When technology thinks fast and humans think deep

Pharmacy claims monitoring is the continuous, independent review of pharmacy benefit transactions to verify that claims are adjudicated according to the health plan’s benefit design, contract terms, and regulatory requirements. Unlike periodic audits that sample a subset of claims months after a plan year closes, continuous monitoring evaluates every transaction on an ongoing basis. At Rivera, it is delivered through a combination of proprietary algorithms and clinical expertise — technology that processes every claim, and pharmacists and analysts who validate findings, diagnose root causes, and ensure the underlying issues get corrected.

This article explains how that system works: what the technology does, where clinical judgment enters, and why the combination produces compounding value that neither component delivers alone.

The verification problem continuous monitoring solves

Health plans design their pharmacy benefits with specific intent. Copay tiers are structured around formulary placement. Pricing is negotiated against benchmark rates. Coordination of benefits rules govern how claims interact with other coverage. Clinical programs — step therapy, prior authorization, quantity limits — are configured to manage utilization and cost.

The pharmacy benefit manager administers all of this. The PBM controls claim adjudication, formulary management, network contracting, and rebate negotiation. But the health plan retains the financial risk. This creates a structural verification gap that is unique to pharmacy: the entity designing the benefit is not the entity executing it, and there is no native mechanism within the PBM’s systems to independently confirm that individual claims are being processed according to the plan’s actual terms.

In medical claims, health plans often adjudicate through their own systems or have direct visibility into the adjudication logic. In pharmacy, the PBM operates a separate technology stack, applying its own configuration of the plan’s terms. Independent pharmacy payment integrity exists to close that gap.

Continuous pharmacy claims monitoring is the operational discipline that makes that independence meaningful. It takes the plan’s benefit design, contract terms, and regulatory requirements, encodes them into a structured set of rules, and applies those rules against every pharmacy claim the PBM processes. The output is not a general compliance score. It is a claim-level accounting of where adjudication matched the plan’s intent and where it did not.

How Rivera’s proprietary algorithms work

Rivera’s monitoring platform runs a library of more than 750 proprietary algorithms against 100% of a plan’s pharmacy claims. Each algorithm is tied to a specific question about how the benefit should be adjudicated. These are not generic compliance checks. They are rules encoded from the plan’s actual contract terms, pricing agreements, copay structures, formulary tier assignments, coordination of benefits rules, and clinical program requirements.

The distinction matters. A generic monitoring tool might flag that a claim’s price exceeds a benchmark. Rivera’s algorithms ask whether the claim’s price matches the specific rate the health plan negotiated with the PBM for that drug, at that pharmacy, under those contract terms. The specificity is what converts a flagged anomaly into an actionable, recoverable finding.

The algorithm library compounds over time. When Rivera identifies an error pattern for one client — a pricing misapplication in specialty drug adjudication, a coordination of benefits logic gap, a configuration drift that followed a mid-year formulary change — that finding informs the algorithms applied across every plan Rivera monitors. The system gets more precise as it scales. Each new client relationship and each new finding adds to the collective intelligence of the library, which means plans that come on later benefit from everything Rivera has already learned.

What 100% claim review changes

Traditional pharmacy audits work from a sample. They are designed to estimate the scope of an issue across a population of claims, not to find every instance. For a health plan processing hundreds of thousands or millions of pharmacy transactions, sampling introduces structural blind spots.

Some errors are systemic — they affect a large percentage of claims and surface reliably in any reasonably designed sample. But others are concentrated in specific drug classes, pharmacy types, or member populations. A pricing error that only affects specialty claims filled at a particular network tier will not appear in a random sample unless the sample happens to include those claims. A coordination of benefits issue that applies only to members with secondary coverage through a specific carrier may never surface at all.

Reviewing 100% of claims eliminates that risk. Every transaction is evaluated against the full algorithm library, regardless of drug class, pharmacy, or member. This is also what makes pattern detection possible at a level sampling cannot match. When the system sees every claim, it can identify correlations between error types, track whether a corrected issue is actually staying corrected, and detect new drift as it begins rather than after it has compounded across months of claims. For a deeper comparison of continuous monitoring and periodic audits, see our analysis of how these approaches differ in practice.

Where clinical expertise enters the process

The algorithms surface findings. Clinical expertise determines what to do with them.

This is not a metaphor about humans being “in the loop.” It is a description of a specific operational layer. When Rivera’s algorithms flag a claim, the finding enters a review process staffed by clinical pharmacists and analysts who evaluate it against the full context of the plan’s benefit design, the PBM’s adjudication history, and the clinical circumstances of the claim.

A NADAC variance on a generic claim might be a true overpayment — or it might reflect a timing difference between the pricing database update and the claim’s adjudication date. A duplicate therapy flag might indicate a safety concern — or a clinically justified overlap during a medication transition. A coordination of benefits discrepancy might be an error in the PBM’s configuration — or a reflection of a secondary payer’s non-standard processing timeline.

The clinical team makes these determinations. They separate true errors from explainable variances. They distinguish one-time configuration issues from systemic problems that will generate recurring overpayments if left uncorrected. And they translate validated findings into documented, quantified recoveries that the health plan can pursue with its PBM — not a list of flags, but an accounting of what was paid incorrectly, why, and what the correct amount should have been.

The cycle that compounds: recover, correct, prevent

Every validated finding initiates a three-part cycle.

Identification and documentation is the immediate output. Claims that were paid incorrectly are validated, quantified, and documented with the detail the health plan needs to pursue recovery with its PBM. This is the number that appears on a board report and, for most health plans, the recoveries they pursue exceed the program’s cost in the first year.

Root-cause correction follows. Rivera works with the health plan and the PBM to identify the underlying configuration, pricing logic, or contract interpretation that produced the error. The goal is not just to give the plan the documentation it needs to recover dollars already lost but to fix the mechanism that generated the loss in the first place.

Recurrence prevention is the compounding layer. Once the root cause is corrected, the same error stops appearing on future claims. The program’s value shifts from finding problems to eliminating them. Over a multi-year engagement, this means the plan’s total pharmacy spend becomes progressively more accurate — not because errors stop occurring entirely, but because each one that surfaces gets resolved at the source rather than persisting undetected.

This is where the algorithm library’s compounding effect and the clinical team’s diagnostic work converge. The algorithms catch the error. The clinical team validates it and identifies the root cause. The correction prevents recurrence for that plan. And the finding gets encoded into the algorithm library, where it prevents the same error across every plan Rivera monitors. For a detailed look at how this translates into measurable ROI, see our analysis of recovery and recurrence prevention value.

What continuous pharmacy claims monitoring does not do

Clarity about what continuous monitoring is also requires clarity about what it is not.

Rivera does not intercept claims before they are paid. Pharmacy claims are adjudicated by the PBM at the point of sale, and Rivera reviews them after payment. The value is not in preventing the initial transaction but in identifying where adjudication did not match the plan’s terms, documenting the overpayment so the plan can recover it, and correcting the root cause so it does not recur. This retrospective model is what allows Rivera to operate independently of the PBM’s adjudication infrastructure — an important structural distinction, because independence is what makes the verification credible.

Rivera does not replace the PBM. It does not adjudicate claims, manage formularies, contract with pharmacies, or negotiate rebates. It verifies that the PBM is executing the health plan’s benefit design and contract terms correctly at the claim level. Plans invest significant resources negotiating those terms and designing those benefits. Rivera’s role is to confirm that the work the plan already did is reflected accurately in how every claim gets processed.

And Rivera does not make benefit design decisions for the plan. The algorithms are configured from the plan’s existing design — they evaluate execution, not strategy. When findings surface, the plan decides how to act on them. Rivera provides the data, the clinical context, and the documentation. The decision authority stays with the plan.

Frequently asked questions

How does continuous pharmacy claims monitoring differ from a PBM audit?

A traditional PBM audit is periodic, typically annual, and works from a sample of claims. It produces a true-up months after the plan year closes. Continuous pharmacy claims monitoring reviews 100% of claims on an ongoing basis, which means errors are identified closer to when they occur and root causes can be corrected before they compound across a full year of transactions. The audit becomes a confirmation step rather than the primary detection mechanism.

What types of errors does continuous claims monitoring find?

Common findings include pricing misapplications where contracted rates are not correctly applied, coordination of benefits errors, misapplied copay and coinsurance logic, incorrect generic or brand adjudication, rebate shortfalls, and configuration drift that occurs after mid-year benefit changes or formulary updates. Rivera’s 750+ proprietary algorithms are each tied to a specific question about how the benefit should be adjudicated, covering the full spectrum of plan design and contract terms.

Does Rivera review claims before they are paid?

No. Rivera operates retrospectively. Pharmacy claims are adjudicated by the PBM at the point of sale, and Rivera reviews them after payment. This retrospective model is what allows Rivera to operate independently of the PBM’s adjudication infrastructure. When an error is identified, Rivera documents the overpayment and provides the plan with the information it needs to pursue recovery. Rivera then works with the plan and PBM to correct the root cause so the error does not recur on future claims.

How long does it take to see results from pharmacy claims monitoring?

Most health plans see measurable recoveries within the first months of engagement. At Rivera, 100% of clients recover their program investment in the first year. Recurrence prevention value begins accruing as soon as root-cause corrections are implemented and compounds over the remaining contract period.

How does Rivera’s algorithm library improve over time?

The library compounds. When Rivera identifies a new error pattern for one client, the finding is encoded as an algorithm that runs across every plan Rivera monitors. This means each new client relationship and each new finding adds to the collective intelligence of the system. Plans that engage later benefit from everything Rivera has already learned across its entire client base.