Cardiology billing: component splits, global periods and monitoring intervals

Diagnostic cardiology studies have two halves. The technical component is the equipment, the supplies and the staff time to acquire the study. The professional component is the physician's interpretation and written report. Whether you bill one, the other, or both depends entirely on what your practice actually performed and who owns the equipment.

Bill globally — no modifier — only when the practice performed both halves. Append modifier TC when you acquired the study but someone else interpreted it. Append modifier 26 when you interpreted a study acquired elsewhere, which is the common case for hospital-based reads.

The expensive error is billing globally by default. A cardiologist reading a study performed at the hospital who bills without modifier 26 has claimed the technical component too, and the hospital has already billed it. The claim denies as a duplicate, or worse, pays and is recouped later.

Stress testing is the exception that catches people who have learned the 26/TC rule and apply it everywhere. The components are not modifiers here; they are separate codes.

93015 is the complete service. Where the work is split, 93016 covers physician supervision, 93017 covers the tracing only, and 93018 covers interpretation and report only. Appending modifier 26 to 93015 is not the correct way to bill a physician-only stress test, and payers reject it.

The practical control is knowing, per location, which half of each study your practice actually performs. That is a configuration question answered once per site and per payer contract, not a decision a coder should be making claim by claim from the note.

Interventional cardiology procedures carry global periods, and during that window related evaluation and management services are included in the procedure's payment rather than separately billable.

This produces a denial pattern that reads as a coding error but is really a modifier omission. A patient returns during the global period, is genuinely seen, and the visit is billed without establishing why it falls outside the package. The payer bundles it, correctly, and the practice writes off work that was performed.

Three modifiers carry almost all of this. Modifier 25 identifies a significant, separately identifiable evaluation on the same day as a procedure. Modifier 24 identifies an unrelated evaluation during another procedure's postoperative period. Modifier 57 identifies the visit at which the decision for a major procedure was made. Each asserts a different exception, and substituting one for another denies as reliably as omitting all three.

Cardiology is unusually exposed because patients are managed continuously. A cardiology patient does not have one episode and disappear; they return for medication management, for an unrelated arrhythmia, for a chronic condition entirely separate from the stent placed six weeks ago. Every one of those visits sits inside somebody's global period.

Remote cardiac monitoring is billed per interval, and the interval must have fully elapsed before the code is submitted. This is the most mechanical denial category in the specialty and the easiest to eliminate.

Remote interrogation of an implanted pacemaker or defibrillator covers a period of up to 90 days. Remote monitoring of physiologic cardiovascular data from an implantable cardiovascular monitor covers up to 30 days. Submitting on day 82 of a 90-day interval denies, and resubmitting on day 91 frequently denies again because the payer now has a prior claim on file for the same period.

Advanced cardiac imaging and interventional procedures frequently require prior authorization, and the requirement is set per payer and per plan rather than by the procedure itself. The same CPT code can require authorization under one plan and not another from the same carrier.

That variability is why a practice-wide rule of thumb fails. Staff learn that a given study needs authorization, apply it consistently, and are correct most of the time — which means the failures are rare, unpredictable and discovered only after the service has been delivered and the authorization can no longer be obtained.

The workable control is checking the requirement against the specific plan at scheduling, when the appointment can still be moved, rather than relying on institutional memory. Registration and eligibility errors are already the largest denial category industry-wide at 24.3%, and authorization sits close behind at 12.8% — in a specialty running high-value procedures, those percentages attach to unusually large claims.

Premier Inc. measures the average administrative cost of fighting a denied claim at $57.23, and roughly 70% of denied claims are eventually overturned and paid. In lower-value specialties that arithmetic makes some denials uneconomic to pursue. In cardiology it almost never does.

A denied interventional claim or advanced imaging study is worth multiples of the cost to appeal it, which means the correct policy is to work essentially all of them. Practices that write off cardiology denials are usually doing so for capacity reasons rather than economic ones — the claims are worth pursuing and nobody has the hours.

The larger opportunity is still prevention. Optum's analysis of 124 million claim remits found 84% of denials are potentially avoidable, and 22% of those are not recoverable at all once they occur. Component splits, global-period modifiers and monitoring intervals are all determined before submission, which puts them squarely in the avoidable category — and the 22% figure is the reminder that appealing is a worse strategy than not being denied.