PhD Scientific Days 2026

Poster Session 3.W - Pharmaceutical Sciences and Health Technologies

CAVE: Real-World Safety Signal Reference Standard for Computational Methods in Pharmacovigilance

Előadó neve

Mr. Csernák, Áron Márk

Neptune code

B0M0PS

Előadó munkahelye

Department of Pharmacology and Pharmacotherapy, Semmelweis University

Előadó telefonszáma

+36301860435

Előadó e-mail címe

csernak.aron@stud.semmelweis.hu

Az előadás címe

CAVE: Real-World Safety Signal Reference Standard for Computational Methods in Pharmacovigilance

Szerző(k) neve és munkahelye

Áron Márk Csernák^1,3, Mátyás Pétervári^1,2,3, Andrea Simon^1,3, Bernadett Bicskei², Olivér Márton Balogh^1,3, Eszter Puhl^1,3, Péter Ferdinandy^1,3,4, Bence Károly Ágg^1,3,4

1: Department of Pharmacology and Pharmacotherapy, Semmelweis University
2: Sanovigado Kft
3: Center for Pharmacology and Drug Research & Development, Semmelweis University
4: Pharmahungary Group

Bemutatás módja

Poszter

Szekció

Poster Session 3.W - Pharmaceutical Sciences and Health Technologies

Language of the presentation

English

Preferred session

Pharmaceutical Sciences and Health Technologies

Összefoglaló szövege

Introduction: Pharmacovigilance is hindered by reference standards that are outdated, limited by size, scope, or lack reliable negative controls. Safety signals have been made publicly available by the Pharmacovigilance Risk Assessment Committee (PRAC) since its establishment in 2012. These offer a reliable resource to establish a benchmark for computational methods in the most challenging cases of causality assessment of adverse drug reactions.
Aims: We aimed to process regulatory documents of investigated safety signals to establish a novel reference standard consisting of drug–event combinations for benchmarking computational methods in pharmacovigilance.
Methods: PRAC meeting minutes from the European Medicines Agency were reviewed by medical doctors and pharmacovigilance experts for regulatory decisions made during causality assessments. In ambiguous cases, the summary of product characteristics was reviewed as well. Individual drug–event combinations were assembled into a standardized dataset and compared with other six established reference standards from literature (Coloma et al., 2013; LePendu et al., 2013; Ryan et al., 2013; Harpaz et al., 2014; Osokogu et al., 2015; Ito & Narukawa, 2024). Reporting Odds Ratio (ROR), the industry-standard signal detection method, was calculated for all seven datasets. ROR performance was evaluated with specificity, accuracy, and area under the receiver operating characteristic curve (AUROC).
Results: Our reference standard (CAVE) includes 1,615 drug–event combinations (983 positives, 632 negatives), 611 different drugs and 456 different adverse events. CAVE posed a more difficult challenge to ROR compared to six other established reference standards (specificity: 0.63 vs 0.75–0.88, accuracy: 0.47 vs 0.58–0.76, AUROC: 0.49 vs 0.63–0.79).
Conclusions: By processing regulatory documents we established CAVE, the first large-scale open-access reference standard for causality assessment with both positive and negative controls based on real-world regulatory decisions. CAVE is ready to facilitate the safe use of drugs by contributing to the development of more reliable computational methods in pharmacovigilance.

University

Semmelweis University

Supervisor

Olivér Márton Balogh, Bence Károly Ágg

Publication of my abstract

I do not give consent to the publication of my abstract on the website of the congress.

phd.section.field

before finishing undergraduate studies (TDK, MD-PhD)

Kind

Szabad

Status

elfogadva

Accepted presentation method

poszter

Előadás fájl jóváhagyás

nem rendelkezett róla

Előadó

9055

Start

14:30

End

14:33