Table 2

Quality assessment of included trials and potential sources of bias

Blinding to treatment allocationDrop outs/losses to follow upOther
StudyRPPtPrOATotal (n)IG (n)CG (n)Reasons given?PC?HS?Comments
RP, randomisation process used; Pt, patient; Pr, provider; OA, outcome assessor; IG, intervention group; CG, control group; PC, power calculation reported; HS, hypothesis stated a priori; NR, not reported or unclear.
Taylor et al9NRYesYesNR8/57 (14%)NRNRYesYesYesPost hoc power calculation shows that study was powered to detect a difference of 0.9 in cough score which is equivalent to natural resolution of cough at day 3. Authors argue that smaller reductions in cough scores are unlikely to be clinically important.
Nespoli et al10NRYesYesYesNRNRNRNoNoNoHigh likelihood of bias.
Antihistamine–decongestant combinations
Clemens et al11NRNRYesYesNRNRNRNoYesYesPatients stated to be “randomly assigned in a double blind fashion”.
Hutton et al12NRYesYesYes12/96 (12.5%)6/36 (16.6%)3/27 (11.1%)YesYesYesSome children in open treatment group (n=9) and drug group (n=3) took other medication.
Other combinations
Reece et al13NRYesNRNRNRNRNRNoNoNoHigh likelihood of bias, medication sponsored by medical director of a laboratory who also performed the analyses.
Sakchainanont et al14NRYesYesYes7/150 (4.6%)NRNRYesNoYesBonferroni correction for multiple comparisons used.