Orthopedic Revenue Cycle Management: The Failure That Pays You Anyway

The orthopedic groups I have built systems for do not have a denial problem on their global-period claims. They have a payment problem, which is worse, because a denial at least generates a report.

I work on the engineering side, which means I usually meet a revenue cycle through its data rather than through its staff. And the pattern I keep finding in orthopedics is not the one practices ask me to look for. They ask about denials. The denials are usually fine. What is not fine is a category of claim that processed cleanly, paid less than it should have, and left no trace anywhere a human would look.

This piece is about that category, and about the four other things orthopedic billing gets wrong that are easier to see. If you want the general shape of the cycle first, our revenue cycle management covers it and the 13 steps of the revenue cycle is the plainer version.

Modifier 58 vs 78 vs 79: Which Applies During a Global Period

When a procedure is performed during the postoperative period of an earlier procedure, one of three modifiers usually has to be present. Modifier 58 identifies a staged, planned, or more extensive related procedure. Modifier 78 identifies an unplanned return to the operating room for a related procedure. Modifier 79 identifies an unrelated procedure, typically at a different anatomic site.

These definitions come from the CMS Global Surgery Booklet, MLN907166, dated December 2025, which is the current canonical reference rather than a billing vendor’s summary of it. The American Academy of Orthopaedic Surgeons has published surgeon-facing guidance on applying these modifiers during the global period. That piece dates from 2013, and it is cited here because the modifier definitions it describes have not changed, not because it is current. Where the two differ, the CMS booklet governs.

• 58 What it identifies: Staged or related procedure during the postoperative period · The distinguishing test: Was it planned, staged, or more extensive than the original?
• 78 What it identifies: Unplanned return to the operating room, related procedure · The distinguishing test: Was it explicitly not planned, staged or anticipated?
• 79 What it identifies: Unrelated procedure by the same physician · The distinguishing test: Different anatomic site, unrelated to the original

And the rule that makes all three load-bearing: Medicare Administrative Contractors do not allow separate payment for an additional procedure furnished within a prior procedure’s global period if it is billed without 58, 78 or 79.

Omit the modifier and the payment does not happen. That is the version of this problem that practices already know about, because it is visible.

The Failure That Pays You Anyway

Here is the one practices do not know about.

Omitting the modifier is the loud failure. There is a quiet one, and it is the reason I wrote this piece.

Apply modifier 78 to a procedure where 58 was correct, and nothing rejects. The claim is well-formed. It processes. It pays. It pays at a reduced rate, because 78 and 58 do not carry the same payment treatment, and 58 opens a fresh global period while 78 does not.

So two things go wrong at once and neither surfaces. You are underpaid on that claim, and the global-period clock for the rest of that episode of care is now set from the wrong date, which means the modifier decision on the next procedure is being made against a wrong baseline.

The mechanism matters more than any figure. On a surgical fee schedule, a systematic partial underpayment applied to a recurring claim type, invisible to every report the practice runs, compounds quietly for as long as nobody goes looking.

Why Your Denial Report Cannot Show You This

This is the part that is genuinely a systems problem rather than a coding one, and it is where I spend most of my time with orthopedic groups.

A denial report answers one question: which claims did not pay. It is built on rejection data. It is very good at what it does.

An underpaid claim is not in it. The claim paid. The payer did what it was going to do, the remittance posted, the account closed. Every downstream report treats it as a success, because by the only definition those reports use, it was one.

To see the quiet failure you need a different question entirely: for each claim, what did we expect to be allowed, and what was actually allowed, and why do those two numbers differ. That is expected-versus-allowed reconciliation, and it is a different data structure from denial analytics. It requires the contracted rate, the modifier applied, and the payment treatment for that modifier, joined per line.

Most practices do not have it. Not because it is difficult to build, but because nobody asked for it, because the problem it detects does not appear anywhere that would prompt someone to ask.

I would put the diagnostic plainly: if your revenue cycle reporting can only tell you about claims that failed, you cannot see this class of problem at all, and its absence from your reports is not evidence that it is absent from your revenue. A revenue cycle analytics view built on expected-versus-allowed is what turns that from an unknown into a number.

Implants Are a Second Revenue Stream, and They Are Billed Separately

The other place orthopedic groups leave money behind is more straightforward, and it has nothing to do with modifiers.

Implants and devices are reported separately from the surgical procedure, under HCPCS Level II codes rather than the surgical CPT code. The surgical code covers the work. The device code covers what was implanted. They are two different things and they are two different lines.

The gap between those two lines is where groups consistently write off reimbursement they were entitled to. Not through any dramatic failure, but because the device line requires information that lives in the implant log or the vendor invoice rather than in the operative note, and the person building the claim is reading the operative note.

I am deliberately not listing specific device codes here. The vendor material that ranks for this topic does, and I could not source those code numbers to CMS or to a specialty society, so publishing them would be exactly the kind of borrowed precision this article is arguing against. What I can say confidently is structural: if your claim has a surgical line and no device line for a procedure that implanted something, check whether that was a decision or an omission.

Where Orthopedic Claims Actually Die

Five mechanisms, and note that the first one does not always kill the claim.

1. Global-period modifier errors. Both varieties. Omission denies; wrong selection underpays. The CMS National Correct Coding Initiative Policy Manual governs the bundling interactions underneath them.
2. Missing device lines. Covered above. Silent.
3. Modifier 24 on postoperative evaluation and management. An unrelated evaluation during a postoperative period needs to be identified as such, or it is treated as included in the global package.

Prior authorization, concentrated in advanced imaging and implant-bearing procedures. The federal Interoperability and Prior Authorization rule, CMS-0057-F requires affected payers to run a Prior Authorization application programming interface by January 1, 2027. Our prior authorization services covers what that integration involves.

Conservative-care-first medical necessity policies. Many payers require documented conservative treatment before authorizing surgical intervention. The requirement is knowable in advance and the documentation is knowable in advance, which makes this more tractable than it usually gets treated as.

I am not going to repeat the reimbursement-lift percentages that the billing services ranking for this topic advertise. They are published without methodology or source, and reprinting them would undercut the argument this piece is making about borrowed numbers.

Four of the five are addressable with rules. Denial management built on those four, plus expected-versus-allowed reconciliation to catch the fifth, covers this specialty better than a general accuracy program, and a claims processing configuration is where the rules actually live.

What Your Systems Have to Know

Four things, and the fourth is the one nobody asks for.

The global period tracked as a clock per episode, with a start date, a length, and the events that reset it. Not looked up per claim.

The modifier suggested from the relationship between procedures, not selected from memory. Whether the second procedure was planned, unplanned, or unrelated is knowable from the record of the first. That is the input the 58-versus-78 decision needs, and it exists.

Device line completeness checked against the surgical code. If the procedure implanted something, the claim should not leave without a device line or an explicit reason it has none.

Expected-versus-allowed reconciliation as a standing report. This is the detection layer for everything above that does not deny. Without it you are relying on the absence of complaints as evidence of correctness.

The engineering reality is worth stating, since I am the one who would have to build it. Connecting to a practice management system or an electronic health record is well-trodden. The hard part is the last mile: reconciling episode boundaries across systems that record them differently, getting implant-log data into a workflow that was designed around the operative note, and writing the reconciliation output back somewhere a biller will actually see it rather than into a report nobody opens. Anyone who tells you that part is routine has not shipped one.

The Clock Is Episode-Level and Most Systems Are Claim-Level

One structural note, because it explains why this problem is durable rather than merely common.

A global period is a property of an episode of care. It starts with a procedure, runs for a defined length, and certain events restart it. Reasoning about it correctly requires knowing what happened before the claim you are currently looking at, and what that earlier event did to the clock.

Most billing systems reason one claim at a time. The claim is the unit of work, the unit of validation and the unit of reporting. That is a sensible design for almost everything else in the revenue cycle and it is the wrong shape for this.

It also explains why the wrong-modifier error compounds. A misapplied 78 does not reset the clock. Every subsequent modifier decision in that episode is then evaluated against a global period that ended, or did not end, on the wrong date. One error at the start of an episode is not one error.

Outsource, Buy, or Build

Almost every page ranking for this topic sells one of the three. Here is the version that does not.

1. Outsourcing is the right answer for a lot of orthopedic practices. Surgical coding is genuinely specialized, and a partner who does it all day will outperform a generalist biller. Outsourcing revenue cycle management covers where that line usually falls, and our list of revenue cycle management companies is a starting point for a shortlist. If the distinction between billing and the full cycle is still unclear, medical billing versus revenue cycle management sets it out.
2. Here is the orthopedic-specific caveat and it is uncomfortable. A billing partner compensated as a percentage of collections has a structural incentive to collect what is collectible, and very little incentive to investigate claims that already paid. The quiet failure in this article is precisely a claim that already paid. I am not suggesting anyone is acting in bad faith. I am pointing out that the detection work sits outside what the arrangement rewards, so if you outsource, expected-versus-allowed reconciliation is a thing to ask for explicitly rather than assume.
3. Building earns its keep when your volume of global-period episodes is high enough that a systematic percentage matters, and when the detection layer has to live somewhere you control. Orthopedics meets the first condition more often than most specialties, because surgical episodes with staged procedures are routine rather than exceptional.

Practices weighing a platform change alongside this should treat system selection as a separate question. Our guide to the best EMR for orthopedics covers that side. Rural and critical access organizations have a different reimbursement profile again, covered in critical access hospital reimbursement.

Each specialty fails in its own shape, which is worth knowing if you run more than one line of business. Cardiology revenue cycle management covers a code-set migration that broke edit rules silently, gastroenterology revenue cycle management a payer-dependent modifier decided mid-procedure, and oncology revenue cycle management a mandatory attestation on the drug line. Orthopedics is the one where the claim pays anyway.