Prasad 2005

Methods	RCT. Parallel design. Location: single centre in UK. Duration: 12 months.
Participants	30 participants (20 girls, 10 boys matched). Age range 6 ‐ 15 years; mean age 11.5 years. 15 to each treatment arm. BMI, LCI and FEV₁ were well matched. One from each group withdrew because either they preferred their previous device, or they found it too fiddly to clean.
Interventions	PEP versus cornet. The treatment was used as their main ACT for 12 months.
Outcomes	FEV₁; LCI; pulmonary exacerbations; health perception; quality of life. FEV₁ and LCI were measured at start 6 months and 12 months. Quality of Well‐Being Scale, health perception and frequency of exacerbations measured at beginning and end.
Notes	Abstracts only, no full paper published as yet. Blinding not possible.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Randomisation was stratified for age, sex and FEV₁, further details not given.
Allocation concealment (selection bias)	Unclear risk	Not discussed.
Blinding (performance bias and detection bias) All outcomes	High risk	Not possible.
Incomplete outcome data (attrition bias) All outcomes	Low risk	One child from each group dropped out after randomisation, reason given.
Selective reporting (reporting bias)	High risk	Ongoing study which spanned 3 abstracts, but no full paper as yet identified.
Other bias	High risk	The authors themselves questioned whether quality of life measures were reliable in children as they may be unable to accurately compare current health to that experienced the previous year.