Why referral management is critical to infusion service line success
At a glance
Insights from two infusion leaders on simplifying the provider referral experience and routing care deliberately across sites of care.
As payer site-of-care requirements become more complex, health systems are expanding infusion capabilities beyond the hospital outpatient department (HOPD). But adding ambulatory infusion centers, ambulatory infusion suites or home infusion does not ensure appropriate referrals will reach those settings. Without a coordinated intake and routing process, providers may face repeated rework and patients may be directed outside the health system.
Erinn Rowe, PharmD, MS, BCSCP, Pharmacy Director, Enterprise Infusion Services at UNC Health, and Matt Hinchey, Vice President, Portfolio Strategy, Infusion and Market Access at Accelerate Pharmacy Solutions, examined this challenge during a breakout session at the 2026 Health Systems National Summit. Their discussion explored how health systems can simplify the provider referral experience, coordinate site-of-care decisions, and establish pharmacy as an orchestration layer across intake, benefits verification, prior authorization, product selection and scheduling.
Their central message was that a referral should enter through a consistent process, while the ultimate site of care should be determined through a deliberate strategy based on payer requirements, patient needs, product considerations and available capacity.
Building an infusion service line, not simply adding sites
Hinchey framed the market challenge: payer site-of-care programs are placing greater pressure on health systems to provide clinically appropriate lower-cost options beyond the HOPD. When a health system cannot meet a payer restriction, an infusion referral may be routed outside the organization. Building alternate sites is therefore important, but it addresses only one side of the problem.
The other side is patient access. A broader site portfolio does not determine where a referral ultimately goes. If HOPD, home infusion and alternate sites rely on separate intake processes, each new option can add friction rather than reduce it. A denied HOPD referral may return to the provider for rework instead of moving to another appropriate setting within the system.
Rowe described UNC Health’s starting point as a predominantly HOPD-based model with approximately 10 oncology infusion suites, three therapeutic infusion sites and an established home infusion operation. As the organization brought those capabilities into a more coordinated leadership structure and added its first ambulatory infusion suite, referral management became a priority. Without a shared approach to intake and routing, the organization risked adding new sites without giving providers a practical way to use them as one connected service line.
Referral management is therefore the operating backbone that connects provider intent with payer requirements, patient access, and the health system’s infusion assets.
Designing referral management around the provider
The work started with provider observation. Rowe’s team visited specialty clinics, spoke with embedded pharmacists, nurses and providers, and watched referrals being created in real time. The goal was to identify where the process created work, where referrals left the organization and what freestanding infusion centers were doing differently.
One important difference was ease of use. Hinchey noted that freestanding infusion centers often offered a simple promise: send the referral once, avoid repeated bounce-backs and receive confirmation that the patient was treated. By comparison, UNC Health’s existing process could be lengthy and was designed primarily around the HOPD.
UNC Health responded by creating a simpler, drug-specific request-for-order workflow in Epic for the 10-15 therapies expected at its new ambulatory infusion suite. Much of the form was prepopulated so the ordering clinician could complete it with fewer steps. The referral then entered a work queue where the appropriate team could evaluate benefits, coverage and site-of-care requirements. As Rowe explained, the goal was to create an “easy button” for providers: let the referral come to pharmacy, determine benefits and the covered product, including the plan-preferred biosimilar, and send it back to the provider for co-signature.
“A referral is a referral. Where you service it is a strategy.”
This model separates two decisions that are often blended together. The provider initiates the referral through a consistent, easy-to-use process. The health system then determines the clinically appropriate site based on payer requirements, product considerations, patient needs, and available capacity. The provider should not need to understand the ownership, revenue cycle, or workflow differences among sites before placing an order.
Pharmacy as the orchestration layer
That front-end role includes intake, authorization, product selection, documentation and determining the appropriate site of care. Rowe connected those decisions directly to financial performance, noting that getting authorization and documentation right on the front end helps the revenue-cycle process work more effectively on the back end.
Hinchey emphasized that pharmacy does not have to own every person, system or operational step. In many organizations, responsibility will remain distributed. Pharmacy can still serve as the orchestration layer by bringing the right teams together, defining handoffs and influencing the inputs required to be accountable for the outcome.
That operating model should establish decision rights and shared measures across pharmacy, access, revenue cycle, IT, nursing, clinic operations, and scheduling. The objective is not ownership for its own sake. It is coordinated execution across functions that currently solve similar referral problems in separate workflows.
Starting with data and building for scale
A centralized referral model can raise understandable concerns, from questions about existing processes to fears that alternate sites will shift established volume. Technology and staffing requirements can make the effort appear larger than it needs to be.
Rowe and Hinchey described data as the foundation for the discussion: which referrals are leaving, where the workflow breaks down and which therapies or clinics represent the clearest opportunity. At UNC Health, provider feedback and leakage data helped establish the need for change.
Rowe also pointed to capabilities UNC Health already had in place. A centralized medication assistance and prioritization team was already serving part of the HOPD footprint, creating an opportunity to extend proven work across more sites. Pharmacy’s record of delivering on earlier business plans also helped establish credibility for the next proposal.
That distinction also helped address concerns within UNC Health about how alternate sites could affect existing HOPD volume. Rowe said HOPD teams were concerned they might lose patients. She explained that the objective was not to move patients already being appropriately served in the HOPD, but to recapture referrals already leaving the system and create additional options for patients who could appropriately be served at a different location. Existing HOPD capacity remained an important part of the strategy.
The first step does not have to be an enterprise-wide transformation. UNC Health started with a focused set of high-priority drugs for its new ambulatory infusion suite. Hinchey emphasized the value of launching a workable process, learning from provider behavior and refining the model over time.
As organizations scale, Rowe emphasized two lessons: design the first site with replication in mind, so the next two or three can be built from a focused set of essential requirements and engage providers early enough to identify practical barriers before they become embedded in the operating model.
These lessons also provide a practical starting point for health systems assessing whether their infusion sites operate as a coordinated service line:
- Do referrals enter through one consistent process that allows providers to refer without first determining the final site of care?
- Where do payer restrictions, prior authorization or benefits verification create rework or cause referrals to leave the health system?
- Who owns or influences intake, product selection, documentation, routing and the handoff to scheduling?
- Which therapies or provider groups offer the strongest starting point for a focused pilot?
The answers can reveal whether the health system has a collection of infusion sites or a coordinated infusion service line.
Turning infusion strategy into an operating capability
Payer site-of-care programs make a broader infusion portfolio increasingly important. Yet the value of that portfolio depends on whether referrals can move through it. Without coordinated intake and routing, providers may continue to face rework, patients may be directed outside the organization, and new sites may not receive the referrals they were built to serve.
A consistent front door, pharmacy-informed access processes and deliberate site-of-care routing can translate enterprise infusion strategy into day-to-day execution. Health systems that make referral management easier for providers will be better positioned to use existing assets, adapt to payer requirements and keep appropriate care connected across the system.
Connect referral management to a broader infusion strategy
Whether you're evaluating a new infusion program or optimizing an existing model, Accelerate Pharmacy Solutions can help your team assess opportunities and prioritize what comes next. Contact us to discuss how a broader infusion strategy could support your health system's goals.
