Exploring the Impact of Patient Characteristics and Reporting Period on Adverse Drug Withdrawal Event Signals in the FAERS Database: A Stratified Disproportionality Analysis Study
Author(s)
Zakir Khan, PhD, Frank Moriarty, PhD.
School of Pharmacy and Biomolecular Sciences, RCSI University of Medicine and Health Sciences, Dublin, Ireland.
School of Pharmacy and Biomolecular Sciences, RCSI University of Medicine and Health Sciences, Dublin, Ireland.
OBJECTIVES: To compare how reporting signals for adverse drug withdrawal events (ADWEs) differ based on patient characteristics and over time.
METHODS: A case/non-case study with disproportionality analysis stratified by age (adult, elderly), sex (male, female), and reporting time (2014-2023 in two-year intervals) was conducted in the FDA Adverse Event Reporting System database. Three disproportionality estimators, proportional reporting rate (PRR≥2), reporting odds ratio (ROR>1) and information component (IC>0) with case count ≥3, were used to detect signals between ADWEs and drugs. The pvda package in R was used for disproportionality analysis, comparing top signals by age/sex groups and time period.
RESULTS: Of 14,940,210 reports, 105,026 included drug-event pairs related to ADWEs (reports with preferred terms containing “withdraw”). Higher proportions of these ADWE reports were in females than males (47.2% vs 43.3%) and in adults versus elderly (25.7% vs 5.9%). An increasing trend over time in ADWE reporting was observed, with the highest proportion in 2020-2021 (32.6%), followed by 2022-2023 (28.7%). Baclofen was associated with positive signals for ADWEs in both elderly (n=116; IC=5.02; PRR=38.38; ROR=41.39) and adult populations (n=1060; IC=4.88; PRR=31.14; ROR=37.40). Tiotixene showed a stronger signal in the elderly (n=7; IC=3.81; PRR=208.14; ROR=353.13) versus adults (n=3; IC=2.23; PRR=12.44; ROR=13.32). A sex-based difference was observed with betamethasone/fusidic acid showing a larger positive signal in females (n=12; IC=4.34; PRR=104.94; ROR=312.83) than males (n=3; IC=2.61; PRR=42.79; ROR=67.86). Whereas, hydrocodone had a stronger signal in males (n=1678; IC=5.26; PRR=40.39; ROR=60.81) versus females (n=1201; IC=5.13; PRR=36.63; ROR=47.11). Analysis over time showed that duloxetine and baclofen had strong signals peaking in 2016-2017 (n=4300; IC=6.61; PRR: 136.91; ROR: 321.91) and 2018-2019 (n=517; IC=5.14; PRR=37.88; ROR=44.58), respectively.
CONCLUSIONS: This study suggests reporting signals for ADWEs due to some medications have varied over time, and may differ based on age and sex.
FUNDING: Funded by the EU through the MSCA (Project acronym/number: HEAD-P/101149577)
METHODS: A case/non-case study with disproportionality analysis stratified by age (adult, elderly), sex (male, female), and reporting time (2014-2023 in two-year intervals) was conducted in the FDA Adverse Event Reporting System database. Three disproportionality estimators, proportional reporting rate (PRR≥2), reporting odds ratio (ROR>1) and information component (IC>0) with case count ≥3, were used to detect signals between ADWEs and drugs. The pvda package in R was used for disproportionality analysis, comparing top signals by age/sex groups and time period.
RESULTS: Of 14,940,210 reports, 105,026 included drug-event pairs related to ADWEs (reports with preferred terms containing “withdraw”). Higher proportions of these ADWE reports were in females than males (47.2% vs 43.3%) and in adults versus elderly (25.7% vs 5.9%). An increasing trend over time in ADWE reporting was observed, with the highest proportion in 2020-2021 (32.6%), followed by 2022-2023 (28.7%). Baclofen was associated with positive signals for ADWEs in both elderly (n=116; IC=5.02; PRR=38.38; ROR=41.39) and adult populations (n=1060; IC=4.88; PRR=31.14; ROR=37.40). Tiotixene showed a stronger signal in the elderly (n=7; IC=3.81; PRR=208.14; ROR=353.13) versus adults (n=3; IC=2.23; PRR=12.44; ROR=13.32). A sex-based difference was observed with betamethasone/fusidic acid showing a larger positive signal in females (n=12; IC=4.34; PRR=104.94; ROR=312.83) than males (n=3; IC=2.61; PRR=42.79; ROR=67.86). Whereas, hydrocodone had a stronger signal in males (n=1678; IC=5.26; PRR=40.39; ROR=60.81) versus females (n=1201; IC=5.13; PRR=36.63; ROR=47.11). Analysis over time showed that duloxetine and baclofen had strong signals peaking in 2016-2017 (n=4300; IC=6.61; PRR: 136.91; ROR: 321.91) and 2018-2019 (n=517; IC=5.14; PRR=37.88; ROR=44.58), respectively.
CONCLUSIONS: This study suggests reporting signals for ADWEs due to some medications have varied over time, and may differ based on age and sex.
FUNDING: Funded by the EU through the MSCA (Project acronym/number: HEAD-P/101149577)
Conference/Value in Health Info
2025-11, ISPOR Europe 2025, Glasgow, Scotland
Value in Health, Volume 28, Issue S2
Code
EPH108
Topic
Epidemiology & Public Health, Real World Data & Information Systems, Study Approaches
Topic Subcategory
Public Health, Safety & Pharmacoepidemiology
Disease
No Additional Disease & Conditions/Specialized Treatment Areas