Crossover trials are widely used, but no standards for analysis and have been established. Mills et al. reviewed a representative sample of published crossover trials.
Of 526 identified randomized controlled trials were 116 crossover trials. Only a few trials reported allocation concealment (17%) or sequence generation (7%). Sample size calculations were reported only in 20% of the trials and just 31% of these considered pairing of data in the calculation.
Carry-over effects were addressed in 29% of trial’s methods, but almost all trials (93%) tested for treatment effects using paired data and by-group results (95%). Confidence intervals or standard errors were presented only in 29% of the articles.
The authors conclude that reports of crossover trials often omit important methodological issues, and that guidelines for the conduct and reporting of crossover trials might improve the conduct and reporting of such studies.
Mills EJ, Chan A-W, Wu P, Vail A, Guyatt GH, Altman DG. Design, analysis, and presentation of crossover trials. Trials 2009, 10:27.