Article type
Abstract
Background: An estimated 80 systematic reviews were published/day in 2019. Systematic reviews take ~6 months to complete for experienced teams, with 20% of the work required for title/abstract screening. Previous research has estimated 36 seconds are required to screen each title/abstract. Systematic review management software offers modules/automation features to expedite the screening process.
Objectives: To estimate the utilisation and potential impact of systematic review management software.
Methods: We systematically searched for systematic review publications in PubMed, Scopus, EBSCO and Cochrane databases (1st January-12th May 2023). There were 23,114 references after de-duplication and title and abstract screening. A 10% random sample was selected for full-text screening and data extraction.
Results: We screened 2,312 papers, excluded 362 and extracted data from 1,950 systematic review reports. Based on this sample, we estimate that 148 systematic reviews were published per day in 2023 (totalling 54,020 in 2023).
A mean of 4.6 (range 2-31) databases/registries were searched and the mean number of references imported was 5,728.6 per review (range 11-1,515,595). The mean number of included studies was 39.6 (range 0-2,458) representing 6.9% of original references.
Based on the published estimate of 36 seconds/abstract per reviewer to screen, we estimate that manual screening of title and abstracts takes an average of 114.6 hours (14.3 person-work-days) for dual-reviewer screening (excluding time required for conflict resolution). From a global resource perspective, this equates to over 6.2 million hours (773,836.5 person-work-days) per year.
Systematic review management software was reported in 295/1,950 (15.1%) reviews. Fourteen tools were identified, with the most common being Covidence (53.6%), Rayyan (36.7%), DistillerSR (3.7%) and Eppi-Reviewer (1.7%).
Conclusions: The volume of systematic review publication continues to increase rapidly and is now more than ten times the 2010 estimate by Bastian et al. Reported uptake of systematic review management software/automation was low despite efficiency benefits. There is significant untapped potential to reduce the total global resources required to identify studies for inclusion in systematic reviews.
Objectives: To estimate the utilisation and potential impact of systematic review management software.
Methods: We systematically searched for systematic review publications in PubMed, Scopus, EBSCO and Cochrane databases (1st January-12th May 2023). There were 23,114 references after de-duplication and title and abstract screening. A 10% random sample was selected for full-text screening and data extraction.
Results: We screened 2,312 papers, excluded 362 and extracted data from 1,950 systematic review reports. Based on this sample, we estimate that 148 systematic reviews were published per day in 2023 (totalling 54,020 in 2023).
A mean of 4.6 (range 2-31) databases/registries were searched and the mean number of references imported was 5,728.6 per review (range 11-1,515,595). The mean number of included studies was 39.6 (range 0-2,458) representing 6.9% of original references.
Based on the published estimate of 36 seconds/abstract per reviewer to screen, we estimate that manual screening of title and abstracts takes an average of 114.6 hours (14.3 person-work-days) for dual-reviewer screening (excluding time required for conflict resolution). From a global resource perspective, this equates to over 6.2 million hours (773,836.5 person-work-days) per year.
Systematic review management software was reported in 295/1,950 (15.1%) reviews. Fourteen tools were identified, with the most common being Covidence (53.6%), Rayyan (36.7%), DistillerSR (3.7%) and Eppi-Reviewer (1.7%).
Conclusions: The volume of systematic review publication continues to increase rapidly and is now more than ten times the 2010 estimate by Bastian et al. Reported uptake of systematic review management software/automation was low despite efficiency benefits. There is significant untapped potential to reduce the total global resources required to identify studies for inclusion in systematic reviews.