Table 2

 Methodological quality of trials of 50% nitrous oxide for pain relief

StudyLoss to follow up(n/N (%))Method of randomisation and allocation concealmentBlinding
Castera et al, 2001150/100 (0)Random numbers. Allocation concealment not statedNurse and patient
Forbes and Collins 2000220/102 (0)States randomised. Allocation concealment not statedNone
Triner et al, 1999190/22 (0)Computer-generated sequence, sealed envelopesInvestigators and patient
Burton et al, 1998170/30 (0)Computer sequence generated by nurse not involved in studyInvestigators and patient
Notini-Gudmarsson et al, 1996102/40 (5)Randomly assigned. Allocation concealment not statedNot stated
Saunders et al, 1995160/131 (0)Randomly allocated. Allocation concealment not statedInvestigators and patient
Evans et al, 199510/30 (0)Sealed unmarked envelopes from boxNot stated
Saunders et al, 1994200/89 (0)Stratified block randomisationInvestigators and patient
Lindblom et al, 1994210/50 (0)Block randomisation, blocks of 10. Closed envelopesInvestigators and patient
Henderson et al, 199014Not statedNot statedOutcome assessor
Harrison et al, 1987230/70 (0)Patient’s choice of analgesiaUnclear
Kerr et al, 19751835/116 (30)Block randomisation. Sequentially numbered cylindersInvestigator and patient