Introduction and objective Cough is prominent in pulmonary tuberculosis (TB) and transmits infection, yet no tool has been validated for assessing cough symptoms. We evaluated the Leicester Cough Questionnaire (LCQ) for measuring cough-related quality of life (QOL) in TB.
Method The face validity of the LCQ was evaluated by structured interviews with patients and a multi-disciplinary team (MDT) discussion (respiratory physicians and nurses). Consecutive patients with TB completed the LCQ just before or within 7 days of starting therapy; a subgroup completed a repeat questionnaire approximately two weeks after the first. Internal reliability (inter-relatedness between items), concurrent validity (association with cough severity visual analogue scale [VAS] score and 24-hour cough frequency measured with the Leicester Cough Monitor), and responsiveness were evaluated.
Results The MDT and patients thought the LCQ to be relevant, comprehensive and useful in TB and no modifications were suggested. Forty patients completed the questionnaire before (n = 29) or just after (n = 11) the start of treatment. Internal reliability of responses was high (Cronbach’s α = 0.93). LCQ scores were correlated with both the VAS (Spearman’s ρ = -0.69 [95% confidence intervaI -0.83 to -0.46], p < 0 0.0001) and 24-hour cough frequency (ρ = -0.36 [-0.62 to -0.04], p = 0 .023), and were worse pre-treatment in culture-positive compared to culture-negative disease (median 12.4 [IQR 8.5–17.4] vs 18.7 [17.8–19.6] respectively, p = 0 .052). There was no evidence of association with other markers of disease severity (sputum smear positivity, lung cavities and radiographic extent of disease), but a trend towards worse LCQ scores amongst current smokers than non-smokers (12.6 [8.3–14.4] vs 17.1 [11.1–21.0] respectively, p = 0 .075).
All patients who repeated the questionnaire appeared adherent to TB medication. There were substantial improvements in LCQ responses after a median of 14 (10–14) days’ treatment (n = 12; median [IQR] score 9.1 [8.1–14.5] at baseline, 18.3 [14.5–19.4] at two weeks, median improvement 5.1 [1.8–9.7], p = 0 .003; Figure 1). The effect size of the change in LCQ scores was 1.17.
Conclusion The LCQ is a valid instrument for evaluating cough-related QOL in TB and may be a useful outcome measure to evaluate therapy.