Checklist for statistical assessment of randomised controlled trials

Reports of randomised controlled trials must conform to the CONSORT statement. LJPC statistical reviewers also use the following checklist to assess such papers:

Design features

• Objective of the trial sufficiently described?
• Satisfactory statement of diagnostic criteria for entry to the trial?
• Satisfactory statement of the source of participants?
• Concurrent (not historical) controls used?
• Interventions well defined?
• Random allocation to intervention used?
• Method of randomisation described?
• Acceptably short delay from allocation to start of intervention?
• Potential degree of blindness used?
• Satisfactory statement of criteria for outcome measures?
• Outcome measures appropriate?
• Pre-study calculation of sample size reported?
• Duration of post-intervention follow up stated?

Conduct of trial

• Intervention and control groups comparable in relevant measures?
• High proportion of participants followed up?
• High proportion of participants complete intervention?
• Were participants who dropped out from intervention and control groups described adequately?
• Adverse effects of intervention reported?

Analysis and presentation

• All statistical procedures adequately described or referenced?
• Statistical analyses appropriate?
• Prognostic factors adequately considered?
• Presentation of statistical material satisfactory?
• Confidence intervals given for the main results?
• Conclusions drawn from the statistical analysis justified?

Recommendation on paper

• Is the paper of acceptable statistical standard for publication?
• If "No" , could it become acceptable with suitable revision?