Explore Our Blog Library!

A self-funded plan is a balance sheet risk because the employer pays every claim from its own funds and holds fiduciary duty under ERISA for paying them correctly. Managing it like one means independent claims audits, contractual audit rights, monthly monitoring and documented committee review, rather than relying on a TPA's sample-based accuracy reports.

Eighty percent of covered workers at large firms now sit in a self-funded plan, according to the KFF 2025 Employer Health Benefits Survey. Those employers pay each claim from their own cash. Most also assume someone is checking that claims were paid correctly. In practice, your self-funded plan may be one of the least examined line items on your income statement.

What Makes a Self-Funded Plan a Balance Sheet Risk

A self-funded plan is a health benefit arrangement in which the employer pays employee medical claims directly from company funds instead of buying insurance. The employer carries the financial risk and the fiduciary duty for every claim, even when a TPA processes it. KFF found that 67% of covered workers are in self-funded plans, including 27% at firms with 10 to 199 workers and 80% at larger firms.

Most leaders assume the TPA absorbs the accuracy risk because it handles the paperwork. ERISA says otherwise. Section 404(a)(1)(B) places responsibility for claims accuracy on the plan sponsor, and the sponsor must select and monitor its TPA with prudence and loyalty.

On the balance sheet, claims are a liability that grows with every service your members use. Stop-loss insurance limits catastrophic claims, but it does not fix routine payment errors. Every dollar paid wrongly leaves plan assets for good unless someone recovers it.

Why the Oversight Gap Exists

Delegation feels like transfer, and that belief creates the gap. Once a plan sponsor signs an administrative services agreement, claims processing moves out of sight. The fiduciary duty does not move with it. Self-funded plans are also exempt from most state insurance laws, so no state regulator fills the void.

Contract terms widen the gap. Many agreements limit audit frequency, cap the number of claims an auditor may pull or restrict access to line-level data. Performance guarantees tend to measure turnaround time and self-reported accuracy rather than independently verified results. TPAs often lack the tools to deliver full claim detail on request, so the request quietly fades.

Capacity is the third cause. Benefits committees meet a few times a year, and their members rarely have claims processing backgrounds. Finance sees funding reports and HR hears employee complaints, but neither sees the payment-level detail where errors live. Vendor changes reset the clock as well, because a prior administrator's claims often go unreviewed when the plan moves to a new one.

The Real Cost of Thin Claims Oversight

Weak oversight costs money in the claims budget first. A plan that pays $25 million in claims each year with a 0.5% payment error loses $125,000 annually, and that money does not return on its own. Errors compound because the same incorrect pricing rule or eligibility flag repeats on every affected claim.

Regulatory exposure comes next. The Department of Labor's Employee Benefits Security Administration (EBSA) recovered more than $1.4 billion for retirement, health and welfare plans in fiscal year 2025. It closed 878 civil investigations, and 556 of them produced monetary results or corrective action. EBSA also opened 291 investigations from Benefits Advisor referrals, so participant complaints can start the process.

Litigation risk is real but unsettled. Courts dismissed suits against Johnson & Johnson and Wells Fargo over prescription drug management for lack of standing. A March 9, 2026 ruling in Stern from the Southern District of New York let some similar claims proceed, so sponsors should not treat standing as a permanent shield. Groom Law Group tracks these cases.

What Is Actually Happening Behind the Scenes

Pricing and contract errors

A claim priced under the wrong provider contract or fee schedule pays the wrong amount. The payment still looks plausible on a summary report, which is why these errors can survive for years. One incorrectly loaded rate can misprice every claim from a single facility, and high-dollar inpatient stays magnify the damage.

Eligibility and coordination failures

Claims for ineligible dependents, terminated employees or members with other primary coverage get paid when eligibility files lag behind reality. Coordination of benefits errors leave the plan covering costs that another carrier or Medicare should bear. Life events such as divorce or a spouse's new job often go unreported to the plan.

Duplicate billing and plan document mismatches

Duplicate charges and unbundled procedures can slip past automated edits. Services excluded by the plan document sometimes pay because the system was configured from an outdated version of it. Manual overrides made for urgent member requests add another layer. Each error looks small in isolation, but the annual total does not.

Why Current Approaches Fall Short

A claims audit is an independent review of paid claims against the plan document and provider contracts. Most plans rely instead on the TPA's own quality checks. One audit firm describes a standard TPA audit as a stratified sample of 250 to 400 claims, which equals roughly 0.3% to 0.5% of activity on a plan with 80,000 annual claims. Sampling at that level is like a restaurant inspector tasting one dish in a kitchen that serves eighty thousand meals.

TPA quality checks still catch processing defects, and they belong in the mix. They work best as one layer of a system that also includes independent verification.

How to Fix It

Closing the gap takes seven steps, and most can start within a quarter.

Red Flags That Signal the Problem Applies to Your Plan

A few warning signs reliably point to an oversight gap. Two or more justify a full review this plan year.

The ROI of Doing It Right

Independent oversight often pays for itself in year one. Audit firms report that a comprehensive independent audit typically recovers 1% to 3% of annual claims spend in its first year, often exceeding the audit's cost. On $10 million in annual claims, that range equals $100,000 to $300,000. It is an industry benchmark from audit providers rather than a government figure, and results vary by plan.

First-year recovery is a one-time catch-up. Continuous monitoring turns it into ongoing savings by fixing errors at the source. Clean claims data also strengthens stop-loss renewals and TPA contract negotiations. Recovery windows are set by contract, so earlier audits capture more.

Fiduciary protection is the third return. A documented and repeated review process gives counsel, insurers and regulators evidence of prudence that a vendor contract alone cannot supply. Sponsors should also confirm that fiduciary liability insurance is in force, since recent litigation has targeted health plan oversight. The EBSA monetary results fact sheet shows what enforcement looks like in practice.

Conclusion and Next Steps

A self-funded plan turns health claims into a balance sheet exposure that only the employer can manage. The TPA processes the claims, but the fiduciary record belongs to you. Sponsors who verify payments independently recover dollars, tighten contracts and build a defensible file.

Start this month. Ask your TPA for line-level claims data and a copy of its audit clause. Then request an independent claims audit assessment to size the opportunity on your plan. Contact our team to schedule a claims oversight review.

Frequently Asked Questions

How often should a self-funded plan be audited?

‍Many sponsors start with a full retrospective audit, then move to quarterly or ongoing reviews so errors surface while still small.

Does HIPAA prevent an independent claims audit?

‍No. Auditors typically work under business associate agreements, which permit access to the protected health information needed for the review.

Do level-funded plans carry the same duty?

‍Yes. Level-funded plans are still ERISA plans, and KFF reports 37% of covered workers at firms with 10 to 199 workers are in one.

Can a claims audit strain the TPA relationship?

‍Rarely. Most large TPAs expect independent audits, and a clear audit clause sets expectations before any review begins.

What did the 2026 PBM reforms change?

‍DOL proposed PBM fee disclosure and audit rules for self-funded plans on January 30, 2026, and Congress enacted related reforms February 3, 2026.

Does stop-loss cover claims paid in error?

‍Often not. Stop-loss contracts typically reimburse only claims paid according to the plan document, so payment errors can jeopardize reimbursement.

‍

‍

‍

‍

Healthcare claims leakage below stop-loss thresholds is money a self-funded plan pays out in error on claims too small to reach the stop-loss carrier. Because the employer funds every dollar under the deductible, duplicate payments, coding errors and eligibility mistakes typically escape a stop-loss carrier's review.

KFF's 2025 survey found that 80% of covered workers at larger firms are in self-funded plans, which means their employers pay claims from their own funds. Picture a 600-employee manufacturer with a $150,000 specific stop-loss deductible. A duplicate $9,000 facility payment clears on a Tuesday, and no stop-loss carrier ever sees it. That is healthcare claims leakage, and it sits in the layer most CFOs assume someone else is watching.

What Healthcare Claims Leakage Below Stop-Loss Thresholds Actually Means

Healthcare claims leakage is money a self-funded plan pays in error and never gets back. Stop-loss insurance protects an employer from large claims only, so mistakes on everyday claims land entirely on the employer.

Most sponsors assume a stop-loss policy means someone is checking the work. The contract says otherwise. A specific deductible is the per-person amount an employer pays before reimbursement begins, and SHRM explains that the insurer covers eligible claims only above that deductible.

Most claims by count sit beneath that line, and each one is small enough to escape notice. Think of stop-loss as a smoke detector on the top floor. It sounds for a large fire and stays silent about a slow pipe leak two floors below.

Why the Problem Exists

Claims leakage persists because no one in the payment chain is set up to check small claims one at a time. TPAs process claims at high volume with automated edits built for speed, and many lack the tools or contract scope for line-by-line financial review. That is a design limit and not a shortcoming of any one administrator.

Sampling widens the gap. A 300-claim sample from a plan that processes 80,000 claims reviews under 0.4% of activity. Everything outside the sample is assumed to resemble it.

Contracts add friction. Some administrative services agreements limit who may audit, how often and with what notice. Sponsors also assume the TPA or the stop-loss carrier already verifies payments, so nobody asks the question.

The Real Cost of Healthcare Claims Leakage

Unchecked payment errors drain plan funds dollar for dollar because no carrier covers the loss. CMS estimated that 6.55% of Medicare payments for individual healthcare services were improper in fiscal 2025, totaling $28.83 billion. Medicare is different from an employer health plan, and the figure includes underpayments and payments without enough documentation. Still, it shows how common payment errors can be, even in a system with extensive automated checks.

Apply simple arithmetic to your own plan. A self-funded plan that pays $10 million in claims loses $100,000 for every 1% paid in error. At KFF's 2025 average family premium of $26,993, that equals nearly four family policies.

Consequences reach beyond cash. The DOL states that fiduciaries who miss the basic standards of conduct may be personally liable to restore losses to the plan. Overpaid claims can also inflate the claims experience that sets next year's funding rates.

What's Actually Happening Behind the Scenes

Claims leakage traces back to four repeatable failure modes.

Duplicate Payments

A duplicate payment is a second payment for a service already paid. Resubmitted claims and altered dates of service are common triggers.

Coding and Unbundling Gaps

Unbundling means paying separately for parts of a procedure that a single code already covers. CMS designed its National Correct Coding Initiative edits to prevent improper payment when incorrect code combinations are reported. A plan without comparable edits can pay for the same work twice.

Eligibility and Coordination of Benefits

Coordination of benefits errors occur when a plan pays first although Medicare or another plan should pay first. Payments for members whose coverage has ended create the same problem.

Pricing and Plan-Term Mismatches

A claim paid above the contracted rate or beyond a plan limit looks normal in a summary report. It surfaces only when each line is compared with the contract and the plan document.

Why Current Approaches Aren't Enough

Today's oversight tools examine a thin slice of claims, so claims leakage often goes unmeasured. The table compares that status quo with a fiduciary-grade approach.

Speed metrics such as turnaround time show how fast claims move. They do not show whether the dollars moved correctly. None of this means TPAs are careless, because the standard toolset was never designed to answer the question a fiduciary must answer.

How to Fix It

A benefits committee can start these seven steps this quarter.

Red Flags That Signal the Problem Applies to Your Plan

The ROI of Doing It Right

Independent claims oversight pays for itself when it recovers even a small share of claims leakage. Audit firms report that full reviews typically recover 1% to 3% of annual claims spend. Those are vendor-reported figures and not government data, so ask any reviewer to document results from comparable plans. On $10 million in paid claims, that range equals $100,000 to $300,000.

Recovery is only the first return. Corrections to system edits and contract terms stop the same error from recurring in later plan years.

Fiduciary protection is the third return. ERISA Section 404(a)(1)(B) sets the prudence standard, and the DOL notes that prudence focuses on process, so decisions and their basis should be documented. A written monitoring record is the evidence that meets it.

Conclusion and Next Steps

Healthcare claims leakage below stop-loss thresholds is measurable, recoverable and preventable. The plans that address it start with one question at the next benefits committee meeting: what was our overpayment rate last year? If no one can answer with a number, request an independent assessment and record the request in the minutes.

Frequently Asked Questions

What is a claims audit?

‍A claims audit is an independent comparison of paid claims against plan terms and provider contracts to find payment errors.

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

‍The plan sponsor. DOL guidance says employers may delegate administration to a TPA but must monitor that provider periodically.

How often should a self-funded plan review claims?

‍ERISA sets no fixed schedule. DOL guidance calls for review at reasonable intervals, so more frequent checks catch errors while they are small.

Does the gag clause attestation prove claims were paid correctly?

‍No. It attests that contracts do not restrict data access. It does not verify that claims were paid correctly.

Is a sample audit enough?

‍Not for recovery. A sample estimates an error rate, but only the claims actually reviewed can produce a recovery.

Does healthcare claims leakage affect small employers too?

‍Yes. KFF reports 27% of covered workers at small firms are in self-funded plans, and every employer-funded dollar carries the same exposure.

Will independent review strain the TPA relationship?

‍It need not. Framed as a routine fiduciary process with agreed data requests, review gives both sides a shared record of payments.

‍

‍

‍

‍

Pharmacy data oversight means using your plan’s prescription claims to check that your PBM prices, pays and reports claims as your contract requires. Under ERISA, self-funded plan sponsors must carefully monitor service providers and make sure their compensation is reasonable. Reviewing detailed pharmacy claims regularly and recording the results helps fiduciaries show they met that responsibility.

The three largest pharmacy benefit managers handled nearly 80 percent of the roughly 6.6 billion prescriptions filled in the United States in 2023, according to the Federal Trade Commission. Your self-funded plan probably uses one of them.

Could you prove to a regulator this week that your PBM priced every claim according to your contract? Pharmacy data oversight helps you answer yes, and the Department of Labor is creating rules that expect you to have that proof.

What Pharmacy Data Oversight Means for a Self-Funded Plan

Pharmacy data oversight means regularly checking your plan’s prescription claims against your PBM contract and plan documents. A PBM summary report shows what the vendor chooses to share, while detailed claim data shows what your plan actually paid. Many sponsors assume an annual report or discount guarantee provides the same level of detail.

Relying on a summary report is like checking your bank balance without reading the statement. The total may look right while individual charges are wrong. A pricing guarantee can be met on average even when some claims are priced incorrectly, and a summary report will not show you which ones.

The responsibility falls on sponsors because self-funding means every claim dollar comes from the plan. The KFF 2025 Employer Health Benefits Survey  found that 67 percent of covered workers were in self-funded plans, including 80 percent at larger firms. Stop-loss coverage limits major losses but does not check whether your PBM priced everyday claims correctly.

Why the Oversight Gap Exists

The gap exists because the data, the contract terms and the internal expertise all sit outside the sponsor's reach. The Department of Labor says fiduciaries often cannot see the full payments that PBMs and their affiliates receive from drug manufacturers and pharmacies. That visibility problem is structural, so it does not point to bad intent by any one party.

Contracts add a second barrier. A benefits advisory firm notes that the audit rights DOL is proposing go beyond what most PBMs currently agree to. Sponsors who signed years ago may hold narrower rights than they assume, which matters most when a problem surfaces.

Internal capacity is the third barrier. Benefits committees often lack a pharmacist or data analyst, and many TPAs lack dedicated pharmacy analytics tools. Renewal talks then focus on headline discounts instead of claim behavior.

The Real Cost of Skipping Pharmacy Oversight

The cost shows up in plan dollars first. An FTC staff report found that pharmacies affiliated with the three largest PBMs generated more than $7.3 billion above estimated acquisition cost on specialty generic drugs from 2017 to 2022. PBM industry representatives have disputed the report's accuracy. The full findings are in the FTC's January 2025 release.

Legal exposure is the second cost. ERISA Section 409 makes a breaching fiduciary personally liable to restore plan losses. Recent suits against Wells Fargo and Johnson & Johnson allege that sponsors failed to manage their pharmacy benefit prudently, and a similar case targets JPMorgan Chase.

Two of those suits were dismissed for lack of standing, while a New York federal court let parts of the JPMorgan case proceed in March 2026. Standing rulings say nothing about whether the sponsors' oversight met ERISA's prudence standard. The Wells Fargo plaintiffs alleged, based on Form 5500 data, that the plan paid over $25 million in administrative fees to its PBM.

What Is Actually Happening Behind the Scenes

Spread Pricing

Spread pricing occurs when a PBM bills the plan more for a drug than it reimburses the pharmacy and keeps the difference. The FTC estimated that the three largest PBMs earned $1.4 billion this way on specialty generics. A plan cannot see that gap unless someone compares plan-paid and pharmacy-paid amounts claim by claim.

Affiliated Pharmacy Markups

Affiliated pharmacy markups arise when a PBM directs prescriptions to a pharmacy it owns and pays that pharmacy more than unaffiliated ones. The FTC found the largest PBMs reimbursed their own pharmacies at higher rates on nearly every specialty generic it examined. Those affiliated pharmacies captured 68 percent of specialty drug dispensing revenue in 2023, up from 54 percent in 2016.

Rebate Gaps and Adjudication Errors

Rebate leakage happens when contract terms allow some manufacturer payments to stay outside the rebates your plan should receive. A PBM may advertise a strong discount while the rebate definition excludes many of the dollars the plan could otherwise receive. Adjudication errors are more straightforward, such as paying claims for uncovered drugs or paying more than the contract allows. Milliman notes that pharmacy claims audits can review 100 percent of claims to identify these types of recurring problems.

Why Current Approaches Are Not Enough

Annual PBM reviews can fall short because they may focus on the vendor’s summary instead of the underlying claims. The DOL proposed a rule on January 29, 2026, under ERISA Section 408(b)(2) that would require PBMs to provide initial and semiannual disclosures about their compensation and financial arrangements. The proposal would also give plan fiduciaries the right to audit those disclosures. The DOL fact sheet explains how the proposal would work.

The final rule could change, but the direction is clear. Congress has also acted. The Consolidated Appropriations Act, 2026 adds new PBM reporting, disclosure and compensation requirements that generally apply to plan years beginning on or after August 3, 2028, or January 1, 2029 for calendar-year plans.

Waiting for those dates would leave two full plan years without a better process. The table shows what closes that gap.

How to Fix It in Seven Steps

A sound oversight routine starts with data you already own and a calendar you can keep. These seven steps build the evidence a fiduciary needs.

Red Flags That Your Plan Is Exposed

Seven warning signs suggest your pharmacy oversight would struggle under scrutiny. Any one of them justifies a closer look this quarter. Several together suggest the committee is relying on trust instead of evidence.

The ROI of Doing It Right

The return comes from recovered dollars, better contract terms and a defensible fiduciary record. Segal reported that a large county public employee plan covering 45,000 people recovered $5.4 million through PBM audits, with annual spend above $550 million. That recovery equals roughly 1 percent of total spend.

No public benchmark sets a typical recovery rate. Results depend on contract language, drug mix and how much data the PBM releases. Audit findings can also inform renewal negotiations on pricing, rebates and audit terms.

Documentation is the fiduciary protection. The DOL proposal would give relief to fiduciaries who reasonably relied on PBM disclosures and took corrective action after finding noncompliance, though the PBM would not share that relief. A written record of what you tested and what you did is what supports that position.

Conclusion and Next Steps

Pharmacy data oversight turns a broad fiduciary duty into a routine you can prove on demand. Sponsors who begin now will hold a dated record long before any final rule or statute reaches their plan year. That record is what a committee, an auditor or an EBSA investigator will ask to see first.

Start with three actions this quarter. Request claim-level files from your PBM or TPA. Read your audit clause carefully, then add pharmacy to the next committee agenda.

If you want an independent view, schedule a pharmacy claims review with our team. We test your claim file against your contract and deliver findings your committee can act on.

Frequently Asked Questions

Does ERISA require self-funded sponsors to monitor their PBM?

‍Yes. ERISA Section 404(a)(1)(B) requires prudent administration, and Section 408(b)(2) requires reasonable service contracts with reasonable compensation.

Is the DOL's PBM fee disclosure rule final?

‍Not at last check. The DOL proposed it January 29, 2026 and extended comments to April 15. Confirm current status on DOL.gov.

What data should we request from the PBM?

‍Claim-level files showing drug, quantity, dispensing pharmacy, plan-paid amount, pharmacy-paid amount, fees and rebate allocation for every prescription.

What does a pharmacy claims audit test?

‍A pharmacy claims audit checks each claim against contract pricing and plan coverage rules to find overpayments and errors.

How often should we review pharmacy claims?

‍Review quarterly for trends and audit independently once a year. That cadence surfaces drift before renewal talks begin.

Can our TPA handle oversight for us?

‍A TPA can supply data and analysis, but you remain responsible for monitoring the work. Pairing a TPA with an independent auditor adds objectivity.

Can the DOL investigate our health plan?

‍Yes. EBSA has authority to investigate ERISA-covered health plans, including how fiduciaries selected and monitored service providers.

What if the PBM refuses to share claim data?

‍Review your contract audit clause first. Escalate through counsel or your consultant and document every request. Refusal is itself a finding worth recording.

‍

‍

The Consolidated Appropriations Act requires self-funded plan sponsors to review PBM payment disclosures, bans gag clauses that prevent access to claims and pricing data, and holds plan fiduciaries personally responsible for checking that PBM contracts are fair. Ignoring these disclosures can itself become a fiduciary breach under ERISA.

A 1,400-employee manufacturer conducted its first independent claims audit in 2024, eighteen months after starting a new relationship with a pharmacy benefit manager. The audit uncovered $812,000 in overpayments, including a specialty drug billed at 240% of the agreed contract rate. The TPA's internal reviews had missed every one of these problems.

Most HR leaders and CFOs assume their broker or PBM is already providing the information they need to manage their health plan properly. But legally, the responsibility still falls on the plan sponsor and its fiduciaries. ERISA can hold individual fiduciaries personally responsible when they fail to meet their duties, and those duties include making sure the plan's pharmacy benefits are managed properly.

What PBM Transparency Actually Means Under the CAA

A plan sponsor's fiduciary duty under the CAA is not satisfied by simply hiring a PBM. It requires an ongoing, documented process of reviewing what that PBM is paid and how it is paid. Many benefits leaders assume signing a PBM contract discharges their obligation. The law treats that moment as the start of the duty, not the end of it.

The CAA amended ERISA to require certain covered service providers to disclose their compensation. The CAA 2021 disclosure requirements apply to persons who provide brokerage services or consulting to ERISA-covered group health plans who reasonably expect to receive $1,000 or more in direct or indirect compensation. This disclosure has to happen before the contract is signed or renewed, not buried in a report months later.

A newer law raises the bar further. The Consolidated Appropriations Act of 2026 treats PBMs as covered service providers under ERISA Section 408(b)(2), subject to compensation disclosure requirements, and requires 100% rebate and remuneration pass-through to ERISA plans with limited exceptions for bona fide service fees.

It also mandates semiannual reporting of detailed drug pricing, spread pricing, rebate, and compensation data to group health plans. Most of these newer provisions phase in over the next several years, but the direction of travel is unmistakable: PBM relationships are moving from private contract terms to statutory obligations.

Why the Problem Exists

PBM compensation has historically been structured so plan sponsors could not see the full picture, even when they asked. Spread pricing is the clearest example. A PBM bills the plan one amount for a drug and pays the pharmacy a lower amount, keeping the difference without disclosing it as compensation.

State Medicaid audits give a sense of scale. Pennsylvania found that taxpayer payments to PBMs for Medicaid enrollees more than doubled from $1.41 billion to $2.86 billion between 2013 and 2017, and Ohio's state auditor found PBMs pocketed $224.8 million through spread pricing alone in a single year, out of $2.5 billion spent annually. Employer plans are not immune to the same dynamics. They simply have less regulatory scrutiny forcing the numbers into daylight.

Contract terms compound the visibility problem. Many PBM agreements historically included gag clauses that restricted a plan's ability to see claims-level pricing data or compare it to market rates. The CAA specifically targeted this practice, but old habits and vague contract language still linger in many renewals.

The Real Cost and Impact

Undetected PBM and claims errors translate directly into inflated plan spend, and the dollars involved are rarely trivial. Carrier post-pay sampling reviews typically cover only 3 to 5 percent of claims, while independent analysis of 100 percent of claims has consistently identified 5 to 12 percent error rates. That gap between sampled review and full review is where money disappears.

Industry estimates put the overpayment error rate at 2 percent to 5 percent of overall medical claim costs each year, even at the best claims administrators. For a plan spending $20 million annually on claims, a 3 percent error rate translates to roughly $600,000 a year in avoidable losses. Multiply that across a five-year contract term and the number becomes difficult to ignore in a board meeting.

Litigation risk adds a separate cost layer. In one closely watched case, plaintiffs alleged an employer breached its fiduciary duty by agreeing to pay its PBM higher prices for generic drugs when those same drugs were available at lower prices, and by steering beneficiaries to the PBM's mail-order pharmacy where prices were routinely higher than retail. The claims were dismissed twice on standing grounds rather than on the merits, but the legal exposure and the defense costs were real regardless of outcome.

What's Actually Happening Behind the Scenes

Spread Pricing Without Disclosure

A PBM can charge a plan more for a drug than it pays the dispensing pharmacy and record that markup as revenue rather than as compensation. Without an independent audit comparing PBM invoices to actual pharmacy reimbursement, a plan sponsor has no way to see this gap.

Rebate Retention

Manufacturer rebates are negotiated by the PBM using the plan's purchasing volume, but the PBM does not always pass the full rebate back to the plan. The PBM retains a portion or all of the negotiated rebate as its compensation, and the exact split is frequently unclear in contract language.

Formulary Steering and Specialty Drug Markups

PBMs influence which drugs are preferred on a formulary, sometimes in ways tied to their own rebate economics rather than the lowest net cost to the plan. Vertically integrated PBMs that own specialty pharmacies or mail-order channels have an added incentive to steer volume toward their own affiliates.

Claims Processing Errors Hiding Inside "Accurate" Reports

TPAs and PBMs typically self-report high accuracy rates. A TPA's self-reported rate is often around 100 percent, while an independent review found actual financial accuracy closer to 96.8 percent and payment accuracy closer to 96.1 percent, both below the 98 percent service level standard. The gap between self-reported and independently verified accuracy is where dollars quietly leak out of a plan.

Why Current Approaches Aren't Enough

Most plan sponsors rely on their TPA or PBM's own quarterly reporting and treat the annual renewal conversation as sufficient oversight. That approach reviews a small, PBM-selected slice of the data and rarely includes an independent comparison to actual pharmacy reimbursement.

How to Fix It

Red Flags That Signal the Problem Applies to Your Plan

The ROI of Doing It Right

Independent claims audits routinely identify overpayments in the mid to high single digits as a percentage of total claims spend. ClaimInformatics' independent analysis consistently identifies 5 to 12 percent error rates across 100 percent of claims reviewed, well above what carrier sampling ever surfaces.

Recovered dollars are only part of the return. A documented, ongoing fiduciary review process is also the primary defense if a plan is ever investigated or sued. EBSA announced a significant overhaul of its national enforcement priorities for fiscal year 2026, with a pronounced shift of investigative resources toward health and welfare plans and the service providers who operate them. Plans that can show board minutes, disclosure reviews, and audit reports are in a materially different position than plans that cannot.

Think of PBM oversight like an annual physical rather than a one-time checkup. Skipping it does not mean nothing is wrong. It only means no one has looked yet, and problems that go unmeasured tend to compound quietly until the bill comes due all at once.

Conclusion and Next Steps

PBM transparency is no longer a negotiating preference. It is a statutory expectation backed by an ERISA fiduciary duty that falls on the plan sponsor, not the vendor. The plans best positioned for 2026 and beyond are the ones that treat disclosure review, contract audit rights, and documented committee oversight as a standing part of how the plan runs, not as a once-a-year renewal task.

Start with one question in your next benefits committee meeting: when was the last time someone outside your TPA or PBM independently verified the numbers you were given. If no one can answer that with a date and a report, that is the place to begin.

Frequently Asked Questions

What does the CAA require of self-funded plan sponsors regarding PBMs?

‍It requires covered service providers, including PBMs, to disclose direct and indirect compensation to plan fiduciaries before contracts are signed or renewed.

Is a plan sponsor personally liable for PBM pricing problems?

‍Individual fiduciaries can be held personally liable under ERISA for breaches of fiduciary duty, including inadequate PBM oversight.

How often should a self-funded plan audit its PBM contract?‍

Most fiduciary advisors recommend an independent audit every one to two years, with ongoing monitoring in between.

What is spread pricing?

‍It is when a PBM bills the plan more for a drug than it pays the pharmacy, keeping the difference as undisclosed revenue.

Does the CAA apply to fully insured plans too?

‍The disclosure and gag clause provisions apply broadly to group health plans, though self-funded plans carry more direct fiduciary exposure.

What happens if a plan ignores 408(b)(2) disclosures?

‍Failing to review or act on required disclosures can itself constitute a fiduciary breach, separate from whether pricing was unreasonable.

What is the difference between CAA 2021 and CAA 2026 PBM provisions?

‍CAA 2021 introduced compensation disclosure requirements; CAA 2026 adds mandatory rebate pass-through and detailed PBM reporting obligations, signed into law February 3, 2026.

Can a broker or consultant also trigger CAA disclosure requirements?

‍Yes. The CAA 2021 mandated transparency improvements including broker and consultant compensation disclosures, not PBMs alone.

‍

‍

‍

A fiduciary intelligence layer is an independent, ongoing system of claims data review, TPA performance monitoring, and documentation that sits above a self-funded health plan’s third-party administrator. It exists because ERISA makes the plan sponsor, not the TPA, legally responsible for ensuring that claims are paid correctly.

A mid-size manufacturer with 650 covered employees paid the same six-figure inpatient claim twice. Nobody caught it for eleven months. The TPA's quarterly report showed 98% payment accuracy the entire time. That gap between what a TPA reports and what a plan sponsor actually owes under ERISA is the reason "fiduciary intelligence" is becoming a distinct discipline rather than a compliance afterthought.

What a Fiduciary Intelligence Layer Actually Is

A fiduciary intelligence layer is an independent, ongoing system of claims review and TPA performance monitoring that sits above the plan's day-to-day administration. Most employers assume their TPA's internal quality checks satisfy their oversight duty. In reality, TPA self-audits typically sample 250 to 400 claims out of tens of thousands processed annually.

That sample size covers a fraction of a percent of total claims volume. The remaining claims move through the system without independent review. A benefits committee that only reads the TPA's summary report is reviewing the TPA's opinion of itself, not the plan's actual financial and legal exposure.

The distinction matters because ERISA does not treat "we trusted our vendor" as a defense. [internal link: what is a fiduciary intelligence layer explainer] The fiduciary duty to monitor service providers is ongoing, not a one-time contract signature.

Why the Oversight Gap Exists

The gap exists because claims administration was built for throughput, not scrutiny. TPAs are compensated to process claims fast and keep denial rates low enough to avoid participant complaints. Neither incentive rewards catching an overpayment after the fact.

Volume compounds the problem. A plan processing 80,000 claims a year cannot be meaningfully checked by a quarterly spot audit, no matter how skilled the reviewer. Automation helps but does not close the gap. The 2025 CAQH Index estimates roughly $21 billion in remaining administrative automation savings across the industry, with claims processing still one of the least automated functions for complex, high-dollar claims.

Plan sponsors also underestimate how contractual language works against them. Service level agreements typically guarantee 98% accuracy against the TPA's own error definitions, not an independent standard. [internal link: TPA performance guarantees guide] A guarantee measured by the party being guaranteed is not oversight. It is a marketing figure.

The Real Cost of Skipping Independent Oversight

Willis Towers Watson puts the industry-standard TPA error rate at 1% to 3% of total claims processed, and some independent audits find error rates running higher on complex claim types. On a plan spending $10 million a year on claims, even the low end of that range represents real dollars leaving the plan every month. One Baker Tilly audit example found a TPA's actual financial accuracy at 96.8% and payment accuracy at 96.1%, both below the 98% SLA the contract required.

The cost is not only financial. Under DOL and EBSA guidance, fiduciaries who fail to monitor claims administration expose themselves and the plan to personal liability, not just plan-level penalties. EBSA closed 878 civil investigations in FY2025, with 63% producing monetary recoveries or corrective action, totaling $714.4 million.

A useful comparison: nobody would let a company's outside payroll vendor self-certify that every paycheck was correct with no independent check. Health claims run through a similar external vendor relationship, at a much larger dollar volume, with far less routine verification.

What's Actually Happening Behind the Scenes

Claims Leakage Hides in Ordinary-Looking Payments

Most overpayments are not fraud. They are duplicate payments, incorrect coordination of benefits, out-of-network claims paid at in-network rates, and specialty pharmacy claims billed above the negotiated rate. Each one looks routine in isolation.

Coding and Billing Drift

Upcoding and unbundling shift costs upward gradually, one claim at a time, without triggering any single red flag. A TPA's automated adjudication system approves claims that pass basic logic checks even when the underlying billing does not match the service rendered.

Eligibility and Dependent Drift

Dependents who should have been removed from coverage, such as an ex-spouse or an adult child who aged out, continue generating claims for months or years. Nobody outside a dedicated eligibility audit typically catches this pattern.

Why Current Approaches Aren't Enough

The status quo relies on the TPA to grade its own work once a year. A fiduciary intelligence layer replaces that with continuous, independent review built specifically to satisfy the plan sponsor's fiduciary duty rather than the vendor's contract renewal.

How to Build a Fiduciary Intelligence Layer

Red Flags That Signal Your Plan Needs This Now

The ROI of Getting This Right

A full independent claims audit typically recovers 1% to 3% of annual claims spend, according to industry benchmarking. On a $15 million plan, that is $150,000 to $450,000 in a single review cycle, before counting the ongoing savings from corrected processes going forward.

The fiduciary protection matters as much as the dollars. Documented, continuous oversight is the evidence a plan sponsor needs if EBSA opens an inquiry or a participant lawsuit alleges a breach of fiduciary duty. Kaiser Family Foundation's 2025 survey found that 67% of covered workers are now enrolled in self-funded plans, meaning this exposure applies to a majority of the employer-sponsored market, not a niche segment.

Employers that treat oversight as a line item rather than a favor from their broker tend to catch problems earlier and pay auditors less over time, because the error patterns get fixed instead of repeating every quarter.

Conclusion and Next Steps

Self-funded employers carry the fiduciary weight of every claim paid on their plan, whether or not anyone reviewed it. A fiduciary intelligence layer turns that exposure into a managed, documented process instead of an open question. The employers who build this now, before EBSA or a participant lawsuit forces the issue, are the ones protecting both their claims dollars and their fiduciary standing.

Start with one claims-level data pull from your TPA and a conversation with your benefits committee about when the plan was last independently audited. That single step usually reveals how much oversight is actually happening today.

Frequently Asked Questions

What is a fiduciary intelligence layer?

‍An independent, ongoing system of claims review and TPA monitoring that documents plan sponsor compliance with ERISA fiduciary duties.

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

‍The plan sponsor, under ERISA Section 404(a)(1)(B), even though the TPA processes the claims.

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

‍Continuous or quarterly review is best practice; annual sampling alone leaves most claims unreviewed.

What percentage of claims typically contain errors?

‍Industry benchmarks from Willis Towers Watson put TPA error rates at 1% to 3% of total claims processed.

Can a TPA's self-reported accuracy be trusted alone?

‍No. Self-reported figures use the TPA's own error definitions and sample only a small fraction of claims.

What does EBSA look for in claims oversight enforcement?

‍Documented, independent monitoring of TPA performance, not just a signed administrative services contract.

How much can an independent claims audit recover?

‍Typically 1% to 3% of annual claims spend, based on published audit benchmarks.

Is claims oversight only relevant for very large employers?

‍No. Sixty-seven percent of covered workers nationally are in self-funded plans, spanning a wide range of employer sizes.

‍

‍

‍

Pre-pay claims oversight is the review of a medical claim for accuracy, medical necessity, and coding integrity before the plan pays it. It catches duplicate billing, unbundling, and eligibility errors before money leaves the plan, unlike post-payment "pay and chase" recovery, which is slower, more adversarial, and recovers only a fraction of what it should.

In fiscal year 2025, the Medicare Fee-for-Service improper payment rate was 6.55 percent, or $28.83 billion, according to CMS's FY2025 Improper Payments Fact Sheet. That is the government's most heavily audited payment system, backed by decades of federal oversight infrastructure.

Most self-funded employer health plans have no comparable review layer at all. A claim gets billed, the TPA's system auto-adjudicates it, and the check goes out, often with nobody outside the TPA looking at the claim before that happens.

What Pre-Pay Claims Oversight Actually Means

Pre-pay claims oversight is the review of a medical claim for coding accuracy, medical necessity, and billing integrity before the plan releases payment. Most employers assume their TPA already does this through standard claims processing. In practice, TPA systems apply automated edits built for speed and volume, not independent scrutiny of every line.

That assumption collides with the numbers. Willis Towers Watson has repeatedly found that 1 to 3 percent of total claims spend flows out in preventable overpayment errors even at well-run TPAs, largely because auto-adjudication is tuned to move claims fast rather than question them. On a $20 million plan, that range alone represents $200,000 to $600,000 a year.

Claims adjudication and claims oversight are not the same function. Adjudication decides whether a claim is payable under plan terms. Oversight asks a second, independent question: is this specific claim priced, coded, and billed the way it should be, before that payment becomes final.

Why the Problem Exists

TPAs process claims at enormous volume, often millions per year across a book of clients, and their systems are built to keep pace with that volume. Structural throughput, not intent, is what limits how deep any single claim gets reviewed. The 2025 CAQH Index reports that a large majority of claims move through straight-through auto-adjudication with minimal human review, industry estimates commonly place that share between 80 and 85 percent.

Plan sponsors compound the gap by treating TPA selection as a one-time decision rather than an ongoing fiduciary duty. Many benefits committees never ask what percentage of claims receive a true prepayment review versus automated pass-through. ERISA Section 404 requires a prudent expert standard of ongoing monitoring, and claims payment accuracy sits squarely inside that duty.

Incentive structures rarely reward deeper review either. A TPA's contract is typically priced on a per-employee-per-month basis tied to processing volume, not on claims accuracy outcomes, so oversight becomes a service the plan sponsor has to actively request rather than one built into the base relationship.

The Real Cost or Impact

The dollar figures here are not theoretical. Willis Towers Watson's 1 to 3 percent error benchmark, applied to a typical $30 million self-funded plan, translates to $300,000 to $900,000 a year in preventable claims errors. A single Baker Tilly audit example found actual claims accuracy running at 96.8 percent and 96.1 percent against a contracted 98 percent SLA, a gap that sounds small until it is priced across a full year of claims.

Vendors marketing prepayment editing report meaningful upside from closing that gap. Cotiviti has publicly cited medical cost savings of up to 4 percent of annual claims spend for clients using its prepay claim editing programs, a figure that should be read as vendor-reported rather than an independent benchmark. Even a conservative reading of the range between Willis Towers Watson's audit findings and vendor-reported prepay results points to real, recoverable money.

The cost is not only financial. The DOL's Employee Benefits Security Administration recovered more than $1.4 billion for plans, participants, and beneficiaries in fiscal year 2025, with $714.4 million of that coming directly from enforcement actions. Fiduciary exposure from unmonitored claims payment practices is not a hypothetical risk category anymore.

What's Actually Happening Behind the Scenes

Auto-adjudication blind spots

Automated claims systems are excellent at catching hard stops like invalid procedure codes or missing eligibility. They are far weaker at catching claims that are technically clean but substantively wrong, such as unbundled procedures billed separately or services billed at a higher intensity than documentation supports.

Volume over verification

A claims examiner reviewing thousands of claims a week cannot meaningfully scrutinize each one. Coordination of benefits errors, duplicate billing across providers, and dependent eligibility drift routinely slip through simply because nobody had the bandwidth to check.

Post-payment recovery friction

Once a claim is paid, recovering an error becomes a negotiation rather than a correction. Providers dispute post-payment recoupment requests, appeals stretch for months, and plans often settle for partial recovery just to close the file.

Dependent eligibility drift

A dependent who ages out, divorces, or gains other coverage does not always get removed from the plan promptly. Claims for that dependent still adjudicate cleanly because the eligibility system was never updated, and the plan pays claims for someone who should no longer be covered at all.

Why Current Approaches Aren't Enough

Most plans still rely entirely on the TPA's built-in adjudication engine and treat any deeper review as an occasional post-payment audit. That model reacts to errors long after the plan's money has already moved.

How to Fix It

Red Flags That Signal the Problem Applies to Your Plan

The ROI of Doing It Right

Applying the Willis Towers Watson 1 to 3 percent benchmark, a $25 million plan carries $250,000 to $750,000 a year in preventable claims errors under a pay and chase model alone. Closing even the lower half of that range funds a prepayment oversight program many times over in its first year.

The fiduciary protection matters just as much as the dollars. A documented, ongoing claims oversight process is exactly the kind of prudent expert conduct that DOL/EBSA and plaintiffs' attorneys now scrutinize in ERISA fiduciary litigation. With 67 percent of covered workers now enrolled in self-funded plans according to KFF's 2025 Employer Health Benefits Survey, this exposure now touches the majority of the employer market, not a narrow slice of it.

Conclusion and Next Steps

Pre-pay claims oversight moves error correction to the only point where it actually saves money: before the payment leaves the plan. The benchmarks are consistent across sources, from Willis Towers Watson's error data to DOL/EBSA's enforcement recoveries, and they all point the same direction. Waiting for an annual post-payment audit is no longer a defensible substitute for ongoing review.

Start with one question at your next benefits committee meeting: what percentage of our claims receive true prepayment review before they are paid. If nobody in the room can answer that with a number, that is the clearest red flag covered in this guide.

Treat pre-pay oversight the way you treat any other fiduciary duty on the plan: something you revisit on a schedule, document in committee minutes, and hold your TPA accountable to with real numbers. The plans that build this habit early are the ones that catch the $300,000 error before it clears, not the ones still negotiating it back eighteen months later.

Frequently Asked Questions

What is pre-pay claims oversight?

‍Review of a claim for accuracy and medical necessity before the plan pays it, catching errors before money leaves the plan.

How is pre-pay oversight different from a claims audit?

‍A claims audit reviews claims after payment. Pre-pay oversight reviews claims before payment executes.

Does my TPA already do this?

‍Most TPAs run automated adjudication edits, but few provide independent prepayment review layered on top.

How much can a self-funded plan save with pre-pay oversight?

‍Industry benchmarks point to 1 to 3 percent of annual claims spend in preventable errors caught before payment.

Is claims oversight an ERISA fiduciary requirement?

‍ERISA Section 404 requires ongoing prudent monitoring of plan administration, which includes claims payment accuracy.

Which claim types benefit most from pre-pay review?

‍Specialty pharmacy, high-cost inpatient claims, and out-of-network billing typically carry the highest error concentration.

Can pre-pay oversight run alongside my existing TPA?

‍Yes. Independent oversight layers on top of the TPA relationship without disrupting claims processing timelines.

Who should introduce pre-pay oversight to a benefits committee?

‍Brokers, consultants, or fiduciary advisors typically bring this forward as part of ongoing plan governance.

‍

‍

‍

A conflict of interest in a self-funded benefits ecosystem exists whenever a broker, consultant, TPA, or PBM earns compensation tied to decisions that should serve the plan and its members. Under ERISA, plan sponsors carry fiduciary responsibility to identify these conflicts, demand disclosure, and verify claims performance independently rather than relying on vendor self-reporting.

A 1,400-employee manufacturer ran its first independent claims audit in 2024, eighteen months into a new TPA relationship. The review found $812,000 in overpayments, including a single inpatient claim paid twice at $47,000 and a specialty drug billed at 240 percent of the contracted rate.

None of it appeared in the TPA's own quarterly reports, which had shown claims processing performance near 100 percent all year. That gap between reported accuracy and actual accuracy is where conflicts of interest in self-funded health plans live, quietly, for years.

What a Conflict of Interest Actually Looks Like in a Benefits Plan

A conflict of interest exists whenever a party administering or advising on your plan has a financial incentive that competes with your plan's financial interest. Most plan sponsors assume their broker works for them because the broker sits across the table during renewal meetings. In reality, brokers and consultants frequently earn commissions, overrides, and bonuses from the same carriers, TPAs, and PBMs they are recommending.

The assumption that "our advisor is on our side" collides with how compensation actually flows. A broker who receives a volume bonus from a specific carrier has a reason, even a subtle one, to steer business there. This is not necessarily dishonesty. It is structural incentive, and structural incentives shape recommendations whether anyone intends them to or not.

TPAs sit in a similar position. They process claims and report their own accuracy, then hand employers a scorecard showing they performed well. The employer rarely has an independent way to check that scorecard against the underlying claims data.

Why the Problem Exists in the First Place

The root cause is that most self-funded plans outsource claims administration and then stop watching. Benefits committees spend months negotiating stop-loss terms, network discounts, and PBM rebates. Once the contract is signed, oversight often ends there.

TPAs are not financially responsible for the plan they administer. The plan sponsor bears the cost of every claim paid, correct or not, so the TPA has limited financial incentive to catch its own errors. This is not a character flaw in TPAs; it is a structural mismatch between who bears the risk and who controls the process.

Regulatory history compounds the gap. Before the CAA took effect in December 2021, compensation disclosure rules under ERISA's prohibited transaction provisions applied primarily to retirement plans, leaving health plan brokers and consultants largely exempt from formal disclosure requirements. Health plan sponsors got used to not asking, because for years there was no legal mechanism forcing the conversation.

The Real Cost of Unmanaged Conflicts

Unmanaged conflicts of interest translate directly into overpaid claims and unrecovered dollars. Most TPAs self-report claims accuracy above 96 percent, yet independent audits routinely find 1 to 10 percent of claims dollars paid in error. That gap, multiplied across a plan paying tens of millions in annual claims, produces real money.

Most self-funded employer health plans review fewer than 5 percent of their claims, and industry-documented TPA error rates run between 3 and 10 percent, so a meaningful share of overpayments simply never surfaces. One claims analytics firm reports finding average per-employee overpayments of $500 to $1,200 per year across its client base, with error detection rates between 5 and 15 percent once claims are independently reviewed.

The consequences are not hypothetical. In September 2025, Aetna and Optum settled for $8.4 million over allegations of fabricated billing codes that concealed administrative fees inside medical charges, inflating out-of-pocket costs for members for nearly a decade. Cases like this show why "the TPA says everything is fine" cannot be the end of a fiduciary's inquiry.

What's Actually Happening Behind the Scenes‍

Broker and Consultant Compensation That Isn't Fully Visible

The CAA requires most brokers and consultants serving ERISA group health plans to disclose, in writing, all direct and indirect compensation they receive for their services. Before this rule, an employer could pay a broker a stated fee while the broker collected additional override commissions or bonuses from carriers behind the scenes.

Disclosure improved visibility, but it did not eliminate the incentive. A broker can disclose a compensation arrangement in full and still be nudged, quarter after quarter, toward the carrier that pays the richest override.

PBM Rebate and Spread Pricing Arrangements

Pharmacy benefit managers often retain a portion of manufacturer rebates rather than passing the full amount to the plan. Some PBMs also engage in spread pricing, charging the plan more for a drug than they reimburse the pharmacy and keeping the difference.

Plan sponsors frequently cannot see these mechanics without contract-level audit rights. Without a spread pricing and rebate audit clause, a plan sponsor is trusting the PBM's own math on money the PBM itself is collecting.

TPA Self-Reporting Versus Independent Verification

One large carrier's own administrative services contracts acknowledged a 1.4 percent claims processing error rate while simultaneously guaranteeing 99 percent accuracy to plan sponsors. That is not necessarily fraudulent; it reflects how narrowly "accuracy" gets defined in a TPA's own performance guarantee versus how an independent auditor defines it.

A weighted financial accuracy rate of 99.73 percent on a sampled audit can still mean thousands of dollars in errors depending on claim volume and dollar concentration. High percentage accuracy and low dollar impact are not the same thing, and TPA scorecards rarely make that distinction obvious.

Why Current Approaches Aren't Enough

Most plan sponsors believe they are covered because they receive quarterly TPA reports and completed a CAA compensation disclosure form. Those steps satisfy a paperwork requirement. They do not verify that claims were actually paid correctly or that compensation influenced vendor selection in ways the plan sponsor would object to if it saw the full picture.

Research comparing audit methodologies found that random-sample audits missed a significant share of claim errors, ranging from $200,000 to $750,000 in value, that a full claims review would have caught. Sampling has a place, but treating it as sufficient oversight leaves real money on the table.

How to Fix It: A Fiduciary Action Plan

Red Flags That Signal a Conflict May Be Affecting Your Plan

The ROI of Doing It Right

A comprehensive independent claims audit typically recovers 1 to 3 percent of annual claims spend in its first year, often exceeding the cost of the audit itself. For a plan paying $20 million in annual claims, that range represents $200,000 to $600,000 in first-year recoveries alone.

Beyond direct recovery, ongoing quarterly monitoring tends to reduce the error rate going forward because vendors know the plan is watching. This is the same reason a store places a mirror near the register: not because it catches every incident, but because visible oversight changes behavior before an incident occurs.

Fiduciary protection is the less visible but equally important return. Maintaining board minutes, committee charters, and audit reports demonstrates that the plan sponsor followed a prudent review process, which matters enormously if a participant or the Department of Labor ever questions how the plan was managed.

Conclusion and Next Steps

Conflicts of interest in a self-funded benefits ecosystem rarely announce themselves. They show up as a slightly favorable renewal recommendation, a claims report that always lands near 100 percent, or a PBM contract that never quite specifies where the rebate money goes. None of these are automatically evidence of wrongdoing, but all of them are questions a prudent fiduciary should ask and document.

The plan sponsors who avoid six-figure surprises are the ones who treat compensation disclosure and claims auditing as ongoing fiduciary practice, not one-time compliance checkboxes. Start with a compensation disclosure review this quarter and schedule an independent claims audit before your next renewal cycle.

Frequently Asked Questions

What counts as a conflict of interest in a self-funded health plan?

‍Any arrangement where a broker, TPA, or PBM earns compensation tied to decisions affecting your plan's costs, rather than solely to your plan's outcomes.

Does the CAA require TPAs to disclose compensation, or just brokers?

‍The CAA's Section 408(b)(2) disclosure requirement applies to brokers and consultants providing brokerage or consulting services, not TPA claims processing generally.

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

‍At minimum, annually. Continuous or quarterly monitoring catches errors faster and often costs less over time than a single large annual review.

Who holds fiduciary liability if a TPA makes claims errors?

‍ERISA places fiduciary responsibility for claims accuracy on the plan sponsor, regardless of whether claims administration has been delegated to a TPA.

What is spread pricing in a PBM contract?

‍It is when a PBM charges the plan more for a drug than it reimburses the pharmacy, keeping the difference without disclosing the markup.

Can a broker legally receive compensation from more than one source?

‍Yes, but the CAA requires written disclosure of all direct and indirect compensation before the plan fiduciary can determine it is reasonable.

What is a reasonable TPA claims error rate?

‍Industry standard estimates put typical administrator error rates at roughly 1 to 3 percent of total claims processed, though independent audits often find higher rates than TPA self-reports suggest.

Should audit rights be negotiated before or after signing a TPA contract?

‍Before. Unrestricted audit rights are far harder to add after a contract is executed than to negotiate during initial terms.

‍

‍

A TPA performance review is a documented, periodic evaluation of a third-party administrator's claims accuracy, financial payment accuracy, service levels and regulatory compliance measured against the plan document and contract terms. Plan sponsors conduct these reviews, often alongside an independent claims audit, to satisfy ERISA fiduciary duty and control health plan spending.

A manufacturer with 1,400 employees ran its first independent claims audit in 2024, eighteen months after switching TPAs. The review turned up $812,000 in overpayments, including one inpatient claim paid twice for $47,000 and 63 dependents who should have been dropped from the plan years earlier.

None of it showed up in the TPA's own quarterly reports, which had shown claims processing at close to 100% accuracy the entire time. That gap between what a TPA reports and what actually happened in the claims system is the whole reason a real TPA performance review matters.

What a TPA Performance Review Actually Involves

A TPA performance review is a structured comparison of what your administrator promised against what actually happened inside the claims system. Most HR leaders assume this already happens because the TPA sends a quarterly report. But that’s not really a performance review.

Those quarterly reports are usually built from the TPA's own internal quality assurance process. That process checks a sample of claims the TPA itself selects, using criteria the TPA itself sets. It's a bit like asking a contractor to grade their own home inspection. The answer will look fine on paper even when the wiring underneath is a mess.

A real review independently checks whether the TPA followed your plan's rules, paid the right amounts, processed claims correctly, and handled them within the promised time.

Why the Oversight Gap Exists

The gap exists because TPAs aren't financially responsible for the health plan, so accuracy isn't tied to their bottom line the way it's tied to yours. A carrier selling fully insured coverage eats the cost of its own claims mistakes. A TPA administering a self-funded plan does not; the plan pays for every claim, correct or not, and the TPA gets paid its administrative fee either way.

That's not a knock on any single TPA's intentions. Most administrators run large books of business, process claims through automated adjudication systems, and genuinely try to get it right. But the financial incentive to catch every dollar of leakage simply isn't as sharp as it would be if the money came out of their own pocket.

Staff turnover at the TPA, changes to plan documents that don't always get reflected correctly in the claims system, and coordination of benefits cases that require manual decisions can all create opportunities for errors. Industry estimates suggest that claims administrator error rates typically fall between 1% and 3% of claims processed, with higher rates possible for more complex plans.

The Real Cost of Skipping the Review

Claims errors are not a rounding error. On a plan paying tens of millions of dollars a year in claims, even a 2% error rate adds up to real money leaving the plan. Independent estimates suggest overpayment error rates run somewhere between 2% and 5% of total medical claim costs annually, even at well-regarded claims administrators. That's before counting dependent eligibility errors, which are a separate and often larger category of leakage.

The manufacturer example above isn't an outlier. One claims analytics firm reports finding error detection rates of 5% to 15% across its client base, with average findings between $500 and $1,200 per covered employee per year. For a 1,000-life plan, that's potentially half a million dollars a year sitting in unreviewed claims.

There's a regulatory cost too. The Department of Labor's Employee Benefits Security Administration recovered more than $1.4 billion for retirement, health and welfare plans in fiscal year 2025, with more than half of that coming directly from enforcement actions. EBSA has named cybersecurity, mental health parity, surprise billing compliance and benefit distribution integrity as enforcement priorities heading into 2026, all of which touch claims administration directly.

What's Actually Happening Behind the Scenes

Benefit Determination Errors

This is the most common error category in claims processing. It happens when a claim is adjudicated against the wrong plan provision, the wrong deductible accumulator, or the wrong coordination of benefits order. A dependent still covered under a divorced spouse's plan, or a retiree who should have moved to Medicare, can generate months of misapplied payments before anyone notices.

Coding and Data Entry Errors

Errors also creep in through how information gets entered into the claims-processing system, including dates of service, diagnosis codes and procedure codes, any of which can change the reimbursement amount owed to a provider. These are rarely intentional. They're the kind of small keystroke mistake that compounds across thousands of claims a month.

Contract Rate Mismatches

Specialty drugs and out-of-network services are common trouble spots. A claim billed at a rate that doesn't match the negotiated contract, or a specialty pharmacy claim priced against the wrong fee schedule, can slip through automated systems that weren't built to catch every pricing exception.

Dependent Eligibility Drift

People change jobs, get divorced, and age off coverage, but plan sponsors rarely run a dedicated eligibility audit separate from claims review. Ineligible dependents quietly stay on the plan and keep filing claims until someone specifically goes looking.

Why the Status Quo Isn't Enough

Most plan sponsors already have some version of TPA oversight in place. The problem is that it's usually built around the TPA's own reporting, not an independent check.

How to Fix It: A Step-by-Step Approach

Red Flags That Signal a Problem on Your Plan

The ROI of Doing It Right

A properly scoped independent claims audit tends to pay for itself well within its first year, sometimes several times over. A comprehensive independent claims audit typically recovers 1% to 3% of annual claims spend in its first year, and that figure often exceeds the cost of running the audit itself. On a plan paying $20 million a year in claims, that's a recovery range of $200,000 to $600,000.

The savings don't stop at recovery either. Once a TPA knows its work is being independently checked, error rates on future claims tend to drop, because the incentive structure finally has some teeth. Think of it the way a restaurant behaves differently once it knows the health inspector shows up unannounced instead of on a predictable schedule.

There's also a fiduciary protection dividend that's harder to put a dollar figure on but matters just as much. A documented, prudent review process is exactly what ERISA asks of a fiduciary, and it's the strongest defense a benefits committee has if a claim, a lawsuit or a DOL inquiry ever puts that process under a microscope.

Conclusion and Next Steps

A meaningful TPA performance review isn't about assuming the worst of your administrator. It's about replacing an assumption with actual evidence, the same way any other vendor relationship involving tens of millions of dollars would get checked. The plan sponsors who skip this step aren't doing so out of negligence. They just haven't had a reason to look closely yet.

Start small if you need to. Pull your plan document, ask your TPA for claim-level detail instead of summary dashboards, and put a review date on the calendar for the next 90 days. If it's been more than three years since anyone independently checked your claims data, that's the clearest sign the review is overdue.

Frequently Asked Questions

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

‍Quarterly service-level check-ins plus a full independent claims audit every one to two years is a reasonable baseline for most plans over 500 lives.

Who is legally responsible if a TPA pays claims incorrectly?

‍The plan sponsor. ERISA places fiduciary responsibility on the plan sponsor regardless of whether claims administration was delegated to a TPA.

What's the difference between a TPA's internal audit and an independent claims audit?

‍The TPA's internal audit uses its own sample and standards. An independent audit uses outside criteria and a sponsor-selected sample.

How many claims need to be reviewed to get a valid result?

‍A stratified random sample of roughly 200 to 400 claims, weighted by claim type and dollar amount, is generally considered statistically valid.

Does a claims audit clause exist in most TPA contracts?

‍Most contracts allow some audit rights, but many restrict frequency, scope or which firms can perform the audit.

Can a claims audit pay for itself?

‍Often, yes. First-year recoveries in the 1% to 3% range of annual claims spend commonly exceed the cost of the audit.

Should dependent eligibility be part of the same audit?

‍It can run alongside a claims audit, but treating it as a distinct workstream tends to surface more issues faster.

What documentation protects a plan sponsor in a DOL inquiry?

‍Audit reports, benefits committee minutes, a written review process and evidence that findings led to corrective action.

‍

‍

‍

‍

TPAs struggle to deliver true claims oversight at scale because they process high claims volume across many client plans with limited staffing, rely on automated systems tuned for speed over scrutiny, and operate under contracts that rarely fund independent verification. These are structural limits, not signs of poor intent, and employers need a separate oversight layer to close the gap.

A single mid-market TPA can process tens of thousands of claims a week across dozens of client plans, each with its own plan document, network contract, and benefit design.

That volume runs through the same claims examiners and the same rules engine, day after day. True claims oversight, the kind that catches a mispriced claim before it pays, needs time and attention that high-volume processing was never built to give.

What "Claims Oversight at Scale" Actually Means

Claims oversight at scale means verifying claims accuracy across an entire book of business without every claim slowing down to a manual crawl.

Most plan sponsors assume their TPA already does this by default. In practice, a TPA's job is to process claims correctly under its own workflow, not to independently audit every payment against your specific contract terms.

Those are related tasks, but they're not the same task. A TPA's quality team checks whether the claim followed the system's rules. It doesn't typically ask whether the rules themselves still match your negotiated rates, your plan document's current language, or a dependent's changed eligibility status.

That distinction matters more as a book of business grows. The bigger the TPA, the more plans, provider contracts, and plan-document variations its systems have to track correctly, all at once.

Why the Scale Problem Exists

The math is the real driver here. A TPA managing claims for 40 employer groups isn't tracking one rulebook, it's tracking 40, and every renewal or plan change adds a new variable to the system. Staffing budgets rarely grow in proportion to that complexity.

Claims examiners are also stretched across client plans rather than dedicated to one. When a queue backs up, the priority becomes clearing it, not slowing down to double-check pricing logic on claims that already passed the automated rules engine. That's a reasonable response to volume pressure, not a lapse in judgment.

Contracts add a third layer. Most ASO agreements pay TPAs for processing speed and service-level compliance, not for catching their own errors. Oversight tools and independent audits usually sit outside the base fee, which means they only happen when someone specifically asks for them.

The Real Cost of the Scale Gap

The 2025 CAQH Index, built on data from more than 600 provider organizations and health plans covering 63% of insured lives, found a remaining $21 billion savings opportunity tied to manual and partially manual healthcare transactions still running today. That gap exists because full automation and full accuracy checking haven't caught up with claims volume industrywide.

Willis Towers Watson's benchmarking work puts typical TPA claims processing errors at 1% to 3% of total claims volume, a figure that holds even at TPAs operating within their service-level targets. On a plan spending $10 million a year, that range translates to $100,000 to $300,000 in claims that were priced, duplicated, or coordinated incorrectly and never flagged.

There's also a documentation gap. When claims accuracy isn't independently verified, plan sponsors have a harder time showing they met the ERISA Section 404(a)(1)(B) prudent expert standard, since a fiduciary can't rely solely on a vendor's own self-reported numbers as proof of oversight.

What's Actually Happening Behind the Scenes

The Rules Engine Handles Speed, Not Judgment

Auto-adjudication systems commonly hit accuracy benchmarks in the 80% to 85% straight-through range industrywide, according to claims operations benchmarking sources. The remaining claims route to manual review, where staffing constraints, not intent, determine how thoroughly each one gets checked.

Plan-Specific Rules Drift Out of Sync

Every time a plan sponsor updates a benefit, changes a network, or amends the plan document, someone has to update the TPA's system to match. On a large book of business, that update cycle sometimes lags behind the actual plan terms, especially mid-year.

Exception Queues Grow Faster Than Review Capacity

Claims that don't auto-adjudicate get pushed into a manual queue, and that queue competes for the same examiner hours across every client plan the TPA services. When volume spikes, the queue grows before staffing catches up.

Multi-Client Servicing Limits Depth

A TPA examiner working across a dozen employer groups in a single week builds broad familiarity with claims processing, not deep familiarity with any one plan's specific contract terms and history. That's a structural tradeoff of scale, not a skills gap.

Why TPA-Only Monitoring Isn't Enough

TPA self-monitoring genuinely catches process errors and keeps claims moving. What it wasn't designed to catch is the slower-moving, plan-specific error that only shows up when someone checks payments against your actual contract terms

How to Close the Gap

Red Flags That Signal a Scale Problem on Your Plan

The ROI of Fixing It

Employers who add an independent oversight layer alongside their TPA relationship typically recover 1% to 3% of annual claims spend in the first year, which on most self-funded plans covers the cost of the review several times over. That recovery comes from catching the specific errors a high-volume TPA system structurally has less bandwidth to catch on its own.

There's a fiduciary benefit too. A documented, separately funded oversight process gives a benefits committee something concrete to point to if a claim or audit question ever escalates, rather than relying entirely on a vendor's internal numbers.

With self-funded plans covering 67% of workers nationally and 80% of workers at large firms, according to KFF's 2025 Employer Health Benefits Survey, TPA claims operations are now a mainstream employer concern rather than an issue limited to the largest self-funded plans.

Conclusion and Next Steps

TPAs manage high-volume operations across many client plans at once. Their teams are responsible for processing claims accurately and efficiently, while plan-specific oversight has to compete for the same limited time and resources.

The fix isn't finding a better TPA. It's recognizing that independent oversight serves a different purpose and, by design, should sit outside the day-to-day claims operation.

Start by asking your TPA a few direct questions about examiner workload and your plan's specific auto-adjudication rate. Then consider an independent review if it's been a while since anyone outside the TPA relationship has checked your plan's claims.

‍Frequently Asked Questions

Why do claims errors happen even with a well-run TPA?

‍High claims volume across many client plans means examiner time and rules-engine attention are shared resources, limiting plan-specific depth.

What's a good auto-adjudication rate for a health plan?

‍Industry sources commonly cite 80% to 85% straight-through processing as a strong benchmark, though it varies by plan complexity.

Does TPA size affect claims oversight quality?

‍Larger TPAs manage more client plans per examiner team, which can reduce plan-specific depth even as processing speed improves.

Can a plan sponsor request its own claims accuracy rate?

‍Yes, and it should be measured against your specific contract terms rather than a TPA-wide or industry average figure.

Is claims oversight the TPA's contractual responsibility?

‍TPAs process claims per their systems and service agreements, but ERISA fiduciary responsibility for accuracy stays with the plan sponsor.

How often do plan documents fall out of sync with TPA systems?

‍It varies, but updates following mid-year plan changes or renewals sometimes lag, especially across large multi-client books.

Does adding independent oversight mean replacing the TPA?

‍No, independent oversight typically runs alongside the existing TPA relationship as a separate, dedicated review layer.

What does an independent oversight layer usually recover?

‍Employers commonly recover 1% to 3% of annual claims spend in the first year of an independent review process.

‍

‍

Employers should ask their TPA for claims-level data access, documented error rates against contract guarantees, coordination-of-benefits procedures, and independent audit rights. Under ERISA Section 404(a)(1)(B), the plan sponsor, not the TPA, carries fiduciary responsibility for claims accuracy, so self-reported accuracy numbers alone don't satisfy that duty.

A manufacturer with a 500-person health plan, referenced in industry audit reports, went eighteen months without ever pulling a claims file from its TPA. When it finally did, an independent review found a six-figure inpatient claim paid twice and a specialty drug billed well above the negotiated rate.

None of it showed up in the TPA's own quarterly report. That gap between what a TPA reports and what actually happened to your money is the entire subject of this article.

What Claims Oversight Actually Means

Claims oversight is the ongoing, independent verification that a TPA paid claims correctly under the plan document and the negotiated provider contract. Most employers assume their TPA's internal quality controls are the oversight. That assumption is where the trouble usually starts.

TPAs self-report claims accuracy in the high 90s, often above 96%. That number describes procedural accuracy, meaning the claim was processed and paid on time using the correct fields. It says almost nothing about whether the dollar amount was actually correct under your contract.

Those are two different questions, and only one of them protects the plan financially. A plan can hit every processing benchmark in its service agreement and still leak six figures a year in duplicate payments, wrong contracted rates, and eligibility errors nobody caught.

Why the Oversight Gap Exists

The gap exists because TPAs don't carry the financial risk of a self-funded plan. In a fully insured plan, the carrier eats the cost of its own mistakes. In a self-funded arrangement, the employer pays the claim either way, so the TPA has limited financial incentive to hunt down every overpayment on its own book.

Add to that the sheer volume. A mid-sized employer's plan can generate tens of thousands of claims a year, and a human reviewer processes maybe 50 a day by hand. Sampling became standard practice because full manual review was never realistic, which is exactly why most plans only see a small slice of their own claims data.

Contracts compound the problem. Many ASO agreements don't guarantee the employer access to detail-level claims files, and some TPAs restrict which outside firms can review their work or require advance notice before an audit. If you can't get the data, you can't verify the number.

The Real Cost of Not Asking

Industry claims audits routinely find payment errors in the range of 1% to 10% of claims dollars, well above the 96% to 98% accuracy TPAs typically self-report, according to benchmarking work published by Willis Towers Watson and independent audit firms. WTW puts the industry-standard error rate at 1% to 3% of total claims processed. On a plan spending $10 million a year on claims, even the low end of that range is real money walking out the door quietly.

One audit example published by Baker Tilly on a client system found the TPA's actual financial accuracy at 96.8% and payment accuracy at 96.1%, both below the 98% service level agreement the contract required and well below the TPA's own self-reported 100%. That's not a rounding error. That's hundreds of thousands of dollars in unrecovered claims for a system of that size.

Beyond the direct dollars, there's fiduciary exposure. The DOL's Employee Benefits Security Administration recovered more than $1.4 billion for benefit plans in fiscal year 2025, with $714.4 million of that coming directly from 556 enforcement investigations. Plan sponsors who can't document independent claims oversight are the ones EBSA investigators zero in on.

What's Actually Happening Behind the Scenes

Duplicate and Overlapping Payments

The same procedure code, same date of service, same patient gets paid twice, often because a provider resubmits a claim or because a system migration reprocesses something already settled. These errors cluster in specific claim types rather than spreading evenly across the book.

Contracted Rate Mismatches

A claim gets priced off an outdated fee schedule, or a specialty drug bills at a percentage above the negotiated rate that nobody flags because the system doesn't cross-check it in real time. Pricing negotiated in a contract only matters if someone confirms the payment matched it.

Coordination of Benefits Failures

When a member has secondary coverage, the primary payer should reduce its liability accordingly. TPAs frequently miss these adjustments, especially for dependents whose other coverage changed mid-year.

Eligibility Drift

Terminated employees, ineligible dependents, and COBRA participants who should have rolled off coverage sometimes keep getting claims paid on their behalf for months. This is one of the most common findings in independent audits and one of the easiest to prevent with a clean data feed.

Why TPA Self-Reporting Isn't Enough

TPA self-audits aren't worthless. They catch process drift and give a rough baseline. But they're not a substitute for an outside party confirming the plan only paid what it actually owed.

How to Fix It

Red Flags That Signal a Problem on Your Plan

The ROI of Getting This Right

Independent claims audits typically recover 1% to 3% of annual claims spend in the first year, often well above the cost of the audit itself. On a $10 million claims book, that's $100,000 to $300,000 recovered in year one alone, plus the ongoing savings from fixing whatever process caused the errors.

There's a second return that doesn't show up on a spreadsheet: fiduciary protection. A documented, recurring oversight process is your best defense if a participant complaint ever escalates into an EBSA inquiry or litigation.

With 67% of covered workers now on self-funded plans nationally, and 80% at larger firms according to KFF's 2025 Employer Health Benefits Survey, this isn't a niche concern anymore. It's the default arrangement for most mid-size and large employers, and oversight needs to scale with that.

Conclusion and Next Steps

Claims oversight isn't a compliance box to check once and forget. It's a recurring discipline, and most self-funded employers are behind on it simply because nobody told them how far behind they were.

Start small if you need to. Ask your TPA for claims-level data access, get a real answer on error rates measured against your contract, and put an independent audit on the calendar if it's been more than two years since the last one.

Your benefits committee doesn't need to become claims auditors. It needs a documented process and a partner who can run the review independently. That's the difference between hoping your TPA got it right and being able to prove you checked.

Frequently Asked Questions

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

‍It's the independent, ongoing verification that a TPA paid claims correctly per the plan document and provider contracts, separate from the TPA's own reporting.

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

‍Most fiduciary advisors recommend an independent audit every one to two years, with claims data monitored continuously between audits.

What claims error rate is normal for a TPA?

‍Industry benchmarks put typical processing errors around 1% to 3% of claims volume, though independent audits sometimes find higher rates.

Who is legally responsible if a TPA pays a claim incorrectly?

‍The plan sponsor. ERISA Section 404(a)(1)(B) places fiduciary responsibility for claims accuracy on the employer, regardless of delegation to a TPA.

Can employers request raw claims data from their TPA?

‍Yes, and the right to detail-level claims files should be written into the ASO agreement, ideally covering 12 to 24 months of history.

What's the difference between a TPA self-audit and an independent claims audit?

‍A self-audit is internally graded with undisclosed methodology. An independent audit uses a statistically valid outside sample and has no financial stake in the TPA's results.

Does a claims audit cost more than it recovers?

‍Usually not. Recoveries commonly run 1% to 3% of annual claims spend, often exceeding audit fees, especially with contingency-based firms.

What should employers do if their TPA resists sharing claims data?

‍Flag it as a contract issue for the next renewal. Restricted data access is itself a red flag worth raising with the benefits committee.

‍

‍

‍

Fiduciary intelligence means continuously verifying that a self-funded plan's TPA is paying claims correctly, not assuming a signed contract guarantees it. Most plans audit under 5% of claims, while independent reviews routinely find 1% to 10% in payment errors. Wellness programs shape employee behavior. Fiduciary intelligence protects the plan sponsor's legal duty and its money.

A 1,400-employee manufacturer ran its first independent claims audit eighteen months into a new TPA relationship. The audit turned up $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 shown up in the TPA's own reporting.

Meanwhile, the same employer's wellness vendor had just completed a step challenge with 62% participation and reported the results in its year-end review.

One of these programs got budget, attention and a slide at the benefits committee meeting. The other one, the one actually protecting the plan's money and the sponsor's legal exposure, didn't exist yet. That gap is the subject of this article.

Fiduciary Intelligence Isn't Another Program. It's a Missing Function

Fiduciary intelligence means the plan sponsor actively verifies that claims are paid correctly, priced correctly and administered according to plan terms, on an ongoing basis rather than once every few years. Most benefits leaders assume their TPA's internal accuracy reporting covers this. It doesn't.

TPAs typically self-report financial and payment accuracy above 96%. Independent audits at the same plans often find results below the 98% service level agreement standard, sometimes closer to 96.1% payment accuracy against a 100% self-reported baseline, according to claims audit work published by Baker Tilly. The difference between what a TPA reports and what an independent reviewer finds is where fiduciary risk lives.

Wellness programs address employee behavior: smoking, weight, chronic disease management. Fiduciary intelligence addresses a completely different question. Is the money already being spent, spent correctly? Those aren't competing priorities, but only one of them carries personal legal liability for the people signing off on the plan.

Why the Problem Exists

TPAs process claims at volume, and their compensation model doesn't reward catching their own errors. According to Willis Towers Watson, industry-standard TPA error rates run 1% to 3% of total claims processed, with other independent audit benchmarks finding a wider band of 1% to 10% depending on plan complexity and claim type.

Administrators aren't financially responsible for the plan's spend. The cost of a missed error or an improperly applied discount lands on the employer, not the TPA. That misalignment isn't malicious. It's structural, and it's been built into standard ASO agreements for decades.

Most administrative services agreements define the audit deliverable as a small sampling review, not a comprehensive one. Employers accept this because it's what's offered, and because "audit" language on a TPA report reads as reassurance. It rarely is.

The Real Cost or Impact

Sixty-seven percent of covered workers nationally, including 80% at firms with more than 200 employees, are enrolled in self-funded plans, according to KFF's 2025 Employer Health Benefits Survey. That's a lot of employers writing checks directly out of company funds for care that mostly goes unverified.

Most self-funded plans independently review fewer than 5% of paid claims, typically through the TPA's own sampling process. A full independent claims audit with comprehensive review typically recovers 1% to 3% of annual claims spend in its first year, often exceeding the cost of the audit itself several times over.

For a plan spending $20 million a year on claims, that recovery range translates to $200,000 to $600,000 left on the table annually. Multiply that across a multi-year TPA relationship and the number stops looking like a rounding error.

What's Actually Happening Behind the Scenes

Duplicate and Overlapping Payments

Large claims move through multiple systems and multiple hands. A hospitalization split across facility and professional billing can generate duplicate payments that a 5% sample audit is statistically unlikely to catch.

Coordination of Benefits Failures

When a dependent has coverage under two plans, the TPA is supposed to determine primary payer status and bill accordingly. Trilogy Consulting's audit case work has documented recurring failures to properly coordinate benefits, along with missed contractual discounts, as a leading source of overpayment in union and employer self-funded plans.

Contract Rate Drift

Negotiated network discounts don't always make it into claims processing accurately. A specialty pharmacy claim billed at 240% of a contracted rate, as found in one independent audit, isn't a one-off. It's what happens when pricing negotiation and claims payment verification are treated as separate functions instead of one continuous process.

Eligibility and Plan Design Errors

Benefits paid at the wrong percentage, exclusions not applied and outdated eligibility records all show up repeatedly in comprehensive audits, according to documented claim audit case studies. None of these require fraud. They require nobody checking.

Why Current Approaches Aren't Enough

Wellness programs aren't worthless, but their financial returns are mixed and often take years to materialize. That makes the contrast worth examining: wellness spending competes for the same budget as fiduciary oversight, which can identify recoverable claims dollars in the first year.

How to Fix It

Red Flags That Signal the Problem Applies to Your Plan

The ROI of Doing It Right

Comprehensive independent audits typically recover 1% to 3% of annual claims spend in year one, frequently covering the audit's cost several times over. That recovery is direct and near-term, unlike most wellness ROI claims, which depend on multi-year behavior change holding steady across an entire population.

There's a second return that doesn't show up on a savings report: documented fiduciary protection. EBSA closed 878 civil investigations in FY 2025, with 556 (63%) producing monetary results or corrective action, and recovered $714.4 million through enforcement alone. A documented, ongoing oversight process is the single best defense a plan sponsor has if that investigation lands on their plan.

Think of it like a family that diets carefully every January but never reconciles the bank statement showing money quietly draining out through an autopay they forgot to cancel three years back. The diet feels virtuous. The autopay is the actual math.

Conclusion and Next Steps

Wellness programs aren't the enemy here. They're just not the function carrying your fiduciary exposure, and treating them as interchangeable with claims oversight leaves real money and real legal risk unattended. Fiduciary intelligence is the unglamorous, unbudgeted work of actually checking whether the plan's biggest expense line is being handled correctly.

Start by finding out what percentage of your claims get reviewed today. If nobody on your committee can answer that with a number, that's your starting point.

Frequently Asked Questions

What is fiduciary intelligence in the context of a health plan?

‍It's the ongoing practice of independently verifying claims accuracy and TPA performance, not just negotiating good contract terms upfront.

Who is legally responsible if a TPA pays claims incorrectly?

‍The plan sponsor. ERISA places fiduciary duty on the employer regardless of delegation to a third-party administrator.

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

‍Annually at minimum, with continuous or quarterly monitoring recommended for plans over 500 employees.

What percentage of claims do TPAs typically review internally?

‍Under 5%, usually through a small statistical sample rather than a comprehensive review.

Can wellness programs and claims oversight coexist in the same budget?

‍Yes. Many plans fund oversight from year-one audit recoveries, then sustain both.

What triggers a DOL or EBSA investigation of a self-funded plan?

‍Participant complaints, referrals from benefits advisors and patterns identified across service providers are common triggers.

Does a signed TPA contract protect the plan sponsor from fiduciary liability?

‍No. Selecting a TPA is only half the duty. Ongoing monitoring is the other half, and it's the half most often missing.

What's a realistic first-year recovery from an independent claims audit?

‍Documented recoveries typically run 1% to 3% of annual claims spend, often exceeding the audit's cost.

‍

‍

‍

Fiduciary intelligence is the ongoing practice of auditing claims data, TPA performance, and plan spend to prove healthcare dollars are managed prudently under ERISA. For brokers and captives, offering it as a standing service, rather than a renewal-season extra, is becoming the clearest way to differentiate and retain self-funded clients.

A mid-market manufacturer with 340 employees spent three years assuming its TPA had claims accuracy handled. Nobody had looked at a claim file directly since the plan moved self-funded. When a broker finally pulled an independent sample, the review found six figures in duplicate payments and dependents who should have been dropped two open enrollments ago.

Family health premiums hit $26,993 on average in 2025, up 6% for the third year running, and 67% of covered workers are now in self-funded plans. Every dollar of that spend sits on the plan sponsor's books, and under ERISA, the sponsor is on the hook for how it's managed, not the TPA. Brokers who can prove that oversight is happening, continuously and independently, are starting to win business that pure renewal negotiation can't touch.

What Fiduciary Intelligence Actually Means

Fiduciary intelligence is the continuous review of claims data, TPA performance, and plan spend to demonstrate that a self-funded plan is being run prudently under ERISA. Most people in this industry hear "claims audit" and picture a one-time project: a consultant pulls a sample, writes a report, and everyone moves on until the next renewal cycle.

That's not what fiduciary intelligence is. It's closer to a standing discipline, similar to how a CFO doesn't audit the books once every three years and call it done. The plan's claims data gets reviewed on a rolling basis, TPA performance gets benchmarked against contract terms, and the plan committee has a documented trail showing they actually looked.

Here's the part most sponsors get wrong: they assume their TPA's self-reported accuracy numbers are the audit. TPAs often report 98% to 100% payment accuracy on their own claims. Third-party reviews of the same claims routinely find something different, because TPAs are grading against their own processing rules, not against the plan document itself.

Why the Problem Exists

TPAs process claims fast because speed is what they're measured on internally. Accuracy against the specific plan document, the one with your custom exclusions, your dependent eligibility rules, your coordination of benefits language, isn't usually the metric that gets watched day to day.

Most TPA contracts include a self-reported accuracy guarantee, and most plan sponsors never verify it independently. That's not negligence exactly. It's a resourcing gap. HR teams running benefits alongside a dozen other responsibilities don't have the bandwidth to pull claim files and check them against plan language line by line.

Brokers, historically, haven't filled that gap either. Renewal negotiation and open enrollment support have been the job. Ongoing claims oversight sat outside the traditional scope, and nobody was pricing it as its own service line.

The Real Cost or Impact

Numbers make this concrete. Independent studies of self-funded plans put TPA payment error rates in the 2% to 6% range, with some reviews finding rates as high as 10% depending on plan complexity and audit method, according to Baker Tilly. Willis Towers Watson pegs the industry standard for financial accuracy, the share of total claim dollars paid incorrectly, at roughly 1%, which still translates into millions of dollars for a large plan, as WTW notes.

Run the math on a $20 million annual claims spend. Even a conservative 1% to 2% error rate represents $200,000 to $400,000 a year, money that may be recoverable but can easily go unnoticed without independent review. One Baker Tilly review of tribal self-funded plans found that independent testing identified accuracy gaps in 60% of cases, highlighting why claims oversight should extend beyond TPA reporting.

Beyond the dollars, there's regulatory exposure. DOL's Employee Benefits Security Administration recovered more than $1.4 billion for retirement, health, and welfare plans in fiscal year 2025, and over half of that, $714.4 million, came directly from enforcement actions rather than voluntary corrections, per DOL's own fact sheet. Nearly 300 of those investigations started because participants complained repeatedly about the same plan or service provider, a pattern that's entirely avoidable with proactive oversight.

What's Actually Happening Behind the Scenes

Claims Leakage Nobody's Tracking

Duplicate payments, coding errors, and out-of-network claims processed at in-network rates rarely show up on a TPA's own dashboard, because the dashboard is measuring what the TPA chose to measure. An independent review checks against the actual plan document instead.

Dependent Eligibility Drift

Divorced spouses, aged-out dependents, and employees who left the company months ago quietly stay on plan rosters. Nobody catches it until an audit specifically checks eligibility files against HR records, and by then it's often been years.

Pharmacy and Specialty Drug Spend

Specialty pharmacy claims are complex enough that errors hide easily inside them. Coordination of benefits failures, meaning the plan pays first when another payer should have, are especially common on pharmacy claims involving Medicare-eligible dependents.

Vague or Missing Documentation

When the DOL or a plan participant asks how a claims decision was made, plans without a documented review process often can't produce one. That absence of a paper trail is itself a fiduciary problem, separate from whatever the underlying claim showed.

Why Current Approaches Aren't Enough

Most plans rely on whatever the TPA offers as standard, and that's rarely built for the sponsor's protection. The comparison 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

Comprehensive claims audits with full population review typically recover 1% to 3% of annual claims spend in the first cycle, according to industry-documented ranges. For a $15 million plan, that's $150,000 to $450,000 recovered in year one alone, before counting the savings from fixing the process going forward.

The ongoing value compounds. Once dependent eligibility is cleaned up and the TPA knows it's being watched, error rates tend to drop on their own. Several audit firms report that plans moving from periodic to continuous oversight see meaningfully lower error rates within a year or two of implementation.

Then there's the fiduciary protection, which is harder to put a dollar figure on but matters just as much. A documented, defensible process is the single best protection a plan committee has if a participant or the DOL ever challenges how the plan was run.

Conclusion and Next Steps

The self-funded market isn't getting simpler, and the fiduciary exposure that comes with it isn't going away either. Brokers and captives that treat fiduciary intelligence as a real, priced service, not a favor thrown in at renewal, are the ones building relationships that survive past the next RFP cycle.

If your current broker relationship stops at renewal negotiation, that's worth a direct conversation. Ask what independent claims oversight looks like for your plan and what it would take to build a documented, defensible fiduciary process starting now.

Frequently Asked Questions

What is fiduciary intelligence in employee benefits?

‍It's the ongoing practice of auditing claims data and TPA performance to prove a self-funded plan is managed prudently under ERISA, not a one-time audit.

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

‍The plan sponsor holds fiduciary responsibility under ERISA, even though the TPA processes the claims day to day.

How often should a self-funded plan be audited?

‍Best practice is continuous or quarterly review, not the once-every-two-to-three-years cadence most plans still use.

What's a typical TPA claims error rate?

‍Industry studies put it between 2% and 10% of paid claims, depending on plan complexity and how the review is conducted.

Can a broker offer fiduciary intelligence as a paid service?

‍Yes. Leading agencies now scope it as a standalone PEPM service rather than bundling it free into renewal work.

Does fiduciary intelligence apply to group captives too?

‍Yes. Claims oversight data feeds directly into loss experience, which affects captive renewal terms and member pricing.

What triggers a DOL/EBSA investigation into a self-funded plan?

‍Repeated participant complaints about the same plan or service provider are a common trigger, along with informal inquiry patterns.

What's the difference between a TPA self-audit and an independent claims audit?

‍A TPA self-audit measures against its own internal rules. An independent audit measures against the actual plan document and contract terms.

‍

‍

‍

‍

Fiduciary oversight becomes a core service, not an add-on, when brokers provide it as an ongoing service with clear deliverables, such as regular claims reviews, documented TPA performance checks and reports prepared for fiduciary committees. This turns a responsibility plan sponsors already have under ERISA Section 404 into a defined, recurring and billable service.

A mid-size manufacturer with 340 covered employees discovered that its TPA had been using the wrong payment rate for out-of-network claims for 18 months. The error was found only after a new CFO ordered an independent claims audit before renewal. The audit helped the plan recover nearly $210,000. The broker had never provided ongoing oversight beyond mentioning it in the annual stewardship presentation.

This happens every year across the self-funded health plan market. It shows why fiduciary oversight is often treated as a free courtesy instead of a separate service that can provide real financial value.

Willis Towers Watson research shows that administrator errors typically affect 1% to 3% of total claims, while financial accuracy errors can be around 1% of paid claims, even when administrators meet industry standards.

For a plan spending $20 million a year, even a 1% error rate could mean $200,000 in potential overpayments or errors that go unnoticed.

What "Positioning Fiduciary Oversight as a Core Service" Actually Means

Positioning fiduciary oversight as a core service means clearly defining it, explaining what it includes and charging for it separately instead of burying it in a renewal presentation. Most brokers already do parts of this work: they review claims trends, flag stop-loss issues or mention TPA performance guarantees. But these actions alone do not create documented, ongoing oversight or give the plan sponsor evidence of a prudent process if the DOL asks for it.

The common assumption is that TPAs handle accuracy internally and that a broker's job ends at plan design and carrier negotiation. The reality is that ERISA places the fiduciary burden on the plan sponsor, not the administrator, for every dollar the plan pays out. A TPA's self-reported accuracy rate is not oversight. It is the vendor grading its own homework.

Why Fiduciary Oversight Keeps Getting Treated as an Add-On

The root cause is structural, not a lack of awareness. Broker compensation has historically been tied to placement and renewal, so revenue flows from the sale, not from ongoing monitoring, and monitoring gets deprioritized by default.

A second root cause is capability. Genuine claims oversight requires access to raw claims data, analytics tools and clinical or coding expertise that a generalist broker team was never built around, so the work gets waved off as "the TPA's job" rather than built out as a service.

A third factor is inertia inside plan sponsor organizations. HR and finance leaders assume that because a TPA is contractually obligated to pay claims correctly, someone is already checking that they do. Baker Tilly notes it is common for organizations to perform a claim audit only once every three years, which leaves long windows where nothing is being verified at all.

The Real Cost of Leaving Oversight Unpriced

Unreviewed claims dollars do not disappear. They compound, and the compounding is the real cost most plan sponsors never see on a single line item.

An independent claims analytics firm's client data across a large self-funded book found 5% to 15% error detection rates once claims were fully reviewed, with average findings landing between $500 and $1,200 per employee per year.

A peer-reviewed study found that sample-based claims audits can miss significant errors. In two Fortune 100 companies, random sampling failed to detect errors worth $200,000 to $750,000 because only a portion of claims were reviewed. The takeaway is simple: a small sample can leave significant dollars undetected.

What's Actually Happening Behind the Scenes

TPA Sampling Covers a Sliver of the Plan

Standard TPA-conducted audits typically review a stratified sample of a few hundred claims out of tens of thousands processed annually, then extrapolate an accuracy score from that sample. The extrapolation looks clean on a stewardship slide, but it was never designed to catch every category of error, only the categories most likely to show up in a small, structured sample.

Financial Accuracy Is Not the Same as Payment Accuracy

A plan can hit its financial accuracy target, meaning the dollar amount paid was close to correct, while still failing payment accuracy, meaning the claim was processed against the wrong plan rule, provider contract or coordination of benefits determination entirely. Baker Tilly's audit example showed exactly this split, with financial and payment accuracy landing at different rates against the same set of claims.

Broker Compensation Structures Rarely Reward Vigilance

The CAA 2021 rules require brokers and consultants to disclose how they are compensated, giving plan sponsors more visibility into broker fees. But compensation tied to placing or renewing coverage does not necessarily reflect the time spent on ongoing claims and TPA oversight. This gives brokers an opportunity to clearly define fiduciary oversight as a separate service and charge for the work they perform.

Why Current Approaches Aren't Enough

Most plans default to whatever oversight the TPA offers as part of the base contract, and that default carries structural conflicts that a standalone, independently priced oversight engagement does not.

How to Fix It: A Framework for Positioning Oversight as a Core Service

Red Flags That Signal Your Plan Needs This Now

The ROI of Doing It Right

Independent claims reviews can uncover far more money than they cost. ClaimInformatics client data shows 5% to 15% error detection rates in full-population reviews, with average findings of $500 to $1,200 per employee per year.  For a 500-employee plan, even the low end could mean $250,000 in findings that a sample-based audit might have missed.

The value goes beyond recovering money. A documented, ongoing oversight process gives plan sponsors evidence that they are actively monitoring their plan if the DOL asks questions. EBSA's FY 2025 enforcement activity shows why this matters. The agency closed 878 civil investigations, with 556 resulting in repayments or corrective action, and recovered more than $1.4 billion for benefit plans overall.

Making fiduciary oversight a defined, paid service helps brokers deliver measurable value while giving plan sponsors a stronger record of prudent oversight.

Conclusion and Next Steps

Fiduciary oversight was never meant to be a courtesy add-on to a broker renewal. It is a standing legal obligation under ERISA Section 404, and the plans that treat it as core, priced and documented are the ones building a real defense against both financial leakage and regulatory exposure. Positioning it that way is also the clearest path for brokers and consultants to build a recurring, defensible revenue line instead of competing on placement alone.

Start with one step: pull your plan's last claims audit and ask who performed it, what percentage of claims it actually reviewed and what happened to the findings. If you cannot answer all three, fiduciary oversight is still an add-on in your organization, not a core service. [internal link: TPA performance guarantees guide] can help you evaluate whether your current administrator's contract even supports the level of oversight your plan needs.

Frequently Asked Questions

What does "fiduciary oversight as a core service" mean?

‍It means naming, scoping and pricing claims and TPA oversight as its own engagement rather than bundling it free into a broker or consulting contract.

Is fiduciary oversight legally required under ERISA?

‍Yes. ERISA Section 404 requires plan fiduciaries to act as a prudent expert would in monitoring how plan assets, including claims payments, are handled.

How is fiduciary oversight different from a standard TPA claims audit?

‍A TPA audit reviews the TPA's own work using a small claims sample. Independent oversight reviews the TPA using outside tools and a much larger claims population.

How often should a self-funded plan review its TPA's claims accuracy?

‍Quarterly reviews aligned to benefits committee meetings build a stronger, more defensible record than the once-every-few-years cadence many plans currently use.

What is a typical TPA claims error rate?

‍Industry sources place standard error rates around 1% to 3% of total claims, with some independent full-population reviews finding 5% to 15% error detection rates.

Can brokers charge separately for fiduciary oversight services?

‍Yes, and the CAA 2021 compensation disclosure rule makes that separate pricing more transparent, not less viable, since compensation must already be disclosed.

What documentation proves a plan sponsor met its fiduciary duty?

‍Committee meeting minutes, written TPA performance reviews, claims audit findings and corrective action logs together form the paper trail regulators expect to see.

What happens if a self-funded plan skips independent oversight entirely?

‍The plan sponsor carries undocumented fiduciary exposure, and EBSA's FY 2025 enforcement data shows regulators are actively pursuing exactly these gaps in health and welfare plans.

‍

A fiduciary intelligence advisor is a broker who goes beyond annual renewal negotiation to monitor TPA claims accuracy, flag ERISA compliance gaps, and help plan sponsors document prudent oversight. The shift responds to rising DOL enforcement of health plans and growing recognition that a good renewal doesn't satisfy a sponsor's fiduciary duty.

In 2024, a 1,400-employee manufacturer switched TPAs after eighteen months of unremarkable renewal cycles. Its broker had negotiated a competitive rate every year without incident. Nobody had checked whether the claims underneath that rate were being paid correctly.

An independent audit, run separately from the broker relationship, found $812,000 in overpayments across eighteen months, including a $47,000 inpatient claim paid twice and 63 ineligible dependents still active on the plan.

What "Fiduciary Intelligence Advisor" Actually Means

A broker's traditional renewal cycle was never designed to catch claims payment errors. Most employers assume that once a broker negotiates favorable rates and a strong network, the plan is being watched. In reality, renewal work happens once a year and focuses on price, plan design, and carrier or TPA selection.

Claims payment accuracy is a separate discipline entirely. It requires reviewing how individual claims were adjudicated against plan documents, contracted rates, and coordination of benefits rules, month after month, not once a year.

A fiduciary intelligence advisor is a broker who has added that discipline to the relationship. They monitor claims data on a recurring basis, flag patterns that suggest overpayment or compliance risk, and help the plan sponsor build the documentation record ERISA Section 404 requires of a prudent fiduciary.

Why the Renewal-Only Model Persists

Broker compensation has historically been tied to placement and renewal, not ongoing claims monitoring. Commission structures reward closing a deal, and claims oversight work sits outside that transaction entirely.

The Consolidated Appropriations Act of 2021 requires brokers and consultants who receive $1,000 or more from a group health plan to disclose their direct and indirect compensation in writing to the plan fiduciary. This makes potential conflicts more visible and is pushing some brokers to provide broader fiduciary support.

Historically, few brokers had claims-level expertise on staff. Reviewing adjudication logic, contracted rate tables, and coordination of benefits data requires a different skill set than plan design or carrier negotiation, and many brokerages simply never built it.

That capability gap, more than a lack of will, has kept most broker relationships confined to the renewal calendar.

The Real Cost of Watching Only at Renewal

Claims errors are not rare edge cases. Across one national analytics platform's client base, claims analysis identified error detection rates of 5% to 15%, with average findings of $500 to $1,200 per employee per year. On a 1,000-employee plan, that range alone represents hundreds of thousands of dollars in annual leakage.

Industry-standard estimates put administrator error rates at 1% to 3% of total claims processed, and other audit firms report a wider band of 2% to 6%. The manufacturer's $812,000 finding sits comfortably inside that range once you apply it to real claims volume.

The regulatory cost is rising alongside the financial one. EBSA recovered more than $1.4 billion for retirement, health and welfare plans in fiscal year 2025, and announced a shift of FY 2026 enforcement resources toward health and welfare plans, service providers, and plan sponsors after decades of retirement-plan focus.

What's Actually Happening Behind the Scenes

Compensation structures limit scope

A broker paid on commission has little financial incentive to add unpaid claims oversight work. Fee-based and hybrid arrangements are changing that calculus, but the shift is uneven across the industry.

TPA reporting is not independent verification

A TPA's internal quality assurance measures its own process against its own standards. It is not the same as an outside party checking claims against the plan document and the contracted rates.

Annual cycles miss continuous risk

Claims errors accumulate every pay period, not once a year. A broker who reviews the plan only at renewal is, by definition, looking backward at a year of unmonitored payment activity.

Dependent eligibility rarely gets its own review

Ineligible dependents, ex-spouses, adult children who aged out, individuals added without documentation, tend to stay on a plan for years once enrolled. Most renewal reviews never ask the eligibility question at all, because it falls outside price negotiation entirely.

Why Renewal-Only Advisory Isn't Enough

How Plan Sponsors Can Move Toward Fiduciary Intelligence

Red Flags That Your Plan Is Still Renewal-Only

he ROI of Fiduciary Intelligence

A comprehensive independent claims audit typically recovers 1% to 3% of annual claims spend in its first year, often covering the cost of the audit itself. For a plan spending $20 million annually on claims, that range translates to $200,000 to $600,000 in first-year recoveries alone.

The less visible return is fiduciary protection. A documented, ongoing oversight process is the evidence a plan sponsor needs if EBSA opens an inquiry, and it is the same evidence that has shielded 401(k) plan fiduciaries from personal liability in excessive-fee litigation.

Ongoing monitoring also compounds. Catching a coding error or an ineligible dependent in month three, rather than at next year's renewal, stops that leakage before it repeats twelve more times. There's a talent retention angle too. A benefits committee that can point to a documented, ongoing oversight process has a stronger answer for skeptical CFOs asking why healthcare spend keeps climbing, and a stronger defense if a participant or regulator ever asks the same question in less friendly terms.

Conclusion and Next Steps

A strong renewal has never been proof that a self-funded plan is being watched. The plans avoiding six and seven-figure claims leakage are the ones whose brokers have expanded into ongoing fiduciary intelligence: independent claims review, dependent audits, and documented oversight, not just annual price negotiation.

Ask your broker where their scope actually ends. If claims accuracy monitoring, compliance documentation, and audit rights aren't part of the relationship, that gap belongs to your plan, not theirs.

Frequently Asked Questions

Is my broker legally a fiduciary?

‍Usually not automatically. Brokers generally aren't ERISA fiduciaries unless they exercise discretionary control, though CAA disclosure rules now increase transparency into their role.

Who holds fiduciary liability for claims accuracy?

‍The plan sponsor, typically through its benefits committee or board, regardless of whether claims administration is delegated to a TPA.

How often should claims be reviewed?

‍Ongoing monthly or quarterly review catches errors faster than annual audits and creates a stronger documentation trail for regulators.

What's the difference between a claims audit and TPA quality assurance?

‍A claims audit is performed by an independent party against plan documents and contracts. TPA quality assurance is internal and self-reported.

Does adding claims oversight cost more than it recovers?

‍Typically not. First-year recoveries of 1% to 3% of claims spend usually exceed the cost of an independent audit.

What does CAA 2021 require of brokers?

‍Brokers earning $1,000 or more must disclose direct and indirect compensation to the plan fiduciary in writing before the contract is finalized.

Can a broker perform the claims audit themselves?

‍Some can, but plan sponsors should confirm the reviewer has healthcare claims expertise and no ownership ties to the TPA being reviewed.

How does this connect to 401(k) fiduciary litigation?

‍Courts and regulators increasingly expect health plan fiduciaries to document prudent process, the same standard already tested in retirement plan excessive-fee cases.

‍

‍

A fiduciary paper trail is simply a record of how a benefits committee makes and reviews decisions about a self-funded health plan. It can include meeting notes, vendor reviews, and claims audit results. Under ERISA, this documentation helps show that the committee took a careful, reasonable approach to its decisions, even if one of those decisions later turned out to be wrong.

In fiscal year 2025, the Department of Labor’s Employee Benefits Security Administration recovered more than $1.4 billion for retirement, health and welfare plans. More than half came from enforcement actions, not voluntary corrections. Plaintiff firms filed 155 ERISA fiduciary class actions in 2025, including 39 involving health plans.

That is a major shift from the retirement-plan lawsuits that dominated a decade ago. Most benefits committees can explain what their plan did. Far fewer can show why they made those decisions, and that gap is often where investigators and plaintiffs’ attorneys start.

What a Fiduciary Paper Trail Actually Is

A fiduciary paper trail is the documented evidence that a benefits committee followed a prudent, repeatable process when making decisions about the plan. Most plan sponsors assume that if the plan works reasonably well and the TPA has a good reputation, the fiduciary duty is satisfied. That assumption is wrong under ERISA Section 404, which requires fiduciaries to act with the care, skill and diligence of a prudent expert, not merely to reach an acceptable result.

Courts and the DOL do not ask whether a claim got paid correctly in hindsight. They ask whether the committee had a process for finding out, and whether that process left a record. A plan that overpaid on a handful of claims but can show quarterly TPA reviews, documented vendor comparisons and audit engagement letters is in a fundamentally different legal position than a plan with the same errors and no record at all.

The key difference is between substantive and procedural prudence. Substantive prudence asks whether the decision was reasonable. Procedural prudence asks whether the committee used a careful, reasonable process to make it. Under ERISA, that process matters, which is why documentation, not perfection, is a committee’s strongest fiduciary protection.

Why Most Plans Have a Documentation Gap

The gap exists because benefits committees are staffed by HR and finance professionals whose core job is running the business, not building a compliance record. Claims administration gets outsourced to a TPA, and plan sponsors quietly extend that outsourcing to include oversight itself, even though ERISA does not allow fiduciary responsibility to be delegated away.

TPAs can make things look better than they really are by reporting their own performance numbers. They often report payment and financial accuracy rates above 96%, which may meet their contract requirements. But independent audits of the same claims can find very different results.

The problem is that a benefits committee cannot rely only on a TPA’s own scorecard to prove it provided proper oversight. Reviewing the vendor’s numbers shows that the committee checked the report. It does not necessarily show that the committee independently tested whether those numbers were accurate.

Turnover makes this problem worse. Committee members leave, brokers change, and people forget why certain plan decisions were made years ago. Without clear meeting notes and supporting documents, a plan sponsor facing a DOL investigation or lawsuit may have to piece together what happened from memory instead of showing a clear record.

The Real Cost of an Undocumented Process

When a fiduciary process is not documented, a simple vendor mistake can become a much bigger legal problem. Under ERISA, fiduciaries who fail to meet their duties may have to repay losses suffered by the plan. This liability can apply to the individual committee members involved, not just the employer.

The scale of enforcement shows why this matters. In FY 2025, EBSA’s civil investigations recovered $714.4 million. The agency also closed 253 criminal investigations, leading to 62 indictments and 45 convictions involving the way plan assets were handled and controlled.

Health plan lawsuits are now following a similar path to the 600+ excessive-fee lawsuits filed against 401(k) plans over the past decade. Plaintiff firms are increasingly using the same arguments about excessive fees and poor oversight against health and welfare plans, especially as the Consolidated Appropriations Act, 2021 increased disclosure requirements.

What's Actually Happening Behind the Scenes

Committees That Meet Without Minutes

Many benefits committees hold regular meetings but treat them as informal check-ins rather than fiduciary proceedings. Decisions about plan design, stop-loss renewal or TPA retention get discussed and agreed upon verbally, with no minutes capturing what alternatives were considered or why the chosen option was selected.

Vendor Oversight That Stops at the Contract

Signing a services agreement with a TPA is treated as the end of the oversight process instead of the beginning. ERISA places fiduciary responsibility for claims accuracy on the plan sponsor regardless of delegation, yet many committees have no calendar for reviewing TPA performance against that contract after signature.

Audit Activity That Is Self-Reported, Not Independent

A claims audit conducted by the TPA on its own claims is not independent evidence of prudent oversight. Objectivity is inherently limited when the party being reviewed also produces the review, and reviews conducted this way are typically performed only once every three years, if at all.

Dependent Eligibility and Data Hygiene Left Unchecked

Ineligible dependents remaining on a plan after a divorce, a dependent aging out or a change in employment status is one of the most common and most avoidable sources of claims leakage, yet dependent eligibility is rarely treated as its own documented review workstream separate from broader claims auditing.

Why Current Approaches Aren't Enough

How to Fix It

Think of fiduciary documentation like a flight data recorder. Nobody expects a plan year to run without any turbulence, and regulators do not expect one either. What they want to know after something goes wrong is whether the committee followed procedure the whole way through.

The paper trail does not prove the plan never made a mistake. It proves the committee was flying the plane on purpose.

Red Flags That Signal Your Plan Is Exposed

The ROI of Doing It Right

A defensible fiduciary process pays for itself twice, once in claims recoveries and once in avoided liability. A comprehensive independent claims audit typically recovers between 1% and 3% of annual claims spend in its first year, an amount that regularly exceeds the full cost of the audit engagement itself. On a $15 million claims book, that range translates to $150,000 to $450,000 in first-year recoveries alone.

The liability side of the equation is harder to quantify but larger in scale. EBSA's FY 2025 enforcement activity alone moved $1.4 billion, and individual ERISA breach settlements in the 401(k) space have run into the tens of millions of dollars per case over the past decade. A documented process is inexpensive insurance against exposure of that magnitude, and unlike claims recoveries, its value is realized only when it is needed most.

Good documentation can also make a DOL investigation faster and less expensive. If a plan has its records organized and ready, investigators can quickly see what happened and why. Without those records, the plan may have to spend extra time searching for documents and piecing together what happened, which can lead to more questions and requests.

Conclusion and Next Steps

Fiduciary prudence is not measured by whether a self-funded plan avoided every error. It is measured by whether the committee overseeing that plan can produce a record showing it looked, asked the right questions and acted on what it found. That record, built consistently over time, is what separates an ordinary vendor mistake from a documented fiduciary breach.

Start with what is fastest to fix. Put a committee charter and minutes template in place this quarter [internal link: benefits committee charter template], schedule your next TPA performance review [internal link: TPA performance guarantees guide], and confirm your ASO agreement actually allows independent claims auditing [internal link: independent claims oversight guide]. If it has been more than a year since your plan's last independent claims audit, that is the single highest-leverage next step available.

Frequently Asked Questions

What does ERISA Section 404 actually require of a plan fiduciary?

‍It requires fiduciaries to act with the care, skill and diligence of a prudent expert, solely in the interest of participants.

Is a self-funded plan sponsor personally liable for TPA errors?

‍Yes. Fiduciary responsibility for claims accuracy stays with the plan sponsor even when claims processing is delegated to a TPA.

How often should a benefits committee meet to stay fiduciary-compliant?

‍Quarterly meetings with retained minutes are the common baseline used by fiduciary-grade committees.

Can a TPA's self-reported audit satisfy fiduciary oversight requirements?

‍No. Independent review is needed because a TPA auditing its own claims lacks the objectivity courts and regulators expect.

How long should fiduciary committee records be retained?

‍Retain minutes, audit reports and vendor contracts beyond ERISA's statute of limitations for breach claims, typically six years or longer.

What triggers a DOL investigation of a self-funded health plan?

‍Common triggers include participant complaints, Form 5500 irregularities, and EBSA's targeted enforcement priorities for a given year.

Does a documented process protect against every fiduciary breach claim?

‍No single record eliminates risk, but a consistent, documented process is the strongest evidence of prudence available in litigation or investigation.

What is the difference between substantive and procedural prudence?

‍Substantive prudence judges the decision itself; procedural prudence judges the process used to reach it, and ERISA case law favors the latter.

‍