| Methods |
Randomised placebo controlled cross‐over trial. |
| Participants |
41 children (8 to 18 years). |
| Interventions |
Azithromycin, 250 mg (500 mg if weight > 40 kg) once a day for 6 months versus placebo. |
| Outcomes |
% change in FEV1 (average of 4 and 6 month values, also for FVC and MEF), hearing, sputum bacterial densities, inflammatory markers, exercise tolerance, subjective well‐being. |
| Notes |
Treatment arms not reported individually, these have subsequently been calculated for the first arm from figures provided in the paper and IPD kindly provided by the authors. |
| Risk of bias |
| Bias |
Authors' judgement |
Support for judgement |
| Random sequence generation (selection bias) |
Low risk |
Provided by Statistics Dept at Pfizer, USA. |
| Allocation concealment (selection bias) |
Low risk |
Hospital Pharmacy department, described in detail. |
| Blinding (performance bias and detection bias)
All outcomes |
Low risk |
All parties involved, identical packaging for intervention and placebo. |
| Incomplete outcome data (attrition bias)
All outcomes |
Low risk |
Complete ITT analysis undertaken on primary outcome. Not clear from paper whether primary outcome was a post hoc protocol change, but subsequent correspondence has confirmed that this was determined a priori.
|
| Selective reporting (reporting bias) |
Low risk |
Not clear if primary outcome calculation (months 4 and 6 averaged for relative change) was an a priori decision. |
| Other bias |
Low risk |
Adequate washout, authors have provided IPD. |
