2025 Southeastern Residency Conference

Title: The Implementation of a Passive Bin-Based Perpetual Medication Inventory Model within Ambulatory Clinics at an Academic Medical Center


Authors: Elly Glazier, Jared Gabbert, Grayson Peek


Background: Ambulatory clinics store and dispense high-value medications with little visibility into inventory quantity or product movement. Automated dispensing cabinets are frequently used within inpatient settings to combat this issue but are a large capital investment for the clinic setting. In current state, pharmacies and clinics follow complex workflows to manage the inventory of high-value medications to meet patient demand while minimizing drug waste. This has led  to unrealized opportunities in inventory optimization due to large information gaps. We assessed a passive bin-based inventory model that leverages a system that records every transaction by a clinic team member. The system records these transactions in real time via light sensors and can both register the removal and replacement of medication products. The system then uses artificial intelligence and various algorithms to recommend inventory optimizations via the assessment of the recorded medication transactions and notably requires no electronic health record integration.


Methods: This study was conducted over 10 weeks and included select medications at two ambulatory locations. We assessed if the utilization of a passive bin-based inventory model within ambulatory clinics allowed for a decrease in inventory on-hand valuation. Inventory valuation was assessed prior to implementation and following system recommendations at the conclusion of the pilot. Accuracy of the system was validated via twice weekly manual cycle counts. The primary outcome was the valuation of the inventory on hand change recommended by the system at the conclusion of the evaluation. Secondary outcomes included the accuracy of the system as well as the outcome of a comparison of pre- and post- implementation satisfaction surveys (modified MAS-NAS) offered to nursing staff, as well as the cost of facility modifications required for implementation.


Results: The passive bin-based inventory recorded 3,454 dispenses during the study period. The average days on hand tabulated ranged from 1.5 days to over 30 days. The total inventory valuation decrease across both clinics included 69 product units for a total of $34,000 of average wholesale price. The accuracy of the model was established to be 84.6% at the neurology site and 96.3% at the orthopedics site. Though not extrapolatable due to volume, the results of the modified MAS-NAS nursing satisfaction survey showed a negative change from pre- to post- survey but notably were generally positive regarding the post implementation passive bin-based technology questions.

Conclusions: In this retrospective pre and post implementation study, the utilization of passive bin-based perpetual medication inventory models reduced inventory on hand valuation but was not statistically significant. Additionally, the model offers the opportunity to greatly increase inventory visibility in a difficult to manage care setting. Additional studies that assess benefits of visibility with powered results should be considered as the implementation of this model improves.