Unintended medication discrepancies across key stages of the in-hospital medication process: a retrospective real-world study in hospitalized patients

Hermann, Christian; Amann, Ute; Rüther, Tobias; Kääb, Stefan; Nadal, Jennifer; Rösch, Caroline; Baierl, Sofie; Steinbrech, Julian; Günther, Maximilian; Strobach, Dorothea (2026): Unintended medication discrepancies across key stages of the in-hospital medication process: a retrospective real-world study in hospitalized patients. BMC Health Services Research, 26: 552. ISSN 1472-6963

Abstract

Background

The in-hospital medication process is vulnerable to medication errors, especially during care transitions, when medication discrepancies occur most frequently. However, previous studies focused only on medication errors at single points in this process. This study therefore investigated key stages of the medication process. Also, we examined potential risk factors for unintended medication discrepancies (UMDs) at admission.

Methods

A single-center retrospective study was conducted on seven reference wards from different medical specialties. Technical UMDs at admission, intra-hospital transfer, dispensing, and discharge were evaluated and categorized for patients hospitalized on two predefined days in November 2023. Potential risk factors for UMDs at admission were analyzed using descriptive statistics and logistic regression.

Results

During the study period, 449 UMDs were identified in 215 patients (Median 1 per patient, range 0–15). Of these, 150 (33.4%) were classified in their potential clinical relevance as low severity, 157 (35.0%) as medium, and 142 (31.6%) as severe. The majority of UMDs (206; 45.9%) occurred at admission, of which 104 (50.5%) were still present in discharge letters. Changes in clinical parameters (blood pressure, oxygen saturation, hyper/hypoglycemia) were associated with 37 (18.0%) UMDs for 27 patients at admission. During intra-hospital transfers, 49 UMDs (10.9%) were detected, 4 between general wards and 45 between general and intensive care units. A total of 90 UMDs (20.0%) were identified during dispensing and 104 (23.2%) at discharge.The multivariable logistic regression showed that patients with ≥10 drugs had an increased risk of UMDs at admission (OR 4.0; 95% CI 1.767–9.277; p < 0.004). Medication histories taken by pharmacists were associated with reduced UMDs at admission (OR 0.13; 95% CI 0.042–0.414; p < 0.001) compared to ward staff.

Discussion

Our findings demonstrate that UMDs are frequent across the hospital pathway, with admission as the most vulnerable point. Taking ≥ 10 drugs was a risk factor, while pharmacist-led medication histories reduced the occurrence of UMDs, highlighting the importance of structured reconciliation. By covering admission, transfers, dispensing, and discharge in a cohort, this study provides a comprehensive perspective on medication process, emphasizing the need for targeted interventions during hospitalization and discharge.

• Dokumententyp: Artikel (Klinikum der LMU)
• Organisationseinheit (Fakultäten): 07 Medizin > Klinikum der LMU München
• DFG-Fachsystematik der Wissenschaftsbereiche: Lebenswissenschaften
• DFG: Gefördert durch die Deutsche Forschungsgemeinschaft (DFG) - 491502892