A comprehensive cross-sectional investigation analyzing voluntary incident reports from the Dutch Institute for Rational Use of Medicine has revealed that ninety-four percent of reported unintentional medication discrepancies occurring during patient transitions between healthcare settings are attributable to human error. Published in the International Journal of Clinical Pharmacy by researcher Betsie Limmen and colleagues, the nationwide study evaluated nearly one thousand eligible transition-related medication incidents extracted from a broader registry of more than thirty-two thousand records collected over an eighteen-month surveillance window. Unintentional medication discrepancies, defined as unintended differences between the pharmaceutical regimen actively used by a patient and the official medication record maintained by a healthcare provider, represent a widely recognized threat to clinical safety across transitional pathways. By applying a standardized root-cause classification framework, the investigators identified that breakdown within human verification practices and procedural intervention protocols accounted for the vast majority of documented discrepancies, while technological malfunctions and broader institutional deficiencies represented only a minor fraction of the submitted reports.
Patient transfers across distinct healthcare environments have long been recognized as vulnerable touchpoints in medical continuity, where the accurate exchange of pharmacological documentation among clinicians frequently falters. Incomplete or distorted communication during these transfers can introduce discrepancies that potentially trigger adverse drug events, unexpected hospital readmissions, and elevated operational costs across healthcare systems. While international bodies, including the World Health Organization in its Medication Safety in Transitions of Care initiative, have stressed that unintended discrepancies affect a substantial majority of transitioning patients, healthcare facilities in the Netherlands have systematically mandated medication reconciliation processes since 2008. Dutch hospitals routinely execute reconciliation by cross-referencing patient interviews with dispensing data retrieved from the Nationwide Medication Record System, an integrated infrastructure that exchanges pharmacy dispensing histories following patient consent. Despite these established safeguards, published literature indicates that between seven and fifty-one percent of patients continue to encounter at least one unintended discrepancy upon hospital admission, with post-discharge studies demonstrating residual discrepancy rates ranging from five percent to seventy-eight percent.
To isolate the precise systemic and behavioral mechanisms responsible for these persistent discrepancies, the research team conducted a retrospective analysis using data supplied by the Dutch National Medication Incident Reporting database, administered by the Dutch Institute for Rational Use of Medicine. The central database serves as a national repository where healthcare institutions, individual practitioners, and primary care providers voluntarily log adverse events and near misses. The overwhelming majority of entries in this registry, comprising approximately ninety-eight percent of all submissions, originate from acute hospital environments, with the remaining two percent contributed by community pharmacies and general practice clinics. All records analyzed by the researchers underwent strict privacy protection, with institutional identifiers pseudonymized prior to delivery and individual patient data systematically scrubbed by database administrators to ensure that the study team operated exclusively with fully anonymized administrative materials in compliance with Dutch privacy legislation and ethics review exemptions.
The analytic cohort was drawn from a comprehensive dataset of 32,261 incidents recorded between January 1, 2022, and August 1, 2023. Researchers initially identified 1,933 incidents explicitly cataloged under the procedural category of medication information transfer, representing the initial operational stage where transitional communication failed. From this pool, ninety-five duplicate reports were eliminated alongside 846 incident records that failed to satisfy rigorous study inclusion criteria. Among these excluded records, 550 incidents did not involve an actual unintentional medication discrepancy, 186 reports lacked sufficient clinical and contextual documentation to reconstruct an underlying root cause, and sixty-eight incidents were driven strictly by interpersonal miscommunication between medical personnel and patients. Following this structured screening process, exactly 992 unique incidents, representing 3.1 percent of all submitted database reports from the study period, met the formal criteria for inclusion and formed the final analytic sample subjected to root-cause determination.
To categorize underlying causal pathways objectively, the authors employed the Prevention and Recovery Information System for Monitoring and Analysis model, designated as PRISMA-Medical. This established safety framework structures contributory factors across a hierarchical taxonomy that differentiates technical, organizational, human, and miscellaneous influences before dividing them into twenty specific operational subcategories through recursive root-cause inquiry. In the context of medication reconciliation, the study defined verification errors as failures arising directly during the active reconciliation procedure, such as transcribing an incorrect dose or overlooking a designated medication cessation date. In contrast, intervention errors were classified as procedural failures surrounding the reconciliation environment, including neglecting to query the Nationwide Medication Record System or relying on obsolete pharmaceutical histories. Working under the supervision of an experienced clinical pharmacist, two junior investigators independently evaluated the entire 992-incident dataset following an initial ten percent calibration sample, yielding an interrater agreement of 73.5 percent and a Cohen’s kappa of 0.62, representing moderate statistical concordance that was subsequently reconciled to full analytical consensus.
Evaluation of primary outcomes demonstrated an overwhelming concentration of causative factors within the domain of human behavior, which accounted for 932 incidents, or 94.0 percent of the evaluated sample. Verification errors emerged as the single most frequent causal driver, representing 492 cases or 49.6 percent of all included transitions, frequently manifesting as clinical oversights during manual chart verification or clerical mistakes such as selecting incorrect administration schedules within electronic health record interfaces. Intervention errors formed the second largest category, responsible for 246 incidents or 24.8 percent of cases, typically characterized by procedural non-compliance such as omitting reconciliation entirely or bypassing verification platforms. Coordination errors constituted the third prominent human factor, causing 164 incidents or 16.5 percent of cases, often stemming from parallel uncoordinated activities among pharmacy technicians and attending physicians or uncommunicated medication changes across handoffs. In contrast, organizational factors accounted for forty incidents, or 4.0 percent, primarily involving weekend or off-hour staffing shortages, while pure technical failures, such as software interoperability lapses and electronic record transmission errors, contributed only twenty incidents, representing 2.0 percent.
Secondary outcome evaluations revealed critical patterns across the clinical types of discrepancies and the pharmacological classifications most susceptible to transitional failure. Among 887 incident reports containing complete typological data, a total of 978 discrete discrepancy classifications were documented, with complete drug omissions forming the predominant failure type at 354 occurrences, representing 36.2 percent of identified discrepancies. Errors involving incorrect doses or dosing frequencies followed closely, accounting for 266 reports or 27.2 percent, while inappropriate medication additions represented 243 instances or 24.8 percent. Anatomical Therapeutic Chemical classification of the 1,360 implicated pharmacological agents across 850 informative reports showed that cardiovascular medications were the most frequently involved, encompassing 339 instances or 24.9 percent of all recorded drugs. Agents affecting blood and blood-forming organs represented the second most frequent category at 257 reports or 18.9 percent, followed by therapies targeting the alimentary tract and metabolic functions, which accounted for 245 cases or 18.0 percent.
Analysis of incident transmission paths showed that an alarming proportion of documented transitional errors reached the patient before detection. Among the 599 incident records that contained explicit follow-up information regarding clinical transmission, 444 cases, representing 74.1 percent, actively reached the patient, while only 155 discrepancies, or 25.9 percent, were intercepted prior to clinical contact. Evaluation of transitional timing across 951 reports with available temporal data indicated that the vast majority of reported breakdowns occurred at hospital admission, which accounted for 679 incidents or 71.4 percent of documented handoffs. Hospital discharge represented 229 cases or 24.1 percent, outpatient clinical encounters contributed twenty-three incidents or 2.4 percent, and twelve cases, or 1.3 percent, originated at admission but persisted undetected throughout the entire inpatient stay into post-discharge care. Among the subset of ninety-eight reports detailing specific external institutions, transfers between distinct hospital facilities and discharges directed to community pharmacies were the most frequently documented transitional pathways.
When juxtaposing these findings against broader patient safety literature, the authors observed substantial concordance with transition-focused intensive care research, while highlighting notable divergences from general medication error registries. The high rate of human performance breakdown mirrored transition studies conducted by Bosma and colleagues in intensive care environments, which similarly identified communication breakdowns and human execution as primary drivers in eighty-five percent of transitional failures, alongside comparable omission rates and high patient penetration. However, broader medication error studies conducted by Alqenae and other investigators often attribute eighty-two to ninety-seven percent of incidents to organizational failures. The authors explain that these contrasting distributions stem from varying classification methodologies, the inclusion of community pharmacy datasets, and the critical distinction between general medication errors and specific informational discrepancies. Furthermore, the predominance of cardiovascular medications was considered reflective of general population drug consumption in the Netherlands during 2022 rather than unique pharmacological vulnerability, while the high concentration of admission errors likely mirrors standardized admission reconciliation routines that maximize detection compared to under-resourced discharge workflows.
The investigators emphasized several important methodological considerations, noting that reliance on voluntary incident reporting systems introduces inherent underreporting and reporting bias, as minor, technical, or repetitive organizational issues are less consistently logged than overt human mistakes. Substantial missing data in descriptive variables, including a 39.6 percent absence rate regarding whether incidents reached patients, further constrained clinical outcome tracking, while initial categorizations executed by the reporting agency prior to academic transfer may have introduced potential misclassification. Because reported narratives contained insufficient clinical detail to measure actual physical harm or clinical severity, the authors refrained from assessing resultant patient injury. Concluding their assessment, the authors argued that future prevention efforts cannot rely solely on clinician training, but must focus on designing intuitive transfer interfaces, eliminating manual reconciliation inputs through automated data integration, encouraging patient participation in verification, and expanding reporting infrastructure within primary care environments to capture outpatient transitional risks comprehensively.
Subject of Research: Causes and characteristics of unintentional medication discrepancies across healthcare transitions
Article Title: Causes and types of voluntarily reported unintentional medication discrepancies in care transitions: a cross-sectional study
Article References: Limmen, B., van Broekhuizen, J., Essink, R., de Ruijter-van Dalem, J., van Eikenhorst, L., van den Bemt, P. M. L. A., & Karapinar-Çarkit, F. (2026). Causes and types of voluntarily reported unintentional medication discrepancies in care transitions: a cross-sectional study. International Journal of Clinical Pharmacy, 48(5), 2064-2072. https://doi.org/10.1007/s11096-026-02189-x
Image Credits: AI Generated
DOI: 10.1007/s11096-026-02189-x
Keywords: Care Transitions, Medication Discrepancies, Patient Safety, Medication Reconciliation, Human Factors, Verification Errors, Intervention Errors, Coordination Errors, Hospital Admission, Hospital Discharge, Incident Reporting, PRISMA-Medical, Cardiovascular Medications, Clinical Pharmacy, Health Systems Research
News Source: Ophelia Keating. (November 21, 2026). Human Errors Cause Most Unintentional Medication Discrepancies During Care Transitions. Scienmag.



