Welcome, and thank you for listening to this recording. My name is Cornelia, but I hope you all know me as Connie. I'll be the narrator of this session, which was created with the help of Kim Lavin, the CAF PCI Registry Science Lead. This session was designed to give participants a sneak peek of upcoming enhancements and improvements planned for the CAF PCI Registry in 2021 and 2022. In collecting all the information we wanted to share, and because several updates were in flight or linked to previous efforts, there are three basic sections. We'll review many of the updates that took place this year, look at the plans which are driving quality improvement onward, and finally discuss what is coming that will elevate registry reporting upward and broaden registry value. Kim and I hope you learn many things through this presentation, and that you feel comfortable discussing some of the major initiatives and registry updates with your clinical teams, and can articulate at least one rationale for registry changes. Let's begin. Everyone is very busy, and we know registry updates and enhancements are sometimes overlooked or missed. I'll begin with a look at what has happened so far. In 2021, we saw a host of metric refinements, dashboard enhancements, and the debut of many new things, such as the cost of care report, COVID-19 data collection and reporting, and a brand new hierarchical risk model for mortality. Since the launch of version 5 and the application of the 2016 appropriate use for coronary revascular ization of patients with acute coronary syndromes, and the 2017 RE-BASC AUC for patients with stable ischemic heart disease, or SIHD, Kim and I have been working with key members of the AUC workgroup to address gaps in the AUC. As you know, the appropriate use criteria are unable to account for all the elements that make up each clinical scenario or consider each patient's individual complexities and assign an AUC indication for PCI. However, the benchmark of unclassifiable procedures between the previous AUC and the updated AUC raised concern early on. In metric 30, PCI procedures not classifiable for AUC reporting, the 2018 quarter 1 benchmark was 3.51%. And one year later, with four quarters of version 5 patient records, and in applying the updated 2016 and 2017 AUC, the benchmark for metric 30 was 10.29%. That's a huge jump and a lot of unclassifiable procedures. Together with members of the AUC workgroup and the analytics center, algorithms and distinct patient populations were re-evaluated. These meetings led to strategic updates, which have been implemented, and these include correcting the algorithm to recognize when coronary circulation dominance is reported as left, and that an IFR value of zero is representative of a hemodynamically significant lesion, and associating a reported IFR or FFR value to the vessel being evaluated when determining the AUC indication. As well, accurately identifying a CABG occurred prior to PCI, whether that was a historical event or during the episode of care, and also excluding procedures for patients with a prior cardiac transplant. Patients with 100% stenosis were assigned an FFR value of 0.7. And assigning the appropriateness of PCI from table 3.1, when the AUC rating is the same across multiple indications, despite a missing variable, such as stress test. PCI on patients undergoing renal transplant or a percutaneous valve procedure, which met more than one AUC indication, are now assigned the highest rating within table 3.1. And patients with contraindications to three or more anti-antrenals are removed from AUC reporting. These adjustments have impacted hospital performance in the AUC metrics. In this example, 2019 quarter one performance showed over 14% of procedures for this site were unclassifiable. But in 2020 quarter four, that dropped to just 10.89%. As procedures were accurately assigned an indication or removed from the AUC evaluation altogether, the benchmark for metric 30 PCI procedures unclassifiable for AUC reporting has dropped. In 2019 quarter one, it was over 10%. And by 2020 quarter four, it dropped to just over 8%. Additionally, and this is a great time to point out, the PCI volume metric comparison was reintroduced to the report. If you caught the Quality Summit presentation, Troubleshooting Your Way to Accurate AUC Indications, you heard Kim discuss the value of this important comparison. Another significant enhancement was updating metric 22 and 46, which is the median post-procedure length of stay for patients with STEMI and patients with uncomplicated STEMI, respectively, which calculate the length of stay. In version 4.4, the determination of a day was strictly based on a calendar day. Consequently, the median length of stay appeared very consistent from quarter to quarter as you can see in these trend graphs. To more accurately account for the total time of hospitalization, the data variable time was added to the discharge field. Updating metrics 22 and 46 to consider the discharge time was completed in April. And now we have a more accurate view of the median post-procedure length of stay for patients with STEMI and additionally for patients with uncomplicated STEMI. There's a dynamic trend graph as the median length of stay is impacted by each hour patients are hospitalized. This metric value is reported to a two-decimal precision and the benchmarks for these measures tell a whole new story. In metric 22, the benchmark was 1.75 days and now we see that it is really 2.17 days. And in metric 46, the benchmark shifted from 1.63 to just under two days at 1.99. The metric drill down for these measures can help hospitals evaluate trends between hospital units, providers, and specific patient types, gaining insights into resource utilization and potentially identifying opportunities for improving quality or efficiency. The cath PCI registry steering committee reviewed participant concerns about the patients included in the denominator of metric two, PCI procedures with positive stress or imaging studies and agreed that stress testing was not safe for patients with severe aortic stenosis and proposed to remove these patients as well as patients who were pre-op for solid organ transplant surgeries from the metric denominator. Select hospitals with busy TAVR and organ transplant programs had been acutely affected by the inclusion of these patients in the metric two denominator. They in particular saw a significant shift in performance as demonstrated here between 2020 quarter one and 2021 quarter one, where the percentage of PCI procedures with positive stress or imaging results is now up to 45% from 26.79%. As well, the trend line, once well below the benchmark for four straight quarters, now hovers closer to the benchmark and provides more insights into the patients who would benefit and are appropriate for stress testing. An exciting enhancement for metric 39, PCI in-hospital risk adjusted acute kidney injury for all patients is the addition of a column reporting the contrast volume used during the procedure. In the effort to prevent acute kidney injury following cardiovascular procedures, we know limiting contrast use is the first and most important step. We hope providing each patient's contrast dose along with other key procedure variables and the PCI operator help support your quality improvement efforts focused on minimizing acute kidney injury. Another exciting and long anticipated drill down enhancement was that of being able to pull four quarters of patient metric data when first navigating to the drill down. Here you see the option R4Q in the year and quarter display window. When selected and then clicking on the retrieve button, all patients in this metric for the four quarters defined by the ending timeframe on the dashboard will populate to the drill down. Export the drill down to Excel, where you can filter and sort the information more easily. This will go a long way towards speeding up the time needed for all the data analysis you are doing. For the first time ever, the NCDR sought to provide registry participants with insight into their organization's cost of care for PCI by collaborating with FairHealth, a nonprofit group that maintains a private claims repository. The majority of registry participants received a PCI cost of care study report for their facility as well as several artifacts which provided insights into the place of service, cost of PCI by procedure type, elective or non-elective, and the actuarial risk score of their patient clientele. These data in comparison to regional and national data aim to help clinicians, administrators, and healthcare leaders better understand the total cost of care for PCI at their facility. We hope these reports assist in identifying opportunities to address cost of care while also optimizing outcomes through quality improvement efforts. The development of the auxiliary data collection COVID-19 data set has opened the door for registry participants to learn more about their patients during this unprecedented time and help inform how the coronavirus is impacting patient care, treatment strategies, and affecting patient outcomes. Registry participants collecting this data can navigate to the detail lines of the dashboard to see the aggregated results for the COVID data points and utilizing the brand new first ever patient drill down available from a detail line metric, use the drill down data to understand performance in the subsequent COVID-19 detail lines. In late 2020, Kim convened the CAF PCI registry risk standardized mortality work group. The goal was to update the old risk adjusted mortality model to a hierarchical one and to evaluate which version five data elements or variables were predictive of mortality. The findings of this work group were presented at ACC's 2021 scientific session. In summary, procedural urgency, cardiovascular instability, and level of consciousness after cardiac arrest were most predictive for in-hospital mortality as shown here in table one. The full model performed well with excellent discrimination. The C index was 0.943 in the validation cohort and there was good calibration across different clinical and procedural risk cohorts as demonstrated in figure one. The bedside risk score model contains the variables with the strongest association with mortality and that in combination explained greater than 90% of the risk model as demonstrated in table two and figure two. The median hospital risk mortality rate was 1.9% and ranged from 1.1 to 3.3% as demonstrated in figure three. We are currently absorbing the SAS code and anticipate adding the new hierarchical mortality outcome measures to the 2021 quarter two report. The implementation plan for the new hierarchical mortality model is to continue reporting metrics one, 18, and 19 until 2021 quarter four. This will ensure that participants and any stakeholders in or outside of the hospital who are monitoring these measures have a line of sight on performance through all of 2021. The new hierarchical mortality outcomes measures will include metric 48, which will look at all patients, metric 49, which will look at only those patients without cardiogenic shock and or cardiac arrest. Metric 50, looking at patients without STEMI and metric 51, looking at STEMI patients without cardiogenic shock and or cardiac arrest. And because traditional risk adjusted outcomes are so good at demonstrating the observed event to the expected event rate for your patient population, the risk adjusted mortality metrics for these four patient populations will be added to the detail lines. I see I have a typo under the last bullet. So sorry about that. All of the detail lines should read PCI in-hospital risk adjusted, not standardized for the specific patient population. Of note metric 49, PCI in-hospital risk standardized mortality for patients without cardiogenic shock and or cardiac arrest will assume NQF endorsement and be added to the performance measure set of the report. Such a busy year and we're not done yet. Let's take a look now at the plans which are driving quality improvement onward. The cath PCI registry steering committee has developed and approved new metrics to report on chronic total occlusions and radiation safety. The FDA has approved the cath PCI registry as a post approval study vehicle for the shockwave intravascular lithotripsy IVL device. And the provider dashboard has a new name. And we are loving the new found ability to add drill downs to detail lines. The goal and what the AUC promote is the performance of PCI when the benefit of the procedure outweighs the potential risks to the patient. PCI for chronic total occlusions have seen success rates improve as rapidly evolving treatment strategies, better equipment, and operator skill tilt the odds in favor of undergoing the procedure. Still, research shows that PCI for CTO lesions generally require more resource resources and continue to carry a higher risk of complications and increased resource utilization and the incremental costs associated with complications such as in-hospital mortality, post-procedure stroke, cardiac tamponade, and urgent coronary artery bypass graft surgery result in an increased length of stay and all of these drive up the cost of the hospitalization. In the registry, we see that of the lesions with a hundred percent stenosis, twenty percent are identified as chronic total occlusions. The data also suggests that while high success rates can be achieved at experienced centers, the overall success rate is low. For these many reasons, the Steering Committee has developed two new executive summary metrics evaluating PCI for CTO. The first CTO metric will report on technical success. The denominator is the count of CTO lesions treated with PCI and the numerator will be the count of CTO lesions with a post intervention stenosis of less than 50% and TIMI 2 or TIMI 3 flow. The next executive summary metric will report the procedural success achieved for CTO interventions. The denominator is PCI procedures with CTO and the numerator is the count of procedures where all treated lesions have a post interventions stenosis of less than 50% and TIMI 2 or TIMI 3 flow without having any of the following major adverse events. Coronary artery perforation, cardiac tamponade, myocardial infarction, stroke, discharge status of deceased or an urgent or emergent or salvage cabbage procedure post PCI. Finally, new detail lines providing the median CTO lesion length, the percent of CTO procedures with cumulative erykerma greater than or equal to 5 grays or greater than or equal to 5,000 milligrays. The median cumulative erykerma in grays and in milligrays and the median contrast volume will be added to the section for PCI procedures with or without diagnostic calf where a CTO lesion was treated. These Genova detail lines and metrics will open a whole new area of focus for the calf PCI registry and we are excited to see how they inform and influence us all. Now let's talk about radiation. The 2018 expert consensus document on the optimal use of ionizing radiation in cardiovascular imaging best practices for safety and effectiveness spends considerable time outlining how radiation can harm people. In health care we frequently discuss the risk to benefit ratio of doing a procedure or intervention for a patient and that is never more important than when radiation is part of the patient intervention be it testing or the actual procedure. Radiation has molecular level detrimental effects on exposed human tissue with the potential to injure both patients and medical personnel. PCI guidelines have noted that if the procedural erykerma dose is greater than 5 grays the patient is at risk for skin injury and if greater than 10 grays they are at risk for serious tissue injuries and doses exceeding 15 grays are a joint commission sentinel event. In general radiation exposes patients to tissue reactions and increases their lifetime risk of cancer. Research has shown that cardiovascular specialists are not fully informed about radiation dose and their associated health implications. There is variability in doses dose usage during PCI that is associated with hospitals and there is a need to standardize the level of knowledge and competence within cardiology and this extends to both those who order cardiac imaging procedures and those who perform them. The radiation safety metric and development will include all PCI procedures with or without diagnostic coronary angiography in the denominator. The numerator will identify the percent of PCI procedures with a reported cumulative erykerma that was greater than or equal to 5 grays or greater than or equal to 5,000 milligrays. There are no denominator exceptions or exclusions. As a safety measure the registry hopes to help hospitals understand every instance where the patient was exposed to a radiation dose that is associated with skin injury and where staff were potentially also exposed. The FDA has approved the cath PCI registry as a post-approval study vehicle. Participants can engage with this initiative and help inform about the shockwave coronary IVL system in a real-world setting by completing the shockwave data set. The ACC has provided an auxiliary data collection tool which can be utilized by all participants no matter how they input their base data. Certified cath PCI registry vendors are at liberty to incorporate the shockwave data set if they would like. One of the new adventures in the in development and possibly available by the time of this viewing is designed for participants utilizing our ACC online data collection tool. The cath PCI registry collects data on diagnostic coronary angiography procedures and PCI procedures. The whole of the data set is relevant for patients having PCI but when the cath lab visit only includes a diagnostic coronary angiography then only the data elements highlighted in yellow on the paper data collection form need to be completed. Our data collection tool will have the ability to highlight the DDS fields. As you can see the highlighted fields will help the user quickly navigate to the required data points. Fields that are not highlighted such as cerebrovascular disease, peripheral arterial disease, and chronic lung disease do not need to be completed and can be skipped. In recognition of the expertise our NPI licensed cardiovascular care professionals demonstrate in all the things they do and the targeted treatment strategies they initiate for their patients, the NCDR renamed the dashboard to professional level dashboard. We sincerely hope all members of the clinical team who have data displayed here feel our deep respect for all that they do. We hope this platform continues to serve them well. And now let's look at what's coming that will elevate registry reporting upward and broaden registry value for you and your stakeholders. Our road upward includes analysis, reanalysis, and raising the bar. And we intend to touch all of these items. 30 day risk assessed mortality, new risk standardized model for acute kidney injury, a refinement of the risk standardized bleeding model, develop AUC for diagnostic procedures, develop new detail lines, introduce a new composite measure for public reporting, and last but not least, touch the version 5 data set. The 30 day mortality measure represents one of the first NCDR measures to use a linked data set to report on an outcome which may occur after discharge from the hospital. NCDR records are matched to the National Death Index. A 30 day time period was chosen as it reflects both the inpatient care and transition of care to outpatient setting. It is a clinically meaningful period for hospitals to collaborate with their communities in an effort to reduce mortality. All causes of death are included in the measure as non-cardiac death may still be directly related to the quality of care. And it is often difficult to determine cause of death after discharge. And from the patient's point of view, no matter the cause, death is a poor outcome. With the knowledge gained from the recently updated risk standardized in hospital hierarchical mortality model, the 30 day mortality measure will be updated. Once this report is updated, we hope to provide additional educational opportunities to help participants appreciate the value this knowledge can bring to their programs. The current risk adjusted acute kidney injury metric was developed in 2014 and was built using version 4.4 data elements. Kim has convened a diverse and talented work group who are evaluating definitions of acute kidney injury, determining which version 5 data variables are predictive of AKI, and will reassess the eligibility criteria of the population. There's more to come on this important effort. Risk models require maintenance and the risk standardized bleeding model is due. As the patient population of the registry shifts and morphs in response to arbitrary influencers such as COVID, or is simply reflective of an aging population, as an example, the model, much like an engine, needs a tune-up. Small adjustments to the model parameters will ensure the model output matches as closely as possible data that can be found empirically. As well, during this process, the model variables can be reassessed to see if there are opportunities for increasing the sensitivity and specificity of the metric criteria. As an example, should transfusions given outside of a 72-hour post-procedure window continue to be counted as an observed bleed? This has always been the case, but the version 5 data set now lets us identify when transfusions were given within 72 hours of PCI. These questions and several others will be evaluated during the RSV maintenance effort. Any revisions to the model will be communicated in the release notes of the report cycle in which the updates are applied. The goal of the version 5 data set was to increase registry reporting to include appropriate use criteria metrics for diagnostic procedures. While it has taken us a while, this effort is not forgotten. Kim and I are meeting now with a work group who are mapping the 2013 AUC indications to the version 5 data set. This work is ongoing and we hope to add these important metrics to the report in 2022. In development now, and perhaps by the time this recording airs, these new detail lines are on the way. While the risk-adjusted mortality metric has always given insights into the outcome of patients having PCI, the detail lines will share the unadjusted percentile of all patients seen during each quarter. This data, along with the rate at which patients were enrolled in hospice or are receiving comfort measures only, will help participants trend the complexities of their patients over time in a different way. Additionally, the version 5 data set collects several specific ventricular mechanical support devices. The rate of usage for these will now be reported in the PCI procedure section, as well as the rate of newly diagnosed heart failure and the type of heart failure seen. Lastly, a detail line reporting when a 3-gram hemoglobin drop is observed will help participants understand potential performance in their risk-adjusted and risk-standardized bleeding metrics. This detail line will be supported with a drill-down and report the hemoglobin drop observed for each patient. We expect this will better support the conversations you are having with all members of the clinical team well in advance of the quarterly report cycle. A CAHPS PCI registry composite measure is in development. The composite measure will combine select process and outcome metrics and provide registry participants and external stakeholders a single comprehensive view of a hospital's quality rating. The composite measure is designed for public reporting and hospital performance will be reported as 1, 2, 3, or 4 stars. The launch of this measure will be shared on the registry announcement page and NCDR news channels, which we expect to happen in 2022. Last, but definitely not least, the version 5 data set is under review. Kim has created an elaborate spreadsheet that has organized our recommendations into three buckets. Our three buckets are what we think can be removed from the data set, what needs to stay, and what should be added. The goal is to alleviate extra data collection effort while purposefully keeping what is needed to report our current metrics and strategically identifying what should be added to increase reporting opportunities. In the near future, a workgroup will convene to evaluate these recommendations against current practice, clinical guidelines, and new opportunities and provide their recommendations or any updates to the version 5 data set. As this project gains momentum and passes key milestones, participants will have the opportunity to provide their thoughts through the 30-day comment period, which will be announced on the registry homepage. Thank you so much for joining me. I hope that through this presentation you recognize how the past has shaped the current and future goals of the registry. As well, I hope you have a better sense of the commitment the CATH PCI registry team, the steering committee, and NCDR leadership have to supporting all registry initiatives and ensuring you have the tools that you need to help drive quality improvement initiatives that support the delivery of quality patient care.

CAF PCI Registry

enhancements

improvements

2021

2022

metric refinements

dashboard enhancements