When cath lab directors began reviewing quarterly reports from the ACC-NCDR CathPCI Registry earlier this year, they might have been surprised to find a new benchmark: the ranking of PCI procedures as appropriate, inappropriate, or uncertain.
Shortly afterward an article appeared in the July 6 issue of the Journal of the American Medical Association (JAMA) showing that, nationwide, about half of elective PCIs reported to the NCDR database could be deemed appropriate when judged against appropriate use criteria (AUC) published in 2009. Another 38 percent fell in to the “uncertain” category, and just under 12 percent were deemed “inappropriate.” (Among patients with acute indications, PCI was considered appropriate nearly 99 percent of the time.)
A wave of negative media coverage followed. Nearly all reports focused on the percentage of inappropriate elective cases. In some instances journalists even lumped “inappropriate” cases together with “uncertain” cases to paint a picture of stent overuse. In September, Consumer Reports joined the fray with a cover story on protecting the heart that included the subtitle “Angioplasty: What your doctor might not tell you.” The article cited both the JAMA study and the high-profile allegations of stent overuse by an interventional cardiologist in Maryland.
“It does seem like interventional cardiologist are a bit in the bull’s-eye these days,” said Gregory J. Dehmer, M.D., FSCAI, a SCAI past president and director of cardiology at Scott & White Healthcare in Temple, TX. “We just have to do the best we can to try to educate the public and the media as to what we’re trying to do. The AUC represent an effort on the part of interventional cardiology to make sure we’re always doing things for the right reasons. You don’t see any other professional organization that has gone after this like the interventional community has.”
Also lost in the uproar was acknowledgment that the NCDR considers the new AUC analysis to be a test metric that, by definition, will put cath labs on a learning curve while construct validity and reliability are assessed. Test metrics are not designed for public reporting and should not be viewed as a definitive scorecard on cath lab quality.
“We need tools to help us measure our improvement, and that’s all this is,” said SCAI President Christopher J. White, M.D., FSCAI, who is also system chairman for cardiovascular diseases at the John Ochsner Heart & Vascular Institute in New Orleans, LA. “We’re going to be better physicians and better hospitals, and our patients are going to receive better and safer care. It’s just going to be a little bit painful at the beginning.”
Opportunities for Improvement
The new AUC scoring system represents a clear opportunity for quality improvement on two fronts. In some instances, analyzing cases that are fl agged as inappropriate may reveal problems in clinical decision-making that need to be addressed in a systematic way using, for example, the tools available in the SCAI Quality Improvement Toolkit, which can be downloaded from www.SCAI.org/QIT.
In many cases, however, improving AUC scores may simply be a matter of better communication, documentation, and coding. By filling out the NCDR data sheets more completely—and with AUC in mind—hospitals could see a marked reduction in the number of cases ranked as uncertain or inappropriate.
“We have to make sure we give the PCI database the information it needs to accurately refl ect what’s going on,” said Tom Cook, system coordinator for cardiology quality assurance and performance improvement at Ochsner Medical Center. “Communication is number one. Documentation is right behind it.”
SCAI has compiled tips, culled from discussions with a number of SCAI members and their cath lab teams, are likely to help any cath lab to improve its AUC scores.
1: Study your NCDR quarterly report.
This is the fi rst step to improving AUC scores. Review all cases classified as inappropriate or uncertain to determine why each case was scored the way it was. This type of analysis can be eye-opening, revealing specific targets for improvement at each hospital.
2: Know the AUC.
In order to earn a good AUC score, it’s important to be familiar not only with the appropriate use criteria but also with the most common reasons cases are classifi ed as inappropriate. In the recent JAMA article, Chan et al. found that nearly 89 percent of all inappropriate elective procedures fell into five clinical scenarios (see the sidebar, Common Scenarios for Inappropriate Elective PCI). Just over 82 percent could be described by three scenarios involving, in various combinations, one- or two-vessel disease in combination with few symptoms, a low-to intermediate-risk stress test, and inadequate antiischemic medications.
3: Re-introduce yourself and your team to the NCDR data forms.
Make sure coders understand all of the definitions used in correctly filling out the NCDR data sheets, so that cases are properly categorized.Ask all of the interventional cardiologists in your practice to review the NCDR forms so they know what information the chart abstracter needs. Don’t assume that coders can read between the lines and come up with the correct clinical picture. Some are very experienced and knowledgeable, but others have less expertise—and all are very busy.
4: Become class-conscious.
It ’s important that the pre-procedure work-up or post-procedure case summary includes a clear and concise description of the severity of the patient’s angina, including the Canadian Cardiovascular Society class. “As clinicians we tend to blur the clinical distinction between Class II and Class III, but in the AUC, it makes a big difference,” said James Blankenship, M.D., FSCAI, who is chair of SCAI’s Advocacy Committee and director of cardiology at Geisinger Medical Center in Danville, PA.
5: Open the medicine cabinet.
Be thorough and complete in listing antianginal medications. This can nudge a case from the uncertain category into the appropriate category by showing that the patient is on maximal medical therapy. Specify whether the patient is taking long-acting nitrates, beta blockers, and/or calcium-channel blockers.
6: Embrace the stress test.
The results of noninvasive testing should be stated concisely and clearly. It’s not enough to say the test was positive, negative, or indeterminate. Also indicate whether there is a low, moderate, or high likelihood of future ischemic events.
7: Let the outside in.
Many patients are referred for PCI after having an initial clinical work-up elsewhere. Information on noninvasive stress testing, medications, and clinical history could be housed at another hospital, physician’s office, or imaging center. It’s essential that this information be included in the cath lab’s records and be available to chart abstracters. Otherwise, the NCDR data sheets will be incomplete.
8: Multitask during PCI.
No one likes to finish a case and then sit down to fill out a long form. Instead, assign a tech to fill out the NCDR form during PCI, taking advantage of pauses in the action to clarify questions such as percent stenosis or patient risk level. “There are always slow spots, such as when a balloon is deflating,” Dr. Blankenship said. “And the information is never more readily available than during the case.”
9: When in doubt, turn to FFR.
When angiography reveals a stenosis of 70 percent or less, it’s important to document the lesion’s clinical significance. Inclusion of flow reserve (FFR) ratios helps to do that. This is particularly true if the results of noninvasive stress testing are indeterminate or missing.
10: Get the word out.
Once you’ve identified the key steps your cath lab should take to improve AUC scores, spread the word. Use a variety of tools, including wall posters, checklists, and cath lab protocols. Incorporate reminders into the electronic medical record if possible.
Interventional cardiologists ultimately hold the key to improving AUC scores, stressed Dr. White. “Physicians have to be responsible for the quality of the data that’s recorded,” he said. “You have to take ownership of that. You can’t expect your coders to save you.” At the same time, no one expects all cases to ranked as appropriate.
“Everybody is going to wind up having a few cases that, by the criteria that now exist, are going to be adjudicated as inappropriate,” Dr. Dehmer said. “That doesn’t mean it was bad medical care. It just refl ects the limitations of the AUC. Clinical medicine has so many twists and turns, you can’t pigeonhole every case.”
Common Scenarios for Inappropriate Elective PCI
Chan et al. found that nearly 89 percent of all inappropriate elective procedures fell into the five clinical scenarios listed here.The percentage provided after each descriptive scenario indicates how much of the “inappropriate” elective total was comprised of cases fitting the scenario. The vast majority of the cases deemed inappropriate were described by the first three scenarios.
- One- or two-vessel CAD, no proximal LAD involvement, no prior CABG, CCS class I or II, low-risk stress test, no/minimal anti-ischemic therapy (39.6%)
- One- or two-vessel CAD, no proximal LAD involvement, no prior CABG, asymptomatic, intermediate-risk stress test, no/ minimal anti-ischemic therapy (24.5%)
- One- or two-vessel CAD, no proximal LAD involvement, no prior CABG, asymptomatic, low-risk stress test, no/minimal anti-ischemic therapy (18.3%)
- One or more stenoses in non-CABG territory, all bypass grafts patent, CCS class I or II, low-risk stress test, no/minimal antiischemic therapy (3.4%)
- One or more stenoses in non-CABG territory, all bypass grafts patent, asymptomatic, intermediate-risk stress test, no/ minimal anti-ischemic therapy (2.9%)
Source: Chan PS et al. JAMA. 2011;306(1):53–61.