Open access
the intervention is only performed once or a few times at the beginning of the trial (eg, FMT), the original trial design will be used. Trusting that most trials have been pre-specified and appropriately registered, the trial’s time point for the primary outcome evaluation will be used for the evaluation of effectiveness outcomes in the present meta-analysis. Safety measures will be evaluated as late as possible (ie, while still respecting the original design). Setting There will be no restrictions by type of setting. Language No language restrictions will be imposed, although only studies in languages other than English that can be trans- lated adequately by co-authors or by using Google Trans- late or other free accessible resources will be included, due to resource limits. A list of possibly relevant titles in languages that we did not manage to translate will be provided as an appendix. Publication types Both full-text manuscripts and abstracts will be included, including conference abstracts and abstracts for which the full text could not be obtained despite contacting the authors. Information sources The systematic search will be performed using three bibliographic databases: Embase (Ovid), Medline (Ovid) and Cochrane Library (Central). In addition, a search will be conducted in ClinicalTrials.gov and the WHO Interna- tional Clinical Trial Registry Platform portal (ICTRP) as recommended by the Cochrane Musculoskeletal Group 68 from inception of the data source to the study search date. After the inclusion of eligible studies, a backward and forward citation will be conducted in Scopus for twenty randomly selected reports. If no additional rele- vant records are identified, the search will be considered comprehensive. If new studies are found, full citation tracking will be conducted for all included reports to capture any remaining relevant records. Search strategy The search strategy was developed by BS and MSK in collaboration with an information specialist. An initial search of PubMed was undertaken, followed by an anal- ysis of the text words contained in the title and abstract, and of the index terms used to describe articles. The search strategy was developed using a 15-step method- ology recommended by Bramer et al 74 The search strategy was first tested in Embase as recommended by Bramer et al , and adjusted according to new keywords and subject headings across Medline and Cochrane Central. The search strategy will be reported according to the PRIS- MA-S guideline for reporting searches (https://www. prisma-statement.org/prisma-search). The prelimi- nary full search strategy for Embase is presented in the online supplemental file 1. This strategy will serve as the
template for searches in the other databases, with modi- fications made to reflect database-specific indexing terms (eg, MeSH in Medline) and search syntax.
Study records Study selection process
The search results will be managed in Covidence, which will be used for deduplication and reference screening. Records retrieved through the search strategy will be imported into Covidence by an information specialist and two reviewers. Two reviewers will independently screen titles and abstracts to identify potentially eligible studies or where eligibility is uncertain. Full-text articles will then be reviewed in detail to assess inclusion (figure 1). Reasons for exclusion will be recorded. The two reviewers will meet in person to compare selected reports, and any disagreements will be resolved through discussion or, if necessary, by consulting a third reviewer. Reviewers will not be blinded to journal titles, study authors or institutions. Data collection process Full-text articles deemed eligible after screening will undergo in-depth assessment by two independent reviewer groups (each comprising two reviewers). Data from each included trial will be entered into a customised REDCap database, 75 which will be designed so that data extracted from each trial will be entered into two similar, but separated records. Each group will have access only to its own record until data extraction is complete. The two review groups will independently extract data and assess risk of bias. Reviewers’ disagreements will be resolved through discussion within and between groups, both in person and during online meetings. If consensus cannot be reached, a fifth, blinded reviewer will adjudicate. If necessary, authors of included studies will be contacted to obtain missing or unpublished data. Before commencing formal data extraction, all reviewers will undergo structured training using a small set of randomly selected trials not included in the final analysis. This pilot phase will ensure consistency in data extraction, calibration of the customised REDCap form and harmonisation of interpretation across reviewer groups. Discrepancies identified during the pilot will be discussed collectively, and the data extraction form will be refined as necessary before full data extraction begins. Data items We will extract two main types of information from each included trial: 1. Study characteristics and methods: title, year of pub- lication, journal, author names, trial design (eg, par- allel, cross-over, adaptive), group allocation, blinding methods and processes, duration of follow-up, time and reasons for trial termination (if applicable) and types of statistical analyses performed. Participant characteristics will include age, sex, diagnosis, disease duration, disease status (eg, active disease, remission),
5
Kragsnaes MS, et al . BMJ Open 2025; 15 :e101593. doi:10.1136/bmjopen-2025-101593
Powered by FlippingBook