Gastroenterology Medical Billing and Revenue Cycle Management 

Most GI practices arrive here with one of these five.

Section 3

Why gastroenterology loses money other specialties don't

Because a large share of your procedures change partway through.

A screening colonoscopy becomes therapeutic the moment a polyp comes out, and that happens often enough to be routine rather than exceptional. A biller who codes every colonoscopy the same way is not occasionally wrong. They are wrong on a substantial fraction of your cases, every week, in a direction that costs you money and generates patient complaints.

The landmark Veterans Affairs cooperative study found polyps noted in 23.9 to 35.7 percent of colonoscopies depending on the criteria used, with an apparent polyp detection yield of at least 24 percent (Lieberman and colleagues, The American Journal of Medicine). Contemporary US adenoma detection rates run around 39 percent (National Endoscopy Database).

So conversion is not an edge case to be handled when it arises. It is a routine event that your coding process either handles correctly by default or gets wrong by default.

Four structural leaks follow from the nature of the work:

• Screening to diagnostic conversion, where the code, the modifier and the patient's bill all change
• NCCI bundling, where biopsy and polypectomy in one session collapse into one payment unless the note supports otherwise
• Anesthesia, where monitored anesthesia care for an average-risk patient is denied by policy at several major payers
• Prior authorisation, concentrated in capsule endoscopy and biologics, where the burden is measured in staff hours rather than in claims

None of these are billing errors in the ordinary sense. They are places where general billing competence is not enough.

Section 4

When a screening colonoscopy becomes diagnostic

The code changes, the modifier changes, and what the patient owes changes.

For Medicare, screening is reported with G0121 or G0105 rather than CPT 45378, and a conversion is reported with the therapeutic CPT plus modifier PT. For commercial and Medicaid, screening is 45378 with modifier 33, and a conversion is the therapeutic CPT with modifier 33. PT and 33 are not interchangeable and using the wrong one denies before anyone reads the claim.

The four cases

Why the modifiers are not interchangeable

Modifier PT is a HCPCS Level II modifier meaning a colorectal cancer screening test was converted to a diagnostic or other procedure. It is used on Medicare claims only.

Modifier 33 is a CPT modifier for preventive services tied to USPSTF grade A and B recommendations. It is used for commercial and Medicaid.

The modifier itself functions as a payer-type flag. PT on a commercial claim, or 33 on a Medicare claim, is rejected automatically. This is a common and entirely avoidable denial.

Modifier 33 is also not used on the Medicare G-codes, because screening is already in the code descriptor.

Diagnosis sequencing

Z12.11, encounter for screening for malignant neoplasm of the colon, goes in the first position. The finding goes second. Per ICD-10-CM Coding Clinic guidance, when a screening is performed the screening code is listed first even where a procedure results from it.

Getting this backwards changes the character of the claim from preventive to diagnostic in the payer's eyes, which is precisely what you are trying to avoid.

Where the code itself goes wrong

Four recurring errors, all of which come from nobody reading the operative report:

• Reporting 45378 alongside a therapeutic code. Never correct. 45378 is the base and is dropped when a therapeutic service is performed.
• Reporting the wrong removal technique. Cold forceps biopsy (45380), hot biopsy forceps (45384) and snare (45385) are different codes for different things.
• Reporting multiple units for the same technique. Three polyps removed by snare is 45385 once, not three times.
• Defaulting to the diagnostic code because the claim was built from the schedule rather than from the findings.

That last one is the expensive one, and it is invisible unless somebody audits it.

Multiple techniques at different sites

Where different techniques are used at genuinely different sites, report both, higher-RVU code first, with modifier 59 or XS on the second. Medicare pays 100 percent of the highest-RVU code plus the difference between each additional code and the base endoscopy value.

See Section 7 for why these appeals usually fail.

Free gastroenterology coding tool

Colonoscopy coding and cost outcome tool

Four questions, and it returns the code, the modifier, the diagnosis order, and what the patient will actually be charged. Runs in your browser. There is no field to enter patient information into, by design.

This tool also lives at its own address: /tools/colonoscopy-coding-tool/.

Section 6

Why your patients are getting bills you told them wouldn't come

Because what the patient owes after a screening depends on their payer, and the rules diverge.

Commercial ACA-compliant plans owe nothing when a polyp is removed during a screening. Medicare patients still owe coinsurance on the therapeutic portion, though that is being phased out. The anesthesia coinsurance is a separate matter again and is not fully absorbed by the same phase-down in every scenario.

This is a billing problem that arrives as a relationship problem. The patient was told a screening is free. A polyp came out. A bill arrived. They call your front desk, and so does the referring physician who told them the same thing.

Commercial ACA-compliant plans

No patient cost-sharing when a polyp is removed during a screening. Federal guidance in 2013 clarified that polyp removal is integral to the screening, and 2015 guidance extended that to anesthesia. Grandfathered plans are the narrow exception.

Medicare

Historically the patient owed full coinsurance on the therapeutic portion. The Consolidated Appropriations Act 2021, Section 122, phases that down:

• 15 percent for 2023 through 2026
• 10 percent for 2027 through 2029
• Zero from 1 January 2030

The phase-down applies only where modifier PT is correctly appended. A coding error does not just cost you the claim, it costs your patient money they should not have paid.

The anesthesia note

For a screening that converts, report 00811 with modifier PT. Deductible is waived. Coinsurance treatment follows the reduced schedule when the PT rules are met, but residual patient responsibility still appears in practice and is worth explaining in advance.

After a positive stool test

A follow-up colonoscopy after a positive FIT, FOBT or Cologuard result is covered as screening with no cost-sharing. Effective 31 May 2022 for ACA plans and 1 January 2023 for Medicare. For a Medicare beneficiary in this pathway, report G0105 or G0121 with the KX modifier.

This closed a genuine trap that produced documented surprise bills.

The only real fix

Tell the patient before the procedure, not after. A short, honest explanation at scheduling that a polyp may be found and what that means for their bill under their specific plan prevents almost all of these calls. The billing is downstream of a conversation nobody had.

Section 7

Why your polypectomy and biopsy claims keep getting bundled

NCCI bundles a biopsy with a polypectomy performed in the same session.

On the same lesion, only the higher-valued therapeutic code is reported and that is correct. On genuinely distinct lesions at distinct sites, the biopsy is separately reportable with modifier 59 or XS. Most appeals fail not because the modifier was wrong but because the operative note does not establish distinct sites and instruments.

The denial is usually RARC N19, bundled and not separately payable.

The modifier on its own does not win the appeal. The note has to support it, and this is where practices lose money they are entitled to.

What the note has to say

The ACG's own coding guidance makes the point with an example worth borrowing. A note describing snare polypectomy in the transverse colon and random biopsies obtained in the ascending and descending colon does not establish distinct instruments to a payer.

What does is naming the instrument explicitly: random cold forceps biopsies of the ascending and descending colon were obtained.

Same procedure, same clinical care, different payment outcome, because one version tells the payer what it needs to know and the other assumes it.

Practical steps

• Use the CMS-1500 Box 19 narrative to list anatomic sites per line item
• Prefer the X modifiers, particularly XS for separate structure, over the general 59. CMS prefers them, though many commercial payers still accept 59
• Feed this back to your endoscopists. This is a documentation habit, and once it changes the denials stop rather than being appealed one at a time

Bleeding control

Control of post-polypectomy bleeding is bundled into the therapeutic endoscopy and is not separately billable where the procedure itself caused the bleeding. All bleeding control methods use a single code, 45382.

Section 8

Why anesthesia claims are denied as not medically necessary

Several major payers restrict monitored anesthesia care for average-risk patients undergoing routine endoscopy.

They cover it only where documented risk factors exist. The denial is CARC CO-50. It is a policy position rather than a claim error, which means the fix is in the pre-anesthesia documentation rather than in the billing.

The GI anesthesia codes:

For a screening that converts, report 00811 with modifier PT.

The coverage dispute

Multiple commercial payers deem monitored anesthesia care not medically necessary for ASA Class I and II patients having routine endoscopy. Named policies include Aetna, Anthem, Humana, Highmark, and Blue Cross Blue Shield of Massachusetts Policy 154, which permits MAC only for defined risk factors including ASA class III or higher, sleep apnea, and patients under 18 or over 70. That policy has been enforced through diagnosis-driven claim edits since 1 July 2023.

What establishes necessity

The pre-anesthesia evaluation has to link the specific condition to the need for an anesthesia professional. Not simply record that the condition exists.

Diagnoses that support it include F41.1 for generalised anxiety, E66.01 for morbid obesity, sleep apnea, and documented severe comorbidities. What fails is a pre-anesthesia note that lists conditions without connecting them to the anesthesia decision.

Why this matters more than the claim value

When MAC is denied, the patient frequently gets billed. That lands back on your practice as a complaint about a service delivered by a group they may not distinguish from you. The documentation fix protects the patient relationship as much as the claim.

Section 9

Capsule endoscopy and biologics

Prior authorisation is the largest non-coding burden in gastroenterology, concentrated in capsule endoscopy and biologics.

Most of the delay comes from submissions that do not document what the payer's policy actually requires, particularly the prior workup for capsule and the step therapy history for biologics.

Capsule endoscopy

Codes: 91110 for esophagus through ileum, 91111 for esophagus only, 91113 for whole gut.

Nearly all commercial payers and Medicare Advantage plans require prior authorisation, and most Medicare Administrative Contractors require it for 91110.

Payers position capsule as second-line and expect a defined covered indication: suspected small bowel bleeding, iron-deficiency anemia with no identified upper or lower source, known or suspected Crohn's disease, or polyposis surveillance.

Two things cause most denials. A request stating only that this is a GI bleed workup, without documenting the prior negative EGD and colonoscopy. And treating an authorisation as transferable, when authorisations are code-specific and an approval for 91110 does not cover 91113 at a later encounter.

An authorisation number is also not a guarantee. Claims are denied for medical necessity despite a valid authorisation where the clinical documentation submitted at the authorisation stage was thin.

The scale of the burden

The American College of Gastroenterology's 2021 national survey on prior authorisation burden found that around half of requests relate to prescription refills, that more than half of respondents choose an inferior treatment at least weekly because of the burden, and that 59.5 percent had hired staff working exclusively on prior authorisation. More than half reported at least one patient suffering a serious adverse event from an authorisation-related delay (Shah and colleagues, American Journal of Gastroenterology, 2022).

In inflammatory bowel disease specifically, a pediatric cohort study found 74.2 percent of patients required prior authorisation (Constant and colleagues, Pediatrics, 2022).

Across all specialties, the American Medical Association's 2025 physician survey found physicians completing an average of 40 prior authorisations a week, with 26 percent reporting an authorisation-related serious adverse event.

What the 2026 rule does and does not fix

The CMS interoperability and prior authorisation rule, effective 1 January 2026, requires affected payers including Medicare Advantage plans to decide standard requests within seven calendar days and expedited requests within 72 hours, and to state a specific denial reason.

It does not cover drugs and biologics. Which is exactly the infliximab and vedolizumab gap. Plan accordingly rather than expecting relief there.

Infusion billing

J-codes and units, where errors are expensive because the claims are large:

• Infliximab (Remicade), J1745. One unit equals 10 mg, so a 400 mg dose bills as 40 units. Biosimilars bill under their own Q-codes, not J1745
• Vedolizumab (Entyvio), J3380. One unit equals 1 mg, so a 300 mg dose bills as 300 units
• Administration: 96413 for the initial hour and 96415 for each additional hour, billed as a complex biologic. Some payers direct the therapeutic codes 96365 and 96366 instead, so confirm the policy per payer

The NDC must appear on the claim and must match what was administered. NDC and unit mismatches, and biosimilar substitution conflicts, are common denials.

JW and JZ, which quietly deny claims

The JW modifier reports discarded drug from a single-dose vial and has been required since 1 January 2017. Not reporting wastage forfeits payment for it.

Since 1 July 2023, the JZ modifier is required on single-dose container claims where there is no waste.

Which means a single-dose vial line carrying neither JW nor JZ is a predictable denial. This requirement sits in CMS transmittals rather than anywhere obvious, and it is quietly denying claims at practices that have not noticed.

Buy and bill or specialty pharmacy

Buy and bill means you purchase the drug, administer it, and bill it as a medical benefit. You keep the drug margin and carry the inventory cost and the accuracy risk on high-value claims.

Specialty pharmacy routes the drug through the pharmacy benefit and you bill only the administration. No margin, no inventory risk, more coordination.

Neither is correct in general. It depends on your volume, your payer mix, and whether the acquisition cost sits below reimbursement for the agents you use most. Worth running the numbers on your top three rather than following what the practice down the road does.

Same-day E/M

An evaluation and management service on the day of an infusion is not separately billable unless a significant, separately identifiable problem was addressed and modifier 25 is genuinely supported.

Worth being disciplined about. Modifier 25 on same-day minor procedures is an active payer audit target, and the ACG's own guidance notes payers regularly targeting its overuse.

Section 10

Pathology and the per-specimen rule

GI biopsies are typically CPT 88305, billed per specimen, meaning per separately accessioned container.

Three polyps in three jars is three units. Three polyps in one jar is one unit. This is decided at the point of collection and cannot be fixed afterwards by any biller, however good.

That single sentence is worth more than most of this page to a practice that has never audited it. If your endoscopy suite routinely combines specimens into one container for convenience, you are giving away revenue that no downstream process can recover.

The component split

For practices with in-house histology:

• Global, no modifier. One entity performs both the technical and professional work. A practice with its own lab reading its own slides.
• 88305-TC. Technical component only. Your lab prepares the slides, an outside pathologist reads them.
• 88305-26. Professional component only. An outside lab prepares, your pathologist reads.

Never report 26 and TC on the same line for the same specimen. If one entity does both, bill global.

Place of service drives whether global is payable, and this is where practices get caught. In-house histology captures a technical component you would otherwise cede to a reference lab, which is a real revenue line, but it adds compliance and place-of-service exposure that has to be handled properly.

One thing to tell patients

Even where a screening colonoscopy is cost-shared at zero, patients can still receive separate facility or pathology charges. Another item for the pre-procedure conversation in Section 6.

Section 11

What GI denial benchmarks are actually known

Less than the internet suggests.

The GI-specific denial rates and revenue leakage figures quoted across billing company websites do not trace back to MGMA, HFMA, or any peer-reviewed source. We are not going to repeat them, and you should be sceptical of anyone who does.

What is independently known

Industry-wide rather than GI-specific:

• The American Hospital Association reported average initial denial rates rising to 11.8 percent in 2024
• The Change Healthcare 2024 Revenue Cycle Denials Index put the industry-wide initial denial rate at 11.81 percent, up from 10.15 percent in 2020
• Root causes in that index: registration and eligibility 27 percent, missing or invalid data 16 percent, authorisation 12 percent, medical necessity 11 percent, service not covered 8 percent, provider eligibility 5 percent
• General best practice targets from HFMA MAP Keys and MGMA: denial rate under 5 percent, days in AR of roughly 30 to 40

What is not known

There is no published, independent, ranked table of the most frequent denial reason codes in gastroenterology. Anyone presenting one has built it themselves.

The figures you will see quoted as GI-specific denial rates, as percentage revenue leakage, and as annual dollar leakage per multi-physician group appear only in vendor content with no traceable source.

What to do instead

Benchmark against yourself. Your own denial rate by reason code, trended over months, tells you more than any industry figure, because your payer mix and your case mix are the variables that actually move it.

That is what the free audit produces, and it is more useful than a number from a blog.

Section 12

Should you outsource GI billing or hire a certified coder?

If you have a coder who reads operative reports and understands screening conversions, keep them.

The specific competence that matters in GI is narrow and identifiable, and a practice that has it does not need to outsource. Most practices do not have it, and often do not know whether they do.

How to find out in an afternoon

Pull twenty colonoscopy claims where the operative note records a polypectomy. Check how many were coded to the therapeutic code with the correct modifier for that payer, and how many defaulted to the diagnostic code.

That single audit tells you more about your billing than any conversation will.

In-house is right when:

• Your coder reads operative reports rather than working from the schedule
• Screening conversions are coded correctly and consistently
• Denials are worked and appealed rather than written off
• Your volume is stable and your payer mix is simple

Outsourcing makes sense when:

• The audit above returns a number you do not like
• One person holds all the knowledge and there is no cover
• You are adding providers or an infusion suite and cannot add billing capacity at the same rate
• Prior authorisation has become a full-time job nobody was hired to do

What we cannot do

Worth saying plainly. We cannot make a payer's authorisation department move faster, change a payer's MAC policy, or recover a specimen that was combined into one jar in the endoscopy suite.

What we can do is make sure the coding is right the first time, the documentation supports the appeal, and someone is working the denials rather than adjusting them off.

What we handle

Section 14

Get a free 48-hour AR audit

We look at your claim data and tell you what is not being worked, where the denials cluster, and what it is costing you. For a GI practice that usually starts with the screening conversion audit described above. No obligation and no requirement to change anything. If your coding is clean, we will tell you that.

Section 15

Common questions

Sources

• CMS Medicare Claims Processing Manual, Chapter 18, Section 60.2
• CMS Medicare Coverage Database, Article A55069
• CMS National Coverage Determination CAG-00440N, multi-target stool DNA
• Consolidated Appropriations Act 2021, Section 122
• American Gastroenterological Association, Coding FAQ for Screening Colonoscopy
• American College of Gastroenterology coding resources
• Shah and colleagues, American Journal of Gastroenterology, 2022, national survey on prior authorisation burden
• Constant and colleagues, Pediatrics, 2022, prior authorisation in pediatric IBD
• Lieberman and colleagues, The American Journal of Medicine, polyp detection
• American Medical Association, 2025 Prior Authorization Physician Survey
• American Hospital Association and Change Healthcare 2024 Revenue Cycle Denials Index