Clinical study report use within Cochrane Reviews: recent developments within Cochrane

Tags: Oral
Flemyng E1, Churchill R2, Dwan K1, Lasserson T1, Meerpohl J3, Skoetz N4, Stweart L2, Tovey D1
1Cochrane, 2University of York, 3University of Freiburg, 4University of Cologne

Background: a recent report to Cochrane has highlighted that there is good evidence that journal reporting of randomized controlled trials (RCTs) is often incomplete and affected by reporting bias. It suggested that systematic review authors should therefore also access other sources of information to ensure that, for each eligible trial, they obtain maximum data on review outcomes and other key trial attributes. It highlighted the value of clinical study reports (CSRs). CSRs are documents submitted to regulators to obtain marketing licenses for drugs and biologics and are a key source of detailed trial information, providing much more information than other sources, particularly adverse event data. CSRs can be long (< 10,000s of pages), but they are generally highly structured. Where they exist, CSRs are likely to be the individual most comprehensive and complete source of information about a trial.

In response to this report, and to the growing interest in data sharing and increased opportunities to access CSRs, Cochrane has convened a Clinical Study Report Working Group to consider how CSRs might be incorporated within Cochrane work. The Clinical Study Report Working Group will include Cochrane members who have and have not previously used CSRs, from methodologists and from Review Groups.

Objectives: this presentation will provide an update on recent developments in Cochrane to support CSR use by Cochrane review authors.

1) We will provide an update following an in-person consultation meeting with stakeholders who have used, or will use, CSRs in Cochrane Reviews. This meeting reflected on the experience of those who have used CSRs as the basis for reviews and explored the practical feasibility of using CSRs in Cochrane Reviews.

2) Key areas of feedback will include access to, and major obstacles to, using CSRs, as well as proposals on how Cochrane could help overcome these.

3) Updates on pilots of Cochrane Reviews using CSRs that are initiated following the consultation meeting and what challenges and benefits we have seen so far.

Conclusions: review authors seek to obtain the fullest and most accurate trial data for Cochrane Reviews. CSRs have the potential to improve understanding of and increase the completeness of RCT data that can be included in a review, and thereby improve review reliability. It is therefore important that Cochrane Review authors are clear on what CSRs are, how to use them and in what context.

Patient or healthcare consumer involvement: as this project is currently looking at the feasibility of using CSRs in Cochrane Reviews and processes, and this stage of a Cochrane Review doesn’t generally involve patient or consumers, and patients or consumers do not usually have experience extracting data from CSRs, we have not involved patients at this point in time. However, if stages of the project have implications in terms of conduct or output on patients and consumers, we will be sure to involve them.