Evidence-based clinical practice guidelines (CPGs) are important to promote evidence-based practice at the point of care. Methodological literature on the development of CPGs is growing, but less research has been done on the subject of CPG updates. Literature monitoring consumes great resources, and optimal strategies for balancing the workload of this process with the risks of overlooking important new evidence are not established.
We seek to inform future CPG update strategies through developing and evaluating a method for dynamic update of evidence-based CPGs. We will evaluate the performance of different literature monitoring strategies in a CPG update setting, in order to assess the validity of less work-intensive approaches and the risk of overlooking key references with such restricted strategies. We aim to assess the frequency with which literature updates trigger changes in CPG recommendations, as well as characterizing the type of change elicited. Updates will be published electronically through a novel structured CPG format. We wish to evaluate the feasibility of dynamic updates of CPGs when supported by new technological solutions.
We will conduct a prospective case study on the new Norwegian CPG for antithrombotic treatment published in November 2013. This CPG is digitally structured and published in the MAGIC authoring and publication platform (www.magicapp.org) and will be dynamically updated by the CPG group from the Norwegian Society of Thrombosis and Hemostasis in a systematic and transparent process based on the GRADE system. Throughout the process we will measure the sensitivity and specificity of literature update alert services (e.g. McMaster PLUS) and restricted search strategies (e.g. PubMed Clinical Queries) in terms of capturing “key studies” (i.e., studies that result in the update or modification of a CPG recommendation, using an exhaustive systematic literature search as the reference standard. We will assess the frequency with which key studies trigger modifications in CPG recommendations, categorizing modifications according to a taxonomy based on the GRADE system (e.g. impact on confidence in and magnitude of effect estimates). We will perform survival analyses for individual CPG recommendations changed during the study period. We will also evaluate the feasibility of the process for dynamic updating of CPGs through recording workload (number of hours per month) and requirement for methodological competence and clinical expertise in the CPG group.