How to evaluate a pharmacy payment integrity vendor

Evaluating a pharmacy payment integrity vendor means testing whether a prospective partner can independently verify, at the claim level, that your pharmacy benefit manager (PBM) contract terms and benefit design are being executed as written, and whether it can do so across your full claims volume rather than a sample. The label covers a wide range of offerings, from audit firms that periodically test guarantees to fraud, waste, and abuse tools to continuous claims monitoring programs, and those offerings differ substantially in coverage, methodology, and the kind of findings they produce.

This article lays out the evaluation criteria that separate those offerings, drawn from what Rivera sees health plans weigh when they run this process well. It is written for pharmacy and finance leaders comparing options, whether the comparison is between external partners or between buying a program and attempting to build one.

Start with independence

Independence is the foundation of the entire category, and it is the first thing to verify. A partner affiliated with a PBM, or one with revenue-sharing arrangements or operational dependencies tied to PBM relationships, has a structural conflict with the work. Its findings are compromised before the first claim is reviewed.

Ask directly: does the vendor have any ownership relationship, reseller arrangement, or revenue dependency involving a PBM? The answer needs to be an unqualified no. Everything else in a pharmacy payment integrity program rests on the credibility of an independent second set of books.

Coverage: 100% of claims or a sample

The second structural question is what share of claims the vendor actually reviews. Sample-based approaches extrapolate from a subset, which means errors that fall outside the sample frame go undetected. Low-volume specialty claims, mid-year configuration changes, and edge cases in coordination of benefits logic are exactly the kinds of issues that sampling misses, and they are also where meaningful dollars tend to sit.

A vendor reviewing 100% of claims on a continuous cadence produces a complete picture rather than an estimate. It can tell you which specific claims were affected by an issue, not that an issue probably exists somewhere in the population. That distinction determines whether findings are actionable, which is the next criterion.

Specificity: whose rules are the algorithms testing

Generic compliance checks applied identically to every client will surface generic findings. The question to ask is whether the vendor's analytic library is configured from your plan's actual benefit design, contract terms, formulary structure, and regulatory requirements, so that every claim is being tested against the rules that should actually govern it.

There is a second dimension to this criterion that is easy to miss: whether the vendor's library compounds. When findings from one client inform the algorithms applied across all clients, the program gets more precise as it scales, and each new client benefits from every issue the vendor has ever worked through. A library built fresh for each engagement, or a static rule set licensed off the shelf, does not improve with use. Rivera's library of more than 750 proprietary algorithms was built this way, from years of real-world findings, and the compounding is the differentiator rather than the count.

For plans operating multiple lines of business, specificity has a third dimension. Commercial, Medicare, Medicaid, and self-funded books each carry different federal and state requirements. A vendor should be able to normalize claim-level findings across those lines so patterns surface once and get addressed once, rather than in siloed projects per product.

Validation: who stands behind a finding before you see it

Not every flagged claim is an error. Some variances have legitimate explanations, and a program that forwards raw exceptions to your team is transferring work rather than performing it. Ask how findings are validated before delivery, and by whom. Rivera's findings are reviewed by pharmacists who separate true errors from explainable variances, because the technology identifies patterns and people confirm what they mean.

The quality of documentation matters as much as the validation. A finding you can act on arrives with the specific claims affected, the contractual or benefit design basis for why the adjudication was incorrect, the quantified impact, and the recommended path for pursuing it with your PBM. Ask any prospective vendor for a redacted sample finding and judge whether your team could take it into a PBM conversation as delivered.

Follow-through: identification is half the job

Some vendors stop at identification. The stronger test is what happens after: does the partner diagnose the root cause, work with the plan and PBM to correct the underlying configuration, and then keep monitoring to confirm the correction holds? Errors in pharmacy benefit administration recur, because the environment that produced them keeps changing. A program without recurrence prevention leaves you paying to find the same problems repeatedly.

Ask for evidence of how the vendor tracks recurrence by error category, and how it reports whether corrected issues stayed corrected.

Alignment: how the fee structure shapes behavior

Fee structures are evaluation criteria, not just budget line items, because they determine what the vendor is incentivized to find. Contingency models tied to recovery volume reward the vendor for error volume. A flat per-member-per-month (PMPM) fee gives the plan cost predictability and aligns the partner around findings that are real, documented, and recoverable, rather than around volume that justifies fees.

Whatever the structure, ask the vendor to commit to how performance will be measured in year one and beyond. Across Rivera's client base, 100% of clients recover their investment in the first year of the contract, and a vendor confident in its program should be willing to be measured against a standard like that.

Build versus buy

Some plans consider building this capability internally, and the pharmacy teams asking that question are usually the sophisticated ones. The honest constraint is scale rather than talent. An internal build requires constructing and maintaining an algorithm library across every benefit design permutation, pricing methodology, and regulatory requirement the plan touches, and that library only becomes precise through years of accumulated findings. Teams reviewing claims manually will always be working a sample, and systemic issues do not reliably surface in small data sets. The realistic comparison is not internal team versus vendor, but internal team plus independent infrastructure versus internal team alone. What to expect from an independent program covers how that combination works in practice.

Frequently asked questions

What criteria should health plans use to select a payment integrity partner?

The core criteria are independence from any PBM, review of 100% of claims rather than a sample, algorithms configured from the plan's own benefit design and contract terms, validation of findings by pharmacists before delivery, root-cause follow-through with recurrence tracking, and a fee structure that aligns the partner with plan outcomes. Plans operating multiple lines of business should also confirm the partner can normalize findings across commercial, Medicare, Medicaid, and self-funded books.

How is a pharmacy payment integrity vendor different from an audit firm?

An audit firm tests a sample of claims against contract guarantees at a point in time, typically after the plan year closes. A pharmacy payment integrity partner monitors 100% of claims continuously against the plan's full benefit design, contract, and regulatory requirements, identifies issues while they are current, and works through root-cause correction so errors do not recur. The two can coexist, but they answer different questions on different timelines.

What documentation should a payment integrity vendor provide with each finding?

Each finding should include the specific claims affected, the contractual or benefit design basis establishing why the adjudication was incorrect, the quantified financial impact, and a recommended path for pursuing the issue with the PBM. Documentation at that level of detail shifts the burden of proof in the PBM conversation, because the plan arrives with evidence rather than a suspicion.

Do payment integrity vendors need pharmacy-specific expertise?

Yes. Pharmacy claims carry adjudication logic, pricing methodologies, and regulatory requirements that differ substantially from the medical benefit, and findings must be validated by people who understand how PBM systems actually behave. Vendors that approach pharmacy as an extension of medical payment integrity tend to miss the pharmacy-specific error categories where the largest dollars sit.

Can a health plan pilot a pharmacy payment integrity program before committing?

Structures vary by vendor. A common starting point is an assessment of a historical claims period, which shows the plan what a full program would have identified and gives both sides a factual basis for scoping the engagement. Ask any prospective partner what an initial assessment covers, how long it takes, and what deliverables the plan keeps regardless of whether the engagement proceeds.