Background: Evidence evaluating quality improvement (QI) strategies designed to optimise diabetes management is rapidly growing, with almost 20 new RCTs published in English annually. Using traditional systematic review (SR) methods to synthesise this evidence is no longer sustainable, as reviews are in danger of being out of date by the time they are published. Living systematic reviews (LSR) which are ‘continually updated, incorporating relevant new information as it becomes available’, have been proposed as a solution to ensure rigorous, timely synthesis in rapidly evolving fields.
Objectives: To review our experience in transitioning a large-scale SR into a LSR, and to provide researchers with the information they require to conduct their own LSR.
Methods: A SR of 278 trials evaluating diabetes QI interventions was transitioned into a Cochrane LSR in 2017. Operationalising the transition of this review into a LSR required numerous methodological considerations, including when and how to update our search strategy, what databases to search, what screening platforms to use, when to update analyses, and the role of machine learning. The publication model also required deliberation to balance the need for maximum visibility and new citations/DOI with each publication, while minimising author/editor workload.
Results: We will review decisions that were made to ensure the successful transition of our SR into a LSR. We will reflect on the expert opinions received, and will integrate this knowledge with our own experiences. Methods to facilitate and streamline the process will be discussed, with a particular focus on capabilities of automation/machine learning. We will provide our final recommendations and thoughts, including suggestions on how other research teams might conceptualise the transition of their own SR into a LSR.
Conclusions: By detailing our decisions and experiences in transitioning an existing large-scale SR into an LSR, we hope to contribute to the discussion of the methodology for this novel, emerging field. Furthermore, we hope to provide researchers with the tools they require to make informed decisions for their own LSR.