April 2011
Feature Story

Anne Paxton

Surrounded by anecdotal evidence, nearly everyone seems to believe that the number of molecular pathology procedures performed has been expanding rapidly in recent years. But a central source of hard evidence has been lacking. The American Medical Association’s CPT (Current Procedural Terminology) coding for molecular testing has been geared to an era when these procedures were rare.

In March of this year, however, the AMA announced that after a lengthy development process it has finalized a major change: two sets of new CPT codes covering well over 90 percent of current molecular pathology procedures, and a framework to add new ones. For pathologists, it could well be the most significant change in CPT coding since 1991, when surgical pathology codes were revised.

With the comment period having closed April 15, the AMA plans to make any necessary revisions to the proposed codes, publish them in the CPT 2012 code set, and submit them to the Centers for Medicare and Medicaid Services, which will conduct its own review.

Mark S. Synovec, MD, who co-chaired the first AMA-CPT Molecular Genetics Workgroup in 2003 and now the second-generation AMA-CPT Molecular Pathology Coding Workgroup for the last one and a half years, describes the decision to reshape the CPT coding as a case of reaching critical mass.

“Every year vocal criticism increases toward the AMA for not dealing with molecular pathology coding. In October of 2009 we held a workshop at the annual CPT Advisors meeting to specifically address the growing molecular pathology problem, and heard loud and clear the time for a permanent CPT solution is now.”

Although the trend in molecular pathology has been clear for some time, it has taken two generations of AMA coding workgroups to reach this point. “Eight years ago we discussed the issue and came up with an interim solution, but now we’re moving to a permanent solution,” Dr. Synovec says.

The new codes will replace a cumbersome and complex system for coding molecular pathology procedures that uses “stacking codes” (83890–83914), which focus on methodology rather than specific analytes in conjunction with a set of identifying disease and mutation modifiers that add specificity to this code set.

Under the new system, there are two categories of codes: Tier 1 and Tier 2. The most frequently performed tests will fall under Tier 1 and will have analyte-specific CPT codes. Less common tests have been assigned to Tier 2, where they will be classified by nine resource level codes determined by the level of resources required for performance and interpretation. Tests not considered Tier 1 or Tier 2 will temporarily continue to be reported with the stacking codes, but that code set is slated for retirement in 2013.

The current CPT stacking codes (83890–83914) were conceptualized as a flexible system that would keep pace with the advancing technologies applied to procedural steps involved in molecular pathology services. The stacking codes will not be missed, Dr. Synovec believes. “They represent unbundling in its finest form. It’s like buying a car by getting a steering wheel, four tires, an engine, and so on, and then looking at the list of parts to determine what kind of car it is but not being able to tell.” When codifying molecular pathology procedures with the stacking codes, “a laboratory starts adding up the component parts, for example a DNA extraction, then a DNA amplification step, maybe multiple times, then detection step, and finally a professional interpretation. You add up all the different parts and submit a series of CPT codes.” However, when payers looked at the codes, they would have no idea what was being tested because no indication of the specific analyte being tested was linked to current CPT codes.

Payers have never been happy with the stacking code system, he says, but originally it wasn’t a high-volume problem for them. Insurers and providers tolerated the stacking codes because molecular pathology was considered such a rapidly evolving area. The CPT editorial process can take up to 18 months to develop or modify a code. “There was a fear that definitive code assignment early on would create codes that might already be antiquated by the time they were published,” Dr. Synovec explains.

The original Molecular Genetics Workgroup proposed an interim solution. “After investigating multiple ideas, we came up with the Genetic Testing Modifiers listed in appendix 1 in the CPT coding manual. The concept was based on a numeric-alpha two-digit code that was appended to the meth­odology codes.” Genetic Testing Modifiers were developed for use in conjunction with CPT and HCPCS codes to provide diagnostic granularity of service, enabling providers to submit complete and precise genetic testing information without altering test descriptors. That way there would be some granularity, Dr. Synovec says, so payers would know what was being tested. “The problem was that a limited number of potential modifiers was available, and we knew we would eventually run out of different modifiers for each genetic aberration.” The insurance industry never embraced this temporary solution; hence providers were reluctant to use the modifiers.

Amid the continued pressure to come up with a comprehensive plan, the AMA began developing the two-tier solution, which was proposed originally in a white paper by the Association for Molecular Pathology. In an unusual step, the AMA released the proposed codes for stakeholder feedback in March. It’s the first time the AMA has made an entire set of proposed codes available for review at this stage, Dr. Synovec points out, but it’s part of a more open CPT system. “We realized this is such a dramatic departure from the way molecular laboratories are used to coding for these services. Therefore we wanted to ensure that the proposed codes were made available early so people could have a chance to review and provide necessary comments on the codes and the new system.”

Asked how many molecular pathology procedures are being performed, Dr. Synovec says, “We can’t tell. Right now, with the stacking codes, it’s hard for us to get a grasp on the volume of molecular pathology. You can easily submit 10 or more CPT codes for a specific analysis. We know it’s growing just because of what we’re seeing anecdotally. But we have so little data on the range of the analytes being tested and the volumes for each type of procedure.”

Many organizations worked collaboratively to make the revised coding system a reality, Dr. Synovec says. In addition to laboratory industry groups like AdvaMed and ACLA, representatives from the CAP, the ASCP, the AMP, and other groups in organized medicine, as well as the CMS and the insurance industry, were responsible for the work of hammering out a new system. The two other co-chairs of the AMA-CPT Molecular Pathology Coding Workgroup are from the insurance industry, Dr. Synovec says. “We were very lucky to have a group of very hard-working people. This is not a glamorous job; it’s very laborious. It’s clear that these participants want to build a code set that works for all.”

Insurers were among those most discontented with the existing code stacking system. “If you asked five different labs how they were coding a procedure like BCR/ABL, it wouldn’t be surprising if you got four or five different answers that would equate to different payments.” In some cases, he says, insurers balked at paying for the molecular pathology procedures. “Even though some payers may have had an idea of what test was being provided to their beneficiaries, they realized that payment should not be based on the way the lab coded the test, but rather on the procedure being performed.”

Much more consistency is expected under the new coding system. The relative value unit, or RVU, calculation, which will affect payment rates, is a separate process under AMA procedures and is being carried out by the AMA Relative Value Update Committee where the CAP will be involved in the valuation process.

Molecular infectious agent codes were not included in this round of changes. “Currently there are ways to code most of the molecular procedures within the microbiology section of CPT, so the bulk have already been addressed in some form,” Dr. Synovec says. HPV testing, for example, is one of the high-volume tests for which the 87470–87801 series of codes was developed in 1997 for testing of primary source specimens for infectious agents.

“At that time, laboratory analysis was primarily culture-based, but there were expanding technologies for direct infectious agent detection other than the typical culture. These methods use a direct specimen and test for a virus or bacteria on either antigenic—for example, immunofluorescent or ELISA—or nucleic acid-based platforms. So four code sets were developed for immunofluorescence, enzyme immunoassay, direct optical observation, and for nucleic acid probe technologies.”

It was somewhat accidental that those molecular infectious agent codes were developed, he adds. “CPT wouldn’t necessarily have had these nucleic acid codes, but since they were doing the revision for primary source microbial identification anyway, they were worked in.” In CPT 2010 several additional codes were added for molecular techniques used in conjunction with microbial cultures. An example: “87149 Culture, typing; identification by nucleic acid (DNA or RNA) probe, direct probe technique, per culture or isolate, each organism probed.”

The Molecular Pathology Coding Workgroup is not hoping that the new coding system will drive technology one way or the other, he emphasizes. “It’s true that once you come up with a single payment for a single analyte, people will try to figure out how to do it faster, cheaper, and better. That’s the incentive model we’ve inherently built, but it wasn’t our primary goal.”

The first 20 CPT codes that were developed for Tier 1 likely account for 80 percent of the molecular pathology procedures performed, Dr. Synovec notes. “This analyte-specific list includes procedures like cystic fibrosis testing. But for the lower-volume tests, we knew there was no way we were going to come up with a CPT code for every conceivable gene. We just don’t have enough codes. So the second tier, with its series of nine codes that escalate in the amount of resources and interpretive work, is something that creates a higher level of granularity, but not absolute granularity.” He likens this code set to that of the surgical pathology codes (88300–88309) but notes a significant difference: “In contrast to surgical pathology, each Tier 2 code will contain a list of specific analyte procedures. The provider will not be allowed to self-assign procedures not included in each list.”

Molecular pathology multi-analyte, or MMA, procedure codes remain to be developed, and the AMA workgroup is discussing the issue. “We [the workgroup] don’t have the resources to thoroughly evaluate every existing MMA to justify whether the procedure has sufficient clinical applicability to fulfill the criteria for CPT category 1 code placement,” Dr. Synovec says. As part of the process, the AMA is inviting laboratories with MMAs to submit specific coding proposals to the AMA for review. As far as the multi-analyte assays are concerned, this will be “an ongoing process that will continue after the workgroup is retired,” Dr. Synovec says, noting that when the majority of the workgroup’s work is done, the group will be disbanded. Its last project will be to submit a proposal to retire the stacking codes.

A yet-to-be-developed unlisted molecular pathology code is termed “Tier 3,” he says. It will replace the stacking codes for all analyses not placed in Tier 1 or Tier 2. The AMA-CPT editorial process also allows for a category 3 approach for proposing CPT codes for “emerging technologies.”

Once the workgroup is disbanded, a standing molecular pathology advisory panel serving the CPT editorial panel will be created to provide expertise in managing this code set. “Yet again, the nuances of molecular pathology have necessitated a different system than CPT has previously used,” Dr. Synovec says.

While the specific numeric codes are still in development, the code system is solidly in place. “This molecular pathology system of coding is pretty hard and fast, but we plan to review any suggested changes received during the comment period to ensure the accuracy of the service descriptions,” he says. Another thing laboratories can count on: “There’s a whole new way of coding for molecular pathology that’s going to begin Jan. 1, 2012, and if you’re involved in molecular pathology, you’d better be aware of it—and be prepared.”