TABLE 3

Recommendations for Practice

Ensure the randomization sequence has been generated by using appropriate methods.
Always conceal the allocation sequence to the personnel recruiting for a trial and prospective participants.
Ensure blinding wherever possible of all key study personnel, patients, and outcome assessors. Where it is not possible to blind the intervention, consider having blinded outcome assessors and using objective outcomes as well as reliable, valid measurement tools.
Prespecify the outcome analysis plan including detailing primary and secondary outcomes and how they will be assessed; statistical tests to be used; subgroup or adjusted analyses; and interim analyses.
Always conduct sample size calculations a priori based on the primary outcome of interest and accepted parameters.
Track and report the number and reasons of participants who withdraw or are lost to follow-up. Conduct ITT or sensitivity analyses to account for missing data.
Detail all outcomes to be assessed, including benefits and harms, as well as how and when they will be assessed. Report on all outcomes assessed or justify changes to outcomes between protocol and final report.
Declare any financial support and the role of the sponsor in the design, conduct, analysis, or reporting of the trial.
Consult specialized resources for issues related to specific or advanced trial designs.
Prespecify how baseline imbalances will be handled in the analysis and report.
Register the trial with a recognized trial registry before patient recruitment.
Use existing standards and guidance to identify and report key items at the protocol stage (ie, SPIRIT). Consider publication of the study protocol.
Follow existing standards and guidance for reporting a randomized controlled trial (ie, The CONSORT Statement).