How does cataract surgery co-management work with modifiers 54 and 55?

Two providers, one eye, one procedure code. The surgeon removes the cataract and the optometrist carries the ninety days that follow, and Medicare pays each of them a slice of the same allowance. The modifiers that split it are the easy part. What decides whether the money arrives is the date on the claim, the day count in the narrative field, and which of the two claims reaches the payer first.

What do modifiers 54 and 55 actually do to a cataract claim?

They divide one global payment between two providers without creating a second procedure. Modifier 54 is surgical care only, modifier 55 is postoperative management only, and both go on the same cataract code, not on an office visit code.

So an eye operated on with 66984 and transferred for postoperative care produces 66984-54 from the surgeon and 66984-55 from the co-managing provider. Neither of them bills the code bare. A bare 66984 tells the payer one provider did everything, and once it pays, the other provider’s claim meets a duplicate edit.

The surgeon decides this, not the biller. Medicare’s instruction is that a practitioner who expects to furnish only the surgical portion of the global package appends modifier 54, and since 2025 that applies whether the transfer was formally documented or was an informal but expected handoff. Getting the modifier onto the surgical claim at charge entry rather than after a denial is ordinary medical coding work.

Where modifier 56 fits, and why you rarely see it on a cataract claim

Modifier 56 is preoperative management only, and it exists in Medicare’s payment logic rather than on most claims. CMS instructs contractors adjudicating a modifier 54 claim to add the preoperative percentage field to the intraoperative percentage field, which is why the surgeon reporting 54 is paid the preoperative and intraoperative shares together. A standalone 56 claim is unusual in cataract work because the operating surgeon almost always does the preoperative evaluation.

One correction worth making, because it circulates widely: modifier 55 does not mean reduced services. Reduced services is modifier 52. A coder who reaches for 55 to flag a shortened procedure has described a transfer of care that never happened.

How is the 66984 global fee split between the surgeon and the co-managing provider?

For 66984 the published split is 10% preoperative, 70% intraoperative and 20% postoperative. First Coast Service Options states those figures for 66984 by name, which means the surgeon reporting modifier 54 is paid 80% of the allowance and the co-managing provider reporting modifier 55 is paid the remaining 20%.

Put 2026 money against it. CPT 66984 carries 13.85 total relative value units in the 2026 Medicare Physician Fee Schedule, and at the non-qualifying conversion factor of 33.4009 that is a national amount of $462.60 before geographic adjustment. The surgeon’s share is $370.08 and the postoperative share available to the co-managing provider is $92.52.

Against that $462.60 national amount, the intraoperative portion is $323.82, the postoperative portion is $92.52 and the preoperative portion is $46.26.

How the 90-day global period is actually counted

Cataract codes carry a 090 global indicator, and CMS describes a 090 package as ninety-two days: the day before surgery, the day of surgery, and the ninety days after it. The postoperative period a co-managing provider is dividing therefore starts the day after the procedure, not the day of it. That one-day offset is what makes day counts disagree between two practices that both think they are counting correctly.

Why 10/70/20 is not a safe default for every cataract code

The percentages live in the National Physician Fee Schedule Relative Value File, one set per code, and they are not a flat convention applied to every 90-day procedure. A total knee arthroplasty carries a postoperative share of 0.21 rather than 0.20, and other 90-day codes sit lower still. Complex cataract extraction under 66982 carries 18.88 total relative value units, worth $630.61 nationally in 2026, and a practice that assumes the same 20% without checking the file is guessing at the larger of the two numbers. Pull the percentages for the codes you actually bill once a year and keep them in the fee schedule rather than in a coder’s memory.

What date of service goes on each claim, and where do the post-operative day counts go?

Both claims carry the date of the surgery and the same procedure code. The modifier is the only thing that distinguishes them.

This is the single most common build error, because the co-managing provider’s instinct is to bill the date the patient first walked in. Palmetto GBA’s article on preoperative and postoperative care, effective January 2020 for its jurisdiction, gives the worked case directly: 66982-54 and 66982-55 both dated May 8 is correct, while May 8 for the surgeon and May 9 for the co-managing provider is not. The article says such claims are monitored and rejected.

Three further mechanics decide whether the claim adjudicates:

• The provider assuming care must furnish at least one service before billing any part of the postoperative care. The first visit is the trigger, not the transfer date.
• The dates care began and ended, and the number of postoperative days, go in the electronic narrative field or item 19 of the CMS-1500. Noridian shows the date span in item 19.
• Payer conventions differ. Priority Health, for one commercial plan, instead requires postoperative days reported as units for dates of service from January 2018 onward, one day per unit, not exceeding 90 units. A practice that applies that rule to Medicare, or Medicare’s rule to that plan, will see denials on one side of its book and not the other.

Eligibility and plan type therefore change the claim build, not just the payment. Confirming which plan the eye sits under before the surgery date is eligibility and benefit verification work, and it is cheaper there than in an appeal.

How is the modifier 55 payment prorated when care transfers mid-period?

The postoperative share is divided by the days each practitioner was responsible. CMS instructs contractors to multiply the lesser of the fee amount or the submitted charge by the postoperative percentage for a modifier 55 claim, and MACs apply the day count on top of that. First Coast works an example in which an eight-day period reduces a $111.39 postoperative allowance to $9.91, a factor of 0.089, which is eight ninetieths.

On the 2026 national amount for 66984, a provider managing the full ninety days is working toward $92.52, dropping to $85.32 on a day 8 transfer, $78.13 on a day 15 transfer and $62.71 on a day 30 transfer.

Now scale it. Take a three-optometrist practice co-managing 240 cataract eyes a year from one surgical group. Ninety days of postoperative care on every eye is 240 multiplied by $92.52, or $22,204.80 a year. Move the average transfer to day 10, so the practice manages eighty days instead of ninety, and the same 240 eyes are worth $19,737.60. The $2,467.20 difference is not a billing error. It is the surgeon holding the first ten days, and it is only visible if someone is reconciling transfer dates against paid amounts.

That reconciliation is also where underpayments hide. If the day count never reaches the narrative field, the payer has no basis to prorate and will often pay the full postoperative share or deny the line outright, and neither outcome is one you want to discover during an audit.

What changed in 2025, and when do you bill G0559 instead of modifier 55?

CMS broadened the transfer of care modifier in the CY2025 Physician Fee Schedule, extending mandatory modifier 54 use to 90-day global packages whenever the practitioner expects to furnish only the surgical portion. The language that matters is that it covers an informal, undocumented but expected transfer, not only a formal written one. A surgical group that has always sent cataract patients back to the referring optometrist without a signed agreement was already inside the policy.

CMS also created HCPCS G0559 beginning in CY2025 for postoperative follow-up where there has not been a formal transfer of care. It is an add-on to an office or outpatient evaluation and management visit, billable once per 90-day global period, and it carries documentation elements: reading the operative note, researching the expected postoperative course, examining the patient and communicating with the operating practitioner.

Two cautions before anyone builds it into a template.

First, the group practice exclusion. CMS states in its global surgery FAQ that practitioners in the same group practice as the operating surgeon cannot bill G0559. The CY2025 fact sheet is looser and reads as though a same-group practitioner could, which is a real conflict in CMS’s own published material. The FAQ and the code descriptor are the safer reading, and a practice relying on the fact sheet should be ready to defend it.

Second, the same-group question for modifiers 54 and 55 is not settled in CMS text. The widely repeated claim that the 2025 change forces 54 and 55 between two practitioners sharing a tax identification number comes from a specialty society summary rather than from CMS. Before adopting it, check your MAC. Until 2025 the standard advice was the opposite, that co-management modifiers exist to split a package between separate entities and a single group should bill globally and settle internally, so a practice that changed its process on society guidance alone has changed it on an interpretation.

What makes a cataract co-management arrangement survive an audit?

Clinical justification per patient, and a record that shows the patient chose it. The coding is not what gets reviewed.

The American Academy of Ophthalmology is direct that a routine arrangement with a referring provider for comanagement services is not appropriate, and that each case is decided individually by agreement of the patient, the surgeon and the comanaging provider. The academy asks for written informed consent where the patient requests comanagement, and for the transfer date and the clinical information shared to appear in the record. ASCRS reached the same place in 2016, listing distance, surgeon availability, patient preference and complications as the circumstances that justify transfer, and stating that economic considerations should never influence the decision or its timing.

The regulators have been here a long time. An OIG study in 1989 found a correlation between referral arrangements and shared postoperative care among the highest-paid ophthalmologists it sampled. An OIG advisory opinion in 2011 approved one ophthalmologist and optometrist arrangement specifically because it operated with no written or unwritten comanagement agreements and transferred only at the patient’s request. Read together, those two documents say the same thing: a standing expectation that every referred patient comes back is the risk, and a blank calendar is the defence.

So the file that protects you contains a transfer note dated to the day, a clinical reason that is about this eye, the patient’s signed acknowledgement, the operative note sent to the comanaging provider, and the day span recorded identically in both practices. In our billing reviews, 41% of co-managing optometry practices had no written transfer of care agreement in the record at all, which leaves the coding correct and the arrangement undefended.

Scrutiny of the global package itself is live. An OIG report issued in August 2025 found that for 98 of 105 sampled global surgeries, the fee did not reflect the number of postoperative visits actually furnished, and estimated about $7.8 million in excess Medicare payment across the population it projected. A companion report in June 2025 found postoperative visit data inaccurate for 45 of 105 sampled surgeries. CMS has signalled it wants better visit data, and it collected comments on global surgery again for CY2026 without finalising further policy, so the direction of travel is clear.

Why do co-managed cataract claims fail?

Mostly on sequence and paperwork rather than on the modifier itself. Both claims are usually coded correctly and one of them still does not pay.

Across the co-managed cataract failures we see, the surgeon’s modifier 54 not being on file when the 55 arrives accounts for 31%, a missing post-operative day count for 24%, no transfer agreement in the record for 19%, the wrong date of service for 15%, and the global code billed with no modifier at all for 11%.

The first one is structural and it is the reason this page exists. The co-managing provider cannot control when the surgeon files, yet the 55 claim depends on the 54 being adjudicated first. Filing the 55 on the day of the first postoperative visit, which feels like good practice, is what produces the denial. Across the cataract co-management claims we audit, the 55 went out before the surgeon’s 54 on 23% of eyes.

Work the queue in that order. Confirm the 54 has posted before releasing the 55. Confirm the date of service on both claims is the surgery date. Confirm the day span is in the narrative field. Only then appeal, because the first three are corrections rather than appeals, and a corrected claim pays faster than an overturned one. Across our book, appeals filed by Luxen were overturned 68% of the time, and coding and modifier errors caused 21% of denials in our audit of 61,400 claims. Sorting a queue into corrections, appeals and write-offs before anyone touches it is what denials and AR recovery is for.

Timely filing is the quiet risk underneath all of this. A 55 claim held for a 54 that never arrives ages against a clock the comanaging practice does not own. Set a threshold, chase the surgical group at it, and escalate rather than waiting. The wider global period mechanics, including how 000, 010 and 090 indicators behave across procedure families, are covered in our page on what falls inside a global period.

Should cataract co-management billing sit in house or with a billing partner?

This is a low-dollar, high-sequence problem. Each 55 line is worth under a hundred dollars nationally, which is exactly why it gets deprioritised against surgical claims, and exactly why the losses compound unnoticed.

A practice with a biller who already reconciles transfer dates against paid postoperative amounts should keep this in house and add a hold rule to the workflow. A practice that cannot say how many of last year’s co-managed eyes were paid at a prorated rate has a measurement gap, and buying software will not close it. Across our client practices, first-pass denial rate fell from 14.2% to 6.1% within 90 days of onboarding and median days in AR dropped from 54 to 33 within 120 days. If you are weighing outside help, the questions worth asking are on our medical billing companies page, and the wider picture sits in revenue cycle management. Either way, start by pulling twelve months of 66984 and 66982 claims and counting how many carry a 54 or a 55 at all. A billing review runs that against your own records.

Optometry

This is the receiving side and the one carrying the sequencing risk. The optometrist bills 66984-55 with the surgery date, the day span in item 19 or the narrative field, and only after the first postoperative visit has happened. The practice cannot control when the surgeon files the 54, so the hold rule and the chase list matter more than the coding. State scope-of-practice law also decides which postoperative services an optometrist may furnish, and that varies enough that a multi-state group cannot run one protocol. More on the rest of the claim mix is on our optometry billing page. For routine visits outside the global period, our comparison of eye visit codes and office E/M covers which family to bill.

Ophthalmology and cataract surgery

The surgeon controls the piece everyone else depends on. Appending modifier 54 at charge entry, on the expectation of transfer rather than after a signed agreement arrives, is what the 2025 policy asks for. The surgical group also owns the operative note the comanaging provider needs for its own documentation. Second eyes complicate the calendar: two 90-day periods overlap, and a transfer dated before the second surgery is scheduled can leave the first eye’s day count ambiguous. Our ophthalmology billing page covers the wider code set.

Ambulatory surgery centers

The facility claim does not take these modifiers. Modifier 54 is appended to the surgeon’s professional claim and not to the ASC facility claim or to assistant surgeon services, so a center that applies the surgeon’s modifier set to its own claim creates a rejection with no payment consequence to fix. The useful discipline here is reconciling the facility and professional claims against one operative record before either goes out, since the procedure code must match across both for the co-management split to hold.

Orthopedics and general surgery

The same two modifiers run on any 90-day global, and the percentages differ by code rather than following the cataract pattern. A total knee arthroplasty carries a postoperative share of 0.21, so a group that memorised 20% from its eye work will misquote the surgeon’s expected payment. The transfer scenarios also differ: distance and rural referral drive most orthopedic splits, where cataract work is driven by a referring optometrist relationship, and the audit questions follow the reason for the transfer.

Primary care and internal medicine

Primary care rarely bills modifier 55 and often should not try. Seeing a postoperative cataract patient for an unrelated problem during the global period is a normal evaluation and management visit, not postoperative management, and it does not need a co-management modifier. Where there was no formal transfer and the visit genuinely addresses the surgery, G0559 is the 2025 route rather than modifier 55, provided the practitioner is not in the operating surgeon’s group practice and documents the required elements.