Article type
Year
Abstract
Background: For meaningful comparisons of interventions across systematic reviews (SRs), including network meta-analyses, outcome measures in SRs should be consistently and clearly specified. Previous research has demonstrated substantial variation in specification of outcome measures across SRs addressing glaucoma.
Objectives: To assess variation in outcome measure specification across Cochrane SRs addressing three common eye conditions: age-related macular degeneration (AMD), cataract, and diabetic retinopathy (DR).
Methods: We first determined all outcomes assessed across the SRs addressing these conditions, and ranked the outcomes by frequency. Two abstractors independently extracted information from the methods sections of these SRs about the three most frequently assessed outcomes. For each outcome, we extracted the measure(s) used in the analysis, whether continuous (e.g., mean visual acuity at a follow-up time point) or discrete (e.g., percent of participants falling into a pre-specified category).
Results: We identified 36 completed Cochrane SRs and protocols addressing AMD (n = 16), cataract (n = 16), and DR (n = 4). The most frequent outcomes were visual acuity (n = 35), quality-of-life (n = 32), and contrast sensitivity (n = 13). A variety of outcome measures were specified (Table). For example, for AMD, 8/16 SRs assessing visual acuity used mean visual acuity at a follow-up time point, 4/16 used mean change in visual acuity from baseline to a follow-up time point, 5/16 used number/percent of participants falling into pre-specified categories, and 6/16 reported that they would analyze visual acuity as provided in included studies (Box). Outcome measures were often not specified, for example, in 6/9 SRs on AMD assessing contrast sensitivity and in all 4 SRs on DR assessing quality-of-life.
Conclusions: While certain outcomes (e.g., visual acuity) were assessed frequently in SRs, specific outcome measures were used in various ways and were often not specified. SR authors should be aided by review group editors to be more consistent and explicit in outcome measures used.
Objectives: To assess variation in outcome measure specification across Cochrane SRs addressing three common eye conditions: age-related macular degeneration (AMD), cataract, and diabetic retinopathy (DR).
Methods: We first determined all outcomes assessed across the SRs addressing these conditions, and ranked the outcomes by frequency. Two abstractors independently extracted information from the methods sections of these SRs about the three most frequently assessed outcomes. For each outcome, we extracted the measure(s) used in the analysis, whether continuous (e.g., mean visual acuity at a follow-up time point) or discrete (e.g., percent of participants falling into a pre-specified category).
Results: We identified 36 completed Cochrane SRs and protocols addressing AMD (n = 16), cataract (n = 16), and DR (n = 4). The most frequent outcomes were visual acuity (n = 35), quality-of-life (n = 32), and contrast sensitivity (n = 13). A variety of outcome measures were specified (Table). For example, for AMD, 8/16 SRs assessing visual acuity used mean visual acuity at a follow-up time point, 4/16 used mean change in visual acuity from baseline to a follow-up time point, 5/16 used number/percent of participants falling into pre-specified categories, and 6/16 reported that they would analyze visual acuity as provided in included studies (Box). Outcome measures were often not specified, for example, in 6/9 SRs on AMD assessing contrast sensitivity and in all 4 SRs on DR assessing quality-of-life.
Conclusions: While certain outcomes (e.g., visual acuity) were assessed frequently in SRs, specific outcome measures were used in various ways and were often not specified. SR authors should be aided by review group editors to be more consistent and explicit in outcome measures used.