Hemato-oncology Research Output in Latin America and the Caribbean: A Bibliometric Analysis

Le résumé fourni par la source

This registration pre-specifies the objectives, information sources, eligibility criteria, search strategy, selection process, data items and analysis plan for a bibliometric analysis of the scientific literature on hematologic malignancies — leukemias, lymphomas, multiple myeloma and plasma-cell disorders, myelodysplastic syndromes, and myeloproliferative neoplasms — produced by authors affiliated with Latin America and the Caribbean (LAC). The aim is to map the volume, growth, geographic and institutional distribution, collaboration structure, thematic focus and dissemination of this body of work in order to identify research gaps and inform regional research-capacity building. Design: bibliometric analysis of bibliographic metadata; the unit of analysis is the publication record, with no full-text clinical-outcome screening. Information sources are PubMed/MEDLINE, Embase, Scopus and Web of Science Core Collection, together with the regional indices LILACS (via BVS/BIREME) and SciELO, which are included to capture regionally indexed output that international databases systematically omit. The search combines a disease block (controlled vocabulary exploded where available, plus title/abstract synonyms) with a LAC region block applied through the author-affiliation field; the country list follows the UN M49 definition of Latin America and the Caribbean and is fixed at registration. The full per-database search strings are deposited with this registration. Records are de-duplicated using a hierarchical key (DOI, then PMID, then normalized title + publication year + first-author surname) and screened by two independent reviewers against pre-specified eligibility criteria, with disagreements resolved by a third reviewer; the identification and selection flow is reported per PRISMA 2020. Extracted items cover bibliographic, geographic, institutional, thematic, collaboration, citation/dissemination, funding and source-database fields. Analysis is descriptive and network-based using R (bibliometrix / Biblioshiny) and VOSviewer. Whole-corpus analyses (all sources, including LILACS/SciELO) cover publication trends, country-level output and its concentration (Lorenz curve, Gini coefficient), the most productive institutions and authors, co-authorship and country-collaboration networks, thematic mapping, journal analysis (core journals by article output via Bradford's law; regional-versus-international and open-versus-closed-access split) and database-overlap analysis. Citation-dependent analyses (most-cited documents and journals; impact indicators normalized as citations per year) are restricted to Scopus and Web of Science, which provide citation metadata. Conference abstracts and any grey literature are examined only as sensitivity analyses. Reporting follows PRISMA 2020 for search and selection and Donthu et al. (2021) for bibliometric methods. No bibliographic data have been extracted or analyzed at the time of registration. The study uses only published bibliographic metadata, involves no human participants, and does not require ethics approval. The dataset and analysis code will be deposited openly upon completion.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.