Background: Including abstracts of randomised controlled trials (RCTs) in systematic reviews can help address publication bias, contribute data where none are otherwise available and increase precision in a metaanalysis. Where the question of interest cannot be answered by RCTs, some Cochrane reviews include non-randomised evidence. For these types of reviews, however, there is less certainty as to the benefits or otherwise of also including abstracts of such evidence. Objectives: This study examined the effects of including abstracts of RCTs and non-randomised studies on the results of a series of systematic reviews that were undertaken to assess the safety and efficacy of interventional procedures. Methods: For each review, numbers of included full text studies and abstracts were tabulated. The amount of data contributed by abstracts of RCTs and non-randomised studies to primary efficacy outcomes and adverse effects information was analysed and compared. Sensitivity analysis was undertaken by examining how excluding abstracts of RCTs and non-randomised studies affected the results of the reviews, both in terms of the amount of data subsequently available for primary efficacy outcomes and adverse effects and any impact on the direction of effect. Results: In 13 reviews including a total of 619 studies, 78 (13%) were abstracts, of which 22 (28%) reported RCTs and 56 (72%) reported non-randomised studies. Six reviews contained no abstracts. Across the remaining seven reviews, the median number of abstracts included was 11 (range 1 to 28). These seven reviews addressed a variety of topics, including interventions aimed at preventing medication error (medicines reconciliation) at hospital admission, mesh or grafts for vaginal wall prolapse repair, foam sclerotherapy for venous disease of the lower limbs, electrosurgery for tonsillectomy, sacral nerve stimulation for urinary and faecal incontinence, and photorefractive surgery for the correction of refractive error. Conclusions: The number of abstracts included in the reviews varied considerably. Detailed results will be presented on how abstracts of RCTs and non-randomised studies contributed data to the reviews and the effect that excluding them had on primary efficacy outcomes and adverse effects information, and direction of effect of the results.