CI 0.45-0.97), and sex interaction with BMI-overweight (IRR=3.68, 95% CI 1.36-9.98) were found to be associated with cough frequency. Conclusion: The significant reduction in cough rates over time highlights the potential improvement of the body's immune response since the participants were on treatment, hence reduce inflammation and cough symptoms. Key associations (BMI, Sex*BMI-overweight and Time*Diabetes) with cough frequency emphasize the importance of considering these factors in clinical management and health educating the public about how key associations influence cough frequency, highlighting the need to maintain a healthy weight and manage diabetes properly.