Jean-Baptiste Briere1, Kevin Bowrin2, Vanessa Taieb3, Aurélie Millier4, Mondher Toumi5, Craig Coleman6. 1. a Bayer AG , Berlin , Germany. 2. b Bayer plc , Reading , UK. 3. c Creativ-Ceutical , London , UK. 4. d Creativ-Ceutical , Paris , France. 5. e Aix-Marseille University , Marseille , France. 6. f School of Pharmacy , University of Connecticut , Storrs , CT , USA.
Abstract
OBJECTIVES: To identify existing guidelines, key recommendations, and existing limitations regarding the evaluation and use of real-world evidence (RWE) in meta-analyses (MAs) to generate clinical and epidemiological evidence: a systematic review of existing recommendations. METHODS: A literature search was performed in April 2017 in MEDLINE and Embase using the Ovid platform, the Cochrane Library, and other sources. No specific inclusion and exclusion criteria were applied, and no restrictions in timeframe, language, or geographical scope were imposed. RESULTS: The search strategy identified 1681 citations; 12 references were included in this review. Recommendations within the literature regarding the use of RWE in MAs are: (1) it may be useful to extract and analyze adjusted results because confounding is expected; (2) testing heterogeneity in the MA of RWE is important as it may minimize the potential for bias and generate hypotheses for future research; (3) limiting a search ≤2 bibliographic databases when conducting MAs of RWE will not provide a thorough summary of existing literature; and (4) the Meta-analysis Of Observational Studies in Epidemiology (MOOSE) checklist is a 35-item checklist developed to allow for more standardized reporting of MAs of RWE and address their limitations. LIMITATIONS: (1) No formal guidelines were found regarding the use of RWE in MAs; (2) no consensus was found on a preferred instrument for the assessment of RWE; and (3) critical appraisal of RWE is often omitted from Health Technology Assessment submissions. CONCLUSIONS: The inclusion of RWE in MAs may facilitate the confirmation of conclusions drawn from randomized controlled trials and, thus, reassure decision-makers that findings can be extrapolated to real-world populations. However, qualitative and quantitative bias may co-exist in MAs of RWE. Reviewers should select the most appropriate tools that match the study designs identified in a particular systematic review.
OBJECTIVES: To identify existing guidelines, key recommendations, and existing limitations regarding the evaluation and use of real-world evidence (RWE) in meta-analyses (MAs) to generate clinical and epidemiological evidence: a systematic review of existing recommendations. METHODS: A literature search was performed in April 2017 in MEDLINE and Embase using the Ovid platform, the Cochrane Library, and other sources. No specific inclusion and exclusion criteria were applied, and no restrictions in timeframe, language, or geographical scope were imposed. RESULTS: The search strategy identified 1681 citations; 12 references were included in this review. Recommendations within the literature regarding the use of RWE in MAs are: (1) it may be useful to extract and analyze adjusted results because confounding is expected; (2) testing heterogeneity in the MA of RWE is important as it may minimize the potential for bias and generate hypotheses for future research; (3) limiting a search ≤2 bibliographic databases when conducting MAs of RWE will not provide a thorough summary of existing literature; and (4) the Meta-analysis Of Observational Studies in Epidemiology (MOOSE) checklist is a 35-item checklist developed to allow for more standardized reporting of MAs of RWE and address their limitations. LIMITATIONS: (1) No formal guidelines were found regarding the use of RWE in MAs; (2) no consensus was found on a preferred instrument for the assessment of RWE; and (3) critical appraisal of RWE is often omitted from Health Technology Assessment submissions. CONCLUSIONS: The inclusion of RWE in MAs may facilitate the confirmation of conclusions drawn from randomized controlled trials and, thus, reassure decision-makers that findings can be extrapolated to real-world populations. However, qualitative and quantitative bias may co-exist in MAs of RWE. Reviewers should select the most appropriate tools that match the study designs identified in a particular systematic review.
Authors: Jennifer S Lin; M Hassan Murad; Brian Leas; Jonathan R Treadwell; Roger Chou; Ilya Ivlev; Devan Kansagara Journal: J Gen Intern Med Date: 2020-04-01 Impact factor: 5.128