Imaging groups, IDTFs, interventional practices and teleradiology rosters lose most of their money in the order, the authorization and the charge, long before a claim is ever built.
Get a free revenue cycle assessmentRadiology revenue cycle management is the full financial process behind an imaging study, from the referring order and benefits check through authorization, charge capture, coding, claim submission, payment posting, denials and AR follow up. Medical billing is one stage inside it, starting at the claim; the revenue cycle starts before the patient is scanned.
Radiology is the only large specialty whose revenue cycle begins in somebody else's office. The referring physician writes the order, supplies the diagnosis that establishes medical necessity, and signs the documentation a payer will later ask for. The imaging practice inherits all of it and controls none of it.
A study can be performed, read and reported correctly and still be unbillable because the order carried no signed indication. 42 CFR 410.32(a) requires a diagnostic test to be ordered by the treating physician who uses the result to manage the patient, and 42 CFR 410.33(d) bars an IDTF from billing a test its own supervising physician ordered.
The work then splits across entities. A hospital-based group bills the interpretation while the hospital bills the equipment, so the group's receivable holds half the economics of every study and none of the registration data behind it. A freestanding center bills both halves. An IDTF bills the technical side under its own enrollment while an outside radiologist bills the read. One CPT code lands in three different receivables depending on who owns the scanner.
Finally, the protocol can change after the approval issues, so the approval on file stops matching the study performed. Radiology also carries the heaviest prior authorization load of any diagnostic specialty, with most plans delegating advanced imaging review to a radiology benefit manager.
Recognise three or more of these in your own numbers and the problem is the process, not the payer.
Get a free assessmentRadiology is organised by whose receivable the money lands in, not by practice size.
People Also Ask wants the seven steps of the revenue cycle. Here they are, with the failure mode radiology hits.
The order either carries a signed indication and a supporting diagnosis, or it does not. Failure mode: the study is scheduled before anyone reads the order, so the defect surfaces after the scan.
Coverage, deductible and site-of-service rules are confirmed before the appointment. Failure mode: eligibility is run against the plan named on the referral, not the plan in force on the date of service.
Advanced imaging is approved by the plan or its radiology benefit manager for a specific code, body part and date range. Failure mode: the approval covers the study requested, not the study performed.
The technologist runs the protocol, which is where contrast, sequences and extra body parts get added. Failure mode: the change never reaches the authorization record or the charge.
The charge is built from the signed report and the modality worklist. Failure mode: it runs off the dictation system alone, so anything read outside the normal worklist never generates one.
The claim is scrubbed against payer rules and the NCCI procedure-to-procedure and medically unlikely edits CMS republishes quarterly. Failure mode: scrubber rules are never retired, so dead edits hold clean claims.
Remittances post, underpayments are identified against the contract, denials are worked by root cause. Failure mode: denials are resubmitted rather than corrected. 19% of denied claims were never reworked or appealed (Luxen billing reviews).
Leakage is easier to find organised by cause than by code. Figures are modelled on the reference practice: a six-radiologist group reading 52,000 studies a year on $3.6 million of collections. These seven leak points come to roughly $283,400 a year.
| Leak point | Codes or rule | What goes wrong | Annual dollars at risk | Luxen audit finding |
|---|---|---|---|---|
| Order defects | Rule 42 CFR 410.32(a); CARC 16 | Study performed against an order with no signed indication | $64,000 | 31% of accounts receivable sat in the referring-order category (Luxen billing reviews) |
| Duplicate interpretation | Manual Ch. 13 s.100.1; modifier 77; CARC 18 | Emergency physician and radiologist both bill a read | $18,600 | Both groups billed an interpretation on 5% of emergency department imaging claims (Luxen claim audit) |
| Uncharged post-processing | CPT 76377; NCCI Policy Manual Ch. IX | 3D rendering documented in the report, never charged | $21,800 | Never charged on 12% of CT angiography and volumetric studies reviewed (Luxen claim audit) |
| Enrollment gaps | Payer participation; CARC 16 | A new or cross-state reader bills before the roster reflects them | $37,000 | 9% of claims for radiologists hired in the prior year denied for enrollment reasons (Luxen billing reviews) |
| Authorization drift | CARC 197 | Approval expires or names a different code, and no peer-to-peer is requested in time | $62,700 | 22% required a peer-to-peer review (Luxen client data) |
| Supply capture | HCPCS 78815, A9500, Q9967 | Contrast and radiopharmaceutical logged at the modality, never on the claim | $14,400 | Supply lines were missing from 8% of nuclear medicine and contrast-enhanced claims reviewed (Luxen claim audit) |
| Studies never billed | Charge reconciliation | Order completed, report signed, no charge created | $64,900 | 2.4% of volume across 34 radiology practices (Luxen billing reviews) |
We will tell you which of these leaks is open in your practice, free, in 30 minutes.
Book the reviewEvery page on this search names metrics and quantifies none. The Typical column here is external and never Luxen data: it names the federal dataset covering the metric, or says plainly that none publishes one. The Target column is Luxen client data, 38 client practices, January 2024 to June 2026.
| Metric | Definition | Typical (external) | Target (Luxen client data) |
|---|---|---|---|
| Days in AR | Receivables divided by average daily charges | No free federal dataset publishes an AR benchmark | 33 or fewer within 120 days, from a median of 54 |
| Net collection rate | Payments divided by charges net of contractual adjustments | None; the nearest federal measure is the CERT improper payment rate for Medicare fee-for-service, 6.55% in FY2025 | 97.8% by month six, from 91.4% |
| Clean claim rate | Claims accepted on first submission | None published federally | 97.3% within 90 days, from 89.6% |
| First-pass denial rate | Claims denied on first adjudication | The CMS Marketplace Transparency in Coverage PUF for PY2026 publishes denied claim counts for PY2024, but no rate and no specialty split | 6.1% or lower within 90 days, from 14.2% |
| Cost to collect | Billing operations cost as a share of collections | None; build it from your own staffing, software and clearinghouse lines | 3% to 6% on the Luxen fee, against $250,100 a year in-house |
| Authorization denial rate on advanced imaging | CT, MRI, PET and nuclear medicine claims denied under CARC 197 | CMS PSPS for CY2025 carries denied service counts by HCPCS and specialty code, though suppression makes any computed rate a lower bound | 2% or lower |
Typical values come from the named federal source in the table intro. Target values come from Luxen client data.
Most plans do not review advanced imaging themselves. They delegate it to a radiology benefit manager, and three cover most commercial and Medicare Advantage lives: EviCore by Evernorth, Carelon Medical Benefits Management (renamed from AIM Specialty Health in March 2023) and HealthHelp. Each publishes its own clinical guidelines and applies its own criteria to CT, MRI, PET and nuclear medicine, so a practice that submits every request the same way is guessing at three rule sets.
Under rule CMS-0057-F, the interoperability and prior authorization rule at 89 FR 8758, Medicare Advantage organizations and state Medicaid and CHIP fee-for-service programs owe expedited decisions in 72 hours and standard ones in seven calendar days, with a specific reason for every denial. Medicaid and CHIP managed care phase in by rating period and Marketplace issuers by plan year, and the Prior Authorization API follows on January 1, 2027. The specific denial reason is the operational gain: it turns an unexplained denial into an appealable one.
Advanced imaging prior authorizations took a median 4 business days from request to decision, and 22% required a peer-to-peer review (Luxen client data). The peer-to-peer is the highest-yield step in imaging utilization review, because the reviewer will accept detail the portal form had no field for, and practices lose these on calendar mechanics rather than clinical merit. Three controls do most of the work: match the approved code, body part and laterality against the study performed after the protocol is final; track the approval date range against the scheduled date and re-request on any reschedule; log every peer-to-peer offer with its deadline. Our eligibility and prior authorization team runs all three before the claim goes out.
Some states force plans to exempt high-performing practices from prior authorization. Texas HB 3459 (2021) created Insurance Code Chapter 4201, Subchapter N, exempting a physician with at least 90% of eligible preauthorization requests approved, over a 12-month evaluation period set by HB 3812 effective September 1, 2025. It binds state-regulated HMO, PPO and EPO plans and excludes self-funded ERISA plans and public programs. If you read in Texas, your approval rate is a contractual asset. No such federal exemption exists.
The signature revenue mechanic in radiology is not a code. It is the chain that runs order, protocol, report, charge. Every other specialty generates its own encounter; radiology receives one, and when that chain breaks the money never appears.
42 CFR 410.32(a) requires a diagnostic test to be ordered by the physician treating the patient and used in managing that patient, and 42 CFR 410.33(d) requires an IDTF's order to be written and signed by the treating physician. An order without a clinical indication is an unbillable study. 31% of accounts receivable sat in the referring-order category, held or denied because the order lacked a signed indication or an ICD-10 diagnosis (Luxen billing reviews).
Contrast is added at the scanner, a second body part is imaged, a screening study converts on the radiologist's recall. The reconciliation point has to sit after the technologist finalises the protocol and before the claim is built, comparing three records: the order, the approval and the modality log.
Build charges from signed reports alone and you lose everything read outside the normal flow. Completed studies with a signed report and no charge were 2.4% of volume across 34 radiology practices (Luxen billing reviews), and 3D post-processing was documented in the report and never charged on 12% of CT angiography and volumetric studies reviewed (Luxen claim audit). The CMS NCCI Policy Manual, Chapter IX, sets out where 3D rendering may be reported separately, including that it is not separately reportable for nuclear medicine.
Medicare Claims Processing Manual Chapter 13, section 100.1 pays for one interpretation per procedure. Where an emergency physician's read contributed to the patient's care, a later radiologist claim for the same study is denied, and a second read is payable only in unusual circumstances with documentation. Both groups billed an interpretation on 5% of emergency department imaging claims, and the second interpretation was denied in 92% of those cases (Luxen claim audit). Component assignment, modifiers and rates are covered on our radiology billing services page; our certified coders check report against charge before submission.
Nobody on this search publishes a denial reason code. Denials are not a category, they are a short list. 63% of practice managers could not name their top three denial reasons (Luxen Practice Manager Survey 2026), while the top three denial reasons accounted for 58% of denied dollars in the average practice reviewed (Luxen billing reviews).
| Code | What the payer is saying | The radiology cause | What reverses it |
|---|---|---|---|
| CARC 197 | Precertification, authorization or notification absent | Approval expired or named a different code | Retro-authorization inside the plan window, or appeal with the protocol record |
| CARC 50 | Not deemed a medical necessity | The referring diagnosis does not support the study | Amended order or referrer documentation, appealed against the plan's imaging criteria |
| CARC 16 | Claim lacks information or has a billing error | Missing order detail, enrollment or report | Correct and resubmit, never rebill unchanged |
| CARC 4 | Procedure code inconsistent with the modifier used | Component or place-of-service mismatch on an interpretation claim | Corrected claim with place of service and modifier reconciled |
| CARC 18 | Exact duplicate claim or service | Two interpretations of one study | Appeal under Chapter 13, section 100.1 with evidence the read contributed to care |
| CARC 151 | Information does not support this frequency of services | Screening frequency limits or MUE units exceeded | Appeal with dates and rationale; some MUE values are confidential |
19% of denied claims were never reworked or appealed (Luxen billing reviews), and 42% of practice managers said nobody owns denial follow up full time (Luxen Practice Manager Survey 2026). Median appeal turnaround was 34 days from filing to payer decision (Luxen client data), so an unowned queue converts into timely filing write-offs. Old accounts still pay: we recovered 61% of the dollar value of claims aged 90 to 180 days that practices had stopped working, and 23% of dollar value on claims aged past 180 days (Luxen client data). Radiology is an ancillary service under the No Surprises Act, so what you may pursue from the patient can turn on state process, as in Illinois. Our denials and AR recovery team works the queue by root cause, oldest dollars first.
Dataset: 8,600 imaging claims from the Luxen claim audit, which covers 61,400 claims audited between January 2025 and June 2026, plus 34 radiology practice reviews from the Luxen billing reviews, 410 practice billing reviews over the same period. Counted: where money stopped moving between the referring order and the posted payment.
The first two findings appear nowhere else on this search. Every page here treats radiology revenue loss as a claims problem; the two largest categories here sit upstream of the claim.
A six-radiologist group reading for two hospitals and one freestanding imaging center, collecting about $3.6 million a year, came to Luxen with AR that had stopped moving and no denial reporting by reason code.
| Measure | At onboarding | After seven months |
|---|---|---|
| Days in AR | 58 | 34 |
| First-pass denial rate | 15.8% | 5.9% |
| AR older than 120 days | $196,000 | $41,000 |
| Studies completed with no charge | 2.9% of volume | 0.4% of volume |
$214,000 was recovered over those seven months, most of it from unappealed authorization denials and after-hours studies that never generated a charge. The group kept its RIS, its practice management system and its front desk (Luxen client data).
Company-wide results across 38 client practices (Luxen client data, January 2024 to June 2026):
Same reference practice, arithmetic shown.
The two inputs that move the answer most are the net collection rate and the unbilled study percentage, and most practices have never measured the second. Practices also lost 3.1% of collections to patient balances written off before a second statement (Luxen client data), which our patient billing team works before write-off.
Want this arithmetic run on your own collections and denial rate?
Run my numbersFull revenue cycle work for radiology runs 3% to 6% of collections. Where a practice lands is set by four things: the share of professional-only against global claims, advanced imaging volume and the authorization load it carries, whether there is interventional work with global periods and devices, and how many payer enrollments the roster needs kept current.
Included at every rate: certified coding, claim submission and edits, authorization matching, payment posting and contract variance checks, denial work and appeals, AR follow up, patient statements, and a monthly denial report by reason code, payer and modality. Month to month, 30 days notice, BAA signed before anyone touches your systems.
Costed for the reference practice at the Luxen rate of 4% of collections.
| Line item | In-house | Luxen |
|---|---|---|
| Billing and AR staff, fully loaded | 2.5 FTE at $52,000 plus 28% benefits, $166,400 | Included |
| Certified coder | 0.5 FTE at $61,000 plus 28% benefits, $39,000 | Included |
| Billing software | $13,800 | Your system, no new licence |
| Clearinghouse | $7,300 | Included |
| Coding education and CEUs | $4,200 | Included |
| Denial rework time | 11 hours a week at $34 loaded, $19,400 | Included |
| Annual total | $250,100, or 6.9% of collections | $154,100 |
| When a biller leaves | Coverage gap | Coverage continues |
Practice managers estimated 11 staff hours a week on insurance calls and portal checks, and 34% of practice managers replaced a biller in the past two years (Luxen Practice Manager Survey 2026). Keep it in-house when the study mix is narrow, the payer list is short and one biller has a trained backup. Move it when advanced imaging volume, multi-site reading or a growing roster outruns that. Compare medical billing companies on the ten criteria above.
People Also Ask wants the top five RCM companies in the USA. Wrong question: the ranking depends on what your revenue cycle leaks. Score candidates on these ten, 0, 1 or 2 each; under 14 out of 20 is a vendor you will be managing.
Score us against our full-service medical billing scope.
You already have someone doing this, so the real question is what the gap costs. Work starts with a signed BAA and read access to your RIS and practice management system, then runs about two weeks to the first claims worked. First recovered payments arrived a median of 17 days after work began (Luxen client data). We take the oldest open AR and the unappealed denial queue first, and run new claims in parallel rather than after a cutover. Nothing migrates: your systems, worklist and report templates stay as they are. We take assignment by date of service so no account is worked twice. Month to month, 30 days notice, no exit fee, and your data extract is yours on request, which is the term worth checking in whatever you signed last time.
We work inside what you already run, with no migration: Epic Radiant, Sectra, Fujifilm Synapse, Intelerad and eRAD on the imaging side, Epic, athenaOne or eClinicalWorks on the practice management side, plus the benefit manager portals and your clearinghouse.
Order intake and scheduling, verification of benefits and eligibility, prior authorization and utilization review, the study and protocol reconciliation, report finalization and charge capture, claim submission and edits, then payment posting, denials and AR follow up. Radiology differs at step one: the order comes from a referring physician, and 42 CFR 410.32(a) makes it a payment condition rather than paperwork.
There is no fixed ranking, because the right vendor depends on what your revenue cycle actually leaks. Score candidates on the ten criteria above instead, covering order intake, authorization matching, charge reconciliation, component discipline, denial reporting by reason code, appeals, enrollment, system fit, fee basis and exit terms. Treat anything under 14 out of 20 as a vendor you will be managing.
Twelve-step versions split the same cycle more finely, separating pre-registration, registration, charge capture, coding, scrubbing, submission, posting, follow up, denial management, appeals, collections and reporting. The count matters less than who owns each step, and in radiology the unowned steps are protocol and charge reconciliation. 42% of practice managers said nobody owns denial follow up full time (Luxen Practice Manager Survey 2026).
Luxen bills 3% to 6% of collections, with no setup fee and no exit fee. For a six-radiologist group collecting $3.6 million a year, 4% is about $154,100 annually against $250,100 for a fully loaded in-house team of 2.5 billing FTEs and half a certified coder. The rate moves with advanced imaging volume, the professional to global claim mix, interventional work and roster size.
Plan on about two weeks from a signed BAA to the first claims worked. First recovered payments arrived a median of 17 days after work began (Luxen client data). Denial rate moves first: first-pass denial rate fell from 14.2% to 6.1% within 90 days, and median days in AR dropped from 54 to 33 within 120 days (Luxen client data).
Billing starts at the claim. Revenue cycle management starts before the patient is scanned, with the referring order, the eligibility check and the prior authorization, and ends with posted cash rather than a submitted claim. That gap is where most radiology money is lost: 31% of accounts receivable sat in the referring-order category, held or denied because the order lacked a signed indication or an ICD-10 diagnosis (Luxen billing reviews).
A free 30 minute review of your AR ageing and denial reasons. We tell you what is recoverable and what it would take. No deck, no commitment, no fee.
Book a free revenue cycle assessment%20(1).avif)
