Experience report of two living systematic Cochrane reviews on COVID-19

Up-to-date systematic reviews (SRs) are essential for making evidence-based decisions. During the 2019 coronavirus (COVID-19) pandemic, there was a particular need for up-to-date evidence, making the living systematic review (LSR) approach an appropriate review type. However, this approach poses certain challenges.

We aim to provide practice insights and report challenges that we faced while conducting two Cochrane LSRs on COVID-19 treatments with (i) convalescent plasma and (ii) systemic corticosteroids. We address our objective with an experience report and share challenges of the following components based on Iannizzi et al. (2022): study design, publication types, intervention/comparator, outcomes, search strategy, review updates and transparent reporting of differences between review updates.

Regarding the study design, the plasma LSR included different study designs because RCT data were not available at the beginning of the pandemic, whereas for the corticosteroids LSR, which started several months later, RCT data were already available. The challenges in both LSRs included the publication types (preprints were included with caution) and the intervention/comparator, for instance the unavailability of standard of care for either LSR, or SARS-CoV-2 variants occurrence. Further challenges in both LSRs occurred in the components "outcome sets" (which had to be adjusted) and "literature search". The decision criteria for updating were based on important studies and available resources in both LSRs and policy relevance in the plasma LSR. Transparent reporting of the differences between the various update versions were discussed for both LSRs.

In summary, there are similarities and differences regarding challenges of review components for both LSRs. It is important to keep in mind that the two LSR examples presented here were conducted in the wake of the COVID-19 pandemic. Therefore, many of the challenges are attributable to the pandemic and are not specific to LSRs, such as constant adjustments of the outcome sets or changes in the database search. Nevertheless, we believe that some of these aspects are helpful for LSR authors and are applicable to other LSRs outside the pandemic context, particularly in areas where new evidence is rapidly emerging.