Home/White Papers/The Eligibility-First Revenue Defense
WHITE PAPER · FRONT-END REVENUE CYCLE · EDITION 1, 2026

The cheapest denial to fix is the one stopped at the front door.

Registration and eligibility is the single largest root cause of claim denials, at nearly 27 percent according to MGMA. It is also the only major root cause a provider can close entirely before a claim exists. This paper is the operating manual for that control: the transaction, the cadence, the coordination logic, the 2026 Medicaid rules, and the queue where the exceptions actually get worked.

Denial root cause
~27%
Registration and eligibility, per MGMA
Electronic adoption
96%
Medical 270/271, CAQH 2024 Index
Provider cost saved
$6.57
Per transaction, manual to electronic
Checkpoints per visit
3
Scheduling, T-72 hours, check-in
Coverage caught
99.4%
At the moment of service, ASP-RCM book

Section 01 · the economicsThe busiest transaction in the revenue cycle is also the cheapest to automate.

Before you argue about denial rates, look at volume. Eligibility and benefit verification is not a step in the revenue cycle; it is the majority of the revenue cycle by transaction count, and it is the place where a small unit cost multiplied by an enormous unit volume becomes a real number on a real income statement.

The CAQH 2024 Index, reporting on calendar year 2023, measured 31.5 billion medical eligibility and benefit verifications, up five percent year over year. That one transaction was 51 percent of all medical administrative transaction volume the Index tracks and 53 percent of the tracked spend, roughly 44 billion dollars. Prior authorization, claim status, remittance advice, and claim submission combined do not come close on either axis.

Automation is also further along here than anywhere else. CAQH reports 96 percent of medical verifications ran fully electronic on the ASC X12N 270/271 standard in 2023, up from 94 and 90 percent in the two prior editions. Four percent still ran through payer portals and interactive voice response, and the fully manual residue rounded to zero. That looks like a solved problem, which is exactly where most executives stop reading.

The residual four percent is the expensive four percent. CAQH puts the medical provider cost of a manual check at 8.57 dollars, a portal check at 4.46 dollars, and a fully electronic check at 2.00 dollars, a provider savings opportunity of 6.57 dollars per transaction from manual and 2.46 from a portal. Across the residual national volume that is an 11.7 billion dollar medical industry opportunity, the largest of any transaction CAQH measures and up 27 percent from the prior edition, plus 12 minutes of provider time recovered per transaction converted.

31.5B
Medical eligibility and benefit verifications in 2023, 51 percent of all tracked medical transaction volume
CAQH 2024 Index
$8.57
Medical provider cost of one fully manual eligibility and benefit verification
CAQH 2024 Index
$2.00
Medical provider cost of the same verification run fully electronic on X12 270/271
CAQH 2024 Index
12 min
Provider time saved per transaction converted from manual to electronic
CAQH 2024 Index
EXHIBIT 1 · MEDICAL PROVIDER COST PER ELIGIBILITY AND BENEFIT VERIFICATION, BY MODE $0.00 $2.50 $5.00 $7.50 $9.375 THE AVOIDABLE SPREAD $6.57 per transaction, manual to electronic $8.57 Fully manual PHONE, FAX, MAIL $4.46 Partially electronic PAYER PORTAL, IVR $2.00 Fully electronic ASC X12N 270/271
Exhibit 1 · Source: CAQH 2024 Index Report, medical provider cost per transaction by mode, reporting on calendar year 2023. The shaded band is the per-transaction provider savings opportunity CAQH attributes to moving a manual verification onto the electronic standard.

What the cost table does not say out loud is that the four percent still running by phone and portal is not a random four percent. It is the hard four percent: the Medicaid managed care member whose plan of record is ambiguous, the newborn not yet added to the policy, the patient whose name does not match the payer's member file, the secondary carrier nobody asked about. Those are precisely the encounters most likely to deny. Automation rate measured by transaction count understates the problem because the residue is adversely selected.

WHAT THIS MEANS Do not benchmark your eligibility program on automation rate alone. Benchmark it on what happens to the encounters that automation cannot clear, because those are the encounters that become denials.

Section 02 · the transactionThe anatomy of a 270 and its 271.

You cannot manage what you treat as a black box. The 270 inquiry and the 271 response are structured documents with named loops and segments, and every eligibility denial you will ever receive traces back to a segment that was either not asked for, not returned, or not read.

The ASC X12N 270/271 pair is the HIPAA-named standard for eligibility and benefit inquiry and response, maintained by X12 and operationalized by the CAQH CORE operating rules. The 270 travels outbound from the provider and names four parties in nested hierarchical loops: the information source, which is the payer; the information receiver, which is the provider or clearinghouse; the subscriber; and, when the patient is not the policyholder, the dependent. Inside the subscriber or dependent loop, the inquiry carries identity segments and one or more service type codes that declare which benefit you are asking about.

That last part is where most implementations quietly underperform. A generic inquiry using service type code 30, health benefit plan coverage, answers only "is this person covered by anything." It does not reliably return the deductible status, copay, and prior authorization requirement for the service you are about to render. Asking at the service level is what converts an eligibility check into a verification of benefits. CAQH notes that the National Committee on Vital and Health Statistics recommended to HHS in June 2023 an updated CORE Eligibility and Benefits Data Content Operating Rule set requiring plans to return telemedicine benefits, prior authorization status, remaining coverage benefits, tiered benefits, and procedure-level detail. The direction of travel is more granularity, and providers who ask coarse questions today will still get coarse answers when the rule lands.

EXHIBIT 2 · ANATOMY OF THE 270 INQUIRY AND THE 271 RESPONSE 270 · OUTBOUND INQUIRY Provider asks the plan a question ISA / GS / ST / BHT Envelope and transaction header LOOP 2000A · NM1*PR Information source: the payer LOOP 2000B · NM1*1P Information receiver: provider NPI LOOP 2000C / 2000D · NM1*IL Subscriber, then dependent if the patient is not the policyholder EQ · SERVICE TYPE CODE Ask per service, not just code 30. This decides how useful the 271 is. TRN · DTP · SE 270 271 271 · INBOUND RESPONSE What the plan sends back, segment by segment EB*1 · ACTIVE COVERAGE Plan type, network status, and the benefit lines that follow it EB*C / EB*B / EB*A / EB*G Deductible, copay, coinsurance, and out-of-pocket maximum, with remaining DTP*291 / DTP*307 Plan and eligibility dates. The two dates that CARC 26 and CARC 27 turn on EB*R · OTHER OR ADDITIONAL PAYER The COB breadcrumb. Ignore this and you bill the wrong payer first AAA · REQUEST VALIDATION Subscriber not found, invalid ID, plan not on file. This is the exception queue MSG · free-text remarks the parser must not discard SEGMENT NAMES PER THE ASC X12N 270/271 IMPLEMENTATION GUIDE · ILLUSTRATIVE, NOT A CONFORMANCE SPECIFICATION
Exhibit 2 · The loops and segments that decide whether a verification is useful. Three of them predict a specific denial: the DTP date pair predicts CARC 26 and 27, the EB*R other-payer line predicts a coordination of benefits denial, and an AAA rejection predicts CARC 31 if nobody works it.

Read the response like a claim, not like a checkbox

A well-built verification does five things with a 271 that a checkbox implementation does not. It stores the effective and termination dates rather than a boolean. It stores remaining deductible and out-of-pocket amounts with the timestamp they were true. It keeps the other-payer segment even when a primary is already on file, because a new secondary appearing is a signal. It preserves free-text remark segments rather than discarding what the parser did not recognize. And it treats an AAA rejection as work to be assigned, not as an answer.

The ASP-RCM stack parses the 271 into structured benefit detail per service code: in-network and out-of-network deductible status, copay, coinsurance, out-of-pocket maximum, prior authorization requirement by service code, specialist referral rules, mental health and substance use parity exceptions, and ABA authorization rules where applicable. Round trip runs under 300 milliseconds at the 95th percentile through the gateway, against a clearinghouse baseline near 1.2 seconds. That latency figure is not vanity. When verification is fast enough to run synchronously at intake, it happens on every patient. When it is slow, staff learn to skip it.

Section 03 · the cadenceOne check is a snapshot. Coverage is a moving target.

The most common design flaw in an eligibility program is not the transaction. It is the timing. A verification performed once, at scheduling, answers a question about a date that has not happened yet, and then is never asked again.

Consider what changes between a scheduling call and a remittance advice. A patient changes employers. An employer changes brokers and the whole roster migrates to a new plan identifier. A deductible resets on January 1. A Medicaid renewal comes due and the member does not return the form. A dependent ages off. A secondary policy terminates. Every one of those events is invisible to a verification captured six weeks earlier, and every one produces a denial that looks, on the remittance, like a front-office mistake.

The MGMA Stat poll fielded March 5, 2024 found 60 percent of medical group leaders reported claim denial rates had risen year over year, with 29 percent flat and 11 percent improved. Among the causes leaders named were eligibility verification failures and incorrect identification numbers, and the groups that reduced denials credited stronger front-desk training and more robust eligibility verification. That is the whole argument in one poll: the same control, run better and run more often, moves the number.

The right design is a cadence, not an event. Batch and real-time are not competing architectures, they are different jobs. Batch sweeps are population surveillance, catching change across everyone you are responsible for. Real-time inquiries are decision points, answering a question a human is about to act on. A serious program runs both and knows which failures belong to which.

Exhibit 3 · Verification cadence matrix. Latency targets and coverage figures reflect the ASP-RCM verification stack described at /ai/for/eligibility-verification/. Mode names follow the CAQH 2024 Index taxonomy.
CheckpointModeTriggerWhat it catchesLatency budgetFailure it prevents
Daily population sweepBATCHNightlyPlan migrations, terminations, deductible resets, Medicaid renewal outcomes across the whole active panelOvernightSilent coverage loss between visits
Checkpoint 1: schedulingREAL TIMEAt bookingCoverage exists, network status, prior authorization requirement by service code<300 ms P95Booking a visit nobody will pay for
Checkpoint 2: T-72 hoursBATCH72h outAny change since booking, refreshed deductible and out-of-pocket remainingOvernightStale financial quote at check-in
Checkpoint 3: check-inREAL TIMEAt arrivalSame-day termination, same-day plan switch, final patient responsibility<300 ms P95CARC 27, service after termination
Pre-claim gateBATCHPre-submitCoverage as of date of service versus coverage on file at claim buildOvernightBilling a payer who is no longer primary
Self-pay discovery rescanBATCHMonthlyCoverage on accounts recorded as self-pay, across 1,200 plus payer endpointsMonthlyWriting off a billable account
WHAT THIS MEANS If your eligibility program has one checkpoint, you do not have a program. You have a data-entry step whose accuracy decays every day between the booking and the bill.

Section 04 · the scoreboardFour codes tell you exactly which control broke.

Claim Adjustment Reason Codes are maintained by X12 and returned on every remittance. Four of them are pure front-end eligibility signals, and each one points at a different, specific failure in the verification chain. Read them as a diagnostic, not as a workqueue.

The value of this mapping is that it inverts the usual denial conversation. Instead of asking how quickly a denial was appealed, it asks which control was supposed to catch it and why the control did not fire. A rising CARC 27 count is not an appeals problem; it is a check-in verification problem. A rising CARC 31 count is not a payer problem; it is a demographic-matching problem in your registration workflow. Denial categories that map cleanly to a control are the only ones you can drive to zero.

CARC 26COVERAGE NOT YET ACTIVE

Expenses incurred prior to coverage

The service date falls before the plan effective date returned in the 271. This is the classic newborn, new-hire waiting period, and January plan-year transition denial. It also appears when a patient is enrolled retroactively and the roster you verified against was not yet current.

Control that prevents itStore the DTP effective date, not a coverage boolean, and compare it to the date of service at claim build rather than at registration.
CARC 27COVERAGE ALREADY ENDED

Expenses incurred after coverage terminated

The service date falls after the plan termination date. This is the denial that a single check at scheduling structurally cannot prevent, because the termination happened after you asked. Job change, Medicaid procedural termination, and dependent age-off are the usual causes.

Control that prevents itA check-in verification and a daily background sweep. Termination is a change event, so it needs surveillance, not a one-time lookup.
CARC 31IDENTITY DID NOT MATCH

Patient cannot be identified as our insured

The payer could not match the demographics submitted to a member record. Usually a transposed member ID, a maiden name, a hyphenated surname stored inconsistently, or a dependent submitted under their own name when the policy is held by a parent. On the 271 side this shows up first as an AAA rejection.

Control that prevents itRoute every AAA rejection into a staffed exception queue at intake. An unresolved AAA today is a CARC 31 in 45 days.
CARC 197AUTHORIZATION ABSENT

Precertification or authorization not fulfilled

The service required prior authorization and none was on file at adjudication. The 271 frequently carries the authorization requirement by service code, which means this denial is often preventable at the same moment coverage is confirmed, provided the inquiry asked at the service level.

Control that prevents itSurface the prior authorization flag from the 271 at intake and block scheduling of the service until authorization exists on file.

Two adjacent codes belong on the same monthly report even though they are not strictly eligibility failures. CARC 22 indicates the care may be covered by another payer under coordination of benefits, and CARC 109 indicates the claim was sent to a payer or contractor that is not responsible for it. Both are order-of-benefits failures rather than coverage failures, and both are addressed in the next section. A denial dashboard that separates coverage failures from sequencing failures will assign work correctly. One that lumps them into a single eligibility bucket will send a coordination problem to the person who verifies coverage, where it will not get fixed.

For the broader taxonomy of denial root causes beyond the front end, the companion paper The Denial Prevention Field Manual maps twelve categories to the control and KPI for each.

Section 05 · the ordering problemKnowing who covers the patient is not the same as knowing who pays first.

Coordination of benefits is where correct eligibility still produces a denial. Every payer is confident it is not primary, and the burden of proving otherwise sits with the provider who submitted in the wrong order.

Order of benefits is governed by three different rule sets depending on who is involved, and mixing them up is the most common cause of a coordination denial. Between two commercial plans the ordering derives from the National Association of Insurance Commissioners coordination of benefits model regulation as adopted in each state: a plan covering a person as an employee or subscriber pays before a plan covering the same person as a dependent, and for a dependent child covered by two parents who are not separated or divorced, the birthday rule applies, meaning the plan of the parent whose birthday falls earlier in the calendar year is primary. Court decrees, active-employee versus retiree or COBRA status, and length of coverage break the remaining ties.

Where Medicare is involved, the framework is Medicare Secondary Payer, codified at 42 U.S.C. 1395y(b) and administered through the CMS Medicare Secondary Payer Manual, Publication 100-05. The main triggers are the working aged provision, disability with a large group health plan, the end-stage renal disease coordination period, and situations where workers compensation, no-fault, or liability insurance is responsible. CMS is explicit that running a 270/271 does not relieve a provider of the obligation to ask the MSP questions when insurance information has changed or is uncertain, and directs that MSP eligibility questions go to the plans identified in the 271 response. The HIPAA Eligibility Transaction System, the CMS real-time route for Medicare beneficiary eligibility, has returned non-group health plan MSP diagnosis codes on eligibility responses since the CMS HETS information bulletin of September 6, 2023, so providers can see which payer is primary for a category of care.

Medicaid sits at the end of the chain by statute. Section 1902(a)(25) of the Social Security Act and the third-party liability rules at 42 CFR 433.139 make Medicaid the payer of last resort, so every other liable payer must be identified and billed first. In practice that is a discovery problem rather than a sequencing problem: a Medicaid coordination denial almost never means staff ordered the payers wrong, it means nobody knew the commercial policy existed.

EXHIBIT 4 · ORDER OF BENEFITS DECISION TREE All payers on file for this patient Is the care related to workers comp, no-fault, or liability? YES That carrier is primary 42 U.S.C. 1395y(b) NO Is Medicare one of the payers, with an MSP trigger present? YES Group health plan pays before Medicare Working aged, disability, or ESRD coordination period NO Two commercial plans? Apply the NAIC model order YES 1. Subscriber before dependent 2. Birthday rule for dependent children 3. Active before COBRA NO Medicaid is the payer of last resort Every other liable payer must be billed first SSA 1902(a)(25) · 42 CFR 433.139 Any EB*R other-payer segment on the 271 reopens this tree, no matter how confident the payer file already looks.
Exhibit 4 · Order of benefits, simplified to the decision path a front-end team actually walks. Commercial ordering follows the NAIC coordination of benefits model regulation as adopted by each state; Medicare ordering follows 42 U.S.C. 1395y(b) and the CMS Medicare Secondary Payer Manual, Publication 100-05.
Exhibit 5 · Medicare Secondary Payer triggers a front-end team must screen for. Statutory framework at 42 U.S.C. 1395y(b); operational detail in the CMS Medicare Secondary Payer Manual, Publication 100-05.
MSP situationWho pays firstFront-end question that finds itDenial if missed
Working agedGHP FIRSTIs the patient or spouse actively working, and does the employer group health plan meet the applicable employer size threshold?CARC 22
DisabilityGHP FIRSTIs the patient under 65, Medicare-entitled by disability, and covered by a large group health plan through current employment?CARC 22
End-stage renal diseaseGHP FIRSTIs the patient inside the ESRD coordination period, and when did it start?CARC 22
Workers compensationWC FIRSTIs this visit related to a work injury with an open claim number?CARC 109
No-fault or liabilityCARRIER FIRSTWas there an auto accident or other injury with a liable third party?CARC 109
Veterans Affairs authorizationVA FIRSTWas the care authorized by the VA for this episode?CARC 109

The operational lesson is that MSP screening is a conversation, not a query. The 271 tells you which plans exist; it does not tell you whether the spouse is still actively employed or whether last week's back pain is a work injury. CMS treats the questionnaire as an obligation precisely because the transaction cannot answer those questions. Build the screening questions into intake, record the answers with a date and an initial, and the appeal you eventually write will have something to stand on.

Every payer is confident it is not primary. Coordination denials are not a knowledge failure, they are an evidence failure, and the evidence has to be collected before the visit, not after the remittance.

The Eligibility-First Revenue Defense · Section 05

Section 06 · the 2026 problemMedicaid renewal churn is about to double in frequency.

Two separate federal changes land back to back. One restores normal renewal timelines. The other cuts the renewal interval in half for the largest Medicaid population a typical outpatient practice sees. Together they change the arithmetic of every Medicaid-heavy book of business.

Start with the baseline rules, because the new requirement layers on top of them rather than replacing them. Under 42 CFR 435.916, a state must renew eligibility once every 12 months and no more frequently, for the populations the section covers. It must first attempt an ex parte renewal, determining continued eligibility from reliable information already in the individual's account or otherwise available to the agency, without asking the beneficiary for anything. Only when ex parte fails may the state send a prepopulated renewal form, and the beneficiary must get at least 30 calendar days from the date the form is sent to respond. If coverage is then terminated for failure to return the form, the state must reconsider eligibility without requiring a new application if the individual returns the form within 90 calendar days after the termination date, or a longer period the state elects.

That 90-day reconsideration window is the single most useful fact in this section for a revenue cycle team, and it is routinely wasted. A patient terminated for a procedural reason on March 1 who returns the form on April 10 can be reinstated, frequently with retroactive effect, converting what your system already wrote down to self-pay back into a billable Medicaid claim. If your accounts reach bad debt or self-pay collections faster than 90 days, you are systematically destroying recoverable revenue on the population least able to pay.

Now the change. CMS issued State Medicaid Director letter SMD number 26-001 on March 6, 2026, implementing section 71107 of Public Law 119-21, which amends section 1902(e)(14) of the Social Security Act to require eligibility redeterminations every six months rather than annually for individuals enrolled through the Medicaid adult expansion group under section 1902(a)(10)(A)(i)(VIII), along with comparably enrolled waiver populations. The provision is effective December 31, 2026 and applies to renewals scheduled on or after January 1, 2027. CMS gives states two paths: move an individual's 2027 renewal date earlier to spread volume across the year, or keep the previously scheduled 2027 date and shorten from there. State plan amendments are due no later than March 31, 2027. CMS is explicit that the ex parte requirement still governs.

Jan 1, 2026
Regular timeliness resumes

Renewals initiated on or after this date are processed under the standard federal renewal timeliness rules rather than unwinding-era flexibilities.

CMS guidance · 42 CFR 435.912
Mar 6, 2026
CMS issues SMD 26-001

Guidance to state Medicaid directors implementing the six-month redetermination requirement, with two options for sequencing 2027 renewal dates.

CMS letter SMD 26-001
Dec 31, 2026
Six-month rule effective

Section 71107 of Public Law 119-21 takes effect, amending section 1902(e)(14) of the Social Security Act for the adult expansion group.

Public Law 119-21, sec. 71107
Jan 1, 2027
Renewals double in frequency

The requirement applies to renewals scheduled on or after this date. State plan amendments are due no later than March 31, 2027.

CMS letter SMD 26-001

The operational consequence is straightforward and unpleasant. Doubling renewal frequency for the expansion population roughly doubles the procedural termination events your patients pass through, and procedural terminations are the ones most likely to be reversed inside the reconsideration window. Coverage will flicker: a patient covered at an April visit may show terminated in July and reinstated in August with retroactive effect. Any workflow that treats coverage status as durable for more than a few weeks will produce denials, and any collections workflow that moves a Medicaid balance to patient responsibility inside 90 days will produce avoidable write-offs.

What a practice should actually change

Four adjustments carry most of the value. Put every Medicaid and Medicaid managed care patient on a daily eligibility sweep rather than a per-visit check, so a termination surfaces within a day instead of on a remittance. Hold Medicaid balances out of self-pay routing for the full 90-day reconsideration period and rescan before any write-off decision. Capture the renewal due date as a field whenever the 271 or the state portal exposes it, and treat the 30 days before it as an outreach window. And plan staffing around the state's chosen sequencing option, because a state that redistributes 2027 renewal dates produces a smooth load while a state that keeps existing dates produces sharp monthly peaks.

Medicaid response formats are not uniform, which is the practical obstacle to all of the above. The ASP-RCM stack carries state-specific adapters for the top 30 Medicaid programs, covering 92 percent of national Medicaid enrollment, normalizing raw 271 responses into consistent fields, with schema updates rolled within 72 hours when a state changes its format. Tooling that treats Medicaid as one homogeneous payer loses exactly the fields that matter here.

Section 07 · the found moneyA meaningful share of your self-pay is not actually self-pay.

Self-pay is a status assigned by a conversation at the front desk, and conversations are lossy. Insurance discovery is the practice of asking the payer network directly instead of asking the patient, and it consistently finds coverage the intake interview did not.

Accounts land in self-pay for several distinct reasons and only one of them is genuine. Some patients truly have no coverage. Some have coverage they do not know about: a Medicaid or Medicaid managed care enrollment processed after a hospital-based application, a spousal plan they are not the subscriber on, or a retroactive determination effective for dates already rendered. Some presented a card that failed verification and were flipped to self-pay to keep the schedule moving, with no follow-up. And some are Medicaid members inside a reconsideration window whose coverage will be reinstated retroactively.

Discovery works by running identity attributes against a broad set of payer endpoints rather than a single guess. The ASP-RCM implementation rescans self-pay accounts monthly against more than 1,200 connected payer endpoints, and any coverage found flows into the same continuous 270/271 sweep as the rest of the panel, so a discovered policy stays current instead of going stale the day after it is found. Across the active book the discovery hit rate on previously declared self-pay runs about 11 percent, and self-pay receivables reclassified to a billable payer typically lift 6 to 11 percent annually.

Exhibit 6 · self-pay discovery funnel ILLUSTRATIVE · 1,000-ACCOUNT COHORT
1,000
Accounts recorded self-pay at month end
~110
Coverage found on rescan at an 11 percent hit rate
Monthly
Rescan cadence across 1,200 plus endpoints
6 to 11%
Typical annual lift in self-pay AR reclassified to billable
Hit rate and AR lift figures are ASP-RCM active-book measurements, anonymized. Individual results depend on payer mix, Medicaid share, and prior intake discipline.

Two governance points keep a discovery program honest. First, discovery must feed the continuous sweep, not a one-time flag, or you will bill a policy that terminated between discovery and submission. Second, timely filing is the constraint that decides whether discovery is profitable. A policy found 200 days after the date of service is worth very little if the payer's filing window closed at 180 days. Run the rescan monthly rather than at write-off, and the found coverage is still billable. The full mechanics of working an aged inventory back to cash are covered in The AR Workdown Operating System.

Section 08 · the workflowAutomation's real output is a shorter queue, not an empty one.

A verification stack that clears 90 percent of a panel automatically has not eliminated human work. It has concentrated it. The design question that decides whether the program succeeds is what happens to the remainder.

Every 271 lands in one of three states. Clean, meaning coverage is confirmed with usable benefit detail and it writes back to the practice management system without a human touching it. Rejected, meaning the payer returned an AAA segment because the subscriber was not found, the identifier was invalid, or the provider is not on file. Or ambiguous, meaning something came back but it does not answer the question: a coverage-active response with no benefit detail, a plan name that does not match the card the patient presented, or a Medicaid managed care response naming an organization you are not contracted with.

The failure mode in most shops is that the second and third states get treated as the first. A queue nobody owns becomes a report nobody reads, and every unresolved AAA in that report becomes a CARC 31 denial about 45 days later. The fix is unglamorous: name an owner, set a service level, and make the queue the only path out of the ambiguous state. Clean results never wait on a person, and no exception ever bypasses one.

Classify

Every 271 sorts to clean, rejected, or ambiguous on parse. No fourth bucket, no silent discards.

Route

Rejections and ambiguities queue to a named specialist by reason code, not to a shared inbox.

Resolve

Demographic correction, payer portal lookup, or a call. Resolution is recorded against the reason code.

Re-verify

The corrected identity goes back through a fresh 270 rather than being marked resolved by assertion.

Feed back

Reason-code volumes tune the parser and the registration script. The queue should shrink month over month.

Exhibit 7 · exception queue console ILLUSTRATIVE · ONE DAY, ONE LOCATION
148
Scheduled visits swept at checkpoint 2
131
Clean 271, written back with no human touch
9
Exceptions routed to a named specialist
8
Pending on payer gateway, auto-retried
Illustrative console volumes · the operating rule is that clean results never wait on a person and no exception exits without a fresh 270
Exhibit 8 · Exception dispositions and the denial each one prevents. AAA rejection categories follow the ASC X12N 270/271 implementation guide.
Exception typeSignal on the 271OwnerService levelDenial prevented
Subscriber not foundAAA rejection, identity not matchedRegistration specialistSame dayCARC 31
Invalid or missing member IDAAA rejection on the identifierRegistration specialistSame dayCARC 31
Provider not on fileAAA rejection on the receiver loopCredentialing72 hoursCARC 109
Coverage active, no benefit detailEB active with no financial segmentsBenefits specialistBefore visitBad quote
Other payer namedEB*R other or additional payerCOB specialistBefore claimCARC 22
Authorization requiredPrior auth flag by service codeAuthorization teamBefore visitCARC 197
Termination date in rangeDTP termination on or before service dateRegistration specialistSame dayCARC 27
Payer gateway timeoutNo response inside the windowAutomated retryAutoNone

One further behavior separates a mature program from a busy one. When a plan migration is detected mid-cycle, in-process claims should be rebilled under the corrected policy automatically rather than waiting for the denial to arrive and be worked. Catching the change is only half the value; acting on it before adjudication is the other half, and it is the half that keeps the correction out of the denial statistics entirely.

Section 09 · the measurementSix numbers that tell you the front door is holding.

An eligibility program that cannot be measured will not survive its first budget cycle. These six figures are the ones a CFO can check, each with a target, a warning threshold, and a review cadence.

Two of them are leading indicators that move within weeks of a change, three are diagnostics that tell you which control broke, and one is the lagging financial figure that pays for the program. Report all six monthly on one page. When any of them moves, the mapping in Section 04 tells you where to look.

Leading · 01
Coverage capture at service
Target 99%+ASP-RCM 99.4%

Share of encounters where active coverage was correctly identified at the moment of service. Reviewed weekly by location.

Leading · 02
Verification completion rate
Target 100%Warn below 95%

Share of scheduled visits verified at all three checkpoints. A gap here always precedes a denial spike.

Diagnostic · 03
Eligibility CARC volume
Trend to zero26 · 27 · 31 · 197

Monthly count and dollar value by code. A single code climbing names the broken control directly.

Diagnostic · 04
Exception queue age
Target same dayWarn over 48h

Median time from AAA rejection to resolution. Unworked exceptions convert to CARC 31 about 45 days later.

Diagnostic · 05
Discovery hit rate
Benchmark ~11%Monthly rescan

Coverage found per hundred self-pay accounts rescanned. Falling hit rate usually means intake improved, which is good news.

Lagging · 06
Self-pay AR reclassified
Target 6 to 11%Annual

Dollars moved from self-pay to a billable payer within the filing window. This is the line that funds the program.

WHAT THIS MEANS Report verification completion and eligibility CARC volume on the same page. When completion drops in one month and CARC volume rises in the next, you have proven the causal link inside your own data and the budget conversation is over.

One caution on benchmarking. Coverage capture rates near 99 percent are achievable but they are a function of cadence, payer connectivity breadth, and Medicaid adapter depth, not of buying a tool. A practice with heavy Medicaid managed care exposure and a single once-at-intake check will not reach that number regardless of vendor. Set your first target against your own current baseline, measure monthly, and let the trend rather than the peer benchmark drive the investment case.

SourcesEvery number in this paper, traced.

Industry benchmarks are attributed to their primary publisher and dated. Policy statements cite the regulation, statute, or agency guidance directly. Figures attributed to ASP-RCM are measurements from our active book, anonymized, and are labeled as such wherever they appear.

  1. CAQH 2024 Index Report, reporting on calendar year 2023. Source of transaction volume (31.5 billion medical eligibility and benefit verifications, 51 percent of medical transaction volume), spend (44 billion dollars, 53 percent of tracked medical spend), electronic adoption (96 percent medical, fully electronic), per-transaction provider cost by mode ($8.57 manual, $4.46 portal, $2.00 electronic), per-transaction provider savings opportunity ($6.57), medical industry savings opportunity (11.7 billion dollars), and the 12-minute provider time saving per converted transaction.
  2. MGMA. Registration and eligibility as the leading claim denial root cause at nearly 27 percent. MGMA Stat poll fielded March 5, 2024 for the finding that 60 percent of medical group leaders reported year-over-year denial rate increases, 29 percent flat and 11 percent improved, with eligibility verification failures and incorrect identification numbers among the named causes.
  3. X12. ASC X12N 270/271 Health Care Eligibility Benefit Inquiry and Response, the HIPAA-named standard, and the Claim Adjustment Reason Code list from which CARC 22, 26, 27, 31, 109, and 197 are drawn.
  4. 42 CFR 435.916. Medicaid renewal frequency of once every 12 months, the ex parte renewal requirement, the minimum 30 calendar days to respond to a prepopulated renewal form, and the 90 calendar day reconsideration period after a procedural termination.
  5. CMS State Medicaid Director letter SMD number 26-001, issued March 6, 2026, implementing section 71107 of Public Law 119-21, which amends section 1902(e)(14) of the Social Security Act to require six-month eligibility redeterminations for the Medicaid adult expansion group under section 1902(a)(10)(A)(i)(VIII). Effective December 31, 2026, applying to renewals scheduled on or after January 1, 2027, with state plan amendments due no later than March 31, 2027.
  6. CMS. Guidance that Medicaid and CHIP renewals initiated on or after January 1, 2026 return to regular federal renewal timeliness standards at 42 CFR 435.912 and 457.340.
  7. CMS Medicare Secondary Payer Manual, Publication 100-05, and 42 U.S.C. 1395y(b). Medicare Secondary Payer triggers and the instruction that use of the 270/271 does not relieve a provider of the obligation to ask MSP questions.
  8. CMS HIPAA Eligibility Transaction System (HETS) 270/271 Companion Guide and the CMS HETS information bulletin of September 6, 2023 on non-group health plan MSP diagnosis codes returned on eligibility responses.
  9. Social Security Act section 1902(a)(25) and 42 CFR 433.139. Medicaid third-party liability and payer-of-last-resort requirements.
  10. National Association of Insurance Commissioners Coordination of Benefits Model Regulation, as adopted state by state, for commercial order-of-benefits rules including the birthday rule for dependent children.
  11. ASP-RCM Solutions active-book measurements, anonymized: 99.4 percent coverage caught at the moment of service, sub-300 millisecond 270/271 round trip at the 95th percentile, 1,200 plus connected payer endpoints, 30 state Medicaid adapters covering 92 percent of national Medicaid enrollment, approximately 11 percent discovery hit rate on previously declared self-pay, and 6 to 11 percent annual self-pay AR reclassification. Detail at AI Eligibility Verification and AI for Eligibility Verification.

Common questionsSix questions operators ask us.

What is a 270/271 transaction?

The ASC X12N 270 is the standardized eligibility and benefit inquiry a provider sends to a health plan, and the 271 is the plan's response. The pair is the HIPAA-named standard for confirming active coverage, plan and network status, patient financial responsibility, and prior authorization requirements before a service is delivered. The CAQH 2024 Index reports that 96 percent of medical eligibility and benefit verifications were fully electronic on this standard.

How often should eligibility be verified?

Once is not enough. A single check at scheduling tells you what was true on the day you asked, not on the day of service or the day the claim adjudicates. ASP-RCM verifies at three checkpoints per visit, at scheduling, at 72 hours before the visit, and at check-in, with a daily background sweep on every active patient so material coverage changes surface within 24 hours.

Which CARC codes signal an eligibility failure?

CARC 26 for expenses incurred prior to coverage, CARC 27 for expenses incurred after coverage terminated, CARC 31 when the patient cannot be identified as the payer's insured, and CARC 197 when a precertification or authorization requirement was not fulfilled. Tracking those four monthly is the cleanest scoreboard for a front-end eligibility program.

What changes for Medicaid redeterminations in 2026?

Two things. Renewals initiated on or after January 1, 2026 return to the regular federal timeliness standards. Separately, CMS letter SMD number 26-001, issued March 6, 2026, implements section 71107 of Public Law 119-21, which amends section 1902(e)(14) of the Social Security Act to require eligibility redetermination every six months for the Medicaid adult expansion group, effective December 31, 2026 and applying to renewals scheduled on or after January 1, 2027.

How much does a manual eligibility check cost?

The CAQH 2024 Index, reporting on 2023 data, puts the medical provider cost at 8.57 dollars for a fully manual eligibility and benefit verification, 4.46 dollars through a payer portal, and 2.00 dollars fully electronic. The provider savings opportunity is 6.57 dollars per transaction moving from manual to electronic, and CAQH estimates 12 minutes of provider time saved per transaction.

Is insurance discovery worth running on self-pay?

Usually yes. A meaningful share of accounts recorded as self-pay carry coverage the intake conversation never surfaced. Across the ASP-RCM active book the discovery hit rate on previously declared self-pay runs about 11 percent, and self-pay AR reclassification to a billable payer typically lifts 6 to 11 percent annually.

How many of your denials started at the front desk?

Send 30 days of appointment volume and your top 10 payer mix. A senior partner returns a written audit: coverage capture rate at intake, mid-cycle change detection rate, self-pay reclassification opportunity with dollar values, discovery hit rate prediction, and a 90-day plan. No PHI required.

Aparna Suresh, CPBCertified Professional Biller · President and Founder, ASP-RCM Solutions