Article type
Abstract
Background: Health care professionals require up-to-date and comprehensive evidence to inform medical practice. Living systematic reviews and network meta-analyses (LSRNMAs) have emerged as an innovative approach to produce timely and comprehensive summaries of the evidence. LSRNMAs, however, also pose practical challenges.
Objective: During the COVID-19 pandemic, we led 3 LSRNMAs addressing the effectiveness of treatments and prophylactics for acute COVID-19. Building on this experience, we intend to produce an LSRNMA addressing the effectiveness of interventions for management of long COVID, applying our learned insights to ensure the project's efficiency, feasibility, and methodological rigor.
Methods: We refined the methods and workflow of our LRSNMAs of acute COVID-19 for our upcoming LRSNMA of long COVID.
Results: For our LSRNMAs of acute COVID-19, we performed daily searches of research databases. Conversely, for our upcoming LSRNMA of long COVID, we intend to time searches with scheduled updates of the review or with the publication of seminal trials. This approach enables a single team of reviewers to perform both screening and data extraction sequentially.
In previous LSRNMAs, we employed a workflow where junior reviewers worked independently and in duplicate to extract data and resolved discrepancies through asynchronous email communications. Senior reviewers were tasked with verifying critical data fields. This workflow proved effective and we intend to maintain this approach for our LSRNMA of long COVID. This approach allows involvement of junior reviewers while maintaining data integrity and accommodates reviewers with different schedules.
During the COVID-19 pandemic, trialists reported difficulty publishing negative trials, trials that were unable to achieve their intended sample size, and trials that addressed interventions that had already been studied in previously published trials. To mitigate the impact of publication bias, we will integrate methods for prospective systematic reviews—systematic reviews that include unpublished and interim data from ongoing and completed trials—for select interventions.
Conclusions: LSRNMAs are a promising solution to the limitations of traditional reviews, offering a more efficient way to incorporate emerging evidence into guidelines and clinical practice. Although producing LSRNMAs is time- and resource-intensive, our methods, strategies, and tools make LSRNMAs a feasible approach for up-to-date and comprehensive evidence synthesis.
Objective: During the COVID-19 pandemic, we led 3 LSRNMAs addressing the effectiveness of treatments and prophylactics for acute COVID-19. Building on this experience, we intend to produce an LSRNMA addressing the effectiveness of interventions for management of long COVID, applying our learned insights to ensure the project's efficiency, feasibility, and methodological rigor.
Methods: We refined the methods and workflow of our LRSNMAs of acute COVID-19 for our upcoming LRSNMA of long COVID.
Results: For our LSRNMAs of acute COVID-19, we performed daily searches of research databases. Conversely, for our upcoming LSRNMA of long COVID, we intend to time searches with scheduled updates of the review or with the publication of seminal trials. This approach enables a single team of reviewers to perform both screening and data extraction sequentially.
In previous LSRNMAs, we employed a workflow where junior reviewers worked independently and in duplicate to extract data and resolved discrepancies through asynchronous email communications. Senior reviewers were tasked with verifying critical data fields. This workflow proved effective and we intend to maintain this approach for our LSRNMA of long COVID. This approach allows involvement of junior reviewers while maintaining data integrity and accommodates reviewers with different schedules.
During the COVID-19 pandemic, trialists reported difficulty publishing negative trials, trials that were unable to achieve their intended sample size, and trials that addressed interventions that had already been studied in previously published trials. To mitigate the impact of publication bias, we will integrate methods for prospective systematic reviews—systematic reviews that include unpublished and interim data from ongoing and completed trials—for select interventions.
Conclusions: LSRNMAs are a promising solution to the limitations of traditional reviews, offering a more efficient way to incorporate emerging evidence into guidelines and clinical practice. Although producing LSRNMAs is time- and resource-intensive, our methods, strategies, and tools make LSRNMAs a feasible approach for up-to-date and comprehensive evidence synthesis.