Table 3.

Five randomised controlled trials reporting cough severity scores at different timepoints

Publication	Treatment comparison	Time point, days	Patients analysed/randomised	Outcome definition	Experimental mean (95% CI)	Control mean (95% CI)	Group difference mean (95% CI)
Wang et al, 2014²⁶	Montelukast versus placebo	14	Exp: 137^a/137 Ctrl: 139^a/139	Change in the LCQ between baseline and two follow-up stages	2.7 (2.2 to 3.3)^b	3.6 (2.9 to 4.3)^b	−0.9 (−1.7 to −0.04)^{^c,^d}
Wang et al, 2014²⁶		28	Exp: 137^a/137 Ctrl: 139^a/139		5.2 (4.5 to 5.9)^b	5.9 (5.1 to 6.7)^b	−0.5 (−1.5 to 0.6)^{^c,^d}
Zanasi et al, 2014²⁷	Salbutamol plus ipratropium bromide versus placebo	10	Exp: 41/46 Ctrl: 42/46	The change in both daytime and night time cough severity, as assessed by the verbal category descriptive score	Daytime: 1.32 (0.93 to 1.71)^e Night time: 0.37 (0.17 to 0.57)^e	Daytime: 2.14 (1.73 to 2.55)^e Night time: 0.74 (0.49 to 0.99)^e	NR
Zanasi et al, 2014²⁷		20	Exp: 41/46 Ctrl: 42/46		Daytime: 0.41 (0.17 to 0.65)^e Night time: 0.15 (0.02 to 0.28)^e	Daytime: 0.64 (0.38 to 0.90)^e Night time: 0.17 (0.04 to 0.30)^e	NR
Woodcock et al, 2010(A)²⁸	NOP1 receptor agonist versus placebo	5	Exp: 26/27 Ctrl: 30/30	Change in cough severity scores	−0.57 (NR)	−0.49 (NR)	NR
Woodcock et al, 2010(B)²⁸^{, ^f}	Codeine versus placebo	5	Exp: 33/34 Ctrl: 30/30		−0.72 (NR)	−0.49 (NR)	NR
Ponsioen et al, 2005³⁰	Fluticasone propionate versus placebo	14	Exp: 65/67 Ctrl: 68/68	Cough score	1.4 (0.2 SEM)	1.9 (0.1 SEM)	NR
Pornsuriyasak et al, 2005³¹	Budesonide versus placebo	14	Exp: 15/15 Ctrl: 15/15	Symptom score	3.93 (1.70 to 6.16)^e	4.27 (2.58 to 5.96)^e	NR
Pornsuriyasak et al, 2005³¹		28	Exp: 14/15 Ctrl: 12/15		2.26 (1.00 to 3.52)^e	2.66 (1.00 to 4.32)^e	NR

↵a Missing data imputed by study authors using last observation carried forward: 14 days: 19 (Exp) and 17 (Ctrl). 28 days: 26 (Exp) and 29 (Ctrl).
↵b Mean difference in scores compared to baseline.
↵c Adjusted for numerous baseline variables (baseline scores, age, sex, duration of cough, pertussis status, pertussis immunisation status, atopy, paroxysmal cough severity, and exercise-induced cough severity); unadjusted analysis slightly and not relevantly different. Higher LCQ score indicates higher quality of life. Therefore the result is in favour of the control.
↵d Higher Leicester Cough Questionnaire score indicates higher quality of life, that is, the result is in favour of the control. For all other cough scores, lower values indicate fewer cough symptoms.
↵e 95% CI calculated by authors from standard deviation.
↵f Woodcock A and B refers to the randomised controlled trial (reference ²⁸) which consists of three treatment arms. The first line in Table above lists the NOP1 receptor agonist versus placebo and the second describes codeine versus placebo. Ctrl = control group. Exp = experimental group. LCQ = Leicester Cough Questionnaire. NOP1 = nociception opioid 1. NR = not reported. SEM = standard error of the mean.