Darbee 2005.

Study characteristics
Methods	Quasi‐RCT. Cross‐over design. Location: single centre in USA. Duration: average length of hospital stay was 11 days (range 9 ‐ 15 days); no follow‐up reported.
Participants	15 participants (8 males, 7 females). Aged at least 7 years, mean (SD) age 17.5 (4.2) years. Participants were admitted to hospital for acute exacerbation. All participants performed HFCWO 1 ‐ 3 times daily as outpatients before admission, but none had performed PEP.
Interventions	PEP versus HFCWO. Both treatments were alternated within 48 hours of hospital admission and then reversed prior to discharge. Treatment lasted 30 minutes.
Outcomes	RFTs and SaO₂ measured before and after every intervention. Each intervention was only done twice i.e. day 1 or 2 following admission then day ‐1 or ‐2 prior to discharge.
Notes	Average length of hospital stay was 11 days (range 9 ‐ 15 days). 3 participants discharged while still receiving intravenous antibiotics, for these participants the final measurement was taken within 48 hours of the final dose of antibiotic.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Participants were assigned to treatment order by numbering them consecutively, 1 through 15, at study entry. On the basis of a coin toss at admission, participant 1 and all odd‐numbered participants were randomly assigned to perform HFCWO on day 1 and PEP breathing on day 2, and even‐numbered participants performed PEP breathing on day 1 and HFCWO on day 2. At discharge, participants received treatment in the order opposite the treatment order at admission.
Allocation concealment (selection bias)	High risk	Used alternation.
Blinding (performance bias and detection bias) All outcomes	Unclear risk	No details given.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	No details given.
Selective reporting (reporting bias)	Unclear risk	Not possible to compare original trial protocol with final abstract.
Other bias	Unclear risk	The authors thanked Hill‐Rom for providing the Vest® device.