Brown 1988.

Methods	Prospective comparison. Taping methods were "assigned randomly on admission" for 2 weight groups (<1500g or >1500g) and continued until intentional extubation, death or 28 days postnatal age. Patients were excluded if (for over half the time they were intubated): (1) the incorrect protocol was being used, (2) they were nasally intubated, (3) they were paralysed or (3) they were sedated. 35% were excluded from analysis for these reasons. After 4 months, data analysis revealed that taping method 2 was better, so for the remaining two months only taping method 2 was used and infants were randomised to receive either head restraint or not.
Participants	Infants who were admitted to the centre and required oral endotracheal intubation during a 6 month period
Interventions	Four groups of care practices. Two interventions using different taping methods and tape material. Group separated into 2, depending on whether head restraint was added or not. Tape method 1, with (n=22) or without head restraint (n=36) and tape method 2 with (n=31) or without head restraint (n=53)
Outcomes	Accidental extubation rates per 100 patient days of intubation Accidental extubation rates as affected by the length of intubation by Poisson regression analysis Pre‐extubation activities most likely to cause accidental extubation
Notes	Infants were stratified by birthweight (<1500g or >1500g)
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not stated how the random sequence was generated. Taping methods were "assigned randomly on admission"
Allocation concealment (selection bias)	Unclear risk	Not stated in the methods whether allocation was concealed
Blinding of participants and personnel (performance bias) All outcomes	High risk	Treatment could not be blinded due to the nature of the study
Blinding of outcome assessment (detection bias) All outcomes	High risk	Accidental extubations were picked up from the bedside chart and/or extubation logs. The nurse entering the events into the charts or logs could not be blinded due to the nature of the study. Not stated in the methods whether outcome assessment of the charts or logs were blinded
Incomplete outcome data (attrition bias) All outcomes	High risk	Analysis was not intention to treat. There was a large exclusion rate which affected the overall results
Selective reporting (reporting bias)	High risk	The tables included in the study summarising the results excluded 35% of infants' results because the research protocol was not followed correctly in these infants
Other bias	Unclear risk	35% of infants randomised were excluded due to various reasons including inconsistency in care methods provided. It was not stated whether the treatment group was similar at the start of the trial. It was not stated whether treatment groups were treated equally apart from the intervention. The study was interrupted early because method 2 was thought to be a better method and for the last two months of the study, only method 2 was used to tape the endotracheal tubes. The results included data to suggest there was a large variation in gestational ages and weights.