Document resource
Background Objective cough frequency is widely accepted as a measure of cough. It is commonly derived from the number of explosive cough sounds occurring in ambulatory audio recordings collected using cough monitoring systems such as VitaloJAK®. Cough second frequency has been proposed as an alternative measure whereby a cough second is any second of audio containing at least one explosive cough sound. 1 Here we compare the performance of cough frequency in 24 hours (C/24H) and cough second frequency in 24 hours (CS/24H) to assess the impact on derived measures.Methods A total of 380 24-hour ambulatory audio recordings were obtained using the VitaloJAK from 77 chronic cough subjects (60F/17M) participating in a clinical study. Trained analysts annotated all coughs totalling 236,738 coughs (median 506) and 185,743 cough seconds (median 389) across all recordings. The relationship between C/24H and CS/24H was assessed using linear regression analysis. The number of coughs per cough second was also computed. Furthermore, the percentage decrease in cough frequency resolution between C/24H and CS/24H was computed per recording.Results A statistically significant linear correlation was observed between C/24H and CS/24H (R 2=0.98; p<0.0001; slope=0.76 (95% CI: 0.75–0.77); intercept=17.28). A 21.5% decrease in cough frequency resolution was observed between C/24H and CS/24H across all recordings (median 21.4; range: 0–38.3). It was also observed that 26.2% of cough seconds contained two or more coughs.Conclusions While C/24H and CS/24H are linearly correlated, cough frequency resolution decreased by >20% when using cough seconds. Furthermore, multiple coughs occurred within cough seconds, making cough seconds less sensitive to changes in other measures of interest e.g. time-based cough bouts, which require individual time locations of coughs. Based on our findings, cough seconds are a less useful unit for measurement of cough than explosive cough sounds due to a loss of resolution impacting on measurement of cough frequency and bouts.Reference Morice AH, et al. Eur Respir J. 2007;29(6):1256–1276. https://doi.org/10.1183/09031936.00101006