2025 Southeastern Residency Conference

A PROFILE OF DIRECT PURCHASING
Lauren Flick, Abbi Rowe

Background: For many health systems, almost the entirety of the drug inventory is purchased “indirectly” through a wholesale partner. This traditional model allows for streamlined access to a variety of products, provides an easy-to-use and regulatorily compliant ordering process, offers quick and consistent logistics, and is able to facilitate smaller, single unit of sale purchases. However, for health systems with the appropriate infrastructure, establishing direct purchasing relationships can be beneficial. AdventHealth’s Central Florida Division includes a Central Fill Pharmacy (CFP) that supplies drugs daily via an internal courier service to 18 acute care facilities within the division. CFP is located at an offsite warehouse space with roughly 15,000 square feet of medication storage space, allowing for pallet storage of bulk medication purchases.  This system provides advantages based on the volume of medications purchased, as well as storage solutions for larger orders, which can be leveraged when optimizing cost savings though alternate purchasing streams.

Purpose: Evaluate historic direct medication purchasing trends for process optimization opportunities.

Methods: Direct purchasing data from the CFP facility was compiled using historic invoices and logged in an Excel spreadsheet. All direct commercial medication purchases made from 3 September 2021 to December 2024 by CFP were included. Supplies and 503B purchases were excluded. The invoice date, medication description, NDC, quantity purchased, average blended price available at the wholesaler at the time of purchase, and direct price [LF1] were all recorded. Descriptive analysis of the data, including counts of unique molecules and NDCs, averages of cost savings and quantities purchased, and various pivots of the historic data was performed using Excel. Data visualization charts were created to display the findings and begin development of a comprehensive dashboard for ongoing tracking and optimization of direct pharmacy purchasing.

Results: A total of 757 historic transactions were included in the analysis, with the highest number (n = 353) occurring in 2023.  While the median amount spent per order fluctuated between years ($8,462 in 2021, $9,600 in 2022, $7,150 in 2023, and $7,725 in 2024) the median savings per order decreased each year; from $5,600 in 2021 to $1,944 in 2024.  This suggests a willingness to complete purchases with smaller cost reductions associated with them, which could be related to familiarity and comfort with the process reducing the time investment required.  The largest percent of savings from a single drug came from dexmedetomidine, which accounted for 12.43% of total savings. Heparin, which had the highest number of total units purchased at 413,000 units, accounted for 5.88% of savings, and enoxaparin, which had the most purchases (158) and NDCs (21) accounted for 5.25%.  The drug class with the highest portion of savings was antimicrobials, with 35.32% of overall spendings, followed by cardiovascular drugs, which had the highest number of buys (261), NDCs (55), and individual units (973,006) and accounted for 28.49% of savings.  For analysis of purchasing source, Medigi was the highest in number of buys (399), maximum saving percent (86.5%), NDCs (133), and total units purchased (1,631,072), and accounted for 60.70% of total savings.

Conclusions: Direct purchasing is capable of sufficient procurement cost reduction for AdventHealth Central Fill Pharmacy to continue with the process.  While a few purchases of high-discount drugs can have result in large per-transaction savings, most of the accumulated savings over time comes from lower rates of savings on commonly purchased drugs.  Optimizing this process has the potential to allow for greater cost reduction in the future.

Title: Implementing a Pharmacy Clinical Decision Support Council for a 54Hospital health-system.

Authors: Sabrina Desmond, Pharm.D.; Fei Wang, Pharm.D.; Craig MacDonald, Pharm.D.

Objective: To describe the creation and maintence of a clinical decision support oversight group to manage alerts within a multi-state health system.

Self Assessment Question: True or False: The only purpose of a clinical decision support council is to eliminate as many alerts as possible to reduce alert fatigue.

Background: Pharmacy-driven clinical decision support (CDS) is a critical element for optimizing medication safety and therapeutic outcomes in inpatient hospital systems. Pharmacy teams often manage an extensive array of drug-related alerts that can range from formulary and dosing recommendations to drug interactions and therapeutic duplications. Without focused oversight, these alerts can become overwhelming and lead to alert fatigue which could compromise the quality of patient care. A pharmacy-specific CDS Council can address these challenges by standardizing alert strategies, prioritizing clinically relevant notifications, and driving evidence-based enhancements to medication-related workflows.

Methods: Constructing an advisory group to oversee drug-related alerts required several steps:

• Establishing a workflow for alert review submissions
• Creating specialized workgroups that could provide insight for alert reviews
• Construct a voting structure that allows for diverse assessment and input

Items are entered into the CDS review process by inputting a request into a ticket management platform. After submission, items are reviewed by a workgroup comprised of electronic health record (EHR) analysts and pharmacy informatics experts to evaluate the request and propose solutions for consideration. Following that initial review, solutions are discussed and voted on independently by a medication safety workgroup and a clinical workgroup. Each of these workgroups have at least one representative from each of the eight regions within the hospital system. Finally, solutions and voting results are presented to a final council that includes leaders in clinical, medication safety, EHR, and informatics specialties. All results are discussed, and final actions are determined.

To evaluate the interventions made by this CDS council, requests have been categorized into 4 groups: alert deactivation, alert activation, creation of a custom alert, and no action taken. A secondary analysis was performed to better understand the reasons for taking no action. Therefore, the group "no action taken" was further subcategorized into one of the following: alert is appropriate, alert is not indicated, functionality not available, or exists in current state.

Results: A total of 59 items were tracked from August 1st, 2024 - March 31, 2025.The number of items reviewed were categorized into the 4 groups as follows:

• Alert deactivation: 4 items 
• Alert activation: 11 items
• Creation of custom alert: 12 items
• No action taken: 32 items

Since there were 32 items that were categorized as "no action taken," these items were further subcategorized as follows:

• Alert is appropriate: 18items 
• Alert is not indicated: 5 items
• Functionality not available: 5 items
• Exists in current state: 4 items

Conclusion: A structured workflow is needed in order to perform a systematic review and comprehensive assessment for CDS enhancements within a health system. By gathering insight from diverse specialties and system-wide representatives, balanced adjustments can be made to alert settings. These adjustments potentially lead to management of alert fatigue and other limitations experienced with CDS tools. Further investigation is needed to assess the overall impact that is made when adjusting alert settings. Additionally, further optimization and expansion of the CDS Council process can allow for proactive investigation of potential alert adjustments as well as an assessment of other CDS components beyond alert settings.

Title: Transitioning RFID Medication Tray Solutions: A Nine-Hospital Network Implementation
Authors: Nina Desai, PharmD.; Hiren Shah, PharmD., BCPS; Craig MacDonald, PharmD.
Objective: To address the conversion between RFID systems and the challenges that arise during this process.
Self Assessment Question: True or False. Vendors will always be able to transfer currently tagged products to a new system during a vendor conversion. 
Background:
RFID (radio-frequency identification) in pharmacy enhances the accuracy and efficiency of medication management. By tagging medications with RFID, pharmacies can automate the tracking process and reduce human error. AdventHealth utilizes RFID technology for code trays and anesthesia station trays. Trays and medications are equipped with RFID tags containing unique identifiers for precise tracking and inventory control. The goal of the system is to monitor tray and medication inventory and notify staff when items are nearing expiration or trays need restocking.
AdventHealth is in the process of transitioning vendors for RFID tray solutions. The key benefit of System B is the ability to purchase pre-tagged items directly from AdventHealth’s established wholesaler.
During a brand-new RFID go live a drug library would be built and the items required in the pharmacy would be tagged and entered into the system. With a conversion the existing information and tagged products would ideally be transferable to the new system, removing potential work and redundant inventory. While most published information focuses on implementing new RFID system, this presentation addresses the conversion between RFID systems and the challenges that arise during this process.
Methods: 
AdventHealth’s Central Florida Division is transitioning from System A to System B in a structured three-phase approach. Each phase includes three hospitals and phase three includes sites using mobile solutions for anesthesia-station trays, including the 1,366-bed flagship hospital. The conversion is led by the pharmacy informatics team, vendor representatives and local pharmacy leadership.
System A was not able to transfer existing drug registry information into System B, requiring a conversion process for existing tagged inventory. The preparation of tagged items for the conversion involved commissioning medications previously tagged in System A, purchasing pre-tagged items from the wholesaler, and newly tagging items by the pharmacy.
An action plan was developed for the six weeks preceding the go-live. The first two weeks were spent gathering and preparing tray, NDC, user, and site data. The next week was dedicated to training users, with each site completing buyer, superuser, and general staff training. The following three weeks focused on commissioning to ensure tagged medication was available during go-live. Additionally, sites were encouraged to purchase additional inventory pre-tagged through the wholesaler associated with System B. The existing tagged inventory was separated by lot and commissioned, with daily tracking of commissioned items.
These commissioning weeks provided enough items to reach the project milestones (three-tiered progression):

1. Completed at least one of each tray template.
2. Completed at least three of each tray template.
3. Completed high-priority trays.

Results:
Phase one was completed in December 2024, and phase two will finish by mid-March 2025. The team initially faced a steep learning curve, including commissioning bottlenecks. During some commissioning steps, we are limited to a single System B kiosk, which caused significant bottlenecks. For reference, a smaller site with nearly 200 beds has an inventory of nearly 3,000 tagged items in System A that needed to be commissioned into system B.
The action plan was adjusted after each go-live, leading to significant improvements in preparation and execution. As the flagship hospital transitions to a mobile solution in phase three, the plan will continue to evolve.
Conclusion: In progress