Jump to: RCM & Money Flow ยท Eligibility & Insurance ยท Coding ยท Claims & EDI ยท Payments & Adjustments ยท Denials, Appeals & A/R ยท Compliance & Regulation ยท Systems & Industry ยท Abbreviations (145) →
RCM & Money Flow (11 terms)
RCM (Revenue Cycle Management)
The end-to-end financial process that converts healthcare services into collected revenue.
How it works: RCM is a pipeline with stages: patient registration โ eligibility verification โ service documentation โ coding โ charge entry โ claim submission โ payer adjudication โ payment posting โ denial management โ patient billing โ collections โ reporting. Every stage feeds the next; an error in an early stage surfaces as a denial or underpayment later. RCM teams measure themselves with KPIs: clean-claim rate, days in A/R, net collection rate, denial rate and cost-to-collect. When people say 'billing,' they really mean managing this whole cycle.
Example: A patient visit that was coded wrong in stage 4 shows up as a CO-50 (medical necessity) denial in stage 8 โ three weeks later. Root-cause analysis is what connects the denial back to the coding stage.
A/R (Accounts Receivable)
Money billed to payers and patients that hasn't been collected yet.
How it works: Every claim that leaves the building becomes A/R until it pays or writes off. A/R is measured in days ('Days in A/R' = average time from charge to payment) and in aging buckets: current, 30, 60, 90, 120+ days. Older A/R is harder to collect โ payer appeal windows expire, and patient balances get harder to recover. Billing teams work A/R queues prioritized by dollar value and age; the goal is to keep Days in A/R around 35โ45 days. A/R that ages past timely-filing limits or appeal deadlines often converts to write-offs.
Example: $100,000 billed this month; $62,000 still unpaid = $62,000 A/R. The 120+ bucket with $15,000 gets a sprint week.
Charge Entry
The step where coded services (CPT/ICD/modifiers) are entered into the billing system as billable claim lines.
How it works: Charge entry takes the provider's documentation (superbill or coded encounter) and turns it into claim-ready lines: CPT codes, diagnosis codes with pointers, modifiers, units, place of service, provider NPI, and dollar amounts from the fee schedule. Accuracy here determines downstream fate โ wrong data produces rejections (caught by clearinghouse) or denials (caught by the payer, later and more expensively). Charge entry specialists are measured on accuracy rate and volume; a good operator also catches upstream problems (missing documentation, ineligible patients) at this stage.
Example: Provider performed 99213 + 20610. Charge entry adds both lines with correct ICD pointers (visit โ back pain, injection โ knee) so the payer can verify medical necessity per line.
Clean Claim
A claim that passes all validations and pays on first submission without edits, rejections or denials.
How it works: A clean claim has correct patient demographics, active insurance data, valid payer ID/member ID, proper coding with medical-necessity linkage, required modifiers, and correct POS/provider info. Clean-claim rate (percentage of claims paying first-pass) is the single best quality metric for a billing operation โ it predicts A/R days, denial workload and cash speed. Most 'dirty' claims trace back to front-end data (registration/eligibility), not billing itself.
Example: 95% clean-claim rate means only 1 in 20 claims needs rework โ a healthy operation. 80% means the team spends a day a week fixing preventable errors.
Adjudication
The payer's internal process of deciding how much (if anything) to pay a claim.
How it works: Adjudication runs a claim through rule layers: eligibility check, benefit coverage, prior authorization verification, coding edits (NCCI bundling, medical policy), coordination of benefits, and finally the fee schedule calculation. Output is either payment, partial payment, denial, or 'pending additional information.' The result is communicated via EOB/ERA with CARC codes explaining each decision. Understanding adjudication order explains most denials: claims that fail early gates (eligibility, COB) never even reach the payment calculation.
Example: A claim with valid eligibility but bundled codes denies CO-97 at the coding-edit layer โ the fee schedule never got a chance to price it.
Write-Off
Removing an uncollectible balance from A/R with a documented reason.
How it works: Write-offs come in two flavors: contractual (billed minus allowed โ mandatory, happens on every claim) and non-contractual (denials past deadline, unrecoverable balances, small-balance policies). Non-contractual write-offs require approval and documentation because unmanaged write-offs hide process failures and, if done without cause, can be fraud (waiving patient copays is illegal under anti-kickback rules). Reporting tracks write-offs by reason so the operation can see what it's actually losing.
Example: CO-29 (timely filing) denial at 100 days with no appeal path left = documented write-off, reason 'timely filing,' approval by supervisor. The denial report shows the trend so registration fixes it at the source.
Net Collection Rate
The percentage of collectible revenue actually collected (collected รท allowed, excluding contractual write-offs).
How it works: NCR answers 'of the money we were entitled to, how much did we get?' โ the truest quality metric in billing. A 96%+ NCR means the operation recovers nearly everything; below 92% means underpayments, denials and missed deadlines are leaking real money. NCR is different from gross collection rate (which divides by billed and looks worse because contractual write-offs are included). Leadership dashboards lead with NCR because it captures the effect of every function combined.
Example: Allowed $500,000 this quarter; collected $475,000 = 95% NCR. The 5% gap ($25,000) is where denial recovery and variance programs hunt.
Cost to Collect
Total billing operation cost divided by cash collected โ the efficiency metric.
How it works: Cost to collect includes salaries, software, clearinghouse fees, statement costs and collection agency fees. Healthy operations run roughly 3โ5% of collections. It's the metric leadership uses to justify automation and process investment: if the operation spends 7% to collect, there's a case for better tools. It also frames the prevention argument โ front-end fixes that cut denial rework lower cost-to-collect directly.
Example: $4 million collected annually, $160,000 billing department cost = 4% cost to collect. An eligibility API that prevents 500 denial-rework hours/year is justified against this.
Fee Schedule
The complete list of contracted payment rates a payer will pay for each service code.
How it works: Fee schedules define the allowed amount per CPT/HCPCS code per payer per contract. Claims are priced against them during adjudication, and payment posting validates against them (paid below schedule = variance). Fee schedules vary by payer, contract year, geography and provider specialty. Billing teams keep fee schedules loaded in the PM system so charge entry bills contract rates and posting catches underpayments automatically.
Example: Contract says 99213 = $150 for this payer. The claim bills $150; the payer pays $120 with CO-45 adjustment of $30 โ matching the schedule. If they'd paid $110, that's a $10 variance to pursue.
Gross Charge / Billed Amount
The provider's full list price for a service before any contractual adjustments.
How it works: Every provider maintains a chargemaster (list of gross charges). Claims bill the gross charge; the payer then applies the contract (allowed amount) and the difference becomes the contractual adjustment. Gross charges are typically much higher than allowed amounts โ this is why 'billed $1,000, insurance paid $200' looks shocking to outsiders: the contractual write-off is normal and expected. Gross charges also anchor self-pay discount policies (many states require good-faith estimates based on them).
Example: Office visit gross charge $220; this payer's allowed $150; CO-45 adjustment $70. All three numbers appear on the EOB โ the write-off is contractual, not a discount.
Chargemaster
The provider's internal price list of every billable service and item.
How it works: The chargemaster (CDM โ Charge Description Master) maps each service to a code and a gross price. It feeds charge entry and, in hospitals, drives facility billing. Chargemaster maintenance is its own discipline: prices must be updated annually, new services added, and codes kept aligned with current code sets. Errors here surface as billing at stale rates or claims with invalid codes.
Example: New drug code added to formulary but not to the chargemaster โ claims for it fail with invalid-code rejections until the CDM is updated.
Eligibility & Insurance (13 terms)
Eligibility Verification
Confirming a patient's coverage is active and reading the benefit details before service.
How it works: Verification answers: Is coverage active on the DOS? What plan type? Deductible status, copay, coinsurance, OOP max? Any referral/PA requirements? COB order? It runs through 270/271 transactions (portals, PM systems) or payer phone lines. The critical skill is reading the FULL response, not just 'Active' โ plan type, benefit specifics and COB flags change what the patient owes and whether the visit will pay. Verification should happen 2โ3 days before DOS (not at check-in) so problems get fixed before the visit.
Example: Response shows 'Active' but also 'HMO โ no PCP on file.' Catching that before the visit lets the front desk fix the referral โ after the visit it's a denial.
Subscriber
The person who owns the insurance policy (often not the patient).
How it works: Insurance claims must use the SUBSCRIBER's information (name, DOB, SSN/ID) โ not the patient's โ when the patient is a dependent. Mixing patient and subscriber data is one of the most common first-pass rejection causes. The relationship field (self, spouse, child) tells the payer how the patient connects to the policy. For children, the birthday rule determines which parent's plan is primary.
Example: Child patient covered under mom's plan: subscriber = mother, patient relationship = child, and mom's DOB goes in the subscriber-DOB field โ not the child's.
Coordination of Benefits (COB)
The rules deciding which payer pays first when a patient has more than one plan.
How it works: COB order follows standard rules: employment-based plan primary over personal; dependent children โ the parent whose birthday falls earlier in the year is primary (birthday rule); active employment primary over retiree plans; Medicare secondary to working-patient employer plans under certain size rules. The secondary payer then pays remaining balances up to its own allowed. Billing out of order produces CO-22 denials ('this may be covered by another payer'). Always document the COB order on the account.
Example: Patient has employer plan + spouse's plan: employer plan is primary. Bill it first, attach its EOB to the secondary claim.
Birthday Rule
For dependent children with two plans: the parent whose birthday is earlier in the calendar year has the primary plan.
How it works: When both parents cover a child, the calendar date (month/day, not year) of each parent's birthday decides primary coverage. Exceptions apply (divorce decrees, court orders, active-employment overrides). Getting the order wrong creates CO-22 denials and rework โ verify both plans and document the order before billing.
Example: Dad's birthday March 5, mom's July 12 โ dad's plan is primary for the child, regardless of whose employer covers whom.
Out-of-Network (OON)
Care from providers who don't have a contract with the patient's payer.
How it works: OON care triggers higher patient cost-sharing (higher coinsurance, balance billing of the difference between billed and allowed) and, in some plan types (HMO/EPO), no coverage at all except emergencies. The No Surprises Act (2022) protects patients from surprise OON bills in emergencies and certain facility situations โ providers must give good-faith estimates. Billing teams check network status during verification because OON claims pay differently and sometimes need special submission channels.
Example: Specialist is OON on a PPO: patient owes 40% coinsurance instead of 20% and may be balance-billed. Caught at verification, the patient can choose an in-network provider first.
In-Network / Contracted
Providers who signed contracts with a payer, agreeing to its fee schedules and rules.
How it works: In-network status means the payer pays per contract, patients pay the lower cost-sharing tier, and the provider writes off billed-vs-allowed (can't balance bill). Network status can lapse (credentialing expirations) โ a silent revenue killer, so billing teams watch panel participation. Some payers have multiple networks under one brand (narrow vs broad), so 'in-network' must be verified for the patient's SPECIFIC plan, not just the carrier.
Example: Provider is in Anthem PPO networks but not in Anthem's narrow HMO network โ HMO patients can't see them except as OON.
Primary Care Provider (PCP)
The designated gatekeeper doctor in HMO-type plans.
How it works: HMO plans assign a PCP who manages care and issues referrals to specialists. If a patient sees a specialist without the PCP referral on file, the claim denies (gatekeeping denial). Verification checks PCP assignment; a 'no PCP on file' flag means the front desk must get the patient to designate one before the visit. PPO plans skip the gatekeeper โ specialists are direct.
Example: HMO patient books cardiology directly: verify shows no PCP assigned. Front desk gets PCP assignment + referral card before the visit; the claim pays.
Marketplace / Exchange Plans
ACA-created individual insurance plans (Ambetter, Oscar, Molina Marketplace, etc.) bought on or off the exchange.
How it works: Marketplace plans come in metal tiers (Bronze/Silver/Gold/Platinum) defining the actuarial value โ Bronze ~60% of costs covered, Platinum ~90%. Subsidies (premium tax credits) depend on income. Billing notes: marketplace plans are commercial payers but often have tighter networks, strict PA rules, and high deductibles โ eligibility detail-reading matters more than ever. Coverage can lapse mid-year when subsidies change; re-verification is critical.
Example: Bronze Ambetter plan: $7,000 deductible. Patient owes nearly the full allowed for most services until deductible is met โ the estimate conversation matters at check-in.
Self-Funded (ASO) Plan
Employer pays claims directly; an insurance company only administers (Administrative Services Only).
How it works: Self-funded (ERISA) plans are governed by federal law, not state insurance regulation โ so state mandates (like certain coverage requirements) may not apply. The card often shows a TPA name (UMR, Meritain, Trustmark) even though the carrier brand is on it. Billing implication: plan rules can vary WILDLY between self-funded employers even under the same TPA โ read each plan's specifics rather than assuming the carrier's standard policies.
Example: Two UMR plans: employer A covers telehealth at 100%, employer B excludes it. Same payer ID, different plan documents โ verification and payer-policy reading catch the difference.
Medigap
Private supplement insurance that fills Original Medicare's cost-sharing gaps.
How it works: Medigap plans (letters AโN) cover coinsurance, copays and deductibles that Original Medicare leaves to the patient. Billing flow: bill Medicare first; Medicare auto-crossovers the claim to the Medigap payer, which pays its share. Medigap can't be combined with Medicare Advantage. Plans G and N are the most common; the letter determines exactly what's covered. In billing, Medigap claims usually cross over automatically โ verify the crossover went through when the secondary payment is missing.
Example: Patient has Plan G: after Medicare pays 80%, the G plan covers the 20% coinsurance โ the claim crossovers from the MAC to the Medigap payer automatically.
Prior Authorization (PA)
Advance payer approval required before certain services are delivered.
How it works: Payers maintain PA lists (imaging, procedures, specialty drugs, DME, hospital stays). The provider submits clinical documentation; the payer's medical-management team approves, denies or pends. PA approvals have validity windows and unit counts โ exceeding either denies CO-197. Some payers allow retro-authorization within a few days post-service; others don't. PA specialists track approvals on the account, alert providers before exhaustion, and submit add-on requests when procedures change. Medicare Advantage and commercial plans require PA far more than Original Medicare.
Example: MRI lumbar approved for 1 study, 30 days. Patient reschedules to day 40 โ PA expired โ new request needed. Unit tracking catches this before the visit.
Referral
A PCP's documented authorization for a patient to see a specialist (HMO gatekeeping).
How it works: Referrals specify the specialist, the visit count and often a validity period. The specialist's claim needs the referral number on file; missing referrals deny. Referrals differ from PAs: referrals are internal-network routing (HMOs), PAs are payer medical-management approvals for specific services. Some HMOs have gone referral-free for specialists โ read the plan's rules, they change by product line.
Example: HMO referral card: 'Dr. Chen, 2 visits, valid 90 days.' Patient uses both visits; a third visit needs a new referral or it denies.
Carelon / Evolent / iO / AIM (PA Vendors)
Third-party companies that run prior authorization for payers.
How it works: Many carriers outsource PA decisions to specialty vendors: Carelon (Elevance/Anthem's arm), Evolent, iO (Cigna's imaging arm, formerly AIM/AmeriHealth), Humana Pharmacy Solutions for drugs. The provider portal may be DIFFERENT from the claims portal โ PA requests go to the vendor's portal (e.g., Carelon's provider portal for Anthem imaging). Billing teams learn each payer's PA channel: which portal, which payer ID, what turnaround. This vendor layer is why 'Aetna PA' might actually mean submitting to an entirely separate system.
Example: Anthem MRI PA: submit through Carelon's portal, not Availity claims. Mixing the channels causes 'no PA on file' denials even when you 'submitted' โ to the wrong system.
Coding (11 terms)
CPT (Current Procedural Terminology)
The AMA-maintained code set describing procedures and services (5-digit numeric codes).
How it works: CPT codes tell the payer WHAT was done. They're organized by category: E/M (99202โ99215), surgery (10004โ69990 by body system), radiology (70010โ79999), lab (80047โ89398), medicine (90281โ99607). Each code has official guidelines, bundling rules (NCCI edits), and payer-specific policies. Correct CPT selection comes from documentation โ coding what wasn't documented is upcoding (fraud); coding less than documented is leaving money. CPT updates annually (new, revised, deleted codes) โ billing teams watch the yearly changes.
Example: New patient office visit with moderate complexity = 99213 (established would be the same code โ 'new' vs 'established' patient rules pick 99202โ99205 vs 99212โ99215).
ICD-10-CM
The diagnosis code set describing WHY care was needed (up to 7 characters).
How it works: ICD-10 codes establish medical necessity โ every CPT line must link to a supporting diagnosis via pointers. Codes have specificity requirements: E11.9 (type 2 diabetes without complications) vs E11.65 (with hyperglycemia) change both necessity logic and risk scoring. Unspecified codes (M54.50 low back pain) invite denials when specificity exists in the documentation. ICD-10 updates twice yearly (October major, April minor). Payers map ICD-10s to coverage policies โ a diagnosis outside policy indications denies as not medically necessary.
Example: Knee injection (20610) with ICD M25.561 (right knee pain) โ specific and supported. Coding M25.569 (unspecified knee) when the note says right knee is a missed specificity win.
Diagnosis Pointer
The link (1โ4) on each claim line connecting the CPT to its supporting ICD-10 code.
How it works: CMS-1500 Box 24E (and 837 loop 2300/2400) carry pointer references. Each service line must point to a diagnosis that supports it โ wrong pointers break medical-necessity logic (a knee injection pointing to a back-pain diagnosis is an audit flag). Multiple pointers (1,2) mean the line is supported by two diagnoses together. Pointers are the difference between a clean claim and a CO-50 denial.
Example: Line 1: 99213 โ pointer 1 (M54.50 back pain). Line 2: 20610 โ pointer 2 (M25.561 right knee pain). Each procedure ties to its own diagnosis.
HCPCS Level II
Alphanumeric codes (letters + 4 digits) for supplies, drugs, DME, ambulance and services CPT doesn't cover.
How it works: HCPCS codes cover what CPT can't: J-codes (injectable drugs โ high dollar, require NDC often), E-codes (DME equipment), A-codes (supplies), ambulance (A0425+), G-codes (Medicare-specific services). Drug billing pairs J-codes with units (J3420 vitamin B12 = up to 1000mcg per unit โ units must match the amount given). HCPCS pricing runs through fee schedules and, for drugs, often ASP-based rates.
Example: 10mg of a drug whose J-code covers 1mg per unit = 10 units billed. JW/JZ modifiers report discarded/wasted amounts on separately identifiable units.
Modifier
A two-character code appended to CPT/HCPCS adding context about how a service was performed.
How it works: Modifiers answer payer questions: Was this a separate service (25, 59, XS)? Left or right side (RT/LT)? Professional or technical component only (26/TC)? Repeat procedure (76/77)? Return to OR during global (58/78/79)? Telehealth (95)? Circumstances of non-coverage (GA/GY/GZ)? Correct modifiers prevent bundling denials and support appeals; wrong or missing modifiers are among the most common denial causes. Modifier misuse (59 without distinct documentation) is a compliance risk.
Example: E/M (99213) + lesion excision (11402) same day: modifier 25 on the E/M proves the visit was separately significant and documented.
NCCI Edits
CMS's bundling rules defining code pairs that shouldn't be billed together.
How it works: The National Correct Coding Initiative publishes quarterly edit files: pairs where one code is 'comprehensive' (includes) the other. When both are billed, the payer denies the component. NCCI allows modifiers 59/XS/XP/XS-family to unbundle ONLY when documentation shows the services were truly distinct (different site, different session, different target). Billing teams run claim scrubbers against NCCI edits before submission to avoid CO-97 denials.
Example: 99213 with 20610 same day: not an NCCI pair โ both pay with modifier 25 on the E/M. But 20610 bilaterally billed twice without RT/LT denies.
E/M (Evaluation & Management)
The office-visit and cognitive-services code family (99202โ99215, 99281+).
How it works: E/M selection (since 2021) uses either Medical Decision Making (MDM: problems addressed, data reviewed, risk) or total time โ no more history/exam scoring for office visits. Levels: 99212โ99215 established, 99202โ99205 new. Telehealth E/M uses the same codes with modifier 95. E/M is the highest-volume billing family in most practices, so level accuracy drives revenue: consistent undercoding loses money, consistent overcoding invites audits.
Example: Visit addressing 2 chronic conditions, reviewing labs, prescribing: MDM moderate = 99214. Time 32 minutes with >50% counseling could also justify it โ either method works, document the one used.
Global Period (Surgical Package)
The bundled follow-up window after surgery where routine care pays nothing separately.
How it works: Surgery codes carry global periods (usually 0, 10, or 90 days). During the global, related E/M visits, suture removal and routine care are bundled โ billing them denies (CO-97 family). Exceptions that DO pay: modifier 24 (unrelated E/M), 58 (staged procedures), 78 (return to OR), 79 (unrelated procedure). Know the global length of every surgery code you bill โ the GPOP (Global Periods files) list them.
Example: Cataract surgery (66984, 10-day global): post-op day 5 visit is bundled. A same-day unrelated ankle sprain visit pays WITH modifier 24.
POS (Place of Service)
The two-digit code stating WHERE care was delivered.
How it works: POS affects payment rates and telehealth rules: 11 (office), 02/10 (telehealth โ 10 is patient's home), 22 (outpatient hospital), 23 (ER), 21 (inpatient), 12 (home), 81 (SALP โ independent lab). Wrong POS causes denials or wrong-rate payments โ e.g., telehealth billed as POS 11 denies on payers enforcing telehealth policy. Facility vs non-facial POS also splits technical/professional payment logic in hospital settings.
Example: Telehealth visit, patient at home: POS 10 with modifier 95 (most commercial). Provider-in-office, patient-at-home era rules: POS 02 โ payer policy decides.
Units
The quantity multiplier on a claim line (how many times the code's defined amount was delivered).
How it works: Most CPTs bill 1 unit per service; some codes are BY definition multi-unit (time-based: anesthesia, therapy 15-minute units 97110=1 unit per 15 min; drug J-codes per mg/ml). Wrong units cause denials or overpayment recoupments. Unit discipline matters most in therapy (frequency caps), drugs (dose-to-unit math) and anesthesia (base + time units).
Example: Physical therapy 97110 for 30 minutes = 2 units (15-minute rule). Billing 1 unit underpays; 3 units without documentation overpays โ recoupment risk.
CDT (Code on Dental Procedures)
The ADA's code set for dental procedures (D0000โD9999) โ the dental billing code system.
How it works: CDT is dental's CPT equivalent: diagnostic (D0100โD0999), preventive (D1000โD1999), restorative (D2000โD2999), prosthetics, oral surgery, ortho (D8000s). Dental claims carry tooth numbers, surfaces, quadrants โ payers enforce frequency limits per tooth and missing-tooth clauses. CDT updates annually. Dental billing is its own discipline: annual maximums, predeterminations and downcoding defenses.
Example: Amalgam restoration on tooth #19, occlusal surface: D2160 with tooth 19 and surface O โ tooth/surface detail is what dental payers adjudicate against.
Claims & EDI (8 terms)
CMS-1500 Form
The standard paper claim form for professional/physician services (the electronic version is 837P).
How it works: CMS-1500 boxes map to data elements: Box 1-13 patient/insurance data, 17 referring provider, 21 diagnosis codes, 24A-D service lines (DOS, POS, CPT, modifiers, diagnosis pointers, charges), 24J rendering NPI, 26 patient account, 33 billing provider + NPI. Even in the electronic era, billers read claims as CMS-1500 anatomy โ rejections reference the boxes ('Box 11 missing'). Institutional claims use UB-04 (837I) instead.
Example: Rejection: 'Invalid Box 24J' โ the rendering provider NPI is missing or wrong. Fix it, resubmit.
837 (P/I/D)
The X12 electronic claim file: 837P professional, 837I institutional, 837D dental.
How it works: The 837 is a structured EDI file built from your claim data: loops for billing provider, subscriber, patient and service lines. PM systems generate 837s automatically; clearinghouses validate and transport them. 837P = physician/professional (CMS-1500 equivalent), 837I = facility/UB-04, 837D = dental (J400 form equivalent). Payers read the same data either way โ the 837 is what actually travels.
Example: Your eCW claim screen shows friendly fields; under the hood it's loop 2000B subscriber data and loop 2300 claim info traveling as 837P.
999 / 997 (Acknowledgments)
Transaction-level receipts confirming a submission batch was received and syntactically valid.
How it works: After submission, the clearinghouse/payer sends a 999 (X12) acknowledging the batch. If the file has structural errors (bad segments, invalid codes), the 999 rejects it โ the whole batch or specific transactions bounce back for fixing. 999 errors happen before payer adjudication; they're pure data-format problems. Billing teams watch acknowledgment reports daily during submission windows.
Example: Batch of 400 claims sent; 999 rejects 3 with 'invalid diagnosis pointer segment' โ those 3 claims never reached the payer; fix and retransmit.
277CA (Claim Acknowledgment)
The payer's acknowledgment confirming the claim PASSED front-end edits and entered adjudication.
How it works: After the 999, the payer issues a 277CA per claim: accepted (in adjudication) or rejected (with reason). 277CA rejections are the payer's front-door edits โ member ID mismatches, invalid payer IDs, missing required fields. This is the checkpoint where 'rejected' claims live in your PM system. Rejections aren't denials โ the claim was never adjudicated; fix data and resubmit (no appeal needed, no timely-filing loss if handled fast).
Example: 277CA: 'Claim rejected โ subscriber ID not found.' Fix the member ID from the card, resubmit, next 277CA accepts it.
Clearinghouse
The intermediary that validates, scrubs and routes claims between providers and payers.
How it works: Clearinghouses (Availity, Waystar, TriZetto, Change Healthcare/Optum, Experian) do three jobs: front-end scrubbing (catching errors BEFORE payers see them), routing (translating and delivering to hundreds of payer endpoints), and tracking (status, acknowledgments, rejections in one dashboard). Many PM systems connect through one clearinghouse; large operations use aggregators for multi-payer eligibility too. Clearinghouse rejections are the cheapest fixes in billing โ errors caught before payer adjudication.
Example: Claim sent to clearinghouse โ scrubber flags 'member ID format invalid for this payer' โ fixed same-day โ routed โ payer accepted. The rejection never touched the payer.
Claim Scrubber
Software that checks claims against payer-specific rules before submission.
How it works: Scrubbers validate: code validity, NCCI pairs, payer-required modifiers, member ID formats, POS/modifier combos, medical-necessity mappings. Configurable rule sets per payer catch errors a generic edit would miss โ the best scrubbers are payer-policy-aware. Scrubber catches reduce rejection rates and speed cash; billing teams maintain scrubber rules as payer policies change.
Example: Scrubber rule: 'Payer X requires modifier 25 on E/M when same-day procedure exists.' The claim auto-flags before submission instead of denying later.
X12 / EDI
The standardized electronic data interchange format all healthcare transactions use.
How it works: X12 is the ANSI standard behind every HIPAA transaction: 837 (claims), 835 (remits), 270/271 (eligibility), 276/277 (status), 278 (auth), 834 (enrollment), 999/997 (acknowledgments). Knowing X12 structure (ISA/GS/ST envelopes, segments, loops) matters for EDI specialists troubleshooting raw files โ most billers interact only through PM screens, but the underlying format is universal.
Example: EDI specialist opens an 837 raw file: ISA header, GS group, ST transaction, then CLM (claim) and SVC (service line) segments โ every claim is this structure.
Batch vs Real-Time
The two submission modes: batched files sent periodically vs real-time single transactions.
How it works: Claims mostly travel in batches (PM systems drop claims into clearinghouse batches on a schedule); eligibility (270/271) and claim status (276/277) run real-time. Some payers accept real-time claim submission. Understanding the rhythm explains timing: a claim entered at 3 PM may sit in the next 5 PM batch, then clear the 999/277CA cycle over 24โ48 hours before adjudication starts.
Example: Claim entered Friday 4 PM โ 5 PM batch โ clearinghouse 999 Saturday โ payer 277CA Sunday โ adjudication Monday. Timelines are predictable once you know the pipeline.
Payments & Adjustments (8 terms)
EOB (Explanation of Benefits)
The payer's per-claim report: billed, allowed, paid, adjustments, patient responsibility.
How it works: Every adjudicated claim produces an EOB line-by-line: what was billed, the allowed amount, what the payer paid, contractual and other adjustments (with CARC codes), and what remains for the patient. Providers get the same data via ERA (835); patients get patient-friendly EOBs. Posting reads EOBs and translates them into payments and adjustments in the PM system โ the accuracy of that translation determines whether A/R and patient statements are correct.
Example: EOB line: billed $300, allowed $220, paid $176, CO-45 $80 adjustment, patient coinsurance $44. Posting enters exactly those numbers per line.
ERA (835)
The electronic remittance advice โ the EOB's machine-readable version for auto-posting.
How it works: The 835 travels with the payment (EFT usually) and contains claim-level and service-line-level detail: paid amounts, adjustment groups (CO/PR/OA/PI) and CARC codes. PM systems import ERAs and auto-post clean lines; exceptions (variances, unexpected denials) route to manual queues. ERA discipline: review batch summaries, work exceptions before closing, reconcile posting totals against the bank deposit.
Example: ERA batch: 150 claims, 12 exception lines flagged (paid below contract). Auto-post the 138 clean lines; the 12 variances go to the underpayment queue.
Adjustment
Any reduction from the billed amount โ contractual write-offs, patient responsibility transfers, denials.
How it works: Adjustments carry group codes: CO (contractual obligation โ provider's loss, can't bill patient), PR (patient responsibility โ deductible/coinsurance/copay, bill the patient), OA (other adjustment), PI (payer-initiated). The math always reconciles: billed = paid + adjustments. Posting errors (coding a PR as CO) misstate both sides โ patient gets under-billed or wrongly billed, A/R distorts. Group-code discipline is posting's core skill.
Example: Billed $300, paid $176: CO-45 $80 (contractual), PR-2 $44 (coinsurance). $176 + $80 + $44 = $300 โ. Wrong group code would corrupt the patient statement.
CARC (Claim Adjustment Reason Code)
Standardized numeric codes explaining why amounts were adjusted or denied.
How it works: CARCs are the industry's denial vocabulary: CO-45 (contractual), CO-16 (missing info), CO-18 (duplicate), CO-22 (COB), CO-27 (coverage terminated), CO-29 (timely filing), CO-50 (not medically necessary), CO-97 (bundled), CO-197 (no authorization), PR-1/2/3 (deductible/coinsurance/copay). Every EOB/ERA line carries them; denial management reads CARC first, then chooses the fix. Fluency in the top 15โ20 CARCs is the fastest credibility signal for any billing hire.
Example: Denial shows CO-22: read as 'other payer primary' โ verify COB, get the primary EOB, refile to primary first or correct the order.
RARC (Remittance Advice Remark Code)
Supplemental remark codes adding context to CARC explanations.
How it works: RARCs explain more: N130 ('consult plan benefit documents'), M80 ('not covered when performed during the same session'), MA01 ('if you think this is wrong, appeal'). CARCs say WHY the money moved; RARCs say what to do about it. Denial triage reads both โ a CO-16 with RARC M119 (missing/invalid member ID) tells you exactly what to fix.
Example: CO-16 with N115 ('dependency code invalid') โ the fix is the subscriber/patient relationship field, not the codes.
EFT (Electronic Funds Transfer)
Direct bank payment from payers โ how payments actually arrive now.
How it works: Payers send remittances (835) and money (EFT) together; enrollment in each payer's EFT program is part of credentialing's downstream setup. Posting reconciles EFT deposits against ERA totals โ a mismatch means combined checks or misapplied funds to trace. EFT enrollment errors leave payments mailed as paper checks (slower, losable).
Example: ERA says $12,400 batch total; bank shows $12,400 deposit โ. If the bank shows $11,900, something paid separately or short โ trace before closing.
Underpayment / Variance
Payment below the contracted allowed amount โ recoverable money.
How it works: When the payer pays under the fee schedule, the difference is a variance, NOT a contractual write-off. Detection happens at posting (comparing paid vs contract) or via automated variance tools; recovery runs through reconsideration requests, underpayment queues or escalation for systematic patterns. Variance programs are among the highest-ROI processes in billing โ small shortfalls multiply across payer volume.
Example: Contract $176; payer paid $160 โ $16 variance ร 800 claims/year on this payer = $12,800 recoverable with one escalation.
Takeback / Recoupment
The payer demanding back money it already paid (overpayment recovery).
How it works: Payers recoup for audit findings, coding corrections, or their own errors โ often netting the takeback against future remittances (watch for it on ERAs). Providers can appeal recoupments within deadlines. Compliance rule: identified overpayments must be reported/returned within payer-specific windows (60 days under some rules) โ sitting on known overpayments creates liability. Post takebacks with their original claim reference, never silently netted.
Example: ERA shows -$450 against a claim from 4 months ago with a takeback code โ post it on the ORIGINAL claim's account with the reason, and evaluate appeal windows.
Denials, Appeals & A/R (7 terms)
Denial
The payer's refusal to pay all or part of an adjudicated claim.
How it works: Denials are adjudicated decisions (unlike rejections, which never entered adjudication). Categories: eligibility/registration, coding (bundling, necessity), authorization, COB, timely filing, medical policy. Every denial carries CARC codes and often appeal rights with deadlines. Management has two halves: recovery (fix/appeal this claim) and prevention (root-cause the pattern upstream). Hard denials (necessity, exclusions) need appeals; soft denials (missing info, COB) need corrections and refiling.
Example: CO-197 (no auth) is hard-ish: retro window or appeal. CO-22 (COB) is soft: correct the order and refile. Classify first, act second.
Rejection
A claim error caught BEFORE payer adjudication โ by the clearinghouse or payer front-end edits.
How it works: Rejections mean the claim never processed: fix the data and resubmit (no appeal path needed, and timely filing usually tolerates documented resubmissions). Sources: 999/277CA messages list exact reasons โ invalid member ID, wrong payer ID, missing fields, format errors. Rejection rates above ~5% signal front-end data problems; the cheapest revenue improvement in billing is fixing registration/eligibility data so claims never reject.
Example: 277CA rejection 'invalid subscriber ID' โ fix ID from card image โ resubmit same day โ accepted. Total cost: 10 minutes.
Appeal
The formal process of asking a payer to reconsider a denied claim.
How it works: Appeal levels: reconsideration (informal, often portal/phone), first-level formal appeal (written, evidence-based), second-level, external/independent review. Deadlines vary by payer (often 30โ180 days) โ deadline tracking is the appeal team's backbone. Winning appeals cite the payer's OWN policy language with targeted evidence; appeals win on precision, not volume. Overturn rates measure appeal quality; not every denial deserves one (expected-value discipline).
Example: CO-50 denial: appeal cites the payer's medical policy section that supports the procedure with the attached clinical notes โ two focused pages beat a document dump.
Timely Filing
The payer's deadline for receiving claims โ commonly 90โ180 days from DOS.
How it works: Miss timely filing and the claim denies CO-29, usually irrecoverably. Deadlines vary: Medicare 12 months, commercial often 90โ180 days, Medicaid varies by state. Appeals on timely-filing denials need PROOF of on-time submission: clearinghouse acknowledgment reports, acceptance references. Billing teams monitor filing-deadline dashboards for at-risk claims and escalate stuck ones before the clock expires.
Example: Claim 85 days old with a 90-day limit and still rejected: fix within days and document the original submission evidence โ after day 90 it's a write-off.
Aging Buckets
A/R grouped by age: current, 31โ60, 61โ90, 91โ120, 120+ days.
How it works: Aging buckets drive working strategy: current claims get status checks on cadence; 30โ60 gets first escalation; 60+ gets priority (deadlines approaching); 120+ is triage (appealable, refileable, or write-off candidates). Bucket movement (how much money flows down vs resolves) tells leadership whether follow-up is actually working. Bucket analysis by PAYER is even more powerful โ slow payers become visible and manageable.
Example: One payer holds 40% of the 120+ bucket: a payer-specific escalation case, not a team-performance problem. Data separates the two.
First-Pass Resolution
The percentage of claims resolved on the first follow-up attempt.
How it works: High first-pass rates mean follow-up is thorough (complete call prep, full account knowledge, right channel). Low rates mean ping-ponging โ claims touched repeatedly without resolution. Measured per caller and per payer, first-pass resolution separates productive AR work from queue-filling. The habit that drives it: prepare fully before every call and capture reference numbers so the next touch starts ahead.
Example: Caller A: 60 calls, 35 resolved first-pass. Caller B: 90 calls, 20 resolved. B is busier; A is better. Coaching targets the difference.
Root-Cause Analysis
Tracing denial patterns back to the process stage that created them.
How it works: Group denials by CARC, attribute each cluster to its failing stage: CO-26/27 โ registration/eligibility; CO-197 โ auth process; CO-97 โ coding; CO-29 โ billing timeliness. Then take data to the owning team with example claims and a fix request, and re-measure in 60 days. Root-cause loops are the highest-ROI process in denial management โ recovery without prevention is a treadmill.
Example: CO-197 denials cluster on one procedure family booked without auth checks โ scheduling adds an auth gate โ denial category drops 70% next quarter.
Compliance & Regulation (7 terms)
HIPAA
The federal law governing protected health information (PHI) privacy and security.
How it works: Billing lives inside HIPAA: only minimum-necessary PHI, secure systems and transmission, identity verification before discussing accounts, no PHI to third parties without authorization, and audit trails. Practical biller habits: verify callers before discussing balances, minimal PHI in voicemails, secure file handling, and never discuss cases in shared spaces. Violations cost organizations heavily โ discipline is a core professional skill, not a formality.
Example: Patient's spouse calls about a balance: verify authorization on file before discussing anything โ or politely decline and take the patient's consent.
PHI (Protected Health Information)
Any individually identifiable health information โ names, dates, IDs, diagnoses, payments.
How it works: PHI covers the whole billing surface: claims, EOBs, statements, call notes, even appointment dates. Handling rules: encrypted systems, access on need-to-know, no PHI in emails to personal addresses, and careful document disposal. In billing workflows, PHI discipline shows up in small places โ masking IDs on shared screens, redacting on faxes, and limiting what's said aloud.
Example: Statement printed with full diagnosis + account number left on a printer tray = a reportable HIPAA incident. Workflow fixes (secure print release) prevent it.
42 CFR Part 2
The stricter federal confidentiality rules for substance use disorder (SUD) treatment records.
How it works: Part 2 records (from OTPs, SUD programs) require special written consent for disclosure โ stronger than HIPAA. In billing: SUD claims get careful handling, disclosure only per consent, and claims shouldn't reveal unnecessary SUD detail. Part 2 violations are severe; teams handling behavioral health build explicit procedures and train on the boundary.
Example: Methadone clinic billing: the claim and any follow-up disclosure follow Part 2 consent rules โ standard HIPAA assumptions don't apply.
No Surprises Act
Federal protection (2022) against surprise out-of-network billing in emergencies and certain facility situations.
How it works: NSA rules: emergency care gets in-network cost-sharing regardless of provider network status; balance billing is prohibited in specified scenarios; providers must give good-faith estimates for self-pay/uninsured patients; OON claims to payers follow special negotiation/arbitration (IDR) paths. Billing teams implement: estimate workflows, NSA-compliant notice-and-consent documents (where allowed), and IDR submission processes for OON payment disputes.
Example: OON anesthesiologist at an in-network surgery: patient can't be balance-billed โ the provider pursues the payer through the NSA IDR process instead.
False Claims Act
Federal law imposing penalties for knowingly submitting false claims to government payers.
How it works: Billing fraud territory: upcoding, billing services not rendered, falsifying documentation, kickbacks for referrals, waiving copays routinely to induce business. Penalties are per-claim and severe, plus qui tam whistleblower exposure. The professional line: bill what documentation supports, never alter records, report known overpayments, and refuse to code around policy. 'The true record, always' is the compliance stance that survives audits.
Example: A supervisor asks to add modifiers to unbundle without documentation support: that's a False Claims Act risk โ refuse and escalate through compliance channels.
Anti-Kickback Statute
Federal law prohibiting remuneration for referrals of federal healthcare business.
How it works: Anything of value offered to induce referrals (cash, free services, routine copay waivers) violates AKS for government-program business. Billing touches it through: write-off policies (waiving patient cost-share without financial-hardship assessment), referral arrangements, and marketing deals. Safe harbors define permitted structures; the operational habit is documenting hardship screening before any patient-balance forgiveness.
Example: Front desk waives $40 copays for a referring physician's patients 'as a favor' โ that's a kickback pattern, not customer service.
Audit (Payer/Internal)
Systematic review of claims and documentation to verify billing accuracy.
How it works: Audits come from payers (TPE, CERT, UPIC for Medicare), internal compliance teams, and external consultants. They sample claims and request full documentation: notes, orders, auths, delivery proof. Response discipline: organized files, internal review before submission, deadline compliance, and post-audit fixes. Audit findings feed corrective actions and training โ an audit-ready documentation culture is cheaper than any single audit response.
Example: TPE notice: 20 claim samples requested. Team assembles complete packets (orders, notes, delivery proof), reviews internally first, responds by deadline, tracks outcome.
Systems & Industry (8 terms)
PM System (Practice Management)
The billing software running scheduling, registration, charge entry, claims, posting and A/R.
How it works: PM systems are the biller's workspace: patient demographics, insurance, appointments, charges, claims, payments and reports all live here. eCW dominates independent practices and billing companies; AdvancedMD, Kareo/Tebra, athenahealth, NextGen are other majors; Epic Resolute runs hospital billing. The skill that transfers: workflow concepts (eligibility โ charge โ submit โ post โ work denials) are identical everywhere โ learn the logic, and any PM system takes 2โ3 days.
Example: Biller moves from eCW to AdvancedMD: same eligibility concept, different button names. Productive by day three.
Payer Portal
Each payer's provider website for eligibility, claim status, PA submissions and authorizations.
How it works: There's no universal portal: UHC (One Healthcare ID / UHCprovider), Aetna/BCBS/Humana via Availity, CignaforHCP, Novitasphere (Medicare), state Medicaid portals โ each with its own login and quirks. Billers work multiple portals daily; large operations use aggregators (Availity, Waystar) to pull multi-payer eligibility into one screen. Portal fluency โ knowing which portal answers which question fastest โ is a real productivity skill.
Example: Morning queue: Availity for Aetna/BCBS eligibility batch, UHC portal for two auth checks, Novitasphere for Medicare status. Three logins, one routine.
MAC (Medicare Administrative Contractor)
The company that processes Medicare claims for a geographic jurisdiction.
How it works: Part B claims route to the patient's state's MAC: Novitas, Noridian, CGS, Palmetto, WPS, NGS, FCSO, etc. Each MAC has its own portal (Novitasphere, NGSMedicare), LCD coverage policies, and bulletins. Billing teams check which MAC covers their state and read its LCDs (Local Coverage Determinations) โ LCDs define what Medicare covers for specific services and directly drive CO-50 appeal arguments.
Example: Texas practice: Part B goes to Novitas (JH/JL). Its LCD on knee arthroscopy defines medical-necessity criteria โ the appeal cites the LCD language.
LCD (Local Coverage Determination)
A Medicare MAC's written policy defining coverage for specific services.
How it works: LCDs list covered diagnoses, documentation requirements and frequency limits for procedures โ they're the authority for Medicare medical-necessity decisions in the jurisdiction. Claims with ICD-10s outside an LCD's covered list deny CO-50; appeals that quote the LCD's own criteria (and attach matching documentation) overturn well. LCD lookups belong in the coding/verification workflow for any Medicare-heavy practice.
Example: LCD says knee injection covers only specific arthritis diagnoses: the claim carries M17.11 (covered) instead of an unspecified pain code โ denial avoided at coding.
Payer ID
The numeric identifier routing electronic claims to the right payer endpoint.
How it works: Every payer has an EDI payer ID (e.g., UHC commercial 87726, Medicare varies by MAC); PM systems and clearinghouses use them for routing. Payer ID errors are a top rejection cause โ the claim goes nowhere or to the wrong plan. Confusingly, card brands โ payer IDs: UMR cards route through UHC's ID; state BCBS plans each differ. Always verify the payer ID on the back of the card or payer enrollment lists.
Example: BCBS Texas claims use a different payer ID than Florida Blue โ same brand family, different endpoints. Wrong ID = rejection.
DSO (Dental Service Organization)
A corporate group running multiple dental practices with centralized business functions.
How it works: DSOs centralize billing, credentialing and operations across offices โ which makes dental billing a scale career: standardized claim protocols, per-location dashboards, payer-volume leverage. DSO growth is why dental billing demand keeps rising; the work differs from single-practice billing in volume and standardization, not in CDT fundamentals.
Example: A 60-office DSO's billing team runs payer-pod structure: one pod owns Delta Dental volume across all offices โ specialization creates speed.
BPO (Business Process Outsourcing)
Companies that run billing operations for US providers from offshore locations.
How it works: US medical billing BPOs (India, Philippines, Pakistan, Bangladesh) run AR calling, charge entry, posting, denial management and credentialing for US clients โ often following US business hours. Career structure is strong: caller โ senior โ QA โ team lead โ operations manager. Client work demands accuracy metrics, HIPAA discipline, and US-payer knowledge; the skills are identical to onshore billing with volume scale.
Example: A Lahore billing BPO runs 40 callers working US payer queues nightly; the best caller becomes a QA lead within two years.
KPI Dashboard
The reporting layer tracking billing health: clean-claim rate, days in A/R, denial rate, NCR, collections.
How it works: Good dashboards are few metrics, dollar-denominated, trend-visible, and action-linked โ every red number has an owner and a fix in motion. Bad dashboards drown in activity counts (claims touched, calls made) that invite gaming. Billing managers build weekly one-pagers: the four cash numbers (NCR, A/R days, denial rate with dollars, cost-to-collect) plus exceptions. Reporting in dollars is what leadership funds.
Example: Weekly page: NCR 94% (target 96%) โ variance program escalated on payer X; A/R 42 days stable; denial rate 7.2% with CO-197 cluster going to scheduling.