Padman 1999b.

Study characteristics
Methods	RCT. Cross‐over design (used CPT between therapies as a washout period length of which was not defined). Location: single centre in USA. Duration: each therapy lasted 1 month.
Participants	15 participants aged 5 ‐ 17 years with CF. Not stated how many were males and how many females. Participants were clinically stable and able to perform RFT's, no hospitalisations in the month prior to study. 5 excluded due to hospital admission for acute exacerbation, 4 withdrew (no reason given). 6 participants completed the study.
Interventions	Flutter versus PEP versus CPT/PD. Each therapy was performed for 15 min 3x daily for 1 month. No changes in established medication regimen.
Outcomes	RFTs (FEV1, FEF25-75) performed at beginning and end of each new therapy, SaO₂, participant satisfaction.
Notes	THIS DUPLICATE REFERENCE HAS BEEN CREATED TO ALLOW DATA FOR BOTH TYPES OF OSCILLATING DEVICE TO BE ENTERED IN THE ANALYSIS. This study ID refers to the Flutter versus CPT section of the study.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Each participant was arbitrarily assigned to 1 of 3 groups of randomly sequenced therapies, no further details of method.
Allocation concealment (selection bias)	Unclear risk	Not discussed.
Blinding (performance bias and detection bias) All outcomes	Unclear risk	Not discussed.
Incomplete outcome data (attrition bias) All outcomes	High risk	9 withdrawals after randomisation took place, 5 excluded due to hospital admission for acute exacerbation, 4 withdrew (no reason given).
Selective reporting (reporting bias)	High risk	Gender split was not stated. Participants stated they felt better but there were no criteria given from which to establish this.
Other bias	Unclear risk	Scandipharm provided the flutter devices for the study.