Article Text

Download PDFPDF

Diagnostic carotid and cerebral angiography: a historical summary of the evolving changes in coding and reimbursement in a complex procedure family
  1. William D Donovan1,
  2. Thabele M Leslie-Mazwi2,
  3. Ezequiel Silva III3,
  4. Henry H Woo4,
  5. Gregory N Nicola5,
  6. Robert M Barr6,
  7. Jacqueline A Bello7,
  8. Raymond Tu8,
  9. Joshua A Hirsch2
  1. 1The William W. Backus Hospital, Norwich, Connecticut, USA
  2. 2Neuroendovascular Program, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts, USA
  3. 3South Texas Radiology Group, Department of Radiology, University of Texas Health Science Center at San Antonio, San Antonio, Texas, USA
  4. 4Department of Neurological Surgery and Radiology, Cerebrovascular Center, Stony Brook University Medical Center, Stony Brook, New York, USA
  5. 5Department of Radiology, Hackensack University Medical Center, Hackensack, New Jersey, USA
  6. 6Mecklenburg Radiology Associates P.A., Charlotte, North Carolina, USA
  7. 7Montefiore Medical Center, Albert Einstein College of Medicine, Bronx, New York, USA
  8. 8Department of Progressive Radiology, The George Washington University, Falls Church, Virginia, USA
  1. Correspondence to Dr Joshua A Hirsch, Neuroendovascular Program, Massachusetts General Hospital, Harvard Medical School, Boston, MA 02114, USA, hirsch{at}


Carotid and cerebral angiography have been a mainstay of neurointerventional and neuroradiologic practice for years. Centers for Medicare and Medicaid Services (CMS) and Relative Value Scale Update Committee (RUC) initiatives have compelled the professional societies to bundle component codes under threat of unilateral CMS revision and revaluation. Code bundling usually results in a decrease in the professional Relative Value Unit (RVU) valuation, and thus the MD reimbursement. The year 2013 saw a dramatic revision to the Current Procedural Terminology (CPT) code set that defines carotid and cerebral procedures. This paper reviews the process that led to that code set being revised and estimates the impact on professional reimbursement. We show the current and previous carotid angiography CPT codes and use clinical examples to assess professional RVU valuation before and after code revision.

  • Angiography

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


The ‘component coding’ structure for angiography codes was formalized in the early 1990s, led by the American College of Radiology (ACR) and the Society of Interventional Radiology (SIR, then known as the Society of Cardiovascular and Interventional Radiology). This procedural coding system was adopted into the Resource-Based Relative Value Scale (RBRVS) and was used as the basis for calculating physician reimbursement by the Centers for Medicare and Medicaid Services (CMS), formerly known as the Health Care Financing Association (HCFA). Since that time, payments have been based on the cost of the service—the work by the physician, the cost to provide the service, and the medical malpractice premium to insure the service. The dollar amount paid is calculated by the three component values multiplied by a CMS conversion factor. Physician work contributes to less than half of the total value of the service. The Harvard University study provided the first calculation which took into account procedure time, skill required for the procedure, expertise, and patient risk. These metrics are revised yearly in order to account for practice evolution and are mandated for review at least every 5 years. Practice expense (PE) accounts for less than 50% of the total RBRVS based on a formula using 1991 charges.


Each Current Procedural Terminology (CPT) code is assigned a numeric value in terms of Relative Value Units (RVUs). The process for deciding on RVU values has been described in detail elsewhere and will be revisited later in this article.1–3

The component coding scheme divided billing into procedural (‘surgical’) codes and radiological ‘supervision and interpretation’ (S&I) codes for each angiographic procedure.4 As interventional neuroradiology was and continues to be an evolving field with new procedures, devices, and injected compounds, component coding provided capture of these innovations while providing flexibility for future innovation.

The procedural code referred to the physical work of examining and preparing the patient, obtaining percutaneous arterial access, catheterizing target vessels, injecting contrast, and subsequent angiography-related efforts including removing all catheters and access devices, achieving hemostasis, and discharging or transferring the patient out of the radiology suite.

The radiology S&I code covered the professional work of reviewing the indication for the examination and all previous studies, interpreting the angiography examination, dictating the results, and communicating the results to the referring physician and the patient and family when appropriate.

This component code structure had several advantages: it allowed separate billing for the same patient encountered by different physicians (for instance, one who performed the angiogram, another interpreting and dictating the results); it facilitated granular coding of complicated procedures due to the graduated approach to the catheterization of the arterial tree; and it allowed detailed tracking of work done for resource planning and research purposes.4 ,5 Component coding provides flexibility in capturing innovations in procedures as well as physician work.

Since 1992 the American Medical Association/Specialty Society Relative Value Scale Update Committee (RUC) has served as a volunteer body that assesses the relative value of physician work across the spectrum of medical and surgical services. The majority of the RUC members are representatives of the largest medical and surgical specialty societies. After presentations by advisers and deliberation, the RUC members vote on RVU valuation and then forward their recommendations to CMS.1 On the initial establishment of the RBRVS in the early 1990s, the RUC's efforts focused on valuing CPT codes for new procedures. Although ongoing review of the values of existing codes in order to ensure appropriate relativity in the system has always been a part of the RUC's mission, the RUC spent a minority of its time doing so. Until recently, the RUC's recommendations were almost universally accepted by CMS for the following year's Physician Fee Schedule. The percentage of outright acceptance has declined in recent years, approximating 75% for the past few years.

Since the inception of the RBRVS, a review of the entire RBRVS is required at least every 5 years. In 2007, CMS began to critically re-evaluate the manner and frequency in which RVU values were reviewed. For a variety of reasons, including outside criticism of the infrequency with which the existing code structure was reviewed, CMS began asking the RUC to attend to that mission with greater focus. Given that there are over 7000 CPT codes in regular use, CMS facilitated this process by requesting that the RUC specifically review categories or groups of codes that might be overvalued—or ‘potentially misvalued’, as the catchphrase became known in legislative, CMS and RUC documents. This retrospective review of large numbers of CPT codes was undertaken by a subcommittee of the RUC, the 5-year-Review Workgroup subsequently renamed the Relativity Assessment Workgroup (RAW) in 2009.1 ,3

CMS and the RAW selected these ‘potentially misvalued’ services for review based on various screens or filters. Some examples of these screens included:

  • services that had seen significant increases in utilization (the theory being that overly generous valuation may have led to overuse or preferential reporting of these codes)

  • services in which the site-of-service had changed since its initial valuation (suggesting that services previously performed primarily in the facility (hospital) setting, but which had evolved to being primarily non-facility (office-based) procedures, may still benefit from outdated facility-based PE reimbursement)

  • services in which the primary specialty performing the service had changed (the theory being that its physician work and PE may have evolved as well or that increased technical efficiency may have contributed to that change)

  • codes that dated from the early 1990s that lacked a verifiable data trail to justify their valuation (so-called ‘CMS/Other’ codes).6

Another CMS/RAW screen, one of the most onerous for the radiology community, has been the ‘Codes Performed Together’ screen—that is, CPT codes charged to the same Medicare beneficiary, by the same clinician, on the same day. CMS's theory was that, if services were very frequently reported together, duplication or overlap of work may be present and re-evaluating this work through the RUC process may demonstrate ‘efficiencies’—that is, codes nearly always performed together were really two overlapping parts of one procedure, not two (or more) separate procedures. It was this screen that identified some of the carotid angiography family of codes as being performed frequently together on the same day (see below). Indeed, such bundling of codes has occurred throughout the 70 000 code series in CPT as CT abdomen and pelvis and X-ray myelography and the injection code.

Through analysis of its own utilization data, CMS identified numerous code pairs consisting of vascular and interventional radiology procedural codes and the radiological S&I codes frequently reported with them, which was not surprising since that was the component coding design. The initial ‘Codes Performed Together’ filter was set at a 95% level, shortly followed by a 90% level of concurrence; the RAW, RUC, and targeted specialty societies are still working through the codes subsequently identified at the 75% level. Since this process began, many procedures previously reported together as code pairs have been identified by the RAW. They have been revised into ‘bundled’ codes by the CPT Editorial Panel and have been presented to the RUC in their new ‘bundled’ form for valuation. Examples include CT of abdomen and pelvis, inferior vena cava filter placement, lower extremity percutaneous revascularization, renal angiography, coronary angiography, thrombolysis procedures, and imaging-guided breast biopsy codes.

The code pair of the injection procedure for myelography and the myelography S&I is in the process of being revalued and revised, as are the code families of vertebroplasty and kyphoplasty. In the case of myelography, the bundling of lumbar injection and myelography radiologic S&I was a response to the RAW, with new proposed codes which were authored by the American Society of Neuroradiology (ASNR) and ACR and valued by the RUC earlier this year. So, in 2015, the codes for intrathecal contrast administration via lumbar injection and X-ray myelography radiologic S&I will be bundled into a single set of bundled codes when both procedures are performed by the same physician or other qualified healthcare professional (‘physician’) on the same patient on the same day. Additional radiology code sets that have been identified and are in line to be restructured include percutaneous biliary and renal interventions and the family of hip and pelvis X-ray codes.

It is important to stress that some, but not all, of the carotid and cerebral angiography code combinations were identified in the ‘Codes Performed Together’ screen in February 2010 by the RAW. However, bundling of just the specific codes caught in the screen is no longer sufficient since, in 2010, CMS mandated that all codes in a ‘family’ related to the specific codes identified in a screen must be acted upon, which may involve a simple resurvey but could involve a presentation to the CPT Editorial Panel for revision and subsequently to the RUC for valuing. This mandate avoids the potential for creating ‘rank order anomalies’ within families of codes.

A rank order anomaly refers to a situation where comparable procedures (not identical, but similar enough to be compared intuitively on the basis of magnitude estimation) demonstrate relative values that don't make sense—that is, they are out of RVU rank order. For instance, a non-contrast MRI of the brain should intuitively be valued lower than a contrast-enhanced MRI of the brain. If this was not the case, a so-called rank order anomaly would exist. Revising an entire family at once and presenting all of its codes to the RUC at one meeting allows direct comparison and valuation, minimizing the risk of creating rank order anomalies between newly-valued procedures and previously-valued procedures. This requirement is now a continual work-intensive challenge for those staff and volunteer physicians participating in the CPT/RUC process. Sometimes a single code identified in a RAW/CMS screen leads to wholesale revision and revaluing of a much larger number of related codes.

Consequently, while only a few code pairs of the entire diagnostic carotid angiography family were identified by the ‘Codes Reported Together’ screen, the entire family had to be revised as bundled codes that included the work of both the surgical and radiologic components.

In 2011, representatives from a group of specialty societies including the ACR, SIR, ASNR, and multiple vascular surgery and neurosurgery societies whose members perform these procedures collaborated to create a new carotid angiography code set.

The new carotid angiography code set was presented to and approved by the CPT Editorial Panel in February 2012. Once new codes are approved by the CPT Editorial Panel, they are added to the agenda for the next RUC meeting. CPT and RUC meetings are each held three times a year. For valuation at the RUC, the involved societies survey random samples of their members with a standardized RUC survey instrument which focuses on the amount of time a service requires, the intensity and complexity of the procedure, and relativity compared to a list of other services performed by the specialty. In addition, and perhaps the most important step, the specialty advisors need to compare/translate these survey data into a language the RUC can understand. This requires referencing a RUC approved master list, which helps to establish relativity of code valuation across multiple specialties. This master list, named the Multi-specialty Points of Comparison list (MPC), is the Rosetta Stone of the RUC process. It contains CPT codes either performed by or familiar to a diverse body of physicians, facilitating comparison of seemingly disparate codes. The process of establishing cross-specialty relativity can lead to contentious debates—especially when trying to compare procedure codes with evaluation and management codes—but these debates will theoretically draw the expertise of numerous specialties as the MPC list is purposely made to develop such comparisons. The societies then analyze the data and present their recommendations for professional RVU work values to the RUC. Direct PE inputs for non-facility settings were also presented to the RUC at the same meeting for valuation of technical component reimbursement, as is customary; that discussion is beyond the scope of this paper. The new carotid angiography codes were presented under this format to the RUC in April 2012.

ASNR and the other presenting societies argued that the existing component codes were previously evaluated separately and individually by the RUC; that each had longstanding RVU values in the CMS database; and that the new bundled codes should be equivalent in RVU value to the sum of previously reported code combinations. However, the member survey results for procedure time did not support this premise for all of the new codes. The RUC deliberated and decided on RVU values for each of the new codes, and sent their recommendations to CMS (see tables 1 and 2). For each of the new bundled codes the agreed RUC-recommended values were generally 10–15% less than the sum of the component codes under which the service would have been previously reported.

Table 1

Case examples 1–5 demonstrating coding, RVU balance, and typical reimbursement under the coding prior to 2013

Table 2

Case examples 1–5 demonstrating the changes to coding, RVU balance, and reimbursement under the revised Relative Value Scale Update Committee (RUC) recommendations

CMS published its RVU and PE decisions in the November 2012 Federal Register. CMS further decreased the professional RVU values for several of the new codes beyond the RUC recommendations. The primary rationale for the reductions was that there are ‘efficiencies gained’ when services are bundled, despite the fact that the work of angiography and the interpretation of the images had previously been considered separate and distinct entities and valued as such. The new code valuations and PE inputs took effect on 1 January 2013.

Component coding for diagnostic angiography

Coding prior to 2013

Prior to 2013 a diagnostic carotid angiogram would allow billing of both procedural (‘surgical’) and radiology S&I codes. Surgical codes would usually consist of one or more selective catheterization codes from the 36215–36218 series. The particular code would depend on whether the most distal branch vessel catheterized was a second-order or third-order vessel. Catheterization of additional selective branches would result in adding additional surgical codes to the overall charge. Radiology S&I codes would also be charged, selected from CPT codes 75650–75685 depending on the vessels studied (see table 3).

Table 3

Component coding for diagnostic angiography prior to 2013

New bundled code structure as of 1 January 2013

Each new code from 36221 through 36226 describes the entire procedure from vascular access through selective catheterization, angiography, post-procedure care, and interpretation of the images obtained. 36227 and 36228 are add-on codes reported in addition to one of the stand-alone base codes (see table 4). If bilateral examinations were performed, the second/contralateral examination can be reported with a modifier (−50).

Table 4

New bundled code structure examples for diagnostic angiography from 2013

Effect on professional reimbursement

The vast majority of carotid angiography is performed in the facility (hospital) setting. As such, the physician is reimbursed for the work component, which is often referred to as the professional component. The hospitals bill through separate payment systems depending on whether the patient is an inpatient or outpatient (the Hospital Outpatient Prospective Payment System or Inpatient Prospective Payment System, respectively). This represents the technical component.

Our analysis refers only to the professional component (PC) reimbursement changes incurred by the new code structure, typically appended with a −26 modifier.


We will review example 1 illustrated in the comparator tables 1 and 2.

Example 1, our most straightforward example, is that of a unilateral right internal carotid angiogram. Appropriate CPT codes for that procedure prior to 2013 would include 36217 (the surgical procedure code for catheterization and injection of a third-order aortic branch artery) and 75665 (the radiologic S&I code for a right carotid angiogram). These would combine for a total PC value of 7.60 RVUs. Actual reimbursement would vary on the basis of geographic location.

As of 2013, that procedure would be reported with a single CPT code, 36224. Comparing RVUs and imagining the patient would be in a system that uses the same reimbursement formula, this would result in a decrease of 15% in professional reimbursement (for Medicare). For purposes of comparison, we have not included a further reduction of 2% in payment due to the Federal sequester that took effect on 3 January 2013.

Add-on codes

CPT codes with the global period designation ‘ZZZ’ are ‘add-on’ codes and can only be charged to a patient in conjunction with a base code. In surgical procedure coding they indicate an extra component of a complex procedure, so as to capture the work involved in treating an additional or secondary area of coverage, such as an additional level of spine surgery or a second skin lesion biopsied in addition to a primary lesion. In the diagnostic carotid angiography family, add-on codes were used to indicate catheterization and contrast injection of an additional second or third order branch (or beyond) in addition to the more central vessel feeding these additional selections.

In the case of complex intracranial angiography performed to establish feeding and draining vessels of a vascular malformation, this typically resulted in a significant number of add-on codes being reported. In revising the diagnostic angiography code set, the CPT Editorial Panel elected to limit the number of add-on codes that a practitioner could charge for a procedure. This was designed to promote efficient care and forestall abusive coding practices. Thus, new add-on codes 36227 (selective catheter placement, external carotid artery, unilateral, with angiography and all associated radiological S&I) and 36228 (selective catheter placement, each intracranial branch of the internal carotid or vertebral arteries, unilateral) were limited to one usage per side for 36227 and two uses per side for 36228. Although we did not generate a complex example referencing this new standard, this add-on limitation may also have a tempering effect on reimbursement; specifics would be difficult to quantify due to previous variations in reimbursement for multiple codes charged during a single complex angiography procedure.


Policy changes at CMS have resulted in multiple radiologic and interventional codes being captured by a variety of screens that dictate review by the RAW. While a streamlined and simplified reimbursement system has obvious advantages and warrants support from the medical community, care must be taken that, in streamlining, appropriate value for services rendered is maintained. The ‘red flag’ for angiography was the ‘frequently reported together’ screen. Carotid and cerebral angiography use component coding which explicitly values individual activities within the same procedure. Almost by definition, these codes were frequently reported together. The previous scheme meticulously identified and valued the individual components discretely, suggesting that the new bundled codes should simply be the sum of these parts. CMS has not accepted this point of view and specialty society surveys have not provided supportive data. Nonetheless, multiple societies worked very hard to present a coherent argument to the RUC and were disappointed that the proposed RUC recommended values, which were already lower than previously, were even further diminished by CMS. On 11 September 2012, all of the specialties involved submitted a letter requesting clarification for the reduction of the AMA RUC recommended values. CMS has acknowledged receipt of the letter. Neurobased societies worked hard to clarify the additional complexity and work behind add-on codes for external carotid and superselective cerebral angiography, where one catheterizes the branches of the external carotid artery and/or the intracranial vasculature. While the financial impact of these valuation changes on carotid procedures is difficult to quantify in an authoritative manner due to the wide variation in specific procedures performed and continued decline in their utilization, the authors hope that the process by which these changes are reached is better understood through the effort of this communication.


The authors would like to thank Michael Morrow (AMA staff) and Erika Damico (manager of the MGPO IR coding team) for their manuscript review and input.



  • Contributors WDD and JAH drafted the initial manuscript. All other authors reviewed the draft and made substantial editorial contributions that resulted in revisions to the original manuscript.

  • Competing interests None.

  • Provenance and peer review Not commissioned; internally peer reviewed.