Language restrictions in systematic reviews should not be imposed in the search strategy but in the eligibility criteria if necessary

Language restrictions in systematic reviews (SRs) should be avoided because they may lead to language bias, a bias due the selection of reports published selected languages. It has been shown that studies certain languages are more likely report positive results [[1]Egger M. Zellweger-Zähner T. Schneider Junker C. Lengeler Antes G. randomised controlled trials English and German.Lancet. 1997; 350: 326-329Abstract Full Text PDF PubMed Scopus (783) Google Scholar,[2]Vickers A. Goyal N. Harland R. Rees Do countries produce only results? A review trials.Control Clin Trials. 1998; 19: 159-166Abstract (511) Scholar]. However, would decrease if there shift toward publications over last decades. There is conflicting evidence support this [[3]Galandi D. Schwarzer The demise trial: bibliometric study German-language health care literature, 1948 2004.BMC Med Res Methodol. 2006; 6: 30Crossref (46) Scholar,[4]Jia Y. Huang Wen J. Wang Rosman L. Chen Q. et al.Assessment indexing biases among Chinese-sponsored randomized clinical trials.JAMA Netw Open. 2020; 3: e205894Crossref (6) Nevertheless, when conducting an SR, authors sometimes choose impose restrictions. This can done at two distinct steps process SR—first by using search strategy second including as eligibility criterion during selection. first option, restricting language, means records written language(s) will not included retrieved searching. actively able deal with. For example, one wants limit searches articles searching PubMed, adding filter/restriction developed strategy. whose full-text English; regardless abstract. criterion, screening SR include full texts were predefined use both restrictions, or second. 2016 reported sample 300 SRs, 43% considered all their criteria [[5]Page M.J. Shamseer Altman D.G. Tetzlaff Sampson Tricco A.C. al.Epidemiology reporting characteristics biomedical research: cross-sectional study.PLoS Med. 2016; 13: e1002028Crossref (335) Hereby, we argue SRs while SRs. If justified reason for such restriction (e.g., time constraints, rapid reviews), stage selecting but literature. In case, excluded list studies. allow transparency regarding number eligible other future investigate these checking prior on same topic. also kept mind could easily update involving co-authors with appropriate skills). Listing along required AMSTAR 2 [[6]Shea B.J. Reeves B.C. Wells Thuku Hamel Moran al.Amstar 2: critical appraisal tool non-randomised healthcare interventions, both.BMJ. 2017; 358: j4008Crossref (1878) Scholar] PRISMA [[7]Page Moher Bossuyt P. Boutron I. Hoffmann Mulrow al.PRISMA 2020 explanation elaboration: updated guidance exemplars reviews.2020https://osf.io/preprints/metaarxiv/v7gm2/Date accessed: November 30, 2020Google tools. We acknowledge option might require resources than option. believe additional workload negligible most proportion was found less 5% [[8]Hartling Featherstone Nuspl Shave K. Dryden D.M. Vandermeer B. Grey literature reviews: contribution non-English reports, unpublished dissertations meta-analyses child-relevant reviews.BMC 17: 64Crossref (113) Scholar,[9]Jackson J.L. Kuriyama How often do exclude English?.J Gen Intern 2019; 34: 1388-1389Crossref (19) some topics where play huge role complementary alternative medicine, example [[10]Pham Klassen T.P. Lawson M.L. publication gave different depending whether intervention conventional complementary.J Epidemiol. 2005; 58: 769-776Abstract (141) translations financial resources. planning SR. showed 191 source funding, 20% declared no funding When seeking budget dealing Furthermore, even without funds, Translate viable translating into facilitate data extraction, risk assessment [[11]Jackson Anton Choi Fournier J.P. Geier A.K. al.The accuracy translate abstracting from non-English-language reviews.Ann 171: 677-679Crossref (49) conclusion, aware Ideally, used must used, provide rationale choice, level confirm article read approved named persons who satisfied authorship listed. They further order listed them. have given consideration protection intellectual property associated work impediments publication, timing respect property. so doing, followed regulations institutions concerning understand corresponding author sole contact editorial (including manager direct communications office). He responsible communicating about progress, submissions revisions final approval proofs. provided current, correct email address ( [email protected] ), which accessible author.