The impact of patient information on prescribing errors: Insights from pharmaceutical interventions.
AI interpretation is pending for this paper.
Open original publication →What the AI sees
Not AI summarized yet.
Research significance
Pending deeper interpretation.
Source abstract
BACKGROUND: Medication errors are more likely to occur in patients with complex conditions, where appropriate prescribing requires accurate and comprehensive patient information. Inadequate use of such information, such as overlooking laboratory results or patient weight, can lead to dosing errors or contraindicated prescriptions, even with electronic checking systems. This study aimed to analyze prescribing errors detected through pharmaceutical interventions, focusing on the patient information in the hospital information system. METHODS: A retrospective analysis was conducted on 9774 pharmaceutical interventions performed between April 2019 and March 2020. Of these, 3372 interventions related to patient information stored in hospital information system were included. Prescribing errors were categorized into five patient-specific information categories: allergy information, laboratory data, concomitant drugs, patient weight, and patient status and history. Demographic and prescription data were analyzed, and a heatmap was developed to visualize high-risk areas. RESULTS: Among the included interventions, 1352 (40.1 %) prescriptions involved corrections of prescribing errors, with laboratory data being the most frequently utilized patient information source (n = 2526). Error rates were higher in weekend settings compared to weekday settings (56.2 % vs. 38.3 %, P < 0.001, Cramér's V = 0.111), and prescriptions for patients younger than 20 years exhibited the highest error rates among age groups (66.9 %; P < 0.001). Error rates were significantly varied by drug type (P < 0.001, Cramér's V = 0.395). Among these, digestive drugs showed the highest error rates (69.1 %), particularly those requiring renal dosing adjustments. Conversely, anti-tumor agents demonstrated a lower error rate (15.9 %) due to stringent monitoring processes. The high-risk heatmap highlighted specific risks, such as weight data for younger patients and patient status for anti-inflammatory drugs. CONCLUSION: Laboratory data were the most frequently used information source, to prevent prescribing errors. The risk heatmap demonstrated weekends, pediatric patients, and renal dosing as high-risk areas. These findings suggest that future information systems should enhance the utility of laboratory data and incorporate tailored alerting strategies focused on high-risk patient conditions and clinical settings, such as real-time lab data alerts or weight-based dosing calculators, and potentially explore the use of AI for proactive error prevention.