Article type
Abstract
Background: Evidence-based clinical practice guidelines utilize evidence from systematic reviews to develop practice recommendations. According to GRADE, recommendations are classified by strength (strong or conditional) and direction (for or against). Strong recommendations, usually based on high/moderate certainty evidence, imply that evidence about the relative effect of two options is unlikely to change. Thus, it is also less likely that there is equipoise to justify the conduct of a new trial.
Objectives: To describe the influence of strong recommendations from a living COVID-19 guideline on the initiation and conduct of COVID-19 randomized controlled trials.
Methods: We will use the WHO living COVID-19 treatment guidelines as a source of recommendations. Across all available subpopulations (i.e. non-severe and severe COVID-19), and for each recommendation, we will extract the date at which each recommendation was made. To describe the influence of recommendations on trials, we will use data collected from the living systematic review and network meta-analysis (LSRNMA) which informs these guidelines and conduct searches in trial registries. For each intervention, pairs or reviewers will extract the number of trials investigating that treatment, trial registration date, trial recruitment period, and trial completion status. Within strong and conditional recommendations separately, we will describe whether there were any changes per treatment following the release of a recommendation (e.g., reduction in trials) and compare changes observed between strong and conditional recommendations.
Results: This work is in progress. To date, there have been 14 versions of the WHO living guidelines on COVID-19 treatments. The latest version (10-Nov-2023) included seven strong recommendations for non-severe COVID-19, three for severe/critical COVID-19, and three for all COVID-19 severities. Five treatments have conditional recommendations and four have been recommended only for research settings. Among treatments with a strong positive recommendation, our LSRNMA has identified a median of 29 trials (IQR 9 to 49; range 8 to 51) and a median of 24 trials (IQR 10 to 39; range 4 to 42) for treatments with a negative recommendation.
Conclusions: Our conclusions will provide insight into how recommendations influence trial design, which can have important ethical considerations.
Objectives: To describe the influence of strong recommendations from a living COVID-19 guideline on the initiation and conduct of COVID-19 randomized controlled trials.
Methods: We will use the WHO living COVID-19 treatment guidelines as a source of recommendations. Across all available subpopulations (i.e. non-severe and severe COVID-19), and for each recommendation, we will extract the date at which each recommendation was made. To describe the influence of recommendations on trials, we will use data collected from the living systematic review and network meta-analysis (LSRNMA) which informs these guidelines and conduct searches in trial registries. For each intervention, pairs or reviewers will extract the number of trials investigating that treatment, trial registration date, trial recruitment period, and trial completion status. Within strong and conditional recommendations separately, we will describe whether there were any changes per treatment following the release of a recommendation (e.g., reduction in trials) and compare changes observed between strong and conditional recommendations.
Results: This work is in progress. To date, there have been 14 versions of the WHO living guidelines on COVID-19 treatments. The latest version (10-Nov-2023) included seven strong recommendations for non-severe COVID-19, three for severe/critical COVID-19, and three for all COVID-19 severities. Five treatments have conditional recommendations and four have been recommended only for research settings. Among treatments with a strong positive recommendation, our LSRNMA has identified a median of 29 trials (IQR 9 to 49; range 8 to 51) and a median of 24 trials (IQR 10 to 39; range 4 to 42) for treatments with a negative recommendation.
Conclusions: Our conclusions will provide insight into how recommendations influence trial design, which can have important ethical considerations.