EUVIMEDSkopri EUVIMED
Open access · CC BY 4.0

Semantic Similarity‐Informed Bayesian Borrowing for Quantitative Signal Detection of Adverse Events

Semantic Similarity‐Informierte Bayes’sche Borrowing-Strategie zur quantitativen Signalentdeckung von unerwünschten Ereignissen

François Haguinet iD, Jeffery L Painter iD, Gregory E Powell, Andrea Callegaro iD, Andrew Bate iD

Statistics in Medicine · 2026-10 · 45 (23-24)

DOI: 10.1002/sim.70757 ↗

Auf einen Blick

Die Studie beschreibt eine Bayes’sche Borrowing-Methode zur Signalentdeckung in Spontanmeldesystemen. IC SSM erkannte echte Signale früher als die traditionelle IC-Analyse.

Fragestellung

Verbessert semantisch ähnlichkeitbasiertes Borrowing die quantitative Signalentdeckung unerwünschter Ereignisse im Vergleich zur traditionellen IC-Analyse und zu IC HLGT?

Methode

Bayes’scher dynamischer Borrowing-Ansatz mit robuster meta-analytisch-prädiktiver Prior und semantischen Similarity Measures; Vergleich mit IC und IC HLGT anhand von FAERS-Daten und einer zeitgestempelten Referenzmenge.

Ergebnisse

IC SSM erkannte echte Signale fünf Monate früher als traditionelle IC. Bei 2,5 % Posterior-Quantil: Sensitivität 57,0 % und Präzision 14,5 %; traditionelle IC 50,1 % und 16,1 %, IC HLGT 55,7 % und 13,9 %.

Schlussfolgerung der Autor:innen

IC SSM zeigte bessere Timeliness und Sensitivität als die Standardverfahren; dies spricht laut den Autor:innen für einen skalierbaren, kontextbewussten Ausbau der traditionellen disproportionality analysis.

Kernaussagen

  • Echte Signale fünf Monate früher erkannt
  • Höhere Sensitivität als traditionelle IC-Analyse
  • Niedrigere Präzision als traditionelle IC-Analyse

Deutsche Zusammenfassung: KI-gestützt aus dem Original-Abstract erstellt; alle Zahlen wurden automatisch mit dem Original abgeglichen. Maßgeblich ist der Original-Abstract.

At a glance

Journal article in Statistics in Medicine, published online on 2026-10-15. Written by François Haguinet and 4 co-authors. Institution: GSK, Wavre, Belgium. Major MEDLINE topics: Adverse Drug Reaction Reporting Systems, Semantics, Drug-Related Side Effects and Adverse Reactions. 63 references. The full text is freely available (open access, CC BY 4.0).

Auf einen BlickAstrattSubjectsDettalji biblijografiċiCiteRelated studies

Abstract (original language)

We present a Bayesian dynamic borrowing approach to enhance quantitative identification of adverse events in spontaneous reporting systems. The method embeds a robust meta‐analytic predictive prior and incorporates semantic similarity measures (SSMs) to enable weighted information sharing from clinically similar MedDRA preferred terms (PTs) to the target PT. This continuous similarity‐based borrowing overcomes limitations of rigid hierarchical grouping in disproportionality analysis (DPA). Using data from the FDA Adverse Event Reporting System and a time‐stamped reference set, we evaluate our approach—termed IC SSM—against the current standard approach of “traditional” information component (IC) analysis and borrowing at the MedDRA HLGT level (IC HLGT). The primary IC SSM approach at the standard threshold (2.5% posterior quantile ) detected true signals over five months earlier than traditional IC. It showed higher sensitivity (57.0%) but lower precision (14.5%) than traditional IC (50.1%, 16.1%), while outperforming IC HLGT (55.7%, 13.9%). In subsequent analyses, even when tightening the alerting quantile of IC SSM to 1% to improve precision (16.4%), superior sensitivity (51.4%) was retained compared to traditional IC. Similarly, relaxing traditional IC's quantile to 5% yielded only 55.6% sensitivity with 14.2% precision, failing to match IC SSM's performance or timeliness and confirming its inherent advantage. IC SSM's superior timeliness and sensitivity stem from integrating clinically meaningful information, making it a scalable and context‐aware enhancement to traditional DPA. The method has potential for exploratory analysis in sparse datasets, and refinements to the borrowing strategies and similarity metrics could yield further performance improvements.

Subjects

MeSH terms (MEDLINE)

Adverse Drug Reaction Reporting SystemsSemanticsDrug-Related Side Effects and Adverse ReactionsBayes TheoremHumansUnited StatesUnited States Food and Drug AdministrationSensitivity and Specificity

Author keywords

Bayesian borrowingdrug safetypharmacovigilancesemantic similaritysignal detection

Authors and institutions

  • GSK, Wavre, Belgium.
  • GSK, Durham, North Carolina, USA.
  • GSK, London, UK.
  • London School of Hygiene and Tropical Medicine, London, UK.

ORCID profiles: François Haguinet, Jeffery L Painter, Andrea Callegaro, Andrew Bate

Cite

APA

Haguinet, F., Painter, J. L., Powell, G. E., Callegaro, A., & Bate, A. (2026). Semantic Similarity‐Informed Bayesian Borrowing for Quantitative Signal Detection of Adverse Events. Statistics in Medicine, 45(23-24). https://doi.org/10.1002/sim.70757

Vancouver

Haguinet F, Painter JL, Powell GE, Callegaro A, Bate A. Semantic Similarity‐Informed Bayesian Borrowing for Quantitative Signal Detection of Adverse Events. Statistics in Medicine. 2026;45(23-24). doi:10.1002/sim.70757

BibTeX
@article{Haguinet2026,
  author = {Haguinet, François and Painter, Jeffery L and Powell, Gregory E and Callegaro, Andrea and Bate, Andrew},
  title = {Semantic Similarity‐Informed Bayesian Borrowing for Quantitative Signal Detection of Adverse Events},
  journal = {Statistics in Medicine},
  year = {2026},
  volume = {45},
  number = {23-24},
  doi = {10.1002/sim.70757}
}
Start for free →