340B rebate readiness is the operational capacity to identify, validate, submit, and reconcile claims under a rebate-based 340B model, then confirm that the rebate actually received matches what was owed. It is not the act of uploading a file once a month. Done correctly, it is a continuous cycle of data cleanup, exception handling, and financial verification.
That distinction sits at the center of a recent episode of 340B Pulse, the NorthArc Health podcast powered by PureLogics. Host Muhammad Atif spoke with Celeste Fowler, Executive Director of Pharmacy 340B at Piedmont Healthcare, about what a rebate-style model would actually demand from covered entities operationally, beyond the litigation and manufacturer-policy headlines dominating the public conversation.
Understanding who is making this argument, and from what vantage point, makes the case for 340B rebate readiness concrete rather than theoretical.
Who Is Celeste Fowler, and What Does Her 340B Program Look Like at Piedmont Healthcare?
Celeste Fowler is Executive Director of Pharmacy 340B at Piedmont Healthcare, where she leads the 340B program at the health-system level across multiple eligible hospitals. She did not plan a career in 340B. As a newly installed pharmacy director at a community hospital, her executive told her to get the program going, a term she did not immediately recognize before educating herself through formal 340B training and building the program from the ground up.
Fowler describes running that first program as a one-woman show, supported by a single strong buyer, before the hospital’s affiliation with a larger health system surfaced other eligible facilities and expanded her role into system-wide oversight. Her background spans pharmacy leadership, hospital operations, healthcare IT, compliance, and patient safety, and that operational range is exactly why she pushes back on framing the rebate model as a compliance-only conversation.
What Are Healthcare Leaders Missing on the Operational Side of the Rebate Debate?
Healthcare leaders are missing that data integrity and security, not policy uncertainty, is the more immediate operational risk in a rebate-based model. Fowler is direct that recent, high-profile data breaches across the industry have made her wary of how information moves once it leaves a covered entity’s own systems.
She calls her organization’s current information security setup “Fort Knox,” built through years of vetting IT partners who agree to standard protections. The rebate model complicates that standard. According to Fowler, current rebate-model vendors have taken a take-it-or-leave-it stance on data protection language, without the indemnification clauses her other IS partners have always been willing to negotiate. That gap leaves covered entities exposed if a vendor-side issue ever surfaces.
Why Doesn’t More Data Automatically Mean More Transparency?
More data does not automatically mean more transparency, because covered entities already submit extensive information to their TPA for internal eligibility and qualification purposes. Adding another vendor or another data field does not inherently strengthen that process; it simply adds another set of eyes to data that already exists.
Fowler’s recommendation is to disclose only what a requirement actually needs. If a data field is not necessary for 340B eligibility, she argues covered entities should be cautious about supplying it just because a manufacturer or vendor is asking. That discipline protects both the covered entity and, ultimately, the patient whose information sits behind every claim.
What Actually Happens Operationally From Claim to Confirmed Value?
Operationally, moving a claim from initial identification to confirmed rebate value requires reconciling data that rarely lives in one place. Fowler describes it as a scavenger hunt across systems that were never built to talk to each other.
A hospital’s revenue-cycle software, its pharmacy software, and its EMR often function as separate, disparate systems, especially since medical claims data is not billed in the same format as retail or outpatient claims data. Small inconsistencies compound quickly. A few of the recurring failure points Fowler names directly include the following.
- Units of measure that do not match across systems, such as milliliters recorded in one place and milligrams in another.
- Billing codes, including HCPCS codes, that are not standardized the same way across every source system.
- Revenue-cycle charges that live outside the pharmacy system entirely, requiring separate reconciliation.
- Medical claims that continue changing after submission, through payer-driven updates, denials, approvals, and reversals.
Retail claims are largely one-and-done once a patient picks up a prescription. Medical claims are not. Fowler stresses that a rebate-ready program has to be fluid enough to ingest ongoing changes accurately and compliantly, not treat a submitted file as final the moment it leaves the building.
Why Is Standardization the Core of a Defensible 340B Program?
Standardization is the fixed compliance foundation that lets a covered entity spot an abnormality quickly, even while the rest of the program adapts to local needs. Fowler compares it to baking: the recipe, meaning the core compliance policy, never changes, even though the frosting on top, meaning how each site implements it, can look completely different.
Without that consistent baseline, she explains, it becomes far harder to root-cause a problem. If every site does something slightly differently, a fluctuation in charges could be normal variation, or it could be a real compliance issue, and there is no clean way to tell the difference. Fowler describes every 340B program as its own snowflake: no two look exactly alike, but the fundamentals of compliance and integrity stay constant across an entire health system regardless of how many hospitals or communities it serves.
Where Do AI and Automation Actually Fit in 340B Operations?
AI and automation fit into 340B operations primarily by handling volume, not by replacing clinical or compliance judgment. Fowler is candid that a human eye and human critical thinking cannot be fully replaced, but AI enables something a manual process cannot: a 100 percent audit of claims instead of a random sample.
That shift matters because most covered entities cannot realistically hire enough staff to review every single claim by hand. Fowler frames AI’s value as giving her “eyes where I can’t see,” since AI does not need to rest the way a human reviewer does. Where AI genuinely helps, according to Fowler, includes auditing large claim volumes for black-and-white compliance issues, reconciliation work such as identifying missing encounter data or a dropped file between the EMR and the TPA, and surfacing patterns that would otherwise take a human far longer to find.
Where human review must stay essential is anywhere clinical judgment enters the picture. Fowler points specifically to patient eligibility questions: whether a patient was genuinely treated for the diagnosis a medication addresses often requires a human to read a SOAP note or confirm a referral, because, in her words, each patient is their own case and not a carbon copy of another.
She also flags a real security dimension to AI adoption. Any AI or data vendor needs thorough vetting, because even redacted data raises the question of whether a manufacturer could reverse-engineer it back to an individual patient. NorthArc Health has faced this same tradeoff directly while building a 340B command center for a large PBM client, ultimately training a model on-premises on the organization’s own claims data specifically because patient information could not leave the premises.
What Does “Checking the Checker” Mean for 340B Financial Sustainability?
Checking the checker means a covered entity must independently confirm that a rebate it is owed is the rebate it actually receives, rather than assuming a manufacturer’s payment is automatically accurate. Fowler expects the rebate model to create a massive change in cash flow for covered entities, many of which lack significant liquid assets in the first place, precisely because a payer mix light on commercial coverage is often what made them 340B-eligible to begin with.
Under a rebate model, covered entities pay the higher WAC price upfront on expensive, high-utilization medications and then wait for the manufacturer’s rebate to come back. Verifying that rebate takes dedicated time and staff that lean 340B teams do not have to spare. Fowler’s practical fix is opening a separate financial sub-account solely for 340B rebate funds, which makes them possible to isolate and trace back against what was expected, rather than trying to find them inside a health system’s much larger pool of rebate activity. She also recommends leaning harder on existing TPA relationships to avoid duplicating uploads, since every additional integration is another place where data and accountability can break down.
How Should Health Systems Build Governance and Vendor Accountability at Scale?
Health systems build effective governance by treating 340B as a cross-departmental responsibility rather than a pharmacy-only function. Fowler describes her own role as an ambassador of 340B, actively educating care management, finance, revenue cycle, and IT on how their individual workflows affect program compliance, often discovered through the auditing process itself.
She learned early that a 340B program cannot run in a silo. Growing from a one-woman show into a system-wide role required executive leadership buy-in, and Fowler credits her organization’s continued investment in the program to leadership genuinely seeing its value. That alignment, she argues, is what allows a program to keep functioning as complexity and vendor involvement keep increasing.
What Should Covered Entities Do First to Prepare for a Rebate-Based Model?
The following table summarizes Fowler’s most concrete, immediately actionable guidance for 340B rebate readiness from the episode.
| Priority Area | Fowler’s Guidance |
|---|---|
| First workflow to map | Start with your existing TPA relationship to see what data it already houses and whether it can submit or report on your behalf. |
| Automation and AI | Necessary for small or underfunded teams, since most 340B operators run the program as an added duty on top of another job. |
| Vendor accountability | Ask directly how a vendor protects your data, and treat any vendor unwilling to guarantee real security as a signal to walk away. |
| Financial tracking | Open a separate account for 340B rebate funds so they can be isolated and checked against what was expected. |
| Mindset | Just because a workflow has not been built before does not mean it cannot be built. |
Conclusion
Celeste Fowler’s account of preparing for a rebate-based 340B model makes one point about 340B rebate readiness unmistakably clear: the five-hour myth undersells the work by treating a single upload as the whole job. Real 340B rebate readiness means fixing data integrity and security first, standardizing the compliance recipe underneath every site’s unique implementation, giving AI the volume work while keeping clinical judgment in human hands, and building a financial process that actively checks the checker instead of assuming a rebate payment is correct. NorthArc Health builds custom technology and Agentic AI solutions that help 340B programs manage exactly this kind of data reconciliation and financial verification discipline.
Frequently Asked Questions (FAQ)
What does 340B rebate readiness actually require?
340B rebate readiness requires a covered entity to identify claims accurately, validate underlying data across pharmacy, revenue-cycle, and EMR systems, submit clean files, and then confirm that any rebate received matches what was actually owed, rather than treating submission as the final step.
Why is the five-hour myth misleading?
The five-hour myth is misleading because it describes only the mechanical act of uploading a file, not the data cleanup, exception handling, and reconciliation that surround it, work Celeste Fowler suggests could realistically run closer to five hours a day for a program done correctly.
Does adding more data fields improve transparency in a rebate model?
Adding more data fields does not automatically improve transparency, since covered entities already submit extensive data to their TPA for internal eligibility purposes; additional fields largely add another set of eyes to information that already exists rather than creating new insight.
Where should AI be used in a 340B program, and where should it not?
AI should be used for high-volume, black-and-white audit and reconciliation work, such as identifying missing encounter data or flagging exceptions across a full claims set, while clinical judgment about an individual patient’s eligibility should always remain with a human reviewer.
What does checking the checker mean in a rebate-based 340B model?
Checking the checker means a covered entity independently verifies that a manufacturer rebate payment matches the amount actually owed, instead of assuming the payment is automatically accurate, a process that requires dedicated financial tracking and staff time.
How can a covered entity start preparing for rebate readiness today?
A covered entity can start by mapping its existing TPA relationship to understand what data it already houses, opening a dedicated financial sub-account to isolate rebate funds, and evaluating whether current vendors are willing to guarantee real data security protections.