(71) Assessments of Potential Adverse Drug Reactions According to a Trigger Tool – A Descriptive Analysis

Sunday, August 28, 2022

8:00 AM – 1:45 PM CEST

Location: Hall C

Publication Number: 1116

Presenting Author(s)

Johan Lonnbro, PhD, MD
– Consultant physician, University of Gothenburg, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden

Co-Author(s)

Lina Holmqvist
– PhD/MD, Department of Medicine, Sahlgrenska Academy, University of Gothenburg, Sweden, Sweden
Elisabeth Persson
– MD, Department of Internal Medicine, Sahlgrenska University Hospital, Gothenburg, Sweden, Sweden
Per Thysell
– MD, Department of Internal Medicine, Sahlgrenska University Hospital, Gothenburg, Sweden, Sweden
N. David Åberg
– PhD/MD, Department of Medicine, Sahlgrenska Academy, University of Gothenburg, Sweden, Sweden
Susanna M. Wallerstedt
– University of Gothenburg, Department of Pharmacology, Sahlgrenska Academy, University of Gothenburg, G�teborg, V�stra G�talands l�n, Sweden

Background: Adverse drug reactions (ADRs) are often difficult to distinguish from other medical conditions, and a trigger tool, including adverse symptoms and drugs that could potentially be associated with such symptoms, may facilitate detection.

Objectives: To investigate assessments according to an ADR trigger tool regarding the prevalence of trigger symptoms and potentially associated drugs upon admission to hospital, as well as causality between trigger and drug, preventability, and relationship with admission.

Methods: We used data from a study including 30 randomly selected patients (50% female, median age: 72 years, median number of drugs: 7) admitted to a tertiary hospital in April 2018. Five physicians (three specialists in internal medicine and two residents) independently identified adverse symptoms and potentially associated drugs according to a published trigger tool. Identified trigger-drug pairs were then independently assessed regarding (i) causality according to the World Health Organization, (ii) preventability according to Hallas, and (iii) relation to admission.

Results: In all, 158 trigger-drug pairs were identified in 22 patients, 96 and 5 of which being identified by a single and all assessors, respectively. The most common triggers were fall/fracture (n=21) and confusion (n=17). A total of 30 (18%) trigger-drug pairs were assessed by at least one assessor to have a probable causal relationship, none being categorised by all assessors as at least probable. In all, 69 (42%) trigger-drug pairs were assessed as possibly preventable by at least one assessor, no pairs being concordantly assessed as preventable by all assessors. A total of 74 (47%) trigger-drug pairs were considered related to the admission according to at least one assessor; no pairs being concordantly assessed as related to the admission by all assessors. In all, two pairs were assessed as related to the admission by four assessors. These pairs concerned bleeding-rivaroxaban and neutropenia-fluorouracil.

Conclusions: With the majority of identified trigger-drug pairs detected by one assessor and not the others, and none of the trigger-drug pairs consistently assessed regarding causality, preventability and relation to admission, reliability seems to be an issue when using the trigger tool in a research setting.