Longitudinal Cough Frequency Monitoring in Persistent Coughers: Daily Variability and Predictability

Abstract

Purpose: We determined the cough counts and their variability in subjects with persistent cough for 30 days.

Methods: The Hyfe cough tracker app uses the mobile phone microphone to monitor sounds and recognizes cough with artificial intelligence-enabled algorithms. We analyzed the daily cough counts including the daily predictability rates of 97 individuals who monitored their coughs over 30 days and had a daily cough rate of at least 5 coughs per hour.

Results: The mean (median) daily cough rates varied from 6.5 to 182 (6.2 to 160) coughs per hour, with standard deviations (interquartile ranges) varying from 0.99 to 124 (1.30 to 207) coughs per hour among all subjects. There was a positive association between cough rate and variability, as subjects with higher mean cough rates (OLS) have larger standard deviations. The accuracy of any given day for predicting all 30 days is the One Day Predictability for that day, defined as the percentage of days when cough frequencies fall within that day's 95% confidence interval. Overall Predictability was the mean of the 30-One Day Predictability percentages and ranged from 95% (best predictability) to 30% (least predictability).

Conclusion: There is substantial within-day and day-to-day variability for each subject with persistent cough recorded over 30 days. If confirmed in future studies, the clinical significance and the impact on the use of cough counts as a primary end-point of cough interventions of this variability need to be assessed.

Keywords: Artificial intelligence; Chronic cough; Cough applications; Cough frequency; Persistent cough.

Fig. 1

Boxplots of the daily cough frequencies over the period of observation of 97 individuals who monitored for at least 30 days and had frequencies of over 5 coughs per hour. Subject 80, an outlier with a median frequency of 160.08 coughs per hour, is not shown. For each subject, the frequency quartiles are shown by the bottom edge (first quartile), midline (median), and top edge (third quartile) of the box; the range is depicted by the vertical lines. Cough rates vary considerably among this cohort, both within and between subjects

Fig. 2

Standard deviations (SD) and means of daily cough frequencies of 97 individuals who monitored for at least 30 days and had frequencies of over 5 coughs per hour. Each point represents one subject; the x-coordinate is the average of that subject’s daily cough frequencies, and the y-coordinate is the SD of that subject’s daily cough frequencies. Panel A shows all 97 individuals (correlation = 0.95; Optimized Least Square (OLS) slope = 0.64), while Panel B shows 96 subjects with the exclusion of Subject 80 (correlation = 0.79; OLS slope = 0.40). Lines of best fit are shown

Fig. 3

Daily cough frequencies data are shown as dots with the 30 day average depicted by the dashed line. The 95% confidence interval of cough frequency for each day is represented by the vertical bars. The coverage percentage for each day is shown at the bottom (color coded as yellow for > 80%, darker for between 60 and 80%, darker for between 40 and 60% and red for < 40%). In Panel A, the user’s cough frequency predictability is 95%, in Panel B subject this is 38% and in Panel C, for this subject 30%

Fig. 4

Distribution of overall cough frequency predictabilities of the 97 subjects

Fig. 5

Daily coverage of daily cough frequency predictability results for all users. Each row represents one user and each cell is one day. The color of each cell corresponds to the coverage percentage for that day (yellow for > 80%, slightly darker for between 60 and 80%, darker for between 40 and 60%, and red for < 40%)