Washington State Sees Increase in Appropriate PCI After AUC Introduction
Following publication of the first Appropriate Use Criteria (AUC) for coronary revascularization in 2009, an increase in procedural appropriateness mirrored a decline in PCI for patients with stable disease, according to an analysis of data from Washington state, published online May 28, 2015, ahead of print in Circulation. However, the improvements were limited to a small number of hospitals, which the study authors say points to a need for ongoing education and assessment efforts.
“We observed declining proportions of elective PCI classified as inappropriate or with insufficient preprocedural assessment for appropriateness assessment, suggestive of temporal improvements in the quality of preprocedural assessment and patient selection for elective PCI,” say Steven M. Bradley, MD, MPH, of the VA Eastern Colorado Health Care System (Denver, CO), and colleagues.
They examined data from the Washington State Clinical Outcomes Assessment Program (COAP) on all patients who underwent PCI between 2010 and 2013. All procedures were classified according to the 2012 updated AUC.
Fewer Inappropriate Procedures
Overall, the number of PCI cases decreased by 6.8% between 2010 and 2013 (from 13,267 procedures to 12,367), driven by a 43% decline in elective PCI for stable angina (from 3,818 procedures to 2,193). Rates of PCI for STEMI and unstable angina remained largely unchanged, while PCI for NSTEMI increased by 17% (from 2,747 procedures to 3,226).
When these temporal trends were compared with treatment patterns between 2006 and 2009, the changes were significant for both stable (P = .03) and unstable angina (P < .001). Meanwhile, patient demographics, risk factors, or comorbid conditions barely changed during the same time period.
Among more than 50,000 patients who had PCI between 2010 and 2013, the procedure was considered appropriate in 79%, uncertain in 8%, inappropriate in 4%, and unable to be classified by the AUC in 9%. The proportion of elective PCIs classified as appropriate increased from 26% to 38%, while the segment classified as inappropriate decreased from 16% to 13% (P < .001 for both). The researchers also found that preprocedural assessment became more complete over time, as evidenced by a decline in the proportion of patients who did not have adequate assessment before elective PCI.
However, not every hospital curbed its rate of inappropriate PCI. Instead, only those in the highest tertile had a reduction (from 25% to 12%; P = .03), whereas those in the lowest tertile actually had an uptick in inappropriate PCI (from 12% to 20%; P < .01).
“Although these findings suggest statewide improvements in preprocedural assessment and appropriate patient selection for elective PCI, evaluation at the hospital level demonstrated these improvements were limited to a minority of hospitals, while some hospitals even saw temporal increases in the proportion of inappropriate PCI,” the study authors write.
Dr. Bradley and colleagues say their study is the first to assess trends in appropriateness measures of PCI quality. It also provides some insight into a concern among many in the interventional community, namely attempts to “game” the system by exaggerating the severity of illness to make patients look more appropriate for PCI.
“Although we did observe a small increase in the number of PCIs performed for NSTEMI and unstable angina that may in part reflect an aspect of gaming, this was accompanied by a much larger decline in the use of elective PCI,” they write. “Furthermore, [this decline] was steeper in the years following the onset of appropriateness assessment. These findings suggest the trend in lower proportions of inappropriate and insufficiently assessed PCI observed in our study largely reflect improvements in patient assessment and selection processes, rather than an artificial change due to upcoding of the clinical indication for PCI.”
Among the factors that may have contributed to the positive shifts, the authors cite the COURAGE trial as possibly encouraging less utilization of elective PCI, as well as:
- Feedback and quality improvement efforts occurring through COAP and the National Cardiovascular Data Registry CathPCI registry
- Publications highlighting patterns in PCI appropriateness assessment from national and regional PCI registries
- Publication and attention given to the AUC themselves, including attention by healthcare payers
Ensuring Appropriate AUC Use
In an editorial accompanying the study, Gregory J. Dehmer, MD, of Baylor Scott & White Health (Temple, TX) and Manesh Patel, MD, of Duke University Medical Center (Durham, NC), say the data add to the “growing evidence base” regarding the benefits of the AUC. In addition, the findings demonstrate that the AUC are readily adaptable for use as clinical decision support (CDS) tools.
They further note that last March, Congress passed the Protecting Access to Medicare Act of 2014, which includes a provision mandating that professionals consult with AUC through a CDS mechanism for Medicare patients who receive advanced imaging beginning in January 2017.
“This program will collect information about practice patterns but importantly will not tie payment to the appropriate use score for individual cases,” Drs. Dehmer and Patel write. “This is an extremely important part of this legislation as some states and payers have started using the AUC to adjudicate payments, a function the AUC were never intended to perform.”
Rather, the AUC were intended to “guide practice and prevent or reduce prior authorization mechanisms,” they assert, adding that an updated version “is currently under development and will hopefully address some of the constructive criticisms directed at the earlier versions.” Use of the AUC concept is increasing, they report, and it now has been adopted by a variety of disciplines to address such things as common orthopedic procedures, skin cancer surgery, and placement of Foley catheters.
Parsing the Good and Bad
“One of the good things that has happened with the application of the AUC is that there is better scrutiny on the part of physicians themselves on the indications for cases,” said Ajay J. Kirtane, MD, SM, of Columbia University Medical Center (New York, NY), in a telephone interview with TCTMD. “It’s also very reassuring to see that they did not find ‘indication creep’ or gaming of the system.”
But in order to put these observations in perspective, Dr. Kirtane said, it is also important to note that many asymptomatic patients are not being sent to the cath lab and to remember that an initial goal of the AUC was to focus on underuse as well as overuse of PCI.
“What we are starting to see now is situations where physicians will not treat patients clinically because they are being prescriptively told they shouldn’t because of the AUC,” he said. This issue is being addressed, however, by nomenclature changes such as the change from “uncertain” to “may be appropriate” and from “inappropriate” to “rarely appropriate,” he added.
“It’s clear that the AUC are not 100% and that there are going to be clinical scenarios, including patient preference and the like, where it becomes entirely clinically appropriate to perform PCI even though the rating is ‘inappropriate’ or ‘rarely appropriate,’ and this is an issue for payers as well,” Dr. Kirtane cautioned. “The ideal rate of ‘inappropriate’ is not 0%. A 0% rate really should lead one to question whether there is underuse or gaming of the system.”
1. Bradley SM, Bohn CM,
Malenka DJ, et al. Temporal trends in
percutaneous coronary intervention appropriateness: insights from the clinical
outcomes assessment program. Circulation.
2015;Epub ahead of print.
2. Dehmer GJ, Patel MR. The use of appropriate use criteria are increasing, but what are their effects on medical care [editorial]? Circulation. 2015;Epub ahead of print.
- Drs. Bradley, Dehmer, and Patel report no relevant conflicts of interest.
- Dr. Kirtane reports receiving institutional research grants from Abbott Vascular, Abiomed, Boston Scientific, Eli Lilly, Medtronic, St. Jude Medical, and Vascular Dynamics.