Measuring the impact of automated dispensing cabinets implementation on data and inventory management of human normal immunoglobulin in an acute teaching hospital: A pre- and post- intervention study

Background

Human Normal Immunoglobulin (HNIg) is a complex plasma-derived blood product used to treat a variety of medical conditions. Global supply problems have increased focus on HNIg stewardship, including mandatory recording of HNIg usage on the National Immunoglobulin Database (NIgD). Local departmental audits identified significant inconsistencies in data uploaded to NIgD. Inventory management issues caused a number of stock losses in the last financial year. A paper-based HNIg batch number recording system was replaced with the use of Automated Dispensing Cabinets (ADCs), which generated an electronic batch number report, with matching fields required for electronic upload to NIgD.

Aim

To measure the impact of ADC implementation on the accuracy of data uploaded to NIgD, HNIg stock control and staff time associated with processes of HNIg data and inventory management.

Method

Three months of pre- and post-implementation HNIg dispensing data was compared to the data uploaded to the NIgD for discrepancies. Inventory reports were used to identify unexplained stock adjustments. Time and motion methods were used to quantify staff time associated with HNIg activities.

Results

Pre-implementation: 20.7 % (3762.5 g/18,217 g) of HNIg by volume (23.7 % of dispensing episodes; 66/279) were inaccurately uploaded or absent on the NIgD and three stock adjustments were made (loss of >£15 k). Post-implementation: 12 % (2325 g/19,347 g) of HNIg by volume (10.8 % of dispensing episodes; 31/286) were inaccurately uploaded or absent on the NIgD and zero stock adjustments were observed; Mean time for dispensing per HNIg prescription reduced from 17 min to 8 min; Time spent uploading data to NIgD per month reduced from 5 h to 1.75 h.

Discussion/Conclusion

The use of ADCs improved accuracy of NIgD data upload. Following implementation direct observations have cited unregistered or finished patient episodes, dispensing procedural compliance, and user familiarity with the system as common reasons for incomplete data uploads to NIgD. The new ADC process had consequences of forcing dispensing procedure compliance to improve uploads, whilst reducing dispensing times. It led to improved stock control and removed upload burden from dispensers. ADC stock discrepancy alerts allowed staff to proactively resolve discrepancies in real time. Time taken for additional processes to support management of stocks in ADCs, including monthly stock cycle counts, were important considerations for implementation. The key advantage of using ADC batch number reports is the ability to upload as a single monthly batch without having to manually access individual patient records on the NIgD. It facilitates early identification of failed upload attempts and supports resolution of stock discrepancies.