Explore Our Blog Library!

TPA administration and fiduciary protection are not the same thing. A TPA processes and pays claims under contract, but ERISA Section 404 keeps fiduciary responsibility with the plan sponsor. Signing a TPA contract does not transfer that duty. Only active, documented oversight of claims activity satisfies the prudent expert standard.

Most self-funded plan sponsors sign a service agreement with a TPA and treat the relationship as a handoff. Once claims start getting paid, the assumption is that someone else is now watching the store. Independent research on claims administration tells a different story.

TPA error rates on processed claims commonly run between 1% and 6% of total volume, and on a plan spending $20 million a year on medical claims, even a conservative 2% error rate is $400,000 in avoidable losses. Fiduciary protection for self-funded health plans does not come from the TPA contract. It comes from what the plan sponsor does after the contract is signed.‍

What TPA Administration Actually Covers

Administering a health plan and protecting it as a fiduciary are two different jobs, and most TPA contracts only cover the first one. A TPA agreement typically spells out claims processing timelines, network access and customer service standards. It rarely includes an obligation to catch every coding error, flag every ineligible dependent or independently verify that billed rates match contracted rates.

Plan sponsors often assume that paying a reputable, well-known TPA is itself a form of due diligence. That assumption is understandable. It is also incomplete under ERISA, where the duty to monitor a service provider cannot be delegated away, even when claims decision authority has been assigned to the TPA.

The result is a structural gap. The TPA handles volume and workflow. The plan sponsor, as the named fiduciary, remains on the hook for verifying that the volume was handled correctly. Nothing in a standard administrative services agreement closes that gap on its own.

Why the Problem Exists

TPAs are built and compensated to process claims at scale, not to flag every dollar that could have been billed more cheaply. Their internal accuracy metrics are typically self-reported, drawn from small samples of their own work, and rarely audited by an outside party before being shared with the plan.

Prompt payment timelines add pressure in the same direction. Most administrative agreements require claims to be paid within 21 to 30 days of receipt, which leaves limited room for a detailed line-item review before a check goes out. Speed and accuracy pull against each other, and speed usually wins by contract design.

Plan sponsors compound the gap by treating the TPA relationship as "set and forget." Committees review premium trends and utilization reports at renewal, but few build a recurring, independent claims review into the plan's operating calendar. Without that cadence, errors accumulate quietly for years.

The Real Cost or Impact

A 2% to 6% error rate sounds small until it is applied to a plan's total claims spend. On a mid-size plan paying $15 million a year in claims, even the low end of that range represents $300,000 moving through the plan incorrectly, year after year, without anyone flagging it. Multiply that across a plan's life span and the number stops looking like a rounding error.

The financial exposure is only part of the picture. In FY 2024, the Department of Labor's Employee Benefits Security Administration recovered nearly $1.4 billion for plans, participants and beneficiaries, a large share of it tied to weak oversight of service providers and plan assets. That is federal enforcement money, not internal audit findings, which means it followed a formal investigation into how a plan was being run.

Self-funded plans are now the dominant funding model in the employer market. KFF's 2025 Employer Health Benefits Survey found that 67% of covered workers are enrolled in self-funded arrangements, rising to 80% at larger firms. That scale means the fiduciary exposure sitting inside unreviewed claims data is not a niche problem. It touches most employer-sponsored health coverage in the country.

What's Actually Happening Behind the Scenes

Sampling Audits Instead of Full Review

Most claims oversight today runs on small statistical samples rather than a full look at every payment. A TPA might review a few hundred claims out of hundreds of thousands processed in a year and extrapolate an accuracy rate from that subset. The math works fine for a general trend line. It does very little to catch the specific overpayment sitting in the other 95% of claims that never got a second look.

Coordination of Benefits Gaps

Coordination of benefits (COB) failures happen when a plan pays as primary on a claim that should have been paid by another carrier first. These errors are common in households with two working spouses or dependents aging onto other coverage. TPAs process COB updates reactively, based on member-submitted information, which means stale COB data can sit in the system for months before anyone notices the plan paid more than it owed.

Dependent Eligibility Drift

Dependent eligibility drift describes ineligible dependents, a divorced spouse, an adult child who aged out, staying on the plan because no one re-verified eligibility after enrollment. Every claim paid on behalf of an ineligible dependent is plan money spent outside the terms of the plan document, which is itself a fiduciary concern independent of the dollar amount involved.

Vendor Accountability and Legal Exposure

Courts have started drawing a sharper line around who actually carries fiduciary risk in these relationships. In *Tiara Yachts, LLC v. Blue Cross Blue Shield of Michigan*, the Sixth Circuit held that functional conduct, not contract labels, determines fiduciary status and liability. That precedent matters for any plan sponsor assuming that a well-worded services agreement automatically shields them from claims-related fiduciary exposure.

Why Current Approaches Aren't Enough

Renewal-cycle reviews and TPA-provided performance dashboards give plan sponsors a general sense of direction. They rarely provide the kind of documented, independent evidence that would hold up if a participant or the Department of Labor challenged how the plan was overseen.

How to Fix It

Red Flags That Signal the Problem Applies to Your Plan

The ROI of Doing It Right

Independent audits regularly identify overpayments in the 2% to 6% range of total claims spend, and those recoveries frequently offset the cost of the audit itself within the first review cycle. For a plan spending $20 million annually, recovering even the low end of that range translates into hundreds of thousands of dollars back in plan assets.

The ongoing value compounds once oversight becomes routine instead of occasional. A plan that reviews claims quarterly catches errors before they repeat across multiple pay cycles, which lowers the total leakage over time rather than just the one-time recovery from a single audit.

The fiduciary protection component is harder to price but just as real. Documented, independent oversight gives the plan sponsor evidence of a prudent process, which is the standard ERISA actually asks fiduciaries to meet. That documentation is the difference between a defensible governance record and an unverified assumption that the TPA had it covered.

Conclusion and Next Steps

TPA administration keeps claims moving. Fiduciary protection for self-funded health plans only comes from what a plan sponsor does to verify that movement is accurate, documented and reviewed on a defined schedule. Those are not the same function, and treating them as interchangeable leaves plan assets and personal fiduciary exposure sitting unmonitored.

The next step is straightforward: schedule an independent claims audit, formalize the benefits committee's charter, and build documentation habits before a complaint or investigation forces the issue. Talk to an independent claims audit specialist.

Frequently Asked Questions

Does hiring a TPA transfer fiduciary liability to them?

No. ERISA keeps the duty to monitor service providers with the plan sponsor, even when claims decisions are delegated to the TPA.

What percentage of claims do TPAs typically get wrong?

Industry-documented error rates on processed claims generally fall between 1% and 6%, depending on the source and plan complexity.

How often should a self-funded plan conduct an independent claims audit?

At least annually, with many advisors recommending quarterly or ongoing monitoring for larger plans.

Can a plan sponsor be sued personally for fiduciary breaches?

Yes. Individuals who exercise discretion over plan administration can face personal liability for breaches of fiduciary duty under ERISA.

Is a sampling audit from the TPA enough to satisfy fiduciary duty?

Sampling audits provide limited assurance. Independent, broader reviews offer stronger documentation of a prudent oversight process.

What is coordination of benefits and why does it matter for audits?

Coordination of benefits (COB) determines which plan pays first when a person has multiple coverages. Errors here often cause plans to overpay.

Does the TPA's size or reputation reduce fiduciary risk?

Not directly. Fiduciary risk is tied to oversight practices and documentation, not the TPA's brand or market share.

What should a benefits committee document to show prudent process?

Meeting minutes, audit reports, corrective action plans and vendor monitoring records, generally retained for at least six years.

‍

‍

‍

‍

‍

‍

A 1,400-employee manufacturer ran its first independent claims audit in 2024 after 18 months of routine TPA reporting had shown nothing unusual. The audit found $812,000 in overpayments, including a $47,000 inpatient claim paid twice, 63 ineligible dependents still active on the plan, and a specialty drug billed at 240% of the contracted rate. None of it had surfaced in the TPA's own reviews.

That gap between what a TPA reports and what an independent audit finds is not an isolated incident. It is a structural feature of how most self-funded plans operate today, and it is one that plan sponsors are increasingly on the hook for.

What Independent Claims Oversight Actually Means

Independent claims oversight is a third-party review of paid medical claims, conducted by an auditor with no financial relationship to the TPA that processed them. Most employers assume their TPA already does this work through its own quality assurance process. That assumption is the core misunderstanding this article addresses.

A TPA's internal audit measures the TPA's own performance using the TPA's own sample, methodology, and reporting standards. That is not oversight. It is self-assessment, and self-assessment has an obvious structural weakness: the entity being reviewed controls what gets reviewed.

Independent oversight replaces that arrangement with an outside party who audits a much larger share of claims, applies clinical and coding expertise the TPA's routine review does not, and reports findings directly to the plan sponsor rather than filtering them through the vendor being evaluated.

Why the Problem Exists

The root cause is not incompetence. Most TPAs process millions of claims accurately and efficiently, and their sampling audits are a legitimate part of normal operations. The problem is scope and incentive, not effort.

TPA sampling audits typically examine a small slice of total claims, often under 5%, leaving the remaining volume unreviewed by anyone with genuine independence. Complex errors such as coordination of benefits failures, upcoding, and unbundling require clinical coding expertise that standard sampling processes are not built to catch at scale.

There is also a quieter incentive dynamic. A TPA that repeatedly identifies its own errors invites tougher contract terms and more scrutiny at renewal. That is simply how vendor relationships work, not an accusation of misconduct, but it explains why relying solely on a vendor's self-reported accuracy leaves a structural blind spot that only independent review can close.

The Real Cost or Impact

The financial exposure from unreviewed claims is measured in real dollars, not theoretical risk. One documented audit of a TPA's claims processing found financial accuracy of 96.8% and payment accuracy of 96.1%, both below the plan's contracted 98% service level agreement, even though the TPA had self-reported 100% accuracy (Baker Tilly). At claims volumes of $10 million to $50 million a year, even a 2% to 5% error rate translates into hundreds of thousands of dollars annually (AIM Benefits).

Coordination of benefits failures are a common driver of this leakage. When a plan pays as primary on a claim that should have been billed secondary, the overpayment on that single claim can run 60% to 80% higher than it should. Multiply that pattern across a workforce of several thousand employees and the leakage compounds every renewal cycle.

There is also a regulatory dimension. EBSA recovered more than $1.4 billion for retirement, health, and welfare benefit plans and their participants in fiscal year 2025, with more than half of that total coming from enforcement actions rather than voluntary correction (U.S. Department of Labor). That figure reflects a broader enforcement environment in which undocumented vendor oversight is increasingly treated as a fiduciary gap, not an administrative footnote.

What's Actually Happening Behind the Scenes

Sampling Instead of Full Review

Most TPA audits work from a statistical sample rather than a full claims review. That approach is efficient for catching obvious errors but is not designed to surface the low-frequency, high-dollar errors, like a duplicate inpatient claim or a specialty drug billed above contract rate, that do the most damage to a plan's budget.

Eligibility Drift

Dependents who age out, spouses who gain other coverage, and employees who terminate but stay active in the claims system for months are a quiet but persistent source of leakage. Eligibility files are rarely reconciled against claims payment in real time, so ineligible claims can accumulate for a full plan year before anyone notices.

Coding Errors That Require Clinical Judgment

Upcoding bills a higher-acuity service code than the documentation supports. Unbundling charges separately for procedures that should be billed under one comprehensive code. Both require someone with clinical and coding expertise to catch, which is exactly the kind of review that high-volume, low-touch sampling processes tend to under-resource.

Why Current Approaches Aren't Enough

Annual TPA-led sampling audits satisfy a contractual checkbox, but they were never designed to serve as fiduciary evidence. The table below shows where the gap sits.

How to Fix It

Red Flags That Signal the Problem Applies to Your Plan

The ROI of Doing It Right

Independent claims audits routinely recover overpayments well above the cost of the audit itself, which is one reason the practice has become standard among large, well-governed plans. Beyond direct dollar recovery, ongoing oversight tends to reduce the error rate in future claims cycles because TPAs adjust processing behavior once they know independent review is a permanent fixture, not a one-time event.

The fiduciary protection value is harder to put a single number on, but it is arguably more important. A plan sponsor who can show continuous, independent review of claims payment occupies a materially stronger position in a DOL inquiry or participant lawsuit than one relying solely on an annual vendor-run sample.

Given that EBSA closed 878 civil investigations in FY 2025, with 556 resulting in repayments or required corrective action (per DOL enforcement data), documented independent oversight is no longer a nice-to-have. It is the evidence a plan sponsor needs on file before anyone asks for it.

Frequently Asked Questions

What is independent claims oversight?

‍A third-party review of paid medical claims by an auditor with no financial ties to the TPA, distinct from a TPA's internal quality checks.

Why can't I rely on my TPA's own audit?

‍A TPA auditing its own claims creates a conflict of interest and typically reviews only a small sample of total volume.

Who is legally responsible for claims errors under ERISA?

‍The plan sponsor, not the TPA, carries fiduciary liability for how the plan's claims are paid.

How often should a self-funded plan run an independent audit?

‍Best practice is quarterly or continuous review, not a single annual engagement, since leakage reappears every plan year.

What is the prudent expert standard?

‍ERISA Section 404 requires fiduciaries to act with the care and skill a knowledgeable expert would use, which independent audits help document.

What kinds of errors do independent audits usually find?

‍Duplicate payments, ineligible dependents, coordination of benefits failures, upcoding, and unbundling are the most common categories.

Does an independent audit replace my TPA relationship?

‍No. It adds an accountability layer alongside the TPA, not a replacement for claims processing services.

What size plan needs independent claims oversight?

‍Any self-funded plan, but the dollar exposure grows quickly for employers with 100 or more covered employees and claims spend in the millions.

‍

‍

A self-funded hospital system trusted its TPA's self-reported claims accuracy of 100%, year after year. An independent audit of a stratified sample of roughly 200 medical claims told a different story.

Financial accuracy landed near 96.8% and payment accuracy at 96.1%, both below the 98% service level agreement in the contract. The plan had data. Nobody outside the TPA had checked whether the data was true.

What Claims Data Transparency Actually Means

Claims data transparency is access to claims pricing, utilization or payment information, not verification that the information is accurate. Plan sponsors often equate transparency with oversight because both involve data. They are not the same function.

Transparency answers "what did we pay." Oversight answers "should we have paid that." A plan can have full access to its claims feed and still have no process for catching a duplicate payment, an ineligible dependent or a specialty drug billed above the contracted rate.

The Transparency in Coverage rule illustrates the gap well. It produced what has been called one of the largest government-mandated data releases in history, yet researchers found the machine-readable files inconsistent in structure, thin on index information and often too large for a standard computer to process. Access without usability is not transparency in any meaningful sense.

Why the Gap Exists

The gap exists because TPAs report their own performance, and self-reporting carries no independent verification. A TPA's contract may guarantee 99% payment accuracy while the TPA's own internal data shows a 1.4% claims processing error rate, a discrepancy that surfaced in one carrier's administrative services agreements. The guarantee and the reality are not required to match unless someone checks.

TPAs also lack a direct financial incentive tied to plan cost outcomes. Their compensation structure is generally built around processing volume and service fees, not around minimizing the plan's total spend, so error prevention competes with throughput.

Most plans compound the problem by reviewing claims infrequently. Annual audits built into ASO agreements often sample only 300 to 350 claims out of hundreds of thousands processed in a year. A snapshot that small can miss systemic issues entirely.

The Real Cost of Assuming Transparency Equals Oversight

Unverified claims errors compound into six- and seven-figure losses for mid-size self-funded plans. Industry claims audit findings put TPA error rates between 3% and 10% of processed claims, well above the 1% to 3% error rate typically cited as an industry baseline.

Independent claims analysis across a large multi-employer dataset found error detection rates of 5% to 15%, with average recoverable findings between $500 and $1,200 per covered employee per year. For a plan with 1,000 covered employees, that range alone represents $500,000 to $1.2 million in annual exposure.

The consequences are not only financial. In September 2025, Aetna and Optum reached an $8.4 million settlement tied to fabricated billing codes that concealed administrative fees inside medical charges for nearly a decade before litigation forced disclosure. Transparency did not surface that pattern. Litigation did.

What's Actually Happening Behind the Scenes

Benefit Determination Errors

The most common category of claims error involves how a plan's own provisions get applied: deductibles, coinsurance, copays and plan-specific exclusions. These errors are rarely intentional. They happen because claims processing blends automated adjudication with manual review, and manual steps introduce inconsistency.

Dependent Eligibility Drift

Dependents who age out, divorce out or otherwise lose eligibility often stay on a plan for months or years because no one is cross-checking enrollment against life events. Dependent eligibility audits are frequently the fastest-paying workstream in a claims review because the savings start the moment ineligible members are removed.

Duplicate and Miscoded Claims

Duplicate payments, upcoding and unbundling do not show up in a pricing dashboard. They show up when someone compares the billed procedure codes against the medical record or the plan's coding logic, a step that transparency tools are not built to perform.

Why Current Approaches Aren't Enough

Most plan sponsors rely on the transparency and reporting the TPA already provides. That approach leaves the same party that processed the claim responsible for verifying it, which is a conflict most plans would not accept in any other financial function.

How to Fix It

Red Flags That Signal the Problem Applies to Your Plan

The ROI of Doing It Right

Independent claims audits routinely recover a meaningful share of the errors they identify, and the ongoing savings from corrected TPA behavior often exceed the one-time recovery. [internal link: claims audit ROI calculator] Dependent eligibility reviews in particular tend to pay for themselves within months, since removing an ineligible dependent stops future claims immediately rather than only recovering past ones.

Beyond dollars, a documented audit and monitoring process is the clearest evidence a plan sponsor has that it met the prudent expert standard under ERISA Section 404. That documentation matters most in the moment a plan can least afford to be without it: a DOL inquiry or a participant lawsuit. EBSA alone recovered $1.4 billion for plans, participants and beneficiaries in fiscal year 2025, a reminder that enforcement activity in this space is active and ongoing, not theoretical.

Frequently Asked Questions

Is claims data transparency required by law?

Yes. The Transparency in Coverage rule and the Consolidated Appropriations Act, 2021 require pricing disclosures, but neither requires payment accuracy verification.

Does transparency alone satisfy ERISA fiduciary duty?

No. ERISA Section 404 requires a prudent process for monitoring plan expenses, which access to data does not fulfill by itself.

How often should a self-funded plan audit its claims?

Most fiduciary advisors recommend an independent audit at least annually, supplemented by ongoing quarterly or monthly monitoring.

What percentage of claims typically contain errors?

Independent audits commonly find error rates between 3% and 10%, above the 1% to 3% often cited as an industry baseline.

Who is legally responsible for claims accuracy, the plan sponsor or the TPA?

The plan sponsor. ERISA places fiduciary responsibility on the plan sponsor even though the TPA processes the claims.

Can a plan sponsor audit claims without the TPA's permission?

Audit rights depend on the ASO agreement. Sponsors should negotiate unrestricted audit access before signing or renewing a contract.

What is the difference between a claims audit and ongoing monitoring?

An audit reviews a sample retrospectively. Ongoing monitoring reviews claims continuously, closer to the time they are paid.

Are machine-readable pricing files useful for claims oversight?

They provide pricing context but are not designed to verify individual claim accuracy, and many are difficult to process without specialized tools.

‍

‍

‍

‍

Fiduciary-grade healthcare operations are the claims oversight, documentation and governance practices a self-funded plan sponsor uses to meet ERISA's prudent expert standard. This includes independent claims audits, documented TPA performance reviews and dependent eligibility checks, rather than relying on a TPA's self-reported accuracy numbers.

In September 2025, Aetna and Optum finalized an $8.4 million settlement over fabricated billing codes that had inflated member costs for nearly a decade before litigation forced disclosure. The case started with one retired plan member who noticed a pattern nobody else was checking.

Most self-funded plan sponsors never look closely enough to find that pattern on their own plan, and industry data suggests TPA error rates run between 1% and 10% of processed claims, depending on the audit methodology used. For a plan paying $20 million a year in claims, that range represents anywhere from $200,000 to $2 million in errors that nobody is actively hunting for.

What Are Fiduciary-Grade Healthcare Operations

Fiduciary-grade healthcare operations are the documented, independently verified claims oversight processes a plan sponsor uses to satisfy ERISA's prudent expert standard, rather than the minimum administrative reporting a TPA provides by default. Most plan sponsors assume their TPA's self-reported accuracy rate reflects reality. It often does not.

A TPA's administrative services agreement typically guarantees financial and payment accuracy in the 98% range. Independent audits regularly find actual accuracy closer to 96% to 97%, a gap that sounds small until it is multiplied across a plan paying eight or nine figures in annual claims. The difference between "TPA-reported" and "independently verified" is the entire fiduciary question.

Fiduciary-grade operations are not a single product or vendor. They are a standing practice: audit cadence, documentation, and a governance record that shows the plan sponsor acted as a prudent expert would, consistent with the standard set out in ERISA Section 404.

Why the Oversight Gap Exists

The oversight gap exists because TPAs have no direct financial stake in the plan's cost outcomes, and most plan sponsors lack the internal expertise to audit claims data themselves. A TPA is paid a fee to process claims, not a percentage of savings, so accuracy beyond the contracted service level agreement carries little upside for the TPA.

Plan sponsors, meanwhile, are often HR or finance leaders managing benefits as one responsibility among many. Building the coding expertise needed to catch upcoding, unbundling, or duplicate billing in-house is rarely realistic for a mid-size employer.

Restrictive audit clauses compound the problem. Many TPA contracts historically limited the number of claims a sponsor could audit, barred extrapolation of findings, or restricted which outside firms could perform the audit, though state laws such as Maine's L.D. 1906, effective for contracts issued or renewed after January 1, 2026, are starting to prohibit those restrictions.

The Real Cost of Skipping Fiduciary-Grade Oversight

Skipping independent claims oversight costs plans real money and creates fiduciary liability that follows the plan sponsor personally. A full-scope claims audit with 100% review, rather than a sampled TPA self-audit, typically recovers between 1% and 3% of annual claims spend, based on industry-reported audit outcomes.

On the enforcement side, the Department of Labor's Employee Benefits Security Administration recovered $1.4 billion for retirement, health, and welfare plans in fiscal year 2025, with $714.4 million of that tied directly to civil enforcement actions. EBSA closed 878 civil investigations that year alone, a volume that signals oversight failures are common, not rare.

Litigation risk is following the same curve. The 401(k) excessive-fee lawsuit model, which produced more than $10 billion in settlements over the past decade, is now being applied to health plan administration, with plaintiffs' firms using DOL enforcement patterns as a roadmap for building breach-of-fiduciary-duty claims against plan sponsors.

What's Actually Happening Behind the Scenes

Sampled Audits Miss Most of the Plan

A standard TPA audit typically reviews a stratified sample of 250 to 400 claims out of hundreds of thousands processed annually, then extrapolates an overall accuracy score from that small slice. The unreviewed remainder, often more than 95% of total claims, never gets examined at all.

Coding Errors Hide in High-Dollar Claims

Upcoding, unbundling, and duplicate billing tend to concentrate in complex, high-dollar claims, exactly the claims a small random sample is least likely to catch. Specialty pharmacy and inpatient facility claims carry disproportionate risk for this reason.

Dependent Eligibility Drifts Without Anyone Noticing

Ex-spouses, aged-out children, and other ineligible dependents routinely remain on plan rolls for years because eligibility verification is rarely built into ongoing claims workflows. Each ineligible dependent represents ongoing claims spend the plan should never have paid.

Automated Adjudication Approves Errors at Scale

A growing share of claims are approved through automated adjudication with no human review, which processes errors just as efficiently as it processes legitimate claims. Duplicate claims and coordination-of-benefits failures are especially prone to slipping through automated systems undetected.

Why Current Approaches Aren't Enough

Most plans still treat claims oversight as a periodic compliance task instead of an ongoing operational discipline. The table below shows how that status quo compares with a fiduciary-grade approach.

How to Build Fiduciary-Grade Healthcare Operations

Red Flags That Signal Your Plan Needs This Now

The ROI of Doing It Right

Independent, full-scope claims audits typically recover 1% to 3% of annual claims spend in direct overpayment findings, and dependent eligibility audits frequently pay for themselves within months of completion.

Beyond direct recovery, fiduciary-grade documentation is itself a form of risk transfer. A plan sponsor that can produce board minutes, audit reports, and a documented TPA review process has evidence of the prudent expert standard that ERISA requires, which matters considerably if EBSA or a plaintiff's firm ever comes asking.

Think of it the way a building owner thinks about fire code compliance. The inspection itself does not prevent every fire, but the documented compliance record is what protects the owner when something goes wrong anyway. Claims audit documentation functions the same way for plan fiduciaries.

Conclusion and Next Steps

Fiduciary-grade healthcare operations are becoming the baseline expectation, not a differentiator, as EBSA enforcement activity and health plan litigation both continue climbing. Plan sponsors that wait for a regulator or a plaintiff's attorney to ask the first question have already lost the advantage of showing a prudent, documented process.

Start with an independent audit scoped to review all claims, not a sample, and build the documentation habit alongside it.

Frequently Asked Questions

What does "fiduciary-grade" mean for a health plan?

It means claims oversight and documentation practices that meet ERISA's prudent expert standard, not just a TPA's minimum reporting.

Who is legally responsible for claims accuracy on a self-funded plan?

The plan sponsor, under ERISA Section 404, regardless of which TPA processes the claims.

How often should a self-funded plan conduct a claims audit?

Independent audits should happen continuously or quarterly, not once every two to three years.

What percentage of claims contain errors?

Industry-documented TPA error rates range from roughly 1% to 10%, depending on audit scope and methodology.

How much money can a claims audit typically recover?

Full-scope, 100% claims audits typically recover 1% to 3% of annual claims spend.

Can a TPA restrict how a plan sponsor audits its own claims?

Some contracts historically did, though new state laws are increasingly prohibiting these restrictions.

Should a plan sponsor use the TPA's recommended audit firm?

No. Independent firms with no ownership or referral ties to the TPA reduce conflict-of-interest risk.

What is the fastest way to reduce fiduciary risk right now?

Document the plan's governance process, including committee minutes and retained audit reports, alongside commissioning an independent audit.

‍

‍

‍

‍

‍

‍

‍

‍

‍

Payment integrity is a technical process that checks whether individual claims were coded and priced correctly. Fiduciary intelligence is a governance framework that documents whether the plan sponsor prudently monitored its vendors, as ERISA Section 404 requires. Payment integrity catches claim errors. Fiduciary intelligence protects the people legally responsible for the plan.

Most self-funded plans confuse a claims tool with a legal defense, and the gap between them is where lawsuits start.

A mid-size manufacturer with 1,400 employees paid $340,000 in claims for a coordination of benefits failure that ran undetected for eighteen months (Willis Towers Watson). The plan's payment integrity vendor never flagged it because the claims were coded correctly.

The problem wasn't accuracy. It was that nobody on the plan sponsor's side could show they had monitored the TPA's handling of COB determinations at all. That gap, between a clean claim and a documented oversight process, is the entire subject of this article.

Why Payment Integrity Is Not the Same as Fiduciary Responsibility

Sixty-seven percent of covered workers in the United States are enrolled in self-funded health plans, and that share climbs to 80 percent among firms with 200 or more employees (Kaiser Family Foundation, 2025).

Every one of those employers carries fiduciary duties under the Employee Retirement Income Security Act (ERISA) regardless of whether they know it. Payment integrity is one useful tool for managing plan spend.

It is not the same thing as meeting that legal duty, and treating the two as interchangeable leaves plan sponsors exposed.

What Payment Integrity Actually Does

Payment integrity is the set of technical processes and software that verify whether a submitted claim was billed, coded, and priced correctly before or after payment. It answers a narrow question: did the plan pay the right dollar amount for this specific service.

Most people assume payment integrity is a comprehensive safeguard for the plan. It is not designed to be one. It is a claims-accuracy layer, built to catch duplicate billing, upcoding, and pricing errors against contracted rates. The global payment integrity market has grown to roughly $9 billion, expanding at about 7 percent annually as payers invest more heavily in automated claims review (McKinsey).

The tools are genuinely useful. They catch a meaningful share of overpayments that would otherwise slip through automated adjudication. What they do not do is document whether the plan sponsor exercised the kind of ongoing, independent oversight that ERISA requires of a fiduciary.

A payment integrity report can show a clean claims file and still leave a plan sponsor unable to answer basic questions about vendor monitoring, fee reasonableness, or conflict of interest review.

Why the Gap Exists

The gap between payment integrity and fiduciary protection exists because most self-funded plans were designed for administrative efficiency, not employer oversight. Standard administrative services agreements typically treat a small sampling audit as the complete review, and plan sponsors accept it because it has been considered the industry standard for decades.

TPAs may have limited incentives to identify their own payment errors. Industry studies report claims processing error rates of 2% to 6%, depending on the source and plan complexity (Baker Tilly). At the same time, many TPAs report self-audit accuracy rates close to 99% (ClaimInformatics). This does not necessarily mean either number is wrong. They measure different things. Self-audits often focus on processing accuracy and system rules, not whether the correct amount was ultimately paid.

There is also a structural asymmetry in who has access to the data. Many TPAs release detailed claims files only upon request, often in formats that require technical expertise to analyze (Benosphere). Under ERISA Section 404(a)(1)(B), the fiduciary must act "with the care, skill, prudence, and diligence" of a prudent expert (U.S. Department of Labor). It is difficult to meet that standard when the party being monitored controls the data used to monitor it.

The Real Cost of Treating Payment Integrity as Sufficient

Self-funded employers spending between $10 million and $100 million annually in healthcare claims face $200,000 to $2 million in unrecovered leakage under even a conservative 2 percent error rate (ClaimInformatics). Payment integrity tools recover some of that. Independent claims audits with full review typically recover an additional 1 percent to 3 percent of annual claims spend on top of whatever the TPA's own systems caught (Benosphere).

The financial cost is real but it is not the largest exposure. The fiduciary cost is. ERISA fiduciaries who breach their duties can be held personally liable to restore plan losses, and courts have referred to fiduciary obligations under the statute as among "the highest known to law" (ASPPA). A plan sponsor that relied entirely on a TPA's self-reported payment integrity metrics, with no independent verification, has a documentation problem the moment a participant or the Department of Labor asks how oversight was performed.

Consider the analogy of a building inspector who only checks whether individual bricks meet code. That inspector can sign off on every brick and still miss that the foundation was never surveyed. Payment integrity checks bricks. Fiduciary intelligence checks whether anyone verified the foundation.

What's Actually Happening Behind the Scenes

Sampling Gaps

A standard TPA audit reviews a stratified sample, commonly 200 to 400 claims, against plan documents (Baker Tilly). For a plan processing 80,000 claims annually, that represents roughly 0.3 percent to 0.5 percent of total claims volume (Benosphere). The remaining 99.5 percent goes unreviewed by anyone independent of the TPA.

Coordination of Benefits Failures

COB errors occur when a plan pays as primary when another payer, such as Medicare or a spouse's plan, should have paid first. A COB failure can produce a 60 percent to 80 percent overpayment on the affected claim (Benosphere), and these errors are difficult for automated payment integrity tools to catch because the claim itself may be coded correctly.

Vendor Fee Structures on Recovered Dollars

When TPAs or carrier-affiliated vendors do identify and recover overpayments, they frequently retain a substantial share. Post-pay recovery programs commonly take 40 percent to 50 percent of recovered dollars (ClaimInformatics), which means even a functioning payment integrity process may return less value to the plan than the raw recovery number suggests.

Governance Documentation

DOL guidance emphasizes that fiduciaries must document their decision-making process, not simply achieve a good outcome (DOL elaws Fiduciary Advisor). A payment integrity dashboard, no matter how sophisticated, is not itself governance documentation unless the plan sponsor can show it was reviewed, questioned, and acted upon.

Why Current Approaches Aren't Enough

How to fix it

Red Flags That Signal the Gap Applies to Your Plan

The ROI of Doing It Right

Independent claims audits with full review typically find discrepancies in 3 percent to 5 percent of paid claims costs, with actual recoveries in the 1 percent to 2 percent range (MedInsight). On a $30 million annual claims spend, that is $300,000 to $600,000 in identified overpayments, with $300,000 to $600,000 recoverable even under conservative assumptions.

The larger return is harder to put a number on but matters more in a dispute. A documented fiduciary process, maintained consistently across plan years, is described by the DOL as the strongest defense against audits and litigation (Ascensus). Plans that can produce that record are simply in a different legal position than plans that cannot, regardless of how their payment integrity metrics look in isolation.

Dependent eligibility reviews, treated as a distinct workstream from claims audits, frequently pay for themselves within months (Benosphere) and are one of the fastest ways to demonstrate near-term ROI while the broader fiduciary documentation process is being built out.

Conclusion and Next Steps

Payment integrity and fiduciary intelligence solve different problems, and a self-funded plan needs both. Payment integrity keeps individual claims accurate. Fiduciary intelligence protects the plan sponsor and named fiduciaries by documenting that oversight actually happened, in the way ERISA requires.

The gap between the two is exactly where fiduciary breach claims originate. Closing it does not require replacing your payment integrity vendor. It requires building a separate, deliberate governance record alongside it. [internal link: TPA performance guarantees guide] can help you evaluate whether your current vendor contract measures the right things, and [internal link: claims audit vs payment integrity comparison] walks through how to structure both functions without duplicating cost.

Frequently Asked Questions

Is payment integrity the same as fiduciary compliance?

‍No. Payment integrity checks individual claims accuracy. Fiduciary compliance requires documented, independent oversight of vendors under ERISA Section 404.

Who is legally responsible for claims errors, the TPA or the plan sponsor?

‍The plan sponsor. ERISA places fiduciary responsibility on the sponsor, not the TPA, even though the TPA processes the claims.

What does ERISA Section 404 actually require?

‍It requires fiduciaries to act with the care, skill, prudence, and diligence of a prudent expert familiar with plan administration.

How often should a self-funded plan run an independent claims audit?

‍Industry guidance generally recommends every two to three years, with more frequent monitoring for larger or higher-risk plans.

Can a plan sponsor rely on a TPA's self-reported accuracy numbers?

‍Relying solely on self-reported data, without independent verification, does not demonstrate the prudent oversight ERISA requires.

What percentage of claims does a typical TPA audit actually review?

‍Often well under 1 percent of total claims volume, since standard TPA audits use small stratified samples.

Does the Consolidated Appropriations Act change plan sponsor audit rights?

‍Yes. It strengthened plan sponsors' rights to full claims data access from TPAs and PBMs upon request.

What is the single most common dollar-weighted claims error?

‍Duplicate billing, followed closely by coordination of benefits failures on claims that should have paid secondary.

‍

Fiduciary intelligence is the independent claims oversight layer that sits between a third-party administrator (TPA) and a self-funded employer. It combines ongoing claims monitoring, audit documentation and vendor accountability data so plan sponsors can meet ERISA Section 404 duties instead of relying solely on the TPA's self-reported performance numbers.

A regional manufacturer with 1,400 employees ran an independent claims audit in 2024 and found $812,000 in overpayments across 18 months. The errors included a $47,000 inpatient claim paid twice, 63 ineligible dependents still active on the plan, and a specialty drug billed at 240% of the contracted rate. None of it showed up in the TPA's internal reporting, because none of it was ever reviewed.

That is not a rare story. It is the default outcome for a self-funded plan that treats its TPA's word as the audit. Most employer plans have no independent layer checking whether claims were paid correctly, and by the time anyone notices, the money is gone and the fiduciary exposure has already accrued.

What Fiduciary Intelligence Actually Means

Fiduciary intelligence is the independent oversight layer that verifies TPA claims performance instead of trusting it. Most employers assume their TPA relationship already includes this. It does not.

A TPA's job is to process claims according to plan documents and network contracts. A TPA's incentive is speed and member satisfaction, not necessarily financial accuracy on every dollar. Those two things frequently pull in different directions, and nothing in a standard administrative services agreement forces alignment.

The common assumption is that performance guarantees in the ASO contract already cover this ground. In reality, most guarantees measure turnaround time and claims-processing speed, not whether the dollar amount paid was correct. A TPA can hit every service level target in its contract while still overpaying claims at a rate the plan sponsor never sees.

Why the Oversight Gap Exists

The gap exists because plan sponsors delegate claims processing but rarely build independent verification into the relationship. Self-funded plans took off because employers wanted to bend the cost curve and gain flexibility insurance carriers do not offer. What did not scale at the same pace was internal expertise to monitor what a TPA actually does with that authority.

TPAs are not financially responsible for the plan. Their costs are covered by administrative fees, not by how accurately claims are paid, so they lack the same financial incentive an insurer carrying its own risk would have. That is not a matter of bad faith. It is simply a structural incentive problem plan sponsors need to correct with independent checks, not TPA good intentions.

Carrier and TPA post-pay sampling reviews typically cover only 3% to 5% of claims, and many ASO agreements specify an annual sample of just 300 to 350 claims regardless of plan size. A plan processing 200,000 claims a year can have 99.8% of its payments never independently reviewed by anyone outside the TPA that made the payment.

The Real Cost of an Unreviewed Plan

Unreviewed claims translate directly into unrecovered dollars, and the scale is larger than most benefits committees assume. Independent, full-population claims analysis consistently identifies error rates between 5% and 12% once every claim is checked instead of a sample, according to third-party claims analytics data covering more than $16 billion in reviewed claims. Industry benchmarks that rely on standard TPA sampling report a narrower 1% to 3% error range, largely because sampling catches fewer error types than a comprehensive review.

The dollar impact compounds quickly. A plan spending $20 million a year that carries even a 2% unrecovered error rate is looking at $400,000 walking out the door annually, before accounting for coordination of benefits failures or dependent eligibility errors that carry their own separate cost.

Consider the manufacturer's numbers again. Sixty-three ineligible dependents sitting on a plan for even one plan year can add tens of thousands of dollars in claims paid for people who should never have been covered. A single misapplied coordination of benefits rule, where the plan should have paid secondary but paid as primary instead, often produces a 60% to 80% overpayment on that specific claim.

What's Actually Happening Behind the Scenes

Coordination of Benefits Failures

Coordination of benefits, or COB, determines which plan pays first when a member has more than one source of coverage. When a TPA's system fails to catch a spouse's other employer coverage or a dependent's eligibility for a separate plan, the self-funded plan can end up paying as primary when it should be secondary. That single misconfiguration routinely produces overpayments in the 60% to 80% range on the affected claims.

Upcoding and Unbundling

Upcoding happens when a provider bills a higher-acuity procedure code than the service actually supports. Unbundling separates a single comprehensive procedure into multiple line items billed individually, inflating the total. Both require clinical and coding expertise to catch, which is exactly why standard TPA sampling rarely flags them.

Dependent Eligibility Drift

Employees change marital status, dependents age out, and COBRA windows close, but eligibility files do not always update on schedule. A plan that has not run a dependent eligibility verification in several years is very likely still paying claims for people who are no longer eligible for coverage.

Out-of-Network and Surprise Billing Gaps

Even after the No Surprises Act took effect, out-of-network claims can still slip through with billed charges well above usual and customary rates when a plan is not actively reviewing them. Without active oversight, the plan simply pays whatever the TPA's repricing engine calculates, correct or not.

Specialty Pharmacy Spend

Specialty drugs now account for more than half of pharmacy spend on many self-funded plans, and pricing errors in this category carry outsized dollar impact per incident compared to medical claims. A single misapplied contract rate on a specialty drug claim can run into tens of thousands of dollars.

Why Current Approaches Aren't Enough

Standard TPA performance guarantees were built to measure operational speed, not financial accuracy, and that mismatch is the core problem plan sponsors need to solve.

How to Fix It

Red Flags That Signal the Problem Applies to Your Plan

The ROI of Doing It Right

A full independent claims audit typically recovers 1% to 3% of annual claims spending on its first pass. For a plan spending $20 million a year, that translates to $200,000 to $600,000 recovered, often within the first audit cycle. Ongoing monthly or quarterly monitoring tends to catch errors closer to the point of payment, which both prevents future leakage and creates a documented accountability trail with the TPA.

Audit costs are typically far lower than the amount recovered, particularly for mid-sized and large plans. Beyond the direct dollar recovery, the process strengthens vendor negotiating position at renewal and produces the kind of documentation that demonstrates a prudent fiduciary process, which matters enormously if the Department of Labor's Employee Benefits Security Administration ever opens an inquiry.

EBSA recovered more than $1.4 billion for retirement, health and welfare plans in fiscal year 2025 alone, closing 878 civil investigations with 556 producing monetary or corrective results.

Conclusion and Next Steps

Self-funded plans now cover the majority of American workers with employer health benefits, and that share keeps growing. Every one of those plans carries fiduciary duties that do not pause just because a TPA is handling the paperwork. Fiduciary intelligence, in the form of independent claims oversight, is what actually closes that gap between delegation and accountability.

The next step is straightforward. Pull your ASO agreement and check the audit rights clause, then ask your benefits committee when the plan's claims were last independently reviewed. If nobody has a confident answer, that is the signal to schedule a claims audit and dependent eligibility review before the next renewal cycle.

Frequently Asked Questions

What is fiduciary intelligence in the context of self-funded health plans?

‍It is the independent oversight layer, combining claims monitoring and audit documentation, that verifies TPA performance rather than relying on the TPA's own reporting.

Who holds fiduciary responsibility for claims accuracy under ERISA?

‍The plan sponsor, under ERISA Section 404, regardless of which vendor actually processes and pays the claims. [external link: DOL/EBSA fiduciary responsibilities guide]

How often should a self-funded plan audit its TPA?

‍At minimum every one to two years for larger plans, with ongoing monthly or quarterly monitoring recommended between formal audits.

What percentage of claims does a typical TPA sampling audit review?

‍Roughly 3% to 5%, often limited to a fixed sample of 300 to 350 claims per year regardless of total plan size.

How much can an independent claims audit typically recover?

‍Most first-time audits recover 1% to 3% of annual claims spend, which can total hundreds of thousands of dollars on mid-sized plans.

What is the difference between a TPA performance guarantee and a fiduciary audit?

‍Performance guarantees measure speed and procedural accuracy; a fiduciary audit measures whether the dollar amount paid was actually correct.

Can a plan sponsor be held personally liable for TPA errors it never discovered?

‍Yes. ERISA fiduciary duty requires a prudent ongoing process, and failing to monitor a TPA can itself constitute a breach regardless of who made the underlying error.

What is coordination of benefits and why does it matter for claims accuracy?

‍COB determines which plan pays first when a member has multiple coverage sources; failures here commonly cause 60% to 80% overpayments on affected claims.

‍

‍

The claims that shape a self-funded plan's renewal are rarely the flagged, high-dollar outliers a TPA reviews closely. They are the routine mid-range claims, often between $500 and $15,000, that pass through standard adjudication unchecked. Left unaudited, this volume of small errors accumulates into the claims trend data insurers use to set renewal pricing.

A self-funded employer with a $30 million annual claims spend trusts its TPA's reported 99% payment accuracy rate at face value. An independent audit later shows the real financial accuracy sits closer to 96.8%, with payment accuracy at 96.1% (Baker Tilly, 2026).

That two to three point gap on a plan this size can mean hundreds of thousands of dollars in overpayments the plan never saw coming. None of those errors showed up as outliers. They were ordinary, mid-range claims that simply moved through the system.

What "The Claims Nobody Flags" Actually Means

The claims that damage a self-funded plan financially are not the ones anyone is watching. Most plan sponsors assume risk lives in the big, obvious claims: the $400,000 transplant, the six-figure NICU stay, the catastrophic case that triggers stop-loss reimbursement. Those claims get scrutiny by default because the dollar amount forces it.

The claims nobody flags are different. They sit in the $500 to $15,000 range, look ordinary on their face, and process through standard adjudication rules without triggering any manual review. A duplicate physical therapy session, a specialty drug billed slightly above the contracted rate, a coordination-of-benefits error where the plan paid as primary instead of secondary: none of these look alarming individually.

The reality is that volume beats size. A single $400,000 claim gets audited. Ten thousand $1,200 claims with a 3% error rate do not, and that 3% adds up to real money moving out the door every single month.

Why This Problem Exists in the First Place

TPAs are not financially incentivized to catch every small error, because the cost of the error falls on the plan, not on them. This is the structural root of the issue. A TPA's contract typically ties performance guarantees to speed and procedural accuracy, not to overpayment rates.

Claims adjudication systems are built for throughput. Prompt payment requirements common in TPA contracts, often 21 to 30 days from receipt, push claims through fast rather than carefully. The faster a claim moves, the less time anyone spends confirming eligibility, coordination of benefits, or contracted pricing before payment goes out.

Add to this a sampling problem. Most routine post-payment audits built into Administrative Services Only agreements review a small, fixed sample, commonly 300 to 350 claims per year regardless of total plan volume. On a plan processing 40,000 claims annually, that sample size touches well under 1% of total activity, which means the other 99% simply goes unchecked.

The Real Cost and Impact on Plan Spending

Unflagged mid-range claims errors are not rounding errors. They are a measurable percentage of total plan spend, year after year. Industry benchmarks place typical TPA claims error rates between 1% and 3% of total claims processed, with some sources citing a wider 2% to 6% range depending on plan complexity and TPA maturity (WTW, 2025; Baker Tilly, 2026).

A peer-reviewed comparison of audit methodologies analyzed claim data from two Fortune 100 corporations and found that random-sample audits failed to catch a meaningful share of errors that, in aggregate, ranged from $200,000 to $750,000 per plan.

The same research concluded that reviewing 100% of claims rather than a small sample surfaced significantly more of these errors. That is not a marginal difference. That is the gap between a plan that thinks it is clean and a plan that is quietly bleeding.

Think of it like a slow leak in a tire rather than a blowout. A blowout, the catastrophic claim, gets immediate attention because it is impossible to ignore. A slow leak, the routine claims error repeated thousands of times, goes unnoticed until the plan sponsor is standing at the renewal table wondering why the trend line moved.

What's Actually Happening Behind the Scenes

Duplicate and Resubmitted Claims

Duplicate billing remains one of the most common dollar-weighted error types found in claims audits. It typically happens when a provider resubmits a claim, when a plan migrates administrative systems, or when a secondary insurer's payment is not properly coordinated with the plan's own payment.

Coordination of Benefits Failures

When a plan pays as primary on a claim where another carrier should have paid first, a common scenario for dependents with other coverage or spouses on a second plan, the overpayment on that single claim can run significantly higher than it should. These errors require cross-referencing eligibility data that standard adjudication rarely checks in real time.

Upcoding and Unbundling

Upcoding bills a higher-acuity procedure code than the service actually supports. Unbundling charges separately for components of care that should be billed under one comprehensive code. Both inflate provider revenue at the plan's expense, and both require clinical and coding expertise to catch, which is exactly why a standard TPA sample audit rarely flags them.

Why Current Audit Approaches Aren't Enough

A once-a-year sample audit checks whether the TPA followed procedure. It does not tell a plan sponsor whether the plan actually lost money. These are two different questions, and most benefits committees only ever get an answer to the first one.

How to Fix It: A Practical Action Plan

Red Flags That Signal Your Plan Has This Problem

The ROI of Getting Claims Oversight Right

A full independent claims audit typically recovers between 1% and 3% of annual claims spending in its first year. For a plan spending $20 million annually, that translates to $200,000 to $600,000 in direct recoveries, often exceeding the entire cost of the audit itself many times over.

Beyond direct recovery, the ongoing value compounds. Ongoing monitoring catches new errors before they repeat across thousands of claims, and it builds a documented, defensible fiduciary process that protects plan sponsors named personally in ERISA litigation.

There is also a renewal-specific benefit. A plan sponsor walking into a stop-loss or TPA renewal conversation with independently verified claims data negotiates from evidence, not assumption, which shifts leverage back toward the employer.

Conclusion and Next Steps

The claims that quietly reshape a self-funded plan's renewal are not the ones anyone was watching. They are the ordinary, mid-range claims moving through standard adjudication every day, unflagged and unaudited. A self-funded health plan claims audit is the single most direct way to find out what is actually happening inside that volume before the renewal conversation forces the question.

Plan sponsors carry fiduciary responsibility for this outcome whether or not they have the visibility to act on it. The next step is straightforward: confirm your audit rights, move from annual sampling toward ongoing monitoring, and bring independently verified claims data to your next renewal instead of relying solely on your TPA's own reporting.

Frequently Asked Questions

What is a self-funded health plan claims audit?

An independent review of paid claims to verify accuracy, eligibility, and contract compliance beyond what the TPA reports internally.

How often should a self-funded plan audit its claims?

Most large plans benefit from an annual independent audit paired with ongoing monthly or quarterly monitoring between full reviews.

What is a typical TPA claims error rate?

Industry benchmarks generally place TPA error rates between 1% and 3% of total claims, with some sources citing up to 6%.

Why does the plan sponsor bear fiduciary risk instead of the TPA?

ERISA assigns fiduciary responsibility to whoever exercises discretionary control over the plan, and sponsors typically retain that role even when a TPA administers claims.

Can a claims audit really pay for itself?

Yes. Recoveries of 1% to 3% of annual claims spend routinely exceed audit costs, especially for mid-sized and large plans.

What is the most common type of claims error?

Duplicate billing is consistently cited as the most common dollar-weighted error category in independent audits.

Does a small sample audit satisfy ERISA's fiduciary standard?

A sample audit alone may not demonstrate a fully prudent process, since it leaves the large majority of claims unreviewed.

What should a plan sponsor look for in an independent audit firm?

Look for coding and clinical expertise, no ownership ties to the TPA, and a fee model tied to actual findings rather than flat retainer billing.

‍

‍

‍

‍

Claims oversight for a self-funded health plan means independently verifying that claims were paid correctly under the negotiated contract, not just that a rate was negotiated. Most TPAs self-report accuracy above 96%, but independent audits routinely find 1% to 10% of claims dollars paid in error.

A 1,400-employee manufacturer ran its first independent claims audit in 2024, eighteen months into a new TPA relationship. The audit found $812,000 in overpayments, including a $47,000 inpatient claim paid twice and a specialty drug billed at 240% of the contracted rate. None of it had surfaced in the TPA's own reporting.

This is the blind spot inside most self-funded employer health plans. Benefits committees spend months negotiating stop-loss terms, network discounts, and PBM rebates, then hand claims payment entirely to a TPA and stop watching.

Sixty-seven percent of covered workers, including 80% at large firms, are now enrolled in self-funded plans, according to KFF's 2025 Employer Health Benefits Survey. That is a lot of employer money moving through systems almost nobody independently checks.

Negotiated Pricing Is Not the Same as Paid Correctly

A negotiated rate protects a plan only if each claim is processed using the correct contract, and most plans have no independent way to verify that. Many plan sponsors assume that once a discount or fee schedule is negotiated, every claim is automatically paid at that rate. It is not. Negotiating the price determines what the plan should pay. Claims adjudication determines what the plan actually pays.

Pricing is a static number in a contract. Payment is a live process running through claims adjudication software, manual review queues, coordination of benefits logic, and provider billing codes, any one of which can override the contracted rate without anyone noticing.

A negotiated 40% network discount means nothing on a claim that was coded wrong, paid to the wrong coordination order, or processed against an expired fee schedule.

Why the Gap Exists

The gap exists because TPAs are not financially exposed when claims payments go wrong, so accuracy checking rarely gets the same rigor as claims processing speed. Under most administrative services agreements, the TPA processes claims and gets paid a fee regardless of whether the plan overpaid. The plan bears the financial risk. The TPA does not.

That misalignment shows up in what gets measured. TPA performance guarantees typically track turnaround time and procedural compliance rather than dollar accuracy. A TPA can hit every service level target in its contract while still paying claims incorrectly, because the contract was never written to test for that.

Reporting compounds the problem. TPA accuracy figures are usually self-reported, while independent claims audits often identify payment errors that internal reviews miss. Because self-reported accuracy measures vary by methodology, an independent review is frequently the only way to verify whether claims were processed according to plan documents, provider contracts and payment rules. Studies of claims accuracy show that audit methodology can materially affect reported results.

The Real Cost of Unreviewed Claims

Unreviewed claims create measurable financial losses for self-funded plans, and those losses grow as plan spending increases. Industry benchmarks estimate that claims payment errors typically affect 1% to 3% of total claims dollars, while some comprehensive independent audits have identified substantially higher error rates when 100% of claims are reviewed instead of a small sample.

Even a modest error rate can translate into hundreds of thousands of dollars in unnecessary spending each year for larger self-funded plans.

Coordination of benefits (COB) failures are among the most expensive claims payment errors. When a self-funded plan pays as the primary insurer on a claim that should have been paid second, it may cover costs that another health plan was responsible for first.

The resulting overpayment on a single claim can equal 60% to 80% of the amount your plan paid. Across a workforce with employees and dependents covered by multiple health plans, even a small number of COB errors can create substantial unnecessary costs over the course of a plan year.

What's Actually Happening Behind the Scenes

Duplicate Payments

Duplicate billing is consistently the highest dollar-weighted error category found in claims audits. It happens when a provider resubmits a claim, when a TPA migrates claims systems, or when a secondary payer's response is not reconciled against the primary payment. The same procedure code, same date of service, and same patient can get paid twice without triggering an internal flag.

Eligibility Drift

Plans routinely keep paying claims for people who are no longer eligible, and nobody catches it until an audit runs a full eligibility reconciliation. Terminated employees, dependents who aged out, and spouses who gained other primary coverage all continue generating paid claims until someone cross-checks eligibility files against claims data, which most TPAs are not contractually required to do proactively.

Coordination of Benefits Errors

When a plan should pay secondary but pays as primary instead, the overpayment on that claim is disproportionately large. This failure mode is common where employees have working spouses with their own coverage, and it is one of the few error categories where a single claim can carry a five- or six-figure correction.

Contract Misapplication

Even a well-negotiated rate schedule can be misapplied if the adjudication system was never updated to reflect it. Specialty drug pricing, out-of-network reimbursement caps, and site-of-service differentials are common places where the contracted rate and the paid rate quietly diverge, sometimes for months before anyone notices.

Why Current Approaches Aren't Enough

Most plans rely on the TPA's own performance guarantees and periodic sample audits to confirm accuracy. Both approaches were built to measure process, not dollars, and both leave the plan sponsor with an incomplete picture of what was actually paid.

How to fix it

Red Flags That Signal the Problem Applies to Your Plan

The ROI of Doing It Right

Independent claims oversight typically pays for itself well beyond its cost, both in direct recovery and in reduced fiduciary risk. A first-year comprehensive audit commonly recovers 1% to 3% of annual claims spend, and audit fees are generally far smaller than the amounts identified, particularly for plans in the mid-size to large range.

There is also a fiduciary dimension that does not show up on a savings spreadsheet. A documented, defensible oversight process is exactly what ERISA's duty of prudence requires, and it is the kind of record that matters if a plan is ever the subject of a DOL inquiry or a participant lawsuit.

There is also a fiduciary dimension that does not show up on a savings spreadsheet. A documented, defensible Independent claims oversight process is exactly what ERISA fiduciary responsibilities require, and it is the kind of record that matters if a plan is ever the subject of a DOL inquiry or a participant lawsuit. Courts evaluating an ERISA duty of prudence place significant weight on the fiduciary's decision-making process and documentation.

Conclusion and Next Steps

Negotiating strong contract terms is necessary but not sufficient. A self-funded health plan's real financial exposure lives in the gap between what was negotiated and what was actually paid, and that gap only closes with independent claims oversight. The plans absorbing $812,000 overpayment findings are not unusual. They are simply the ones who finally looked.

Start with a scoping conversation about your plan's current audit rights, your TPA's self-reported accuracy figures, and when your claims data was last independently reviewed. If the honest answer is "we're not sure," that uncertainty is itself the finding your benefits committee needs to document and address.

Frequently Asked Questions

Is a self-funded employer legally required to audit its health plan claims?

ERISA does not mandate a specific audit schedule, but it does require fiduciaries to prudently monitor service providers, including TPAs. Without some form of independent verification, a plan sponsor has no reliable way to demonstrate that monitoring duty was fulfilled.

How often should a self-funded plan conduct a claims audit?

Most claims audit specialists recommend an annual review for larger plans and at least a biennial review for smaller ones. Plans that change TPAs, PBMs, or plan designs should also audit at the transition point before historical claims become harder to access.

What is a normal TPA claims error rate?

Industry benchmarks generally place administrator error rates between 1% and 3% of total claims dollars in routine operations, with some independent full-scope audits finding rates as high as 10% once every claim is reviewed rather than a small sample.

Who is liable if a TPA pays a claim incorrectly?

The plan sponsor generally carries fiduciary and financial responsibility, even though the TPA processed the claim. A contract with the TPA does not transfer ERISA fiduciary liability away from the plan sponsor.

How much does an independent claims audit typically cost?

Costs vary by plan size and audit scope, but many claims audit firms work on a contingency or hybrid fee tied to recovered dollars. Fees are typically well below the amount recovered, especially for mid-sized and large plans.

Can a TPA restrict how much of the claims data a plan is allowed to audit?

Some TPA contracts include language limiting audits to a random sample. Plan sponsors can and should negotiate broader audit rights, including the ability to review any claim at any time using an auditor of their choosing.

What is the difference between a TPA performance guarantee and a claims audit?

A performance guarantee measures whether the TPA hit contractual service metrics, typically speed and procedural compliance. A claims audit independently verifies whether the dollar amount paid on each claim matches what the plan document and contract actually require.

Does the Department of Labor actually investigate self-funded health plans over claims issues?

Yes. EBSA closed 878 civil investigations in FY 2025 and recovered $1.4 billion across enforcement and informal resolution, and has named health and welfare plan oversight a national enforcement priority for FY 2026.

‍

‍

‍

‍

‍

The healthcare claims oversight gap refers to medical claims, usually between $5,000 and $25,000, that often receive little or no review. They are too small to trigger a stop-loss review and too large to be treated as routine low-cost claims.

Together, these claims make up a large share of plan spending, and payment errors in this range are estimated to cost self-funded plans 2% to 5% of their annual claims spend.

A self-funded employer spendizng $15 million on healthcare claims each year could be losing $150,000 to $450,000 annually to payment errors without ever receiving a report that identifies those mistakes.

The Oversight Gap Most Plans Don't Know They Have

Mid-range claims are the most common and least-reviewed category in most self-funded health plans. Most plan sponsors believe their TPA handles auditing. That assumption is expensive. A standard TPA audit reviews a random sample of 250 to 400 claims and uses the results to generate an overall accuracy score. That score becomes the plan's reported error rate.

The claims oversight gap exists because no one routinely reviews a large portion of healthcare claims. Stop-loss carriers focus on very high-cost claims, while low-cost claims are often processed automatically. The claims in between receive the least attention, even though they make up a significant share of healthcare spending.

This middle band, typically claims between $5,000 and $25,000, accounts for a significant share of a typical plan's annual healthcare spending. It includes orthopedic surgeries, outpatient procedures, emergency room visits with facility fees, and specialty infusion claims that quietly add up across hundreds of members. For most self-funded plans, these claims are processed and paid without independent review.

Why the Problem Exists

This oversight gap exists for three main reasons. First, TPAs are paid to process claims, not to find mistakes in their own work. Second, many audits review only a sample of claims instead of every claim. Third, employers often have limited access to detailed claims data. Many plan sponsors also don't realize that the "audit" included in their TPA contract is usually an internal review by the TPA, not an independent claims audit.

Most claims audits review only a small sample of claims. For example, if a health plan processes 40,000 claims in a year and the audit reviews only 350, more than 39,000 claims are never checked. If the same billing mistake appears on hundreds of claims that aren't part of the sample, the audit may never find it.

Data access has historically made the situation worse. Many TPA contracts included gag clauses that restricted employers from accessing detailed claims data or from hiring independent auditors outside the TPA's approved list. Section 201 of the Consolidated Appropriations Act of 2021 (CAA 2021) explicitly prohibited these clauses and required plan sponsors to attest annually to the DOL that their contracts do not contain them. Despite the law, some plan sponsors still face delays and partial disclosures when requesting their own data.

The Real Cost

The financial impact of unreviewed claims is significant. Industry estimates put claims payment error rates at 2% to 5% of annual healthcare spending, even among well-performing TPAs. According to Withum, overpayment recoveries typically represent 1% to 1.5% of paid claims.

For a health plan spending $20 million a year, that could mean $200,000 to $300,000 in recoverable overpayments that go unnoticed.

The risk is not only financial. ERISA Section 404 requires plan fiduciaries to manage plan assets with care and prudence. Regular claims oversight helps demonstrate that the plan sponsor is actively monitoring healthcare spending.

Without documented oversight, employers may face greater legal and fiduciary risk if their claims administration is questioned. Recent litigation has also placed increased attention on whether plan sponsors are adequately monitoring healthcare plan expenses and service providers.

What's Actually Happening Behind the Scenes

Four specific error categories drive the majority of recoverable overpayments in mid-range claims, and none of them require fraud to occur.

Duplicate Payments

Duplicate claims occur when the same medical service is billed more than once. Although TPAs have duplicate detection systems, small differences in billing details can prevent the claims from being flagged, allowing the same service to be paid twice.

Coordination of Benefits Failures

When a member has more than one health plan, one plan should pay first and the other should pay only the remaining balance. COB errors occur when the wrong plan pays first or the TPA doesn't recognize other coverage, causing the employer's plan to overpay.

Medical Coding Errors and Upcoding

Procedure codes determine how much a health plan pays for a service. Upcoding happens when a provider bills for a more complex service than was performed, resulting in a higher payment. These errors are often caused by coding or documentation issues and can go unnoticed without an independent review.

Eligibility Errors

Eligibility errors occur when a health plan pays claims for people who are no longer covered, such as former employees or dependents who have lost eligibility. If these records are not updated on time, the plan can pay claims that should never have been covered.

Why Current Approaches Are Not Enough

The typical audit model was designed for compliance documentation, not financial recovery. The table below illustrates the structural difference between what most plans do and what an independent claims oversight program actually requires.

How to Fix It

Fixing the oversight gap requires six specific actions, not a single vendor change.

Red Flags That Signal This Problem Applies to Your Plan

The ROI of Doing It Right

An independent claims oversight program can pay for itself. First-time audits often recover 1% to 3% of annual claims spend. For a $10 million health plan, that could mean $100,000 to $300,000 in recoveries, while audit fees are typically much lower.

The benefits continue beyond the first audit. Regular claims monitoring helps reduce recurring errors because TPAs know their work is being reviewed. Over time, this can improve claims accuracy and reduce unnecessary spending.

There is also a fiduciary benefit. A documented claims oversight program shows that the plan sponsor is actively monitoring healthcare spending. If the plan is ever reviewed by regulators or challenged by participants, those records can help demonstrate prudent oversight.

Conclusion and Next Steps

Many mid-range healthcare claims are paid without an independent review, leaving plans exposed to unnecessary costs and risk. The good news is that employers now have better access to claims data, making independent oversight easier than before.

Start by requesting complete claims data from your TPA under your CAA 2021 rights. Then conduct an independent 100% claims audit to identify errors and establish a baseline. Use the results to strengthen your TPA agreement and implement regular claims monitoring to prevent future overpayments.

Frequently Asked Questions

What is the healthcare claims oversight gap?

‍The claims oversight gap is the portion of claims that are paid without an independent review. Most self-funded plans audit only a small sample of claims, leaving many payment errors undetected.

What percentage of health plan claims contain errors?

‍Industry estimates suggest that 2% to 5% of annual claims costs contain payment errors, including duplicate payments, coding mistakes, COB failures, and eligibility errors.

Who is responsible for claims accuracy under ERISA?

‍The plan sponsor is responsible for overseeing claims payments. While a TPA processes claims, the employer remains responsible for monitoring the plan.

What does the CAA 2021 require regarding claims data access?

‍The CAA 2021 prohibits contract clauses that prevent employers from accessing claims data or using independent auditors. Plans must annually attest that their contracts comply.

How much can an independent claims audit recover?

‍A first-time independent audit typically recovers 1% to 3% of annual claims spend. On a $20 million plan, that could equal $200,000 to $600,000 in recoverable overpayments.

How is a 100% claims review different from a sample audit?

‍A sample audit reviews only a small number of claims. A 100% review examines every claim, helping identify specific overpayments and recurring billing errors.

How often should a self-funded plan conduct an independent claims audit?

‍Most plans should perform an independent audit at least once a year and monitor claims regularly throughout the year.

What should I look for in an independent claims audit firm?

‍Choose a firm that is independent of your TPA, reviews 100% of claims, has medical coding expertise, and offers a transparent fee structure.

‍

‍

‍

Healthcare claims oversight refers to a plan sponsor's process of systematically reviewing medical and pharmacy claims paid on a self-funded health plan for errors, fraud and overbilling. Most employers rely on their TPA to catch errors, but independent audits routinely find that 3 to 7 percent of paid claims contain recoverable overpayments.

A mid-sized manufacturer in Ohio discovered last year that its TPA had been paying a local hospital system at billed charges for nearly three years. The contract clearly called for network rates. The error cost the plan over $1.1 million before anyone noticed. The plan sponsor was the employer. No one at the company had reviewed a single Explanation of Benefits.

Plan Sponsors Are Flying Blind on Claims

Most employers have no systematic process to verify that the claims paid on their health plan are accurate, appropriate or even contractually permitted.

The assumption is that the TPA handles it. The TPA assumes it is processing claims per the plan document and network contracts. In practice, neither party is actively looking for errors on a sustained basis. That gap between assumption and reality is where overpayments live.

Self-funded plans cover roughly 65 percent of all privately insured workers in the United States, according to the Kaiser Family Foundation. Collectively, employer-sponsored plans pay trillions in claims each year.

The typical process remains largely unchanged: a TPA processes claims, issues EOBs, and coordinates payments, while employers often rely on summary reports rather than reviewing claim-level activity.

Why the Problem Exists

The current self-funded model can make it challenging to maintain continuous, detailed claims oversight.

TPAs are generally compensated for claims administration rather than claims accuracy. As a result, independent claims reviews can provide an additional layer of oversight and help identify issues that may not be apparent through routine claims processing alone.

As a result, many employers rely primarily on vendor reporting and may have limited ability to independently validate claim accuracy.

The Real Cost of Inadequate Claims Oversight

Claims leakage, the portion of plan spend lost to errors, fraud, waste and overbilling, typically runs between 3 and 7 percent of total paid claims, based on industry auditing benchmarks published by auditing firms and benefits consultants.

For a self-funded employer spending $10 million annually on health claims, a 3% to 7% payment error rate could represent $300,000 to $700,000 in potential payment errors or avoidable claims costs, depending on the nature of the errors and whether they can be recovered.

Self-funded employers are facing greater scrutiny over how they oversee health plan spending. Benefits attorneys, including the Groom Law Group, have noted that ERISA investigations increasingly focus on whether plan sponsors have prudent oversight processes.

In Lewandowski v. Johnson & Johnson, participants alleged the employer failed to manage pharmacy spending prudently. Although the court dismissed the fiduciary claims on standing grounds without ruling on the merits, the case highlighted growing expectations for stronger health plan oversight and governance.

What's Actually Happening Behind the Scenes

Claims adjudication is a high-volume, rules-based automated process and automated rules can be misconfigured, outdated or simply absent.

Duplicate Claims

A provider submits a claim. A clearinghouse resubmits it. The TPA pays it twice. Duplicate logic in adjudication systems is supposed to catch this, but logic exceptions exist. Retrospective audits by independent firms routinely recover duplicate payments as one of the most common error categories.

Upcoding and Unbundling

A facility bills a complex evaluation and management visit when the documentation supports a lower-level code. Or a surgical procedure is billed as multiple component codes when a single bundled code should apply. These errors inflate reimbursement and are rarely flagged by TPA automated edits without specific clinical logic overlays.

Coordination of Benefits Failures

When a participant has coverage under two plans, the primary carrier is supposed to pay first. COB errors occur when the TPA lacks current information about secondary coverage or fails to recover from the other carrier. These can result in the self-funded plan overpaying as both primary and secondary payer simultaneously.

Contract Pricing Errors

Network contracts specify allowed amounts, discount thresholds, carve-outs and fee schedules. When a TPA's pricing file is not updated to reflect contract renegotiations, or when a claim routes to an out-of-network provider incorrectly classified as in-network, the plan overpays. This category of error can be large per-claim and is difficult to detect without comparing paid amounts against the actual contract.

Billing Fraud and Abuse

This is distinct from honest errors. Some providers systematically bill for services not rendered, upcode routinely or fabricate diagnoses. FBI and OIG enforcement data show healthcare fraud costs the U.S. system tens of billions annually. Self-funded plans are targets because they often lack the robust fraud analytics that large commercial carriers deploy.

Why Current Approaches Are Not Enough

Most TPAs have internal audit functions and claim editing software. These catch a portion of errors before payment (pre-payment editing) or flag obvious duplicates. But they are not independent, they are not comprehensive, and they are not designed to surface systemic problems in the TPA's own adjudication.

Think of it this way. Asking your TPA to audit its own claims accuracy is like asking a restaurant to grade its own health inspection. The restaurant may have good intentions and internal protocols. But the incentive structure makes objective self-assessment structurally difficult.

How to Fix It: A Practical Oversight Framework

Plan sponsors can establish meaningful claims oversight without replacing their TPA, using a layered approach that combines contractual rights, independent review and ongoing monitoring.

Red Flags That Signal Your Plan Has a Claims Oversight Problem

The ROI of Getting Claims Oversight Right

Independent claims audits return between $3 and $12 for every $1 spent on the audit, based on published recovery ranges from auditing firms and benefits consulting organizations.

A one-time retrospective audit on a plan with $15 million in annual claims typically identifies $300,000 to $900,000 in recoverable overpayments. Recovery depends on contractual provisions with the TPA and the age of the errors found, but most TPA agreements allow recovery of documented overpayments.

The Milliman actuarial consulting firm has documented that plans with active oversight programs spend 2 to 5 percent less on claims over time compared to unmonitored plans, controlling for demographics and benefit design.

Regular claims audits don't just help find billing errors and save money. They also create a record showing that the employer was actively monitoring the health plan. If questions ever arise about how the plan was managed, that documentation can be valuable. In many cases, the cost of dealing with legal disputes or regulatory investigations can be far greater than the cost of maintaining a proactive claims oversight program.

Conclusion and Next Steps

Healthcare claims oversight is not a luxury or a compliance checkbox. It is one of the most financially consequential activities a self-funded plan sponsor can undertake.

The technology to monitor claims accurately in near-real time has existed for years. The legal obligation to do so under ERISA has existed for decades. What has changed is the enforcement environment. DOL investigations are broader. Participant lawsuits are more specific about fiduciary process failures. Stop-loss carriers are paying closer attention to what was reviewed before a large claim was paid.

Start with a baseline retrospective audit. Review your TPA agreement for audit rights. Assign someone in finance or legal to own the oversight process. Those three steps will put your plan ahead of a large majority of self-funded employers in the country.

Frequently Asked Questions

What is healthcare claims oversight for a self-funded plan?

Healthcare claims oversight is the process of checking that medical and pharmacy claims are paid accurately and according to plan rules. It includes claims audits, eligibility checks, coordination of benefits (COB), and monitoring TPA performance.

Is a plan sponsor legally required to audit claims under ERISA?

ERISA does not require a specific audit schedule, but it requires plan fiduciaries to monitor plan operations and service providers. Regular claims reviews help demonstrate prudent oversight.

How much does an independent claims audit cost?

Costs vary by plan size and scope. Retrospective audits often range from $15,000 to $50,000, while some firms work on a contingency basis and only charge if they recover overpayments.

What percentage of claims typically contain errors?

Industry benchmarks suggest that 3% to 7% of paid claims contain recoverable errors, including duplicate payments, pricing mistakes, eligibility issues, and billing errors.

Can the plan recover overpayments from the TPA?

It depends on the TPA agreement and the type of error. Many contracts allow recovery of overpayments, but recovery deadlines and conditions vary.

What is the difference between a retrospective audit and concurrent review?

A retrospective audit reviews claims after they are paid to identify recoverable overpayments. Concurrent review examines claims before or shortly after payment to prevent errors from occurring.

Does the CAA 2021 change claims oversight obligations for employers?

The CAA 2021 increased transparency requirements but did not require routine claims audits. It reinforced that plan sponsors are responsible for understanding and monitoring healthcare spending.

How do I know if my TPA agreement allows an independent audit?

Review your ASO or ASA agreement for audit rights, data access, and recovery provisions. If these rights are unclear, consider negotiating stronger audit language at your next contract renewal.

‍

Waiting 90 days to review claims increases both financial and operational risk. As recovery windows narrow, providers become less likely to return overpayments. Meanwhile, recurring billing and adjudication errors can continue across thousands of claims before they are detected.

The delay not only increases potential losses but also leaves plan sponsors with less documentation demonstrating ongoing oversight of plan expenditures and vendor performance.

A distribution company with 600 employees did not review its healthcare claims until its annual audit. By then, a provider had been overcharging for infusion therapy for five months, and the health plan paid the higher amount on every claim during that period.

Because no one was monitoring claims regularly, the problem continued unnoticed and the losses kept growing. Plan sponsors are expected to take reasonable steps to protect plan funds. Waiting months to review claims can allow small errors to turn into significant costs.

What "Waiting 90 Days" Actually Means for a Self-Funded Plan

A 90-day audit cycle means that for every dollar paid in error, the plan sponsor has no visibility into the mistake for up to three months. Most plan sponsors think of a quarterly or annual audit as standard practice, even responsible practice. The assumption is that claims get checked eventually, so the system works.

The reality is that "eventually" is doing a lot of work in that sentence. A TPA processing claims for a 600-life plan might process 2,000 to 4,000 claims per month. Over 90 days, that is 6,000 to 12,000 claims that go entirely unreviewed by anyone outside the TPA's own internal quality process. Errors do not wait politely for the audit calendar. They compound.

The Government Accountability Office has documented that improper payments in health benefit programs are persistent precisely because detection systems are not built to catch errors close to the point of payment. A self-funded plan operating on a 90-day or annual review cycle is replicating that same structural weakness inside its own claims oversight.

Why the 90-Day Gap Exists in the First Place

The 90-day audit lag exists because it was built around TPA reporting cycles, not around fiduciary risk. Several structural habits keep this delay in place across the industry.

Legacy Contract Design

Many administrative services only (ASO) agreements were written years ago when quarterly reporting was the technical norm. TPAs batch claims data into quarterly files because that matched their internal reporting infrastructure, not because it served the plan sponsor's oversight needs.

Data Access Friction

Some TPAs charge additional fees for more frequent claims data extracts, or simply do not offer a faster cadence as a standard service tier. Plan sponsors who never ask for a 30-day data feed never learn that one is possible.

Underestimating compounding risk

A single duplicate payment or repricing error rarely alarms anyone. The risk is not the individual error. It is the same error repeating across every claim that matches the same billing pattern for as long as nobody is looking.

The Real Cost of a 90-Day Blind Spot

The financial cost of a 90-day audit delay scales with claims volume in a way that catches most plan sponsors off guard. Consider a self-funded plan with $9 million in annual claims spend, which is roughly $750,000 per month. A billing error affecting just 2% of claims volume that goes undetected for a full quarter represents $45,000 in overpayments before anyone even opens an audit file.

Now extend the timeline. If the same plan operates on an annual cycle instead of quarterly, that same 2% error rate compounds to $180,000 before review. The The Healthcare Financial Management Association has reported that hospitals lose an average of 4.8% of net revenue to claim denials, and other industry estimates for self-funded claims processing errors range higher still.

The fiduciary cost runs alongside the financial cost. ERISA Section 404(a)(1) requires fiduciaries to act with the care, skill, prudence, and diligence that a prudent person familiar with such matters would use under similar circumstances. A plan sponsor who can produce evidence of monthly claims monitoring has a materially stronger position in a DOL inquiry than one who can only show an annual review. Documented frequency matters because intent and diligence are judged by process, not by outcome alone.

What's Actually Happening Behind the Scenes During the Gap

Error Patterns Repeat Before Anyone Notices

A single coding error from a provider's billing system rarely appears once. If a provider's software misapplies a modifier code or a TPA's adjudication engine mishandles a specific procedure code, that same error recurs on every matching claim until someone catches it. Ninety days gives that pattern three full months to multiply.

Provider Dispute Windows Start Closing

Most state prompt-pay statutes and provider contracts establish specific windows during which a payer can dispute or recoup a payment without escalated friction. As days pass, providers gain stronger legal footing to resist repayment demands, citing reliance on the original payment and administrative finality. A claim flagged at day 15 is a different negotiation than the same claim flagged at day 95.

Dependent Eligibility Drift Goes Unchecked

Life events such as divorce, a dependent aging out, or a spouse gaining other coverage do not pause for the audit calendar. A 90-day gap means a plan can pay three additional months of claims for someone who became ineligible on day one of the quarter, with no mechanism catching it until the cycle closes.

Stop-Loss Reporting Misalignment

Self-funded plans with stop-loss coverage typically need to report large claims promptly to remain compliant with notice provisions in the stop-loss contract. A 90-day internal audit cycle can mean large claims are not flagged for stop-loss notification within the carrier's required timeframe, risking a denied or reduced reimbursement on a catastrophic claim.

Fiduciary Documentation Gaps Widen

Every quarter without a documented review is a quarter without evidence that the plan exercised the procedural prudence ERISA requires. DOL investigators reviewing a plan's fiduciary process look for a paper trail. A 90-day gap, repeated across multiple plan years, creates a pattern of sparse documentation that is difficult to explain after the fact.

Why an Annual or Quarterly Cadence Isn't Enough

A quarterly or annual audit cycle was adequate when claims data was difficult to access. It is not adequate now that near real-time data feeds exist. The table below compares the standard cadence against a tighter monitoring model.

How to Fix It: Shortening the Audit Cycle

Red Flags That Signal Your Plan Is Exposed

The ROI of Moving to a Shorter Audit Cycle

Shortening the audit cycle from 90 days to 30 days or less produces a measurable reduction in both recoverable loss and ongoing exposure. Independent claims integrity data suggests that catching an error within 30 days of payment results in recovery rates roughly two to three times higher than catching the same error after 90 days, largely because provider dispute friction and statutory lookback limits have not yet hardened.

On a plan with $9 million in annual claims spend and a conservative 3% error rate, that is $270,000 in identifiable overpayments per year. Shifting from annual to monthly review does not eliminate errors, but it shrinks the average exposure window from 180 days (the midpoint of an annual cycle) to roughly 15 days (the midpoint of a monthly cycle). That alone can convert a six-figure annual loss into a five-figure one.

The fiduciary protection value compounds over time. A documented monthly review process, sustained over multiple plan years, builds a defensible record of prudent process under ERISA Section 404. The Department of Labor's Meeting Your Fiduciary Responsibilities guidance emphasizes that fiduciaries are judged on the process they followed, not simply the financial outcome. A consistent, frequent, well-documented audit cadence is one of the clearest ways to demonstrate that process.

Conclusion: The Calendar Is Not a Fiduciary Strategy

Waiting 90 days to audit claims is a scheduling habit, not a deliberate risk decision, and that distinction matters under ERISA. Plan sponsors who have never questioned their audit cadence are not negligent by intent. They are simply operating on a calendar inherited from TPA reporting defaults and broker stewardship cycles that were never designed with fiduciary exposure in mind.

The fix does not require new technology budgets or a TPA switch. It requires a written policy, a faster data feed, and a named owner who reviews claims monthly instead of quarterly. The dollar savings are real and the fiduciary protection compounds every plan year the new cadence holds.

Start by asking your TPA one question this week: how quickly can we get claims data, and how often are we currently reviewing it. The gap between those two answers is your fiduciary exposure.

Frequently Asked Questions

What does ERISA require regarding claims audit frequency?

‍ERISA does not specify an exact audit frequency. Section 404(a)(1) requires fiduciaries to act with the care, skill, prudence, and diligence of a prudent person managing similar matters, which courts and the Department of Labor interpret as an ongoing obligation rather than a once-a-year exercise. The absence of a specific number does not mean infrequent review satisfies the standard.

How much does a 90-day audit delay typically cost a self-funded plan?

‍The cost scales with claims volume and the underlying error rate. A plan with $9 million in annual claims spend and a 3% error rate can accumulate roughly $67,500 in unreviewed overpayments over a single 90-day quarter before any audit begins, based on standard industry error-rate benchmarks from claims integrity sources.

Is a quarterly claims audit considered sufficient under ERISA?

‍There is no bright-line rule stating that quarterly review satisfies fiduciary duty. The Department of Labor evaluates the totality of a plan's prudent process, including frequency, documentation, and follow-through on identified errors. A quarterly cadence with strong documentation may be defensible, but a monthly or near-continuous cadence offers a stronger position.

What is the difference between a claims audit and ongoing claims monitoring?

‍A claims audit is typically a discrete, periodic review of a defined claims population, often tied to an annual or quarterly cycle. Claims monitoring refers to a continuous or near-continuous process that reviews claims data on a rolling basis, closer to the point of payment, which reduces the window during which errors go undetected.

Can a plan sponsor be held personally liable for failing to audit claims promptly?

‍Named fiduciaries, including HR leaders or finance executives who exercise discretionary authority over plan administration, can face personal liability under ERISA Section 409 for breaches of fiduciary duty. Failing to establish a reasonable claims monitoring process is a factor courts and the DOL consider when evaluating whether a breach occurred.

How do stop-loss notification requirements relate to audit timing?

‍Stop-loss contracts typically require prompt notification of large or catastrophic claims, often within a specified number of days from the claim reaching a certain dollar threshold. A 90-day internal audit cycle can cause large claims to be flagged for stop-loss reporting later than the contract requires, risking reduced or denied reimbursement on the claim.

What is a reasonable target audit cadence for a self-funded plan with 100 or more employees?

‍Most claims integrity consultants and TPA performance benchmarks recommend monthly review of claims data at minimum, with concurrent or near-real-time review for claims above a defined high-dollar threshold. Plans below 500 employees may reasonably start with monthly review and layer in more frequent high-dollar claim flags as the program matures.

Does shortening the audit cycle require switching TPAs?

‍No. Most plan sponsors can shorten their audit cycle by renegotiating data delivery terms within their existing TPA contract or by adding an independent claims monitoring vendor that integrates with the current TPA's claims feed. Switching TPAs is rarely necessary and introduces its own transition risk.

‍

A manufacturing company with 400 employees received its annual post-pay audit results and learned it had overpaid $340,000 in medical claims over the prior plan year. The TPA recovered $47,000. The rest was gone.

One reason was simple. The TPA's claims system lacked the sophisticated pre-payment detection capabilities needed to identify certain billing errors before funds were released.

What a Post-Pay Audit Actually Does (and Doesn't Do)

A post-pay audit reviews claims after the TPA has already processed and paid them, which means every dollar of error it finds has already been spent. Most employers understand post-pay audits as a quality-control tool.

TPAs process millions of transactions per year. A post-pay audit typically samples 100% of claims above a dollar threshold or a statistical sample of all claims, then flags anomalies for potential recovery. The auditor sends a demand letter to the provider, the provider disputes it, negotiations begin, and the employer eventually recovers a fraction of the original error amount.

Industry data from the Healthcare Financial Management Association suggests that self-funded plans overpay between 3% and 10% of total medical claims spend annually due to processing errors, billing fraud, and repricing failures. On a $5 million annual claims spend, that is $150,000 to $500,000 in potential overpayments. The post-pay audit catches some of it. It permanently loses most of it.

Why Post-Pay Audits Are Structured to Underperform

The fundamental flaw is not execution. It is timing. Post-pay audits were designed for a world where employers primarily wanted to satisfy an annual compliance checkbox. They were never engineered to maximize claims integrity or fiduciary protection.

Several structural problems compound the timing issue:

The Real Cost to Plan Sponsors

The financial exposure from inadequate claims oversight is larger than most HR leaders and CFOs realize, and it compounds year over year. A single plan year of 5% overpayments on a $10 million claims spend equals $500,000. Over five years, with no corrective action, that is $2.5 million in preventable losses.

The cost extends beyond the dollar amount recovered or not recovered. ERISA Section 404(a)(1) requires plan fiduciaries to discharge their duties with the care, skill, prudence, and diligence that a prudent person acting in a like capacity would use.

The Department of Labor has made clear through its audit and enforcement activity that reliance on a TPA does not absolve plan sponsors of fiduciary responsibility. A plan that conducts only an annual post-pay audit and recovers 15 cents on the dollar is not meeting that standard.

The Kaiser Family Foundation's 2023 Employer Health Benefits Survey found that the average annual family premium for employer-sponsored coverage exceeded $23,000. Self-funded plans bear 100% of claims cost directly. Every dollar of claims error flows straight to the plan's bottom line, and ultimately to employee cost-sharing and benefit design decisions.

What's Actually Happening Behind the Scenes

Why Current Approaches Aren't Enough

Annual post-pay auditing has become an industry default, not an industry best practice. The comparison below shows the practical difference between a post-pay approach and a concurrent or pre-pay model.

How to Fix It: A Practical Path for Plan Sponsors

Red Flags That Signal This Problem Applies to Your Plan

The ROI of Getting Claims Oversight Right

The return on investment from upgrading claims audit infrastructure is measurable and consistent across employer sizes. Independent studies and vendor case data suggest the following benchmarks:

Pre-pay and concurrent audit programs typically generate $3 to $8 in recovered or avoided overpayments for every $1 spent on the program. On a plan spending $8 million annually in medical claims, capturing even half of a conservative 3% error rate produces $120,000 in savings. A concurrent audit program for a plan that size typically costs $30,000 to $60,000 annually. The math is straightforward.

Dependent eligibility audits cost $15 to $40 per employee audited and routinely return 10 to 20 times that amount in annual premium savings from removing ineligible dependents. For a plan with 500 employees, the audit cost might be $20,000. If 4% of 900 covered dependents are removed and each carried an average monthly cost of $400, the annual savings exceeds $86,000.

The fiduciary protection value is harder to quantify but significant. DOL investigations of self-funded plans that result in findings of inadequate claims oversight can require the plan to reimburse participants for losses plus interest. Documented, proactive audit programs are a primary defense against that exposure.

Conclusion: Stop Auditing Yesterday's Mistakes

Post-pay audits have a place in a comprehensive claims oversight program, but they cannot be the entire program. Self-funded employers who rely on an annual post-pay review as their primary quality control tool are systematically overpaying their claims, underperforming on their ERISA fiduciary obligations, and leaving recoverable money on the table every month.

The good news is that better tools exist and are accessible to plans well below the Fortune 500 threshold. Concurrent audit programs, dependent eligibility verification, and clinical review of high-cost claims can be layered into most TPA relationships with modest contract adjustments and reasonable vendor investment. The ROI is well-documented. The fiduciary argument is clear.

Start by requesting your full claims data file from your TPA and scheduling an independent review. If your TPA resists providing the data, that resistance is itself a finding.

Frequently Asked Questions

What is a post-pay claims audit?

‍A post-pay claims audit is a review of health plan claims that have already been processed and paid by the third-party administrator. The auditor identifies errors such as duplicate payments, incorrect repricing, or unbundled procedure codes after the funds have transferred to providers.

How much do self-funded employers typically overpay on medical claims?

‍Industry estimates from the Healthcare Financial Management Association and independent payment integrity consultants place the overpayment rate for self-funded plans at 3% to 10% of total annual claims spend.

What is the difference between a post-pay audit and a concurrent audit?

‍A post-pay audit reviews claims after payment has been made. A concurrent audit integrates with the claims payment process in real time and flags suspected errors before the TPA releases payment to the provider. Concurrent audits prevent overpayment rather than attempting to recover it, which produces materially higher net savings for the plan.

Does ERISA require self-funded employers to audit their claims?

‍ERISA does not mandate a specific audit frequency or methodology, but Section 404 requires plan fiduciaries to act with the care, prudence, and diligence of a knowledgeable person managing plan assets.

Can a TPA conduct its own claims audit?

‍A TPA can conduct internal quality reviews, and most do. However, relying solely on the TPA to audit its own claims processing creates a conflict of interest. An independent third-party auditor with access to raw claims data provides a more objective assessment and typically identifies a different, often larger, set of errors than the TPA's internal team.

What should a self-funded employer look for in a claims audit vendor?

‍Look for a vendor that reviews 100% of claims rather than a statistical sample, has direct integration with your TPA's claims system, provides itemized error reporting with CPT code-level detail, covers dependent eligibility as part of the audit scope, and offers a clear fee structure that is not contingency-only (which can create incentives to flag borderline items). Ask for client references from plans of similar size and industry.

How long should self-funded employers retain claims audit records?

‍ERISA Section 107 requires plan records to be retained for at least six years from the filing date of the annual Form 5500 to which they relate. Claims audit documentation, including methodology, findings, and corrective actions, should be treated as plan records subject to this retention requirement.

What percentage of identified overpayments does a post-pay audit typically recover?

‍Recovery rates vary widely, but most independent claims auditors and benefits consultants report effective post-pay recovery of 15% to 35% of identified overpayments. Provider disputes, expired lookback windows, and practical collection limitations account for the gap. Pre-pay and concurrent models avoid this problem because the money never leaves the plan's account.

‍

‍

‍

‍

Claims leakage is the chronic, undetected loss of health plan dollars caused by billing errors, pricing failures, and processing gaps rather than intentional fraud. It typically costs self-funded employers 3 to 5 percent of total annual claims spend. It is addressable through independent claims audits and stronger TPA oversight contracts.

A mid-size manufacturing company with 800 employees discovers, two years into its self-funded health plan, that its TPA paid the same surgical claim three times. The total overcharge: $47,000. No one committed fraud.

The TPA's system missed a duplicate claim after the facility changed a single digit in the billing code, a reminder that many TPAs lack the advanced technology needed to identify complex claims leakage.

According to the Healthcare Financial Management Association, improper payments and billing errors account for an estimated 3 to 5 percent of total health plan spend annually. For a plan spending $8 million a year, that's up to $400,000 leaving through the back door quietly, year after year.

What Claims Leakage Actually Is (And What It Is Not)

Claims leakage is not fraud. It is the slow, compounding erosion of health plan assets caused by errors, process failures, and gaps in oversight that no one catches.

Most HR leaders and CFOs associate financial loss with intentional misconduct. That framing is understandable but costly. It means they do not look for a problem that is almost certainly present in their plan right now.

The term "claims leakage" describes payments that leave the plan incorrectly. They may be duplicates. They may be priced against the wrong contract. They may reflect services that were billed but not rendered. They may include charges for a dependent who aged off the plan six months ago. None of these require bad intent. All of them represent real money paid out that should not have been.

The analogy that makes this stick: claims leakage is like a slow water leak behind your walls. Nothing looks wrong from the outside. There's no burst pipe, no flood. But by the time you notice the damage, the loss has been accumulating for years.

Why Claims Leakage Is Structural, Not Accidental

The root cause is a fundamental misalignment between who processes claims, who audits them, and who bears the financial risk.

Your TPA adjudicates and pays claims on your behalf. Their incentive is speed and throughput. Volume is how they demonstrate value. Catching every nuanced billing error requires the kind of meticulous review that slows throughput.

Most TPA performance guarantees focus on turnaround time and error rates tied to a narrow sample of total claims, often 1 to 3 percent reviewed post-payment.

The employer. the party that actually funds every claim. typically has no internal claims expertise. Benefits staff manage enrollment and vendor relationships. They are not trained to interrogate 835 transaction files or identify miscoded facility charges.

This is not a conspiracy. It is a structural gap that allows billing errors to pass through the system unchallenged. According to the Government Accountability Office, the lack of independent oversight in self-funded plan management is a documented and recurring concern in federal reporting on healthcare payment integrity.

The Real Cost: What the Numbers Show

Employers with self-funded health plans lose an estimated 3 to 5 percent of total annual claims spend to leakage, a figure documented across multiple industry analyses.

The dollar impact scales fast. A plan spending $5 million per year loses $150,000 to $250,000. A plan at $20 million loses $600,000 to $1 million. These figures represent what independent claims auditors routinely find when they examine a full plan year of adjudicated data.

Duplicate claims alone account for a meaningful share of that loss. The Medical Billing Advocates of America estimates that up to 80 percent of medical bills contain at least one error. Not all errors favor the payer, but a significant portion result in overpayments.

Coordination of benefits failures create additional exposure. When a dependent is covered under two health plans, the plan that should pay secondary sometimes pays primary because the eligibility data was never updated. The overpayment is often never recovered unless someone specifically looks for it.

Stop-loss reimbursement accuracy is another underappreciated risk. If your TPA misclassifies claims or applies the wrong deductible accumulator logic, your stop-loss carrier may pay less than your plan is owed. A claim repriced at $180,000 instead of the actual $210,000 allowed amount could cause the plan to miss its specific attachment point entirely.

What Is Actually Happening Inside Claim Adjudication

Repricing Errors

Repricing failures occur when a claim is paid against the wrong network contract, wrong fee schedule, or outdated rate.

This is more common than most plan sponsors realize. Provider contracts update. Network configurations change when TPAs renegotiate. If system tables are not updated promptly, claims process at old rates. In some cases, out-of-network claims are incorrectly routed as in-network. The difference per claim can be tens of thousands of dollars on high-cost procedures.

Duplicate Claim Payments

A duplicate payment occurs when the same service is paid more than once due to minor variations in the claim submission.

Facilities often refile claims after a denial or delay. If the original claim was eventually paid and the refiled version is also paid, the plan has double-funded the same service. Variations in billing code, date of service formatting, or provider NPI can defeat basic duplicate detection logic.

Coordination of Benefits Failures

COB failures result in the plan paying primary when it should pay secondary or not paying at all.

Dependents covered under a spouse's plan, working retirees with Medicare, and children of divorced parents with dual coverage all create COB complexity. When eligibility data is stale or the TPA's COB workflow breaks down, the employer plan overpays. Recovery requires proactive subrogation and COB recovery programs most plans do not have in place.

Unbundling and Upcoding

Unbundling occurs when a provider bills separately for services that should be billed as a single bundled procedure code.

CPT code bundling rules exist precisely to prevent this. But automated claim systems do not always catch every improper unbundling pattern, particularly for surgical assists, anesthesia, and facility fees. Upcoding is the related practice of billing for a more complex service than the one delivered. Both inflate plan costs without triggering fraud detection.

Ineligible Dependent Coverage

Claims paid for dependents who no longer qualify under the plan document represent direct leakage with a recoverable element.

Dependents who age out, ex-spouses who remain on the plan after divorce, and adult children past the plan's cutoff date all generate improper payments. Dependent eligibility audits conducted by independent vendors typically find ineligible dependents on 3 to 8 percent of enrolled employee files.

Why Current Oversight Approaches Are Not Enough

Relying on your TPA to self-report payment errors is structurally equivalent to asking a contractor to audit their own invoices.

Most plan sponsors accept their TPA's claims reports as authoritative. They review aggregate spend by category and surface-level utilization metrics. They do not examine claim-level data for patterns that indicate systematic errors. This is not negligence. It is a knowledge gap reinforced by a lack of independent access to the data.

The Consolidated Appropriations Act of 2021 (CAA 2021) now requires TPAs and brokers to disclose conflicts of interest and compensation to plan fiduciaries. This is a significant development. But disclosure does not equal oversight.

A plan sponsor who receives a compensation disclosure and takes no further action has not fulfilled their fiduciary duty under ERISA Section 404. The DOL's Employee Benefits Security Administration has been explicit on this point.

How to Fix Claims Leakage in Your Plan

Addressing claims leakage requires a shift from passive monitoring to structured, independent verification at the claim level.

Red Flags That Claims Leakage Is Present in Your Plan

The ROI of Getting Claims Oversight Right

Independent claims audits consistently return more than they cost, often by a factor of three to ten.

Recovery rates vary by plan size, audit scope, and how long since the last audit. For plans that have never been independently audited, first-year recovery of 1 to 3 percent of total audited spend is typical. On a $10 million plan, that is $100,000 to $300,000 recovered in year one.

The ongoing benefit is larger than the one-time recovery. Once errors are identified and the TPA corrects underlying system or process failures, forward-looking savings compound annually. A repricing error corrected in year one does not recur in years two through five.

ERISA litigation against plan sponsors has increased significantly in the past five years. Cases like Lewandowski v. Johnson and Johnson and parallel suits against Wells Fargo and JPMorgan Chase demonstrate that courts and plaintiffs will examine whether plan fiduciaries took active steps to control costs and verify claims accuracy.

Conclusion and Next Steps

Claims leakage is not a fringe problem. It is the predictable outcome of a system in which the party that processes payments is also the party responsible for validating them. For self-funded employers, the financial exposure is real, the fiduciary risk is documented, and the fix is actionable.

The first step is accepting that your claims data probably contains errors you have not seen. The second step is gaining independent access to that data. The third is engaging an auditor who owes their loyalty to the plan and its participants, not to your vendor relationships.

ERISA does not require perfection. It requires diligence. An independent claims audit is one of the most concrete demonstrations of that diligence available to a plan sponsor today.

Frequently Asked Questions

What is claims leakage in a self-funded health plan?

Claims leakage is the loss of health plan assets through payment errors, processing failures, and oversight gaps rather than fraud. It includes duplicate payments, repricing mistakes, coordination of benefits failures, and payments for ineligible dependents. It is endemic to self-funded plans and typically goes undetected without an independent claims audit.

How much does claims leakage cost employers?

Industry estimates consistently place claims leakage at 3 to 5 percent of total annual claims spend. For a plan with $8 million in annual claims, that represents $240,000 to $400,000 in annual loss. First-time independent audits frequently recover 1 to 3 percent of total audited spend in identifiable overpayments.

Is my TPA responsible for catching claims leakage?

Your TPA has contractual obligations to process claims accurately, but their internal QA programs typically audit only a small sample of total claims. They are not positioned as independent auditors and have no financial incentive to surface systemic errors. Under ERISA, the plan fiduciary, which is the employer, bears responsibility for oversight.

What does ERISA require of plan sponsors regarding claims accuracy?

ERISA Section 404 requires plan fiduciaries to act with the care, skill, prudence, and diligence that a knowledgeable person familiar with such matters would use. This prudent expert standard means plan sponsors cannot simply defer to their TPA. They must take active steps to verify that the plan is being administered correctly and in the interest of participants.

What is an independent claims audit and how does it work?

An independent claims audit is a systematic review of adjudicated claims data conducted by a firm with no financial relationship to the TPA, broker, or PBM. The auditor receives full claims data in electronic format (typically 835 transaction files), applies rule-based and analytical review logic, and produces a report identifying overpayments, error patterns, and recovery opportunities.

How is claims leakage different from healthcare fraud?

Fraud involves intentional misrepresentation. Claims leakage involves errors, system failures, and process gaps. Both result in improper payments, but fraud requires criminal intent and legal enforcement. Leakage is correctable through operational fixes, contract renegotiation, and process improvement. Most dollar losses in self-funded plans fall into the leakage category, not fraud.

Does CAA 2021 help plan sponsors address claims leakage?

The Consolidated Appropriations Act of 2021 strengthened plan sponsors' data access rights and required TPAs and brokers to disclose compensation and conflicts of interest. These provisions create a foundation for better oversight. But the law gives plan sponsors tools, not guarantees. Sponsors still need to exercise their data rights and act on what the data shows.

How often should a self-funded plan conduct a claims audit?

Annual audits are the standard recommended by benefits attorneys and claims integrity consultants for plans above $5 million in annual spend. Plans that have never been audited should treat the first audit as a priority regardless of size. Plans undergoing TPA transitions should audit the prior TPA's full claims history before the transition closes.

‍

A TPA fiduciary risk is the legal and financial exposure a self-funded employer faces when its third-party administrator processes claims incorrectly and the employer, as ERISA plan fiduciary, is held responsible for the losses. Because TPAs are typically not ERISA fiduciaries themselves, the liability stays with the plan sponsor.

A mid-sized manufacturer in Ohio recently discovered that its TPA had been paying a terminated employee's medical claims for 14 months after the employee left the company.

The total exposure: $340,000. Under ERISA, the employer, not the TPA, bore responsibility for recovering those funds. The Department of Labor does not grade plan sponsors on how trusting they were of their vendor.

What the TPA Relationship Actually Means for Fiduciary Liability

Most employers believe their TPA carries the legal risk when something goes wrong with claims. That belief is incorrect, and it is expensive.

Under ERISA Section 404, the plan sponsor (the employer) is a named fiduciary obligated to act solely in the interest of plan participants, follow the plan document and exercise the skill of a prudent expert.

TPAs are hired as service providers. Unless a TPA contractually accepts discretionary authority over plan assets and explicitly agrees to ERISA fiduciary status, which almost none do, it operates as a vendor, not a co-fiduciary.

Think of it like hiring a contractor to wire your building. If the work is faulty and someone gets hurt, the building owner faces liability. The contractor may owe indemnification under the service contract, but that is a separate civil dispute that takes time and money to resolve. Meanwhile, the DOL or an aggrieved participant is looking at you.

Why the Problem Exists

The TPA model was built for efficiency, not for employer oversight.

When an employer moves from fully insured to self-funded, it gains cost transparency and control. It also inherits accountability.

The administrative services only (ASO) agreement that governs the TPA relationship is typically written by the TPA's legal team. These contracts often include liability caps, indemnification carve-outs and language that limits the TPA's responsibility for errors to a narrow definition of "gross negligence."

The Real Cost of Unchecked Claims

The financial exposure from TPA claims errors is not theoretical. It is documented, recurring and significant.

The Government Accountability Office has reported that improper payments in employer health plans are a persistent problem across both public and private sectors.

Industry benchmarks from claims audit firms consistently show that between 3 and 10 percent of processed claims contain some form of error, ranging from duplicate payments to incorrect member eligibility to miscoded procedures.

For a self-funded employer spending $5 million annually on medical claims, a 5 percent error rate represents $250,000 in potential misprocessed payments. A 2022 analysis by the Healthcare Financial Management Association found that coordination of benefits (COB) errors alone cost employers an average of $350 per affected employee per year.

Beyond the direct dollar loss, there are secondary costs:

What Is Actually Happening Behind the Scenes

Most claims errors are not fraud. They are systemic, predictable and preventable through routine auditing.

Duplicate Claims

A provider submits the same claim twice with minor coding variations. Auto-adjudication systems miss the duplication. Both claims pay. This is among the most common and most recoverable error types.

Eligibility Errors

Dependents age off coverage but are not removed from the system. Former employees remain active in the TPA's eligibility file. Claims pay for individuals who are no longer entitled to benefits. These errors are often months old before anyone notices.

Coordination of Benefits Failures

When a member has coverage under two plans, the primary payer should pay first and the secondary payer should pay only the remaining balance. When COB logic is applied incorrectly or not applied at all, both plans pay in full. The employer's plan absorbs a cost it should never have incurred.

Incorrect Repricing and Network Discounts

A claim is processed at billed charges rather than the contracted network rate. The provider is overpaid. Recovery from a provider after the fact is possible but administratively burdensome and often partial.

Unbundling and Upcoding

Providers submit separate line items for services that should be billed as a single bundled procedure code, inflating the allowed amount. Upcoding, billing for a higher-acuity service than was documented, is an ongoing issue that claims review software sometimes catches and sometimes does not.

Terminated Provider Contracts

A provider's network contract expires or is terminated, but the TPA continues to process claims as if the contract is in force. The employer pays network rates on claims that should have been processed as out-of-network, which may create additional downstream liability.

Why Current Approaches Are Not Enough

Relying solely on TPA internal controls to protect your plan is the equivalent of asking the contractor to inspect their own work.

Most employers receive monthly or quarterly claims reports. Those reports show aggregated spend by category, provider type or member. They are useful for budgeting. They do not reveal individual claim errors.

Some TPAs offer internal audit functions. These are not independent by definition. The TPA auditing its own claims adjudication has an inherent conflict of interest, regardless of how diligent the staff may be.

How to Fix It: A Practical Action Plan

Closing the fiduciary gap requires structure, contract language and independent verification. None of these steps requires replacing your TPA.

Red Flags That Signal This Problem Applies to Your Plan

The ROI of Doing It Right

Independent claims auditing consistently returns more than it costs, often by a multiple of three to five.

Retrospective audits of self-funded plans regularly recover between 1 and 3 percent of total paid claims. On a $5 million claims spend, that is $50,000 to $150,000 in recoveries per audit cycle. Contingency-fee audit arrangements mean the employer pays nothing unless recoveries are made.

The less quantifiable but equally real returns include:

• Documented fiduciary process that withstands a DOL inquiry
• Corrected eligibility files that reduce future claim errors
• Data that reveals patterns requiring TPA system corrections
• Leverage in TPA contract renegotiation backed by actual performance data
• Stop-loss carrier confidence that reduces friction at claim time

One regional health system with approximately 1,200 covered lives conducted its first independent claims audit after a compliance review flagged the absence of any oversight process. The audit recovered $218,000 in overpayments and identified a COB configuration error in the TPA system that had been generating duplicate payments for 22 months.

Conclusion and Next Steps

The fiduciary risk inside your TPA relationship is not a hypothetical. It is a documented, measurable and addressable problem that most plan sponsors have simply not prioritized.

ERISA does not expect perfection. It expects process. The plan sponsors who fare best in DOL audits, stop-loss disputes and participant complaints are those who can show a documented, repeatable approach to monitoring their TPA and correcting errors when they occur. An independent claims audit is the most direct tool available to accomplish that.

If you have never conducted an independent claims audit, that is the place to start. If you have not reviewed your ASO agreement with ERISA counsel, that is the second step. Neither task requires replacing your TPA. Both tasks are within reach for any plan sponsor, regardless of plan size.

‍Frequently Asked Questions

Is my TPA an ERISA fiduciary?

‍Almost certainly not. Most TPAs operate under administrative services only (ASO) agreements and explicitly disclaim ERISA fiduciary status in those contracts. Unless your TPA has signed a written agreement accepting discretionary fiduciary authority over plan assets, ERISA fiduciary responsibility stays with the employer as plan sponsor. Always verify this with ERISA counsel by reviewing your ASO agreement directly.

What does ERISA actually require me to do to oversee my TPA?

‍ERISA Section 404(a) requires plan fiduciaries to act with the care, skill, prudence and diligence of a knowledgeable professional. Applied to TPA oversight, this means having a documented process for selecting, monitoring and, when warranted, replacing your TPA. Courts and the DOL have held that "monitoring" requires more than receiving summary reports. It requires meaningful review of the TPA's actual claims performance.

How often should we conduct a claims audit?

‍Most benefits consultants recommend a full retrospective audit every one to two years, with continuous or quarterly monitoring in between. High-volume plans or plans that have recently changed TPAs, plan designs or eligibility rules benefit from more frequent review. The first audit typically yields the highest recoveries because it establishes a baseline and catches errors that have accumulated over time.

What types of errors does a claims audit typically find?

‍The most common categories are duplicate payments, eligibility errors (covering ineligible members or dependents), coordination of benefits failures, incorrect network repricing, unbundled or upcoded procedure codes and terminated provider contract issues. Eligibility errors and COB failures tend to generate the largest individual recoveries because they often persist for months before detection.

Can we require our TPA to conduct audits on our behalf?

‍You can require it contractually, and many ASO agreements include provisions for TPA self-reporting and internal quality reviews. However, a TPA auditing its own claims adjudication is not independent oversight. It does not satisfy the prudent expert standard under ERISA and will not carry the same weight with the DOL or in litigation as an audit conducted by a firm with no financial relationship to the TPA.

What should we look for in an ASO agreement before signing?

‍Prioritize four provisions: (1) unrestricted access to claim-level data, (2) explicit audit rights allowing independent review at any time, (3) defined liability for claims errors with no cap that effectively eliminates recovery, and (4) performance guarantees with financial penalties tied to measurable claims accuracy metrics. Most standard TPA contracts require negotiation to include all four.

Does our stop-loss carrier care whether we conduct claims audits?

‍Increasingly, yes. Stop-loss carriers are paying closer attention to plan sponsor fiduciary practices because their own exposure depends on the accuracy of underlying claims data. Some carriers now include audit requirements as a condition of coverage or give premium credit to employers with documented audit programs. If your stop-loss carrier has never asked about your audit practices, raise the topic proactively.

What is the difference between a prospective and a retrospective claims audit?

‍A prospective audit reviews claims before payment, typically for high-dollar or complex claims, to catch errors before money leaves the plan. A retrospective audit reviews claims already paid to identify recoverable overpayments and systemic errors. Most employers start with a retrospective audit because it yields immediate recoveries and reveals patterns for the TPA to correct going forward.

‍

‍

‍

A 1,000-employee company spends ~$15 million each year on healthcare claims. The TPA processes payments on time. Employees receive care without disruption. Renewal discussions focus on trend projections and stop-loss premiums.

Three years later, an independent review uncovers hundreds of thousands of dollars in claims errors that nobody noticed.

The surprising part is not that errors occurred. The surprising part is that nobody was explicitly responsible for finding them.

In many self-funded health plans, claims accuracy falls into an accountability gap. Everyone assumes someone else is watching. Few organizations verify whether that assumption is true.

What Is the Core Problem?

The core problem is that claims accuracy often lacks a clearly designated owner within self-funded health plans.

Most employers assume their TPA is fully responsible for ensuring every claim is paid correctly. That assumption sounds reasonable but overlooks an important reality.

TPAs administer claims according to plan documents, contracts, system configurations, provider agreements, and eligibility data. Each component introduces opportunities for mistakes.

Meanwhile, HR teams focus on employee experience. Finance leaders focus on budgets. Brokers focus on strategy and market positioning. Stop-loss carriers focus on large claim exposure.

As a result, no single stakeholder consistently validates whether claims are being paid accurately.

Claims accuracy becomes like a building with multiple security cameras but no one monitoring the screens.

Why the Problem Exists

Claims accuracy gaps exist because responsibility is distributed while accountability remains undefined.

Several structural factors contribute to the issue.

Complexity Continues to Increase

Modern healthcare claims involve network discounts, coding rules, plan provisions, coordination of benefits, eligibility feeds, pharmacy integrations, and provider contracts.

Each transaction depends on multiple systems working correctly.

The complexity of healthcare payment systems continues to grow as reimbursement methodologies and payment integrity requirements evolve across the industry. See resources from Healthcare Financial Management Association (HFMA) for additional guidance on healthcare finance and payment integrity.

Employers Trust Administrative Expertise

Most plan sponsors hire experienced TPAs and reasonably expect professional administration.

That trust often reduces demand for independent verification.

Audits Are Frequently Limited

Many organizations review only small samples of claims.

Sampling can identify patterns but cannot guarantee visibility into every payment.

Performance Metrics Focus Elsewhere

Service metrics often emphasize call center performance, turnaround times, and participant satisfaction.

Accuracy receives less attention than operational speed.

Data Is Difficult to Access

Claims data is often fragmented across vendors, making comprehensive oversight challenging without specialized tools.

The Real Cost and Impact

Even modest claims error rates can create substantial financial exposure.

Healthcare payment integrity studies consistently find that payment errors occur across public and private healthcare programs.

Multiple reviews by the U.S. Government Accountability Office (GAO) have highlighted the ongoing challenge of improper payments across healthcare systems.

A self-funded employer spending $15 million annually may view a 1% error rate as insignificant.

That seemingly small percentage equals $150,000 per year.

Over five years, the cumulative impact exceeds $750,000 before considering trend growth.

The consequences extend beyond direct overpayments.

Financial Leakage

Incorrect payments increase healthcare costs without improving outcomes.

Budget Distortion

Leadership teams make future decisions using inaccurate spending data.

Contract Compliance Risks

Undetected processing errors can indicate deviations from plan terms or administrative agreements.

Fiduciary Exposure

Under ERISA, plan fiduciaries must act prudently and solely in the interest of participants and beneficiaries.

The U.S. Department of Labor's Fiduciary Responsibilities Guidance outlines the standards fiduciaries are expected to follow when overseeing employee benefit plans.Failure to monitor service providers can create governance concerns.

According to the Employee Benefits Security Administration (EBSA), plan fiduciaries have a responsibility to prudently select and monitor service providers acting on behalf of the plan.

Lost Recovery Opportunities

Many overpayments become harder to recover as time passes and contractual recovery windows expire.

What's Actually Happening Behind the Scenes?

Most claims inaccuracies result from ordinary operational breakdowns rather than intentional misconduct.

Eligibility Errors

Employees or dependents may remain active in administrative systems after coverage should have ended.

Claims continue to be paid despite ineligible status.

Duplicate Payments

The same service can occasionally be paid more than once due to billing variations or processing workflows.

Incorrect Plan Provisions

System configurations may apply outdated deductibles, copays, or benefit limits.

Provider Contract Issues

Network discounts may not match contracted reimbursement terms.

Even small deviations can accumulate across thousands of transactions.

Coordination of Benefits Problems

Claims involving multiple coverage sources often create payment discrepancies.

Coding and Pricing Errors

Incorrect coding logic can affect reimbursement calculations.

Automation improves efficiency but can also scale mistakes rapidly.

Stop-Loss Reimbursement Gaps

Large claims may contain payment errors that affect reimbursement calculations and downstream reporting.

Vendor Integration Failures

Eligibility platforms, pharmacy systems, and claims platforms exchange data continuously.

Minor integration issues can create significant downstream effects.

Why Current Approaches Aren't Enough

Relying on TPA system edits and annual vendor reviews is not a claims accuracy program. It is claims processing with a thin layer of fraud detection.

How to Fix It: A Practical Action Plan

The solution is not to distrust your TPA. It is to implement independent oversight as a standard plan governance practice. Here are the steps that work.

Red Flags That Signal This Problem Applies to Your Plan

The ROI of Doing It Right

Strong claims oversight can generate measurable financial and governance value.

Independent claims audits commonly identify recoverable overpayments and process improvements.

Financial benefits often come from multiple sources.

Direct Recoveries

Previously undetected payment errors can be recovered when identified within applicable recovery periods.

Future Savings

Correcting root causes prevents repeated mistakes.

Improved Vendor Performance

Measurement drives accountability.

Vendors generally perform better when accuracy receives consistent attention.

Better Decision-Making

Cleaner data improves forecasting, budgeting, and plan design decisions.

Stronger Fiduciary Position

Documented oversight demonstrates prudent governance practices.

The most valuable outcome may not be the recovered dollars.

It may be the confidence that healthcare spending reflects intended plan design rather than avoidable administrative errors.

Frequently Asked Questions

Who is responsible for claims accuracy in a self-funded health plan?

The TPA processes claims, but plan sponsors retain ultimate responsibility for monitoring plan operations. Effective oversight typically involves HR, finance, consultants, and independent auditors working within a defined governance structure.

How common are healthcare claims errors?

Claims errors occur across all healthcare payment environments. Error frequency varies by plan structure, administration quality, and audit methodology. Even low error rates can create meaningful financial impact when applied to millions of dollars in annual claims spending.

Why isn't the TPA enough to ensure claims accuracy?

TPAs maintain internal controls and quality assurance processes. However, independent verification provides an additional layer of accountability. Organizations routinely audit financial statements despite having accounting teams. Claims oversight follows a similar principle.

What types of claims errors occur most often?

Common issues include eligibility mistakes, duplicate payments, coordination of benefits errors, pricing discrepancies, provider reimbursement issues, and incorrect application of plan provisions.

How often should self-funded plans conduct claims audits?

Many experts recommend periodic independent audits supported by ongoing monitoring. The appropriate frequency depends on plan size, complexity, annual spend, and organizational risk tolerance.

Can claims errors create ERISA fiduciary concerns?

Yes. ERISA requires fiduciaries to act prudently and monitor service providers. Consistent oversight helps demonstrate responsible governance and protection of plan assets.

What is the difference between a claims audit and a financial audit?

A financial audit evaluates financial reporting accuracy. A claims audit examines whether healthcare claims were processed and paid according to plan rules, contracts, and administrative requirements.

What is the business case for investing in claims oversight?

The value comes from recoveries, future savings, improved vendor accountability, better data quality, and stronger governance practices. Many employers view claims oversight as both a financial control and a fiduciary safeguard.

‍

‍