Disproportionality Analysis and Characterisation of Medication Errors in EudraVigilance: Exploring Findings on Sexes and Age Groups

Abstract

Background: While medication errors (MEs) have been studied in the European Medicines Agency's EudraVigilance, extensive characterisation and signal detection based on sexes and age groups have not been attempted.

Objectives: The aim of this study was to characterise all ME-related individual case safety reports in EudraVigilance and explore notable signals of disproportionate reporting (SDRs) among sexes and age groups for the 30 most frequently reported drugs.

Methods: Individual case safety reports were used from EudraVigilance reported between 2002 and 2021. An ME was defined as any Preferred Term from the narrow Standardised Medical Dictionary for Regulatory Activities^® Query. Signals of disproportionate reporting were selected based on a lower boundary of the 95% confidence interval ≥ 1 of the reporting odds ratio, and at least 3 individual case safety reports. Analysed subgroups were female individuals, male individuals, and age groups 0-1 month, 2 months to 2 years, 3-11 years, 12-17 years, 18-64 years, 65-85 years, and >85 years. Heatmaps were utilised as a visual aid to identify striking SDRs.

Results: Of the 9,662,345 EudraVigilance reports, 267,262 (2.8%) contained at least one ME, with a total of 300,324 MEs, for 429,554 drugs. The most reported ME was "Inappropriate schedule of product administration" (52,646; 17.5%), followed by "Incorrect dose administered" (32,379; 10.8%) and "Wrong technique in product usage process" (26,831; 8.9%). Individual case safety reports with MEs were most frequently related to female individuals (148,009; 55.4%), most often submitted by healthcare professionals (155,711; 58.3%), originated predominantly from the USA (98,716; 36.9%), followed by France (26,678; 10.0%), and showed a median reported age of 50 years (interquartile range: 26-68). Most ME individual case safety reports (158,991; 59.5%) were associated with a serious health outcome. A total of 847 SDRs were identified, based on the entire EudraVigilance database; for subgroups, the number of SDRs ranged from 84 for the age group 0-1 month to 749 for female individuals. Signals of disproportionate reporting for female individuals and male individuals were very similar. Most MEs were reported for the vaccine against human papillomavirus (Anatomical Therapeutic Chemical [ATC]: J07BM01; 11,086 MEs, 57% being "inappropriate schedule of product administration"), with reporting odds ratios that range from 1.5 to 47.0 among age groups. The SDR for the live-attenuated vaccine against herpes zoster (ATC: J07BK02) had a reporting odds ratio that ranged from 26.6 to 78.1 among all subgroups. Signals of disproportionate reporting for oxycodone (ATC: N02AA05; 847 cases of "Accidental overdose", 35%), risperidone (ATC: N05AX08; 469 cases "Inappropriate schedule of product administration", 22.3%) and rivaroxaban (ATC: B01AF01; 1,377 cases of "Incorrect dose administered", 34.6%) stood out with higher magnitude SDRs for the age group 2 months to 2 years, with an reporting odds ratio range between 8.2 and 10.7, while for the entire EudraVigilance the reporting odds ratio ranged between 1.3 and 1.6 for the same drugs.

Conclusions: This exploratory research provides an overview of characterised ME individual case safety reports and SDRs from the EudraVigilance database. Most conspicuous SDRs were identified in specific age groups. Signals of disproportionate reporting, not described in the literature, were found for vaccines, oxycodone, rivaroxaban and risperidone, and may prompt further examination by stakeholders. Top-reported MEs ("Inappropriate schedule of product administration", "Incorrect dose administered" and "Wrong technique in product usage process") emerged as a general priority focus to perform a further root-cause analysis involving healthcare providers, manufacturers and regulatory bodies, to improve the understanding and prevention of MEs.

Fig. 1

Methodological overview on the selection of individual case safety reports (ICSRs) for analysis, including examples of analysis based on sexes and age groups. ATC Anatomical Therapeutic Chemical Classification

Fig. 2

Overview of all reports submitted to EudraVigilance (represented by the red bar chart, number of reports on the right y-axis) and the proportion of reports submitted mentioning a medication error (represented by the blue line chart, proportions noted on the left y-axis)

Fig. 3

Thirty most commonly reported drugs with their corresponding code from the Anatomical Therapeutic Chemical Classification System, associated with medication errors, ordered by the magnitude of disproportionality based on the values obtained when analysing the entire EudraVigilance database. Values of the reporting odds ratio are displayed where available for all the subgroups based on sexes and age groups. An empty cell signifies either no medication errors found or disproportionality criteria were not met. Interactive figure available in the ESM

Fig. 4

Fifteen most commonly reported medication errors highlighted and quantified for the 30 most commonly reported drugs associated with medication errors based on the entire EudraVigilance database. Values of the amount of reported cases are displayed where available. An empty cell signifies that the corresponding type of medication error was not reported for a drug. Interactive figure available in the ESM