Global map of biomedical journals reveals stark regional gaps in metadata, fees and impact

A sweeping cross-sectional analysis of nearly 4,800 MEDLINE-indexed journals has uncovered deep geographic stratification in the infrastructure of biomedical publishing, and a surprising twist: much of that regional divide appears to be driven not by geography itself, but by a small number of large publishers. The study, published in Discover Informatics by Abdullah Umer of Aga Khan University, Karachi, simultaneously examined three dimensions that are usually studied in isolation — metadata completeness, article processing charges (APCs) and journal impact — and mapped them across eight world regions.

The research drew on four publicly available data sources, queried programmatically in a single session on April 1, 2026, to ensure temporal consistency. Journal identification came from PubMed/MEDLINE, metadata records from the Crossref REST API, impact values from the SCImago Journal Rank (SJR) 2024 dataset, and pricing information from the Directory of Open Access Journals (DOAJ). Of 92,542 unique journals registered with Crossref, 20,333 were PubMed-indexed and 4,844 carried the more selective MEDLINE designation. After excluding 72 journals lacking geographic data, 4,772 journals formed the final analytical sample, each identified by its International Standard Serial Number.

Geographic classification was based on publisher headquarters, a choice with important implications. Western Europe accounted for 2,362 journals (49.5 percent) and Northern America for 1,896 (39.7 percent), meaning these two regions together hosted 89.2 percent of the sample. The remainder was spread thinly: 202 journals in the Asiatic Region, 95 in Eastern Europe, 88 in the Middle East, 75 in Latin America, 36 in the Pacific Region and just 18 in Africa. The author notes that this reflects the distribution of publishing infrastructure rather than the origin of research content or authorship, and that MEDLINE’s requirement that journals publish or provide abstracts in English contributes to this Anglocentric concentration.

Metadata completeness was scored on a composite scale from 0 to 12, built from eight indicators. Four core elements — abstracts, author affiliations, ORCID identifiers and funder information — were assessed as continuous coverage percentages and transformed into categorical scores, while four repository deposit elements — reference lists, license information, update policies and resource links — were scored as present or absent. The overall median score was 8.0, but regional differences were statistically significant (Kruskal-Wallis χ² = 231.42, p < 0.001). Africa and Eastern Europe consistently sat at the bottom, while the Middle East, somewhat counterintuitively, posted the highest median composite score, driven by high-performing journals in a bimodal distribution.

One finding stands out as near-universal: funder information was the most consistently deficient metadata element in every region, ranging from zero coverage in Africa to a maximum of just 40 percent in Western Europe. This matters because funders increasingly mandate institutional and financial tracking of research outputs. There is a hint of improvement on the horizon — Crossref announced reduced fees for registering grant information on ongoing works and complete waivers for historical records starting January 2026 — but the current picture is one of systemic underreporting. By contrast, references and licenses showed the highest and most consistent coverage, reaching 97 and 95 percent respectively in Western Europe.

Pricing data revealed a different kind of divide. APC information was available for 2,950 journals (61.8 percent of the sample), with an overall median fee of €2,470. Where data existed, Northern America (median €2,535) and Western Europe (€2,451) charged the highest fees, while Africa (€369) and Eastern Europe (€676) charged the lowest. But availability itself was highly uneven: fewer than a quarter of journals in Africa, Eastern Europe, the Asiatic Region and the Middle East reported an APC, compared with 73.6 percent in Western Europe. The author cautions that this missingness limits the comparability of regional figures, and an inverse-probability-weighted sensitivity analysis was used to probe whether selective reporting was distorting the results.

That sensitivity analysis produced one of the study’s most instructive results. In the unweighted complete-case model, metadata completeness appeared to be a significant negative predictor of APC — a counterintuitive suggestion that better-documented journals charge less. Under inverse-probability weighting, that association vanished entirely (β = 0.0007, p = 0.87), indicating it likely reflected which journals bother to report fees at all rather than a genuine pricing relationship. The regional pricing pattern, however, survived weighting largely intact, and APCs remained moderately positively correlated with SJR (ρ = 0.289), consistent with higher-impact journals commanding higher fees.

Journal impact showed the strongest regional stratification of all (χ² = 477.46, p < 0.001). Northern America, Western Europe and the Pacific Region formed a statistically indistinguishable high-impact cluster, each differing significantly from every other region, while Africa, Eastern Europe, Latin America and the Middle East formed a lower cluster. Distributions were right-skewed almost everywhere, suggesting a small number of high-impact journals within each region. In a multivariable model of log-transformed SJR, metadata completeness remained a significant positive predictor even after adjusting for region and APC, though the study’s correlational design cannot establish whether richer metadata causes higher impact.

The most consequential analytical move came when the author formally accounted for the fact that journals are nested within publishers. A multilevel model partitioning variance found that publisher identity alone explained roughly 87 percent of the variance in metadata completeness (ICC = 0.868) and about 45 percent of the variance in log-transformed SJR. Once publisher clustering was modeled, none of the regional coefficients for metadata completeness remained statistically significant. The pattern was stark in the raw data too: major publishers including Elsevier (987 journals), Springer-Verlag (455), Taylor & Francis (337), Springer BioMed Central (147) and S. Karger (62) all showed near-complete deposition of references and licenses (≥98 percent) alongside near-zero deposition of abstracts and/or affiliations. Because these deficits are concentrated among specific well-resourced publishers, resource constraints alone seem an unlikely full explanation, though the author flags this as a hypothesis that cannot be tested from Crossref deposition records alone.

The implications cut in two directions. For metadata, interventions aimed at a handful of large publishers could meaningfully close the global gap, since the regional pattern is substantially a publisher-level phenomenon that happens to cluster geographically. For pricing, the story is different: regional APC differences persisted even after accounting for publisher, suggesting that fee disparities would require region-targeted policy such as expanded waiver programs rather than publisher-targeted reform alone. The study has clear limitations — an ecological design prone to aggregation effects, headquarters location as an imperfect proxy for publishing activity, no disciplinary classification available as a covariate, and exclusion of journals not registered with Crossref, which may be concentrated in lower-resourced regions and could underestimate true disparities. Still, the overall picture is of a stratified biomedical publishing ecosystem in which discoverability, prestige and cost are entangled, and in which a relatively small number of corporate actors hold disproportionate leverage over whether the world’s medical literature can actually be found, linked and trusted.

Subject of Research: Geographic variation in metadata completeness, article processing charges and journal impact across MEDLINE-indexed biomedical journals

Article Title: Mapping geographic variation in metadata completeness, article processing charges and journal impact across MEDLINE journals and the influence of publisher clustering

Article References: Mapping geographic variation in metadata completeness, article processing charges and journal impact across MEDLINE journals and the influence of publisher clustering. (n.d.). https://doi.org/10.1007/s44564-026-00027-w

Image Credits: AI Generated

DOI: 10.1007/s44564-026-00027-w

Keywords: MEDLINE, metadata completeness, article processing charges, journal impact, Crossref, SCImago Journal Rank, open access publishing, publisher clustering, geographic inequality, biomedical publishing, DOAJ, scholarly communication