Living recommendations do not "flip-flop": examining the probability of directional changes to recommendations in...

Living approaches to guideline development enable the maintenance of up-to-date recommendations, ensuring that users are basing health care decisions on the best available evidence. At present, the stability of living recommendations across multiple updates is yet to be established, and there is a risk that the direction of recommendations may change following the frequent inclusion of new evidence. The objective of this article is to examine the probability of living recommendations changing direction multiple times (ie, flip-flopping) following identification and inclusion of new relevant information.

All updates to living recommendations across three exemplar living guidelines focused on COVID-19, stroke, and inflammatory arthritis, were quantified and characterized. Information regarding the recommendation, date of update, and total number of studies and participants included was captured. Changes were classified based on impact on the recommendation, with a particular focus on changes in the direction of recommendations.

There were a total of 241 individual updates across all three guidelines over a period of 13 cumulative years, the majority of which did not affect the strength or direction of recommendation. Although there were 26 instances in which a recommendation changed direction, there were no instances in which a recommendation changed direction twice, in other words, flip-flopped.

Although living recommendations do change direction, the current analysis suggests that this happens infrequently and instances of recommendations flip-flopping are likely rare. Guideline users can be confident that the decisions they make today would not suddenly change following updates to living recommendations.