Table 2

Quality assessment of randomised controlled trials

StudyAdequate generation of allocation sequenceAdequate concealment of allocationKnowledge of allocated intervention adequately preventedIncomplete outcome data adequately addressedReports free of suggestion of selective outcome reportingFree of other problems that could put study at a high risk of biasComments
Allen et al11NoYesYesYesYesYes
Barr et al12NoNoYesYesYesYes
Dilli et al13NoYesYesYesYesNoSmall sample size in subgroup of infants included
Hatfield et al15YesYesYesYesYesYesNon-nutritive sucking part of control condition
Hatfield14YesYesYesNoYesYesData missing for 5 infants at 2nd observation
Lewindon et al16YesYesYesYesYesNoPacifier use differed
Lindh et al17YesYesYesYesYesYesCombination of interventions
Mörelius et al18NoYesYesYesYesYes
Ramenghi et al21YesYesYes for control and sucrose. No for glucoseYesYesYes
Reis et al22YesYesYes for outcome assessors. No for parents and nursesYesYesYes
Soriano and Gomez23NoNoYesYesYesNoInsufficient details given of assessment of infants' liking of the solution
Thyr et al24NoYesYes for outcome assessors. No for nurse giving injectionsYesYesYes