Consideration of patient-relevant outcomes in Cochrane reviews

Tags: Poster
Richter B1, Heise T1
1Metabolic and Endocrine Disorders Group, Germany

Background: Cochrane reviews should include outcomes that are meaningful to patients, physicians and stakeholders to help establish informed decisions on the basis of adequate benefit-risk data. However, primary studies often focus on surrogate endpoints. Authors of Cochrane reviews should therefore make clear which patient-relevant outcomes (such as death from any cause, morbidity, complications, health-related quality of life, adverse events, burdens and economic outcomes) are available and which not.

Objectives: To evaluate whether authors of Cochrane reviews present patient-relevant outcomes and if these are quantified in critical sections of Cochrane reviews such as the abstract and plain language summary. As a secondary objective differences between Cochrane Review Groups (CRGs) will be analyzed.

Methods: All reviews published in The Cochrane Library (issue 2, 2012) will be transferred to a database. A random sample of five reviews per CRG will be drawn. Inclusion criteria comprise CRGs focusing on particular disease areas and reviews published after 2005. We will classify all stated endpoints in the main sections of the Cochrane reviews as (non) patient-relevant outcomes. All endpoints will be investigated whether they were reported as primary or secondary outcomes and if results were quantified. For comparison, we will scrutinize the total number of laboratory, imaging and other outcomes. We will analyze data descriptively with special consideration of the abstract, plain language summary and summary of findings sections of the Cochrane reviews.

Results: We are currently evaluating data and will present results during the Cochrane Colloquium.

Conclusions: The ’Methodological Expectations of Cochrane Intervention Reviews’ (MECIR) project (http://www.editorial-unit.cochrane.org/mecir) specifies that authors should choose outcomes being relevant to stakeholders and avoid trivial, biochemical, interim and process outcomes. However, this is not a mandatory but only ‘highly desirable’ condition. Our project tries to establish reliable information whether this policy should be changed.