Posted 6 August 2019.
The ‘CONSORT 2010 statement: extension to randomised crossover trials’ published last month in The BMJ (the original CONSORT 2010 Statement can be found here). Leading a team of researchers on this work was Dr Kerry Dwan, Statistical Editor from the Cochrane Editorial and Methods Department, who is the first author on the paper.
In a cross-over trial participants receive two or more interventions in a randomised order, i.e. they are randomised to a sequence of interventions. Cross-over trials are often used in chronic conditions such as asthma and sleep apnoea.
Well-reported primary research is important for evidence to be used in practice and policy decisions, and for its use in systematic reviews. This extension of CONSORT is important guidance for improving how key aspects of the design, conduct and analysis are reported in cross-over trials. These items will help systematic review teams to appraise and analyse them in their reviews.
Cross-over trials can present particular challenges when they are being used in a meta-analysis. These include carryover effects, when an intervention has a long-lasting effect that persists in to the second treatment period, and participants dropping out after the first treatment period (see Chapter 23 of the Cochrane Handbook for Systematic Reviews of Interventions for more information – public release date September 2019 or Cochrane members can access it in draft form here). The analysis of cross-over trials should be based on paired data since participants are acting as their own controls, and it should be reported as such. However, many cross-over trials do not report baseline characteristics, or report analyses of outcome measures that fully reflect the within participant design. This has led some review teams to use data as if they were from a parallel group trial, which is a unit of analysis error.
Several studies (see here, here and here) have shown that data are often not reported by cross-over trials in a way that allows them to be included in meta-analysis appropriately. Nolan and colleagues looked at 53 reviews which synthesised evidence from 218 cross-over trials. Thirty-three (63%) Cochrane Reviews described a clear and appropriate method for the inclusion of cross-over data, and of these 19 (56%) were able to implement/use the same method to analyse results. 145 cross-over trials were described narratively or treated as parallel trials in reviews but in 30 (21%) of these trials data existed in the trial reports to account for the cross-over design. At the trial level, the analysis and presentation of results were often inappropriate or unclear, with only 69 (32%) trials presenting results that could be included in meta-analysis.
The extension of 14 items in the CONSORT 2010 statement to cross-over trials will help researchers and journal editors ensure clear reporting. It provides a revised checklist and a modified flowchart and baseline table, along with examples of good reporting. The intention is to improve future reporting of cross-over trials, which will in turn help systematic reviews that include cross-over trials.