Initial assessment of a newly implemented medication administration dashboard at St. Bartholomew Hospital in London, England

Abstract

Introduction: Over 237 million medication errors are made within the medication process every year in England. While the majority represent little to no harm, they are still preventable causes of morbidity and mortality, thus remaining a focus area for quality improvement. St. Bartholomew's Hospital has targeted missed dose dispensing as a key area in their ongoing effort to improve institutional safety. This project's purpose was to evaluate the utility of a newly available, real-time medication dispensing dashboard (Qliksense), to identify quality improvement initiatives focused on missed doses for the institution. Method(s): Data for December 2024 from Qliksense, a program interface with the CernerTM electronic medical record (EPMA) at St Bartholomew's Hospital, was uploaded to a de-identified ExcelTM dashboard. All missed doses from this one month period were captured for analysis. A missed dose was defined as a medication 'not given', 'not done', or given outside of a two hour window. Data cleaning was performed to omit blank entries and doses documented as 'as needed'. The data were crossmatched with medications available from the emergency drug reserve in each ward and the critical need list. Analysis of the following was performed: Hospital wards with the highest rate of missed doses, common medications associated with missed doses and missed dose administration trends over time. A review of potential contributing factors (e.g., ward-specific workflows, medication types, administration timing) was also performed through institutional meetings with providers associated with governance and medication safety. Coded reasons for a missed dose were also analysed. Result(s): A total of 45,808 rows of data were exported to ExcelTM for analysis, representing less than 10% of total doses prescribed during the timeframe. Seventy percent of all medications administered did not meet the definition of a missed dose. After data were cleaned, 13,711 entries were analysed. The majority (99%) of all missed doses had a documented reason. Paracetamol had the highest number of orders marked as "not given" and "not done," followed by nutritional supplements and docusate. There was a low frequency of missed doses for 'critical need' medications. Missed doses peaked at 8 AM, followed by 6 PM and 10 PM which aligns with administration times. Difficult to interpret reporting of reasons for missed doses prevented a better understanding of these descriptors. Limitations to the dashboard assessment included missing date and time values, discrepancies between overall and ward-level assessments, the size of the database, and the descriptive only nature of the data evaluation. The identified quality improvement opportunities with this dashboard include: Expanding analysis of monthly data to validate findings, observation of EPMA documentation to identify necessary changes, alignment with Datix reports for missed doses, proposing strategies to enhance real-time medication administration monitoring use of the dashboard for ward-based evaluation, and implementation of visibility documents to promote interdisciplinary education and engagement. Conclusion(s): This initial analysis of the utility of the dashboard within the hospital was critical in identify data-driven strategic ways to identify the focus areas for quality improvement.