A fatal flaw undermines CMS’s Physician Quality Reporting System, resulting in a failure to improve quality and adequately reward radiologists.

In 2007, the Centers for Medicare and Medicaid Services (CMS) touted its newly created Physician Quality Reporting Initiative (PQRI) as an important first step in a journey that ultimately could transform the way health care is delivered and paid for in the United States. For 2011, CMS shows its continued commitment to this program by changing its name from the Physician Quality Reporting Initiative to the Physician Quality Reporting System (PQRS).

The initiative’s intent was to spur quality improvements and push Medicare closer to pay-for-performance reimbursement by creating financial incentives for physicians to voluntarily report key quality measures. It was anticipated that PQRS eventually would become mandatory and play an increasingly central role in federal health care reform efforts in the years ahead.

Regrettably, 3 years into the program, the reality of PQRS—for radiologists, at least—is a far cry from the initiative’s lofty goals. Many physicians now perceive PQRS involvement as time-consuming, clinically nonproductive, and financially futile.

Worse, a change in the way radiology quality measures are accounted for by CMS has unfairly punished participating radiologists and, more seriously, has raised fundamental questions about the validity and integrity of the data CMS is collecting.

The problem has been compounded by conflicting guidance from CMS regarding appropriate quality reporting. As a result, radiology practices and their billing partners remain uncertain about the documentation, reporting, and coding parameters currently required for compliant PQRS participation.

PQRS may ultimately prove to be an important vehicle for spurring quality and minimizing care variance across the health care system. But without fundamental program changes, greater program transparency, and decisive leadership from CMS, the initiative will remain an exercise in futility for many radiologists and the reporting that does occur will remain highly suspect.

Brian M. Barbeito

Willing Volunteers

Physicians associated with Mid-South Imaging and Therapeutics (MSIT), a 37-physician radiology practice based in Memphis, Tenn, identified the problems with PQRS, including the central flaw in CMS’s reporting protocol.

Shortly after CMS announced plans to implement PQRS in 2007, the group’s leadership initiated a strategy for full participation by the practice. Group leaders concluded that while start-up costs would likely exceed initial returns, pursuing PQRS would nonetheless be worthwhile—to prepare for future, perhaps more lucrative programs, to avoid potential punitive ramifications, and, most importantly, to reinforce a culture of quality compliance and reporting within the group.

Per direction from CMS and with the help of its outsourced billing provider, MSIT established coding and documentation protocols for reporting on three initial PQRS quality measures applicable to radiology. The group participated in the initial PQRS program, from July 1, 2007, through December 31, 2007, as well as the program’s second iteration, which ran through all 12 months of calendar year 2008. A total of 29 physicians from MSIT participated in both the ‘07 and ‘08 cycles of PQRS.

The maximum bonus that could be earned for the first 2 years of the program for successfully reporting 80% of applicable quality measures was 1.5% of total allowed charges for covered Medicare Physician Fee Schedule (PFS) services. CMS subsequently increased the bonus to 2% of total allowed charges for covered Medicare PFS services in 2009 and 2010. For 2011, the Affordable Care Act instituted a 1.0% incentive payment on allowed Medicare charges, but for years 2012?2014 the incentive payment will be lowered yet again to 0.5%. Beginning in 2015, the PQRS program will then become mandatory for eligible providers.

Payment Shortfall

PQRS bonus checks for the calendar year 2008 program—the first full year of the initiative—were received by MSIT in late fall of 2009, or nearly a year after the program concluded.

The payment levels came as a surprise to many. Over the entire two-cycle interval, the mean incentive bonus paid to all participating MSIT physicians amounted to just 0.36% of Medicare claims.

Because physicians, coders, and practice managers believed the group had largely complied with PQRS reporting requirements, there was considerable confusion after the bonuses proved to be far less than anticipated. MSIT consequently asked its outsource billing vendor to conduct a thorough audit to determine exactly how and where the reporting process may have gone awry.

For radiologists, the reality of PQRS is a far cry from the initiative?s lofty goals.

Line-Item Reporting

At the outset of the PQRS program, guidance from CMS for documenting quality measures was relatively straightforward: A specific CPT clinical service code had to be matched to a designated ICD-9 diagnosis code. Each match on a line-item basis constituted a reportable PQRS event. Assuming physicians successfully reported 80% of their total potential CPT-ICD-9 matches over the course of a year, the bonus would be paid.

MSIT’s billing vendor began the audit process by reviewing its claims scrubber, an automated application designed to ensure that all Medicare claims were compliant with the line-item PQRS methodology. No problems with the system were evident.

Claims Commingling

During the course of the audit, MSIT asked for and received from CMS a report that documented in detail each potentially reportable PQRS event associated with group physicians in 2008, as well as the claims events that were successfully reported. A review of the report by the billing vendor produced some startling information: CMS was showing a far greater number of potentially reportable events than had been identified by the billing company’s software.

The vendor dug deeper and eventually identified the origin of the additional opportunities flagged by CMS: In identifying potentially reportable events, CMS was commingling ICD-9 codes that were on the same claim but not necessarily performed on the same date of service. In effect, CMS was scouring claims for a code match, regardless of whether the diagnosis or date of treatment was part of the same service event. And while no evidence was found indicating that different service providers were combined or mixed to achieve a reportable event, at least one CMS representative claimed that such combinations were acceptable.

CMS’s methodology—highly questionable on its face—proved especially problematic for radiologists. Unlike other specialties, radiology groups often bill multiple service events by multiple providers on the same claim. This is primarily due to the high-volume, low-complexity, low-dollar-value nature of the specialty and the frequent delays groups face in receiving patient demographic information from the hospital. Once the information is acquired, backlogged service events are consolidated on a single claim and submitted to Medicare.

A Search for Answers

In light of the surprising shift from line-item PQRS reporting to claims-based reporting, all available guidance from CMS about how the agency intended to adjudicate PQRS claims was reviewed. What this effort revealed only compounded the growing confusion surrounding the program.

On March 15, 2007, CMS created FAQ 8273 in the PQRS section of its Web site. The question specifically asked: “Is the primary diagnosis the only diagnosis that is applicable to the quality measure being reported or will Physician Quality Reporting Initiative [PQRS] consider all diagnoses reported on a claim?”

CMS’s answer was unequivocal: “Only diagnoses linked to the paid service/procedure and quality-data line items are considered for Physician Quality Reporting Initiative [PQRS] reporting.”

However, when CMS published the “PQRI 2007 Reporting Experience” document on December 3, 2008, the agency stated that “When analyzing the 2008 data, we will look at all diagnoses submitted on the claim as well as the line-item diagnosis.” CMS went on to state, “We will further assess the degree of impact that this is likely to have. This modification to the analytics may result in QDCs [quality data codes] being considered valid based on including all diagnoses on the claim rather than limiting analysis to only line-item diagnoses.”

Oddly, the agency also stated: “However, our ability to use this modified analytic may be limited by the requirement to make PQRS determinations at the individual physician level where multiple professionals bill on the same claim for the same patient. Therefore, the business rule for submitting the correct diagnosis code associated with the QDC on the same line on the claim as the QDC will not change; however, since our systems may not capture more than one diagnosis code, we will revise the analytics to look at all diagnosis codes on the claim as well as the line-item diagnosis code.”

What CMS in essence seemed to be saying was that because it lacked the technological capability to identify the appropriate one-on-one match, it would instead look at any and all potential matches on a claim.

FAQ 8975, originally published on February 27, 2008, reinforced the change, yet seemed to hedge with the comment that “The line-item containing the QDC should point to the diagnosis that is relevant to the PQRI measure being reported.”

Setting aside troubling questions about the value of data gleaned by the commingling dates of service, the timing of the change by CMS could not have been more ill-conceived, given that it came at the end of the program year. Because the decision was made, in effect, retroactively, physicians and their billing partners had no way to correct claims already submitted. Even worse, the agency was still reaffirming its original line-item reporting guidance through FAQ 8273 just 3 months earlier, in September 2008.

As of March 30, 2010, CMS again asserted that it intended to look at the entire claim to find a quality measure match. The same question that appeared back in 2007 now had a different response—”Is the primary diagnosis the only diagnosis that is applicable to the quality measure being reported or will Physician Quality Reporting Initiative [PQRS] consider all diagnoses reported on a claim?”

This time, the answer was as follows: “All base claim diagnoses and valid quality-data codes are considered for the analysis of PQRI reporting?. The PQRI quality measure specifications identify the combinations of diagnosis and encounter codes making a claim eligible for each measure.”

Continuing Contradictions

Beginning in 2015, the PQRS program will become mandatory for eligible providers.

In the face of this confusing and conflicting guidance from CMS about whether line-item or claims-based reporting was the appropriate PQRS protocol, a representative of MSIT’s billing company contacted CMS. In an e-mail dated February 17, 2010, and addressed to Pamela Frederick, MSB, Director, Division of Ambulatory Care and Management, Quality Measurement and Health Assessment Group; Michael Rapp, MD, Director of the Quality Measurement and Health Assessment Group; and Sylvia Publ, CMS representative; the MSIT representative noted that she had seen the new guidance and was seeking confirmation if all diagnoses on the claim should be reviewed for eligibility, or whether reporting would remain at the line-item level.

Rapp responded that the inquiry was being sent to the support team for a more detailed response, but to answer the question, he stated: “Yes, we look at all the diagnoses (so-called base claim diagnoses) rather than only a diagnosis indicated by a pointer.” When the QualityNet Help Desk provided the same response in reference to FAQ 8273, the MSIT billing representative asked whether it was still a program requirement that the qualifying diagnosis and service code be from the same physician and same date of service.

The QualityNet Help Desk responded in writing on February 22, 2010, stating: “When information on a claim has the same TIN/NPI, beneficiary and date of service, the information will be considered in combination to determine denominator eligibility. Because date of service is considered, the answer to your second question is: No, CMS doesn’t look at all elements regardless of date of service; the date of service must be the same.”

In a CMS open forum call on March 10, 2010, CMS representatives Michelle Allender-Smith and Merri Lowry were asked a similar question: “So the ICD-9 from one provider is matched with a CPT code from another provider and that’s counting as a qualifying case against the provider?” Amazingly, given the answer provided just a few weeks earlier to the billing company, the representative responded by saying: “That would be correct. The QDCs—the base claim—is [sic] used and the QDCs that are applied on that claim would match up with the NPIs that are on there. That was a change to the program where the line-item diagnosis is no longer used.”

The billing company obtained a transcript of the open forum discussion and subsequently contacted the QualityNet Help Desk on April 12, 2010, seeking clarification. However, despite repeated requests, CMS did not provide further guidance regarding the issue of line-item versus claims-based reporting and stated only that no formal appeals process existed for the PQRS program as a whole.

Repairing PQRS

For MSIT, the PQRS program—entered into in good faith and with high expectations—has proven to be a costly and disheartening fiasco. Physicians who accurately reported quality measures were not rewarded for their efforts. Many were flagged for missed opportunities when, in fact, no such opportunities existed.

Beyond shortchanging the physicians who met program requirements, the problems identified with PQRS underscore a larger, more disturbing fact: The primary objective of the initiative—to improve care quality by offering incentives for evidence-based protocols—is gravely, if not fatally, undermined by allowing claims-based matches. It is impossible to understand how reporting that can involve multiple physicians and multiple dates of service can be in any way considered valid. Even more troubling, CMS appears either unaware of or unresponsive to this fact.

If PQRS were simply a well-intentioned but poorly executed experiment that didn’t work, the program’s deficiencies would not be critical. But PQRS is meant to be a centerpiece of reform efforts in the future. The initiative likely will become mandatory at some point and physicians may be classified and/or penalized based on their PQRS compliance or lack thereof. It is therefore imperative that CMS address and rectify the central contradiction in PQRS and provide far greater program transparency and improved communications going forward. Only then will the agency fulfill its legal and ethical obligations to physicians, patients, and taxpayers, and begin to repair its shattered credibility with the radiology community.

Brian M. Barbeito, MSHA, MBA, is chief executive officer of Mid-South Imaging & Therapeutics, PA.