Nineteen studies were included in the review: six randomised controlled trials (RCTs), seven randomised cross-over trials and six observational studies. The quality of the included trials was variable, most trials used randomisation and reported loss to follow-up, but few studies clearly reported intention-to-treat, allocation concealment or standardisation of the co-interventions. All studies were open-label.
Manual hyperinflation versus standard care (eight RCTs)
The eight trials showed mixed effects, with some positive effects and some null effects. Pulmonary compliance was increased with manual hyperinflation in four trials, but there was no effect in one trial.
Manual hyperinflation versus other strategies (five RCTs; one observational study)
The five trials showed little difference between groups. One observational study found an increase in pulmonary compliance.
Side-effects (two RCTs; six observational studies)
Most studies found that manual hyperinflation had no effect on systemic blood pressure and heart rate. Other results were presented in the review.