Background: The Dutch College of General Practitioners has a guideline programme consisting of approximately 100 evidence-based guidelines. Preferably we base our recommendations on systematic reviews of high quality that specifically provide us with an up-to-date summary of the evidence to answer our PICO-formulated key-questions.

Objectives: To assess the uptake of Cochrane reviews (CR) in guidelines for general practitioners.

Methods: We extracted 42 therapeutic PICO-questions from 3 recent (>2015) guidelines on urinary stones, knee complaints (knee osteoarthritis and patellofemoral pain syndrome) and non-specific low back pain. We searched for potentially relevant CRs in the Cochrane database answering our PICO-questions and whether these were used for answering our questions in the selected guidelines. In addition, we analysed reasons for not using available CRs.

Results: For 32 of the 42 (74%) PICO-questions a CR was available. For 26 (60%) PICO-questions a CR was used to summarise the evidence. When a CR was available, reasons for not using it were that a CR did not fully match the PICO-question (n=1) or an overview/NMA based on Cochrane methodology was available outside the Cochrane library (n=5). If a CR was used they were rarely up to date and it was necessary to perform an update of the literature in most cases (88%).

Conclusion: Although CRs were available for the majority of the PICO-questions in our selective sample of Dutch guidelines for general practice, their use seems not to be optimal. The mismatch between available Cochrane reviews and general practice guideline could be reduced by formulating PICO questions based on knowledges gaps raised from guidelines. To improve the production of relevant Cochrane reviews and to increase their clinical impact, guideline developers and (Cochrane) systematic reviewers should work more closely together.