Consistency of Medical Subject Headings assignment: A test-retest reliability analysis
Bibliographic record
Abstract
INTRODUCTION: Medical Subject Headings (MeSH) are the controlled vocabulary used by the National Library of Medicine (NLM) to index articles covered by MEDLINE. OBJECTIVE: Evaluate the consistency of MeSH assignment using a test-retest analysis of articles published multiple times. METHODS: Three sets of articles that had been published multiple times were selected: Vancouver Group articles, CONSORT Statement articles, and Granada Statement articles. The articles publishing these position papers were searched in PubMed in February 2025, and their records were exported in XML format. The articles' metadata, the assigned MeSH terms, and the indexing methods were extracted. Consistency was assessed using Fleiss' kappa for inter-rater agreement and Krippendorff's alpha for classification reliability, considering each article as a different rater. RESULTS: A total of 6, 8, and 5 articles indexed in MEDLINE were retrieved that had published articles with Vancouver, CONSORT, and Granada statements, with 14, 6, and 10 different MeSH terms assigned, respectively. The first two sets of articles were manually indexed, while the Granada articles were automatically indexed. Fleiss' kappa for the MeSH terms assigned to the Vancouver, CONSORT, and Granada articles were -0.390, -0.370, and -0.333, respectively, and Krippendorff's alphas were 0.178, 0.525, and 0.183, respectively. "Periodicals as Topic" and "Randomized Controlled Trials as Topic" were used in all Vancouver and CONSORT articles, respectively. Except for "Humans," no other MeSH terms appeared in all Granada articles. The most prevalent terms were "Pharmacy" and "Pharmacies" and "Pharmacy Research." Geographic MeSH terms were assigned to the Vancouver and Granada articles. CONCLUSION: A highly inconsistent MeSH indexing pattern was found across the three sets of articles. Automated indexing of the Granada Statements articles did not improve the results.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.002 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".