Background: Coughing is important in the transmission of tuberculosis (TB). It is not known if characteristics of cough in TB are distinct and whether they are clinically useful.
Aims: To compare temporal patterns of coughing in TB to those in COPD and refractory chronic cough.
Methods: Patients with stable COPD, refractory chronic cough and newly-diagnosed smear-positive pulmonary TB wore the Leicester Cough Monitor for 24 h. For TB, treatment was started either during or after the monitoring period. Cough frequency, nocturnal cough rates and cough clustering were calculated with the monitor's automated algorithm and compared between diagnostic groups. Separation of coughs by >2 s defined the limits of a cough episode, consisting of either a lone cough or a bout of ≥2 cough sounds.
Results: Median (IQR) cough counts in COPD (n=14), chronic cough (n=30) and TB (n=25) were 305 (128-462), 578 (215-1002) and 316 (101-700) per 24h, respectively (p=0.02). Nocturnal reductions in cough rates occurred to a similar extent in all groups; a mean (SD) of 24.5 (12.5), 20.4 (11.8) and 26.4 (19.2)% of coughs were detected overnight between 22:00 and 08:00, respectively (p=0.32). The clustering of coughs was not significantly different between groups: a mean (SD) of 42.0 (22.2), 29.2 (18.4) and 35.4 (19.3)% of episodes consisted of lone coughs rather than bouts (p=0.13), and the median (IQR) number of coughs per episode was 1.6 (1.4-2.0), 2.0 (1.7-2.5) and 1.7 (1.4-2.1), respectively (p=0.28).
Conclusions: Coughing was less frequent in TB than refractory chronic cough, but of similar frequency to in COPD. There was no evidence of a difference in the temporal pattern of cough in TB compared to COPD and chronic cough.