Claims Adjudication and Accuracy: How to Measure What Most Employers Never Check

Claims adjudication is the core processing function that converts provider bills into plan payments. The adjudication system receives claims, applies plan terms, calculates member cost-sharing, determines the payable amount, and triggers payment. The quality of adjudication determines whether the plan pays correctly or leaks money through overpayments and underpayments. Industry benchmarks target 97% to 99% financial accuracy. Many small TPAs fall below 95%. A 2% accuracy gap on a $500,000 claims fund is $10,000 in errors annually for a single 25-person group. Most employers never audit their TPA’s claims accuracy. They assume the numbers are correct because they have no way to check.

The Adjudication Process

#

Claims arrive at the TPA through electronic and paper channels. Electronic claims use the 837 transaction format specified by HIPAA administrative simplification rules. Professional claims (physician services) use the 837P format. Institutional claims (hospital and facility services) use the 837I format. Paper claims still exist, particularly from smaller providers, and must be converted to electronic format through data entry or optical character recognition.

The adjudication system matches the claim to eligibility. Is this member covered? On the date of service? For the claimed dependent? If eligibility cannot be verified, the claim stops. If eligibility is confirmed, the claim proceeds to benefit determination.

Benefit determination applies plan terms to the claimed services. Is this service covered under the plan? Does the plan exclude it? Does it require prior authorization that was or was not obtained? Is there a limitation (such as a visit limit for physical therapy) that has been reached? The adjudication system must correctly interpret the plan document and apply it to each line of the claim.

Cost-sharing calculation determines member responsibility. What is the deductible status? Has the member met their deductible for the year? What coinsurance applies after the deductible? What copay applies if this is a copay service? Has the member reached their out-of-pocket maximum? The calculation must track accumulators correctly across all claims for the member and covered dependents throughout the plan year.

Repricing determines the allowed amount for the service. If the provider is in-network, the contracted rate applies. If the provider is out-of-network, an allowable amount is calculated based on plan terms, which may reference Medicare rates, usual and customary amounts, or other benchmarks. Repricing accuracy depends on correct network contract loading and correct application of out-of-network allowable methodologies.

The adjudicated claim produces a payment amount (what the plan pays), a member responsibility amount (what the member owes), and, for out-of-network claims, potentially a balance bill amount (what the provider may bill the member above the allowable amount). This information is communicated to the provider through the 835 remittance transaction and to the member through the explanation of benefits.

Where Errors Occur

#

Errors can occur at every stage of adjudication, and the causes differ by error type.

Eligibility errors occur when the adjudication system pays claims for members who are not covered or denies claims for members who are covered. These errors flow from upstream eligibility management failures. The adjudication system trusts the eligibility file. If the file is wrong, adjudication proceeds incorrectly.

Benefit configuration errors occur when the plan terms are incorrectly loaded into the adjudication system. A deductible entered as $2,000 when the plan document says $2,500. A service coded as excluded when it should be covered. A visit limit of 20 when the plan allows 30. These errors are systematic: every claim for the misconfigured benefit is processed incorrectly until someone identifies and corrects the configuration.

Accumulator errors occur when the system tracks deductibles, coinsurance, or out-of-pocket maximums incorrectly. A claim that should have applied to the deductible is processed as post-deductible. A member who has reached their out-of-pocket maximum continues to be assessed cost-sharing. Accumulator errors are difficult to detect without detailed claim-by-claim review because they depend on the sequence of claims throughout the year.

Repricing errors occur when the wrong contracted rate is applied or when out-of-network allowables are calculated incorrectly. A provider paid at an old contract rate when a new rate should apply. A reference-based pricing calculation that uses the wrong Medicare conversion factor. A usual and customary calculation based on incorrect geographic data. Repricing errors directly affect plan cost and provider payment.

Duplicate payment errors occur when the same service is paid more than once. A provider submits a claim, receives payment, and submits again. The system should detect the duplicate, but detection depends on matching logic that may not catch all variations in how duplicates are submitted. Duplicate payments are pure overpayment.

Coding errors occur when the adjudication system misinterprets the procedure or diagnosis codes on the claim. A procedure code that should trigger a bundling edit does not. A diagnosis code that should indicate a covered condition is misread. Coding errors require clinical knowledge to detect and correct.

How to Measure Accuracy

#

Claims accuracy is measured through audits that compare what was paid to what should have been paid according to plan terms.

Financial accuracy measures the percentage of claims dollars correctly paid. If the plan paid $500,000 in claims and $10,000 was overpaid while $5,000 was underpaid, the financial accuracy rate is ($500,000 - $15,000) / $500,000 = 97%. The industry target is 98% or higher. Many small TPAs operate in the 94% to 96% range without knowing it because no one audits them.

Procedural accuracy measures the percentage of claims correctly adjudicated on all dimensions: eligibility verification, benefit determination, cost-sharing calculation, and repricing. A claim can be procedurally incorrect even if the payment amount happens to be correct. Procedural accuracy targets are typically 95% or higher.

Audit methodology matters. A statistically valid audit samples claims across service types, providers, and time periods. The sample size must be sufficient to produce reliable estimates of the overall accuracy rate. Random sampling avoids bias toward easy-to-audit claims. The audit should be conducted by auditors who understand plan terms and can independently calculate what should have been paid.

Who conducts the audit also matters. The TPA can audit itself, but self-audits have obvious limitations. The employer can hire an independent claims auditor, which produces more credible results but costs money. Some brokers offer claims audit services or can recommend independent auditors. The employer who never audits simply does not know their TPA’s accuracy rate.

What the Metrics Reveal

#

A TPA with 95% financial accuracy is leaking 5% of claims dollars. On a $500,000 claims fund, that is $25,000 annually for a single 25-person group. The leakage takes two forms: overpayments that the plan should not have made, and underpayments that the plan will need to correct when providers or members complain.

Overpayments are money lost. The plan paid more than it should have. Recovery is possible but difficult. The TPA must identify the overpayment, contact the provider, request a refund, and follow up until the refund is received. Many overpayments are never recovered because the cost of pursuing recovery exceeds the amount at stake or because the provider disputes the claim.

Underpayments create relationship problems. The provider who was underpaid will complain. The member who was overcharged on cost-sharing will complain. Correcting underpayments requires rework that consumes TPA staff time and creates friction with providers and members. The plan may also face contractual penalties if network contracts require accurate payment within specified timeframes.

A TPA with a 60% auto-adjudication rate has 40% of claims requiring manual intervention. Manual intervention is slower, more expensive, and more error-prone than automated processing. The 40% of claims that require manual review are likely to have higher error rates than the 60% that auto-adjudicate. A low auto-adjudication rate indicates either complex plan designs that the system cannot handle or inadequate system configuration that forces manual review for claims that should auto-adjudicate.

Denial rates reveal adjudication patterns. An unusually low denial rate may indicate that the TPA is paying claims that should be denied, whether non-covered services, services without required authorization, or claims for ineligible members. An unusually high denial rate may indicate overly aggressive denial policies that require appeals and rework. The appropriate denial rate depends on plan terms and member population, but significant deviation from industry norms warrants investigation.

The Employer’s Audit Decision

#

Most small employers never audit their TPA’s claims accuracy. The reasons are practical: audits cost money, employers do not know how to commission an audit, and employers assume the TPA is processing correctly.

The cost objection is real but should be weighed against the cost of inaccurate processing. An independent claims audit for a small group might cost $5,000 to $10,000. If the audit identifies $15,000 in recoverable overpayments and prevents future leakage, the return on investment is positive. If the audit finds that the TPA is processing accurately, the employer has verification that their plan is being managed correctly.

The knowledge objection is addressable. The broker should be able to recommend claims auditors or conduct broker-level review of claims data. Industry organizations like SIIA publish audit standards and can refer employers to qualified auditors. The employer who wants to audit can find resources.

The assumption of accuracy is the largest barrier. Employers assume that if claims are being paid and members are receiving care, the processing must be correct. They do not realize that systematic errors can persist for years without anyone noticing. The TPA has no incentive to identify its own errors. The employer is the only party with an incentive to verify accuracy, and most employers do not act on that incentive.

The employer who does audit sends a signal. The TPA knows that this employer is checking. That knowledge alone may improve performance. The employer who establishes an expectation of periodic audits at the beginning of the relationship creates accountability that benefits the plan throughout the engagement.

The audit should be part of the TPA contract negotiation. The employer should retain the right to conduct or commission claims audits at reasonable intervals. The TPA should agree to cooperate with auditors and provide access to claims data. Resistance to audit rights in contract negotiation is a warning sign. A TPA confident in its processing quality should welcome verification.

The Continuous Improvement Imperative

#

Claims accuracy is not a static measure. It requires ongoing attention and improvement.

Root cause analysis should follow every identified error. When an audit identifies an overpayment, the question is not just how to recover the money but why the error occurred. Was it a one-time data entry mistake, or is there a systematic configuration problem affecting multiple claims? Was it a training gap that other adjusters also have? Root cause analysis turns individual errors into process improvements.

System configuration should be validated regularly. Plan terms change at renewal. Benefit updates, deductible changes, and new covered services require system configuration updates. Each configuration change creates opportunity for error. A TPA that validates configuration against the plan document after each change catches problems before they produce incorrect claims. A TPA that configures without validation discovers problems months later when someone complains.

Staff training should be continuous. Medical coding evolves. Plan designs become more complex. Regulatory requirements change. Adjusters who were trained five years ago may not know current standards. Ongoing training, competency testing, and feedback loops keep claims staff current and accurate.

Technology investment improves accuracy over time. Modern claims systems with integrated edits, automated accumulator tracking, and configuration validation tools produce higher accuracy than legacy systems with manual processes. The TPA that invests in technology improves accuracy. The TPA that runs on outdated systems struggles to maintain standards.

The employer evaluating TPAs should ask about continuous improvement. Does the TPA conduct root cause analysis? How often is system configuration validated? What is the training program for claims staff? When was the claims system last upgraded? These questions reveal whether the TPA treats claims accuracy as a continuous pursuit or as a static state.