Coronary Intervention Codes and Reimbursement: Two Decades of Effective Advocacy Why have interventional cardiologists’ salaries ranked at or near the top compared to other specialties for the past decade (1)? Long hours under high stress using extreme skills to perform dangerous procedures? Yes, but there is more. Effective advocacy by the Society for Cardiac Angiography and Interventions (SCAI) and American College of Cardiology (ACC) has played a large role. This is the story. Medicare, enacted in 1965, based reimbursement for physician services on the actual charge on the current bill, the customary charge over the past year, or the local medical profession’s “prevailing” charge over the past year, whichever was lowest (2). This system was chaotic and confusing. In response, the Omnibus Budget Reconciliation Act of 1989 switched Medicare to the Resource Based Relative Value System (RBRVS).
This used Hsaio et al’s estimates of physician time and effort to assign Relative Value Units (RVUs) to physician services (3). In 1991, the Center for Medicare and Medicaid Services (CMS) convened a series of Technical Expert Panels (TEP) to refine Hsaio’s initial estimates of work for selected procedures. One of these was percutaneous transluminal coronary angioplasty (PTCA). A representative of the SCAI/ACC convinced the TEP to increase reimbursement for PTCA from Hsaio’s estimate of 9.5 RVU’s to 10.5 RVU’s. The 20 million or so coronary angioplasty and stenting procedures performed in the US since 1992 have all been reimbursed at a rate reflecting that 1 RVU increase granted by the TEP in 1991.
Thus, this one instance of effective advocacy by SCAI/ACC increased reimbursement for these 20 million coronary intervention procedures over two decades. Now jump to 1994 when STRESS (4) and BENESTENT (5) compared elective stenting to balloon angioplasty, and a randomized trial compared then state-of-the-art Palmaz-Schatz and Gianturco-Roubin II stents (6). Elective stenting was just starting; most stents were placed to bail out failed balloon angioplasty. In this milieu a code for coronary stenting was developed.
The expert panel that advised CMS on reimbursement estimated that the average stenting procedure required 120 minutes of physician time from first injection of lidocaine to last catheter withdrawn (diagnostic catheterization not included), 45 minutes of preparation time before the procedure, and 60 minutes of physician work after the procedure, for a total physician work time of 225 minutes per coronary stenting case. Thus, interventionists have been paid for coronary stenting at a rate based on almost 4 hours per procedure for the past 17 years.
New Coronary Intervention Codes and Values
For the past several years, CMS has attempted to curb Medicare expenditures by identifying and reducing payment for over-priced services. In 2011 CMS identified coronary stenting as possibly over-priced and required that it be re-valued. The value of a service depends on the time required to perform it, and to a lesser extent the intensity of the work. SCAI and ACC knew that invasive cardiologists were reimbursed for 4 hours of work per stent case since 1994, and that procedural times might have shortened since then. A re-valuation could significantly decrease the RVUs paid for a coronary stenting procedure. Interventional cardiologists were also keenly aware of problems with the existing coronary intervention codes (Table 1). Reimbursement for an emergency middle-of-the-night ST elevation myocardial infarction (STEMI) stent procedure was the same as for elective stenting of a healthy patient at noon.
Stenting of complex left anterior descending bifurcation lesions requiring 3 stents was valued the same as stenting of a type A lesion requiring 1 short stent. SCAI/ACC experts decided that if interventional procedures were to be re-valued, it was time to get codes that recognized and reimbursed for the extra work of performing complex coronary interventions. SCAI/ACC experts developed a new set of codes that describe interventional procedures with greater detail (Table 2) and won their approval by the AMA Current Procedural Terminology (CPT) Panel. Then they had to be valued. This required several steps. The first step was a survey of practicing interventionists to estimate physician work and time required for each new coronary intervention code.
As expected, practicing cardiologists estimated the skin-to-skin time required for coronary stenting to be much less than original 2 hours – 45 minutes to be exact. Without the new codes, reimbursement for coronary stenting would likely have been reduced proportionately, by over 50%. Fortunately, SCAI/ACC experts convinced the American Medical Association Relative Value Update Committee to recommend to CMS that the new complex coronary intervention codes be reimbursed at higher rates (by up to 25%) than simple coronary stenting. Overall, reimbursement for the family of coronary intervention procedures will drop 18-20%, much less than the 50% that might have occurred without the new codes.
New Coronary Intervention Codes Solve Old Problems
The new codes solve several longstanding problems.
.1. For a decade interventionists have complained that they are not reimbursed for the intensity of STEMI PCI. Now they are. RBRVS rates intensity using units of “RVU’s per minute of procedure time”. The intensity of seeing patients in clinic rates.03, coronary bypass surgery rates.10, and emergency tracheostomy rates.26. Coronary intervention codes were previously rated at.10, but the new code for STEMI PCI has an intensity rating of .18. Intensity of other new coronary intervention codes is raised to the .13 – 15 range. .2. The extra work and stress of PCI of grafts and chronic total occlusions is now recognized and reimbursed higher, by 10% and 25% respectively. .3. Stenting preceded by atherectomy is now reimbursed at a higher rate (by 12%) than stenting alone. Previously there was no differential.
.4. The additional work of performing PCI on multiple branches of a single artery is now recognized with separate codes. CMS refuses to pay for these, and SCAI and ACC are lobbying CMS reverse this decision. The good news is that CMS’ decision does not limit reimbursement because CMS bundled the value of the “additional branch codes” into payment for the base codes. SCAI/ACC still recommends that the “additional branch codes” be used because some private payers may choose to reimburse them. Interventional Coding Examples to Illustrate Basic Principles .1. Problem: Coronary angiography is followed by ad hoc coronary stenting of the right and circumflex coronary arteries. Solution: 93454 (coronary angiography), 92928 (stenting single coronary), and 92928 again (stenting circumflex). Principles: As before, catheterization is coded using the separate cardiac cath codes, which are paid at 50% when performed with coronary intervention.
Also, the base code for coronary stenting (92928) is used for both vessels, whereas previously the base code was used once, along with an “each additional vessel” code which was retired in 2013. .2. Problem: Stenting of the circumflex is performed followed by atherectomy and stenting of the ramus. Solution: 92928 (stenting single coronary), 92933 (atherectomy and stenting single coronary). Principles: Previously CMS recognized and reimbursed for procedures in only 3 arteries (the left anterior descending, the circumflex, and the right) and might have denied reimbursement for the ramus PCI. Starting in 2013, CMS recognizes two additional arteries (the left main and ramus arteries) and will reimburse for PCI in all of them. Also, use the new “atherectomy + stenting” code (92933) offers higher reimbursement than the stent code (92928). .
3. Problem: A patient with non-ST elevation myocardial infarction has a 99% lesion with slow flow stented. Solution: 92941: (stenting of subtotal/total occlusion causing acute MI). Principle: This code can be used for any acute MI patient (STEMI or non-STEMI) with a “total or subtotal” lesion. CPT does not provide a definition of “total or sub-total”, so if the code is used an accurate description of the lesion to support this code should be included in the procedural report. .4. Problem: Bifurcation stenting of the left anterior descending is performed, with PTCA of the sidebranch ostium and stenting of the parent vessel. Distally, a separate diagonal sidebranch is rotationally atherectomized. Solution: 92928 (stenting of the LAD), 92921 (angioplasty, additional branch for the LAD diagonal bi9furcation), 92925 (atherectomy, additional branch).
Principles: PTCA of the diagonal as part of the bifurcation stenting is now recognized. When a separate branch is treated, use a second “additional branch” code. .5. Problem: Intravascular ultrasound (IVUS) shows a significant left main lesion extending into the proximal LAD which is stented. Fractional flow reserve across a distal lesion is measured and is not significant. Solution: 92928 (stenting of the left main/LAD), 92978 (intravascular ultrasound), 93571 (fractional flow reserve). Principle: As before, IVUS and FFR codes are used as “add-on” codes in addition to the base coronary intervention codes. When a single stent is used to treat a lesion in the left main extending into the LAD or circumflex, it is coded with only one code.