Meeting News

Novel TAVR analysis: 1 in 10 sites underperform in patient-centered outcomes

Nimesh Desai

A novel analysis of transcatheter aortic valve replacement in the U.S. determined that 1 in 10 facilities that perform the procedure confer worse than expected mortality and quality of life outcomes.

According to findings from the Society of Thoracic Surgeons/American College of Cardiology Transcatheter Valve Therapies (TVT) registry presented at the virtual American College of Cardiology Scientific Session, researchers identified 30-day outcomes that most affect patients’ quality of life at 1 year, including stroke, severe bleeding, stage 3 acute kidney injury, moderate to severe paravalvular leak and death.

Using these outcomes, researchers assessed 54,217 patients at 301 U.S. facilities that performed TAVR between 2015 and 2017, and they found that 80% of sites had an as-expected rate of 30-day complications while 11% were worse than expected and 8% were better than expected.

Among the patients in the cohort, at 30 days, 3.2% died, 2% had a stroke, 5.8% had a severe bleeding event, 0.6% had stage 3 acute kidney injury, 2.5% had moderate or severe paravalular leak and 85.9% had none of those outcomes.

In comparison, between 2018 and June 30, 2019, researchers observed that 9% of sites had worse-than-expected 30-day outcomes while 88% were as expected and 3% had better-than-expected outcomes.

“We developed a novel patient-centered composite outcome for TAVR based on 30-day outcomes, and their ranked association with 1-year mortality and quality of life,” Nimesh Desai, MD, PhD, cardiac surgeon and associate professor of surgery at the Hospital of the University of Pennsylvania, said during his presentation. “We have identified significant site-level variation, mortality and major complications after TAVR procedures in the United States. The model demonstrated excellent performance, including internal validity and moderate to high reliability, even when including lower volume programs, and this 30-day composite metric is appropriate for high-stakes applications such as public reporting.”

“There’s clearly an opportunity to improve processes and try to better standardize care to decrease variation between different sites,” Desai said in a press release. “The overarching goal of this work is to provide transparency to the public and also to provide feedback to sites so that they can review their practices and develop ways to improve the results in their patients.”

 

Model design

For this novel analysis, researchers analyzed data from the STS/ACC TVT registry, a CMS-mandated registry that includes patient data from virtually every TAVR performed in the U.S., to create a model to assess complications that patients may want to consider when choosing TAVR. After researchers applied this model to STS/ACC TVT registry data and identified average in-hospital complication rates across all TAVR sites, they categorized sites as performing better or worse than expected if their outcomes were outside the 95% CI of that average.

 

Confidence for public reporting

Howard C. Herrmann

In a follow-up discussion of this analysis of the STS/ACC TVT registry, Howard C. Herrmann, MD, FACC, director of the cardiac catheterization laboratories at the Hospital of the University of Pennsylvania, said he questioned the researchers’ confidence in their findings, given missing data.

“It’s clear that we all, both public and payers, want more transparency related to TAVR outcomes, and that we, as operators, want to be able to improve our quality outcomes and prioritize quality over just volume,” Herrmann said during the discussion. “How confident are you in these data, given that it was derived initially from only about 50% of the hospitals that participated in the registry, due to missing data? Are you confident that we can use this for public reporting and that the outcomes are going to be useful?”

“In the end, payers like insurance companies and commercial entities, which perform hospital rankings, already penalize hospitals, either in their reimbursement or in their public rankings for not publicly reporting their STS data for surgical AVR,” Desai said in response to Herrmann. “We’ve already heard that similar incentives likely being performed for TAVR. So not having publicly reportable data for TAVR will likely end up being a significant disincentive for sites for reasons outside of the registry.”

 

Allowing the 11% to improve

“The significance of this is very important in the way that we are now beginning to really see, if you will, the fruits from a high-quality mandated stated registry, which was part of the rollout of TAVR,” Joseph C. Cleveland Jr., MD, professor of surgery-cardiothoracic at University of Colorado School of Medicine in Aurora, said during a press conference. “Particularly, the strength of this model, as Dr. Desai emphasized, we’re really beginning to drill down to patient-centered planning. What do our patients care about? They want to be alive and well.

“Dr. Desai did a very good job of showing that this model has been tested, internally validated, is reliable and is very robust. While no model can be perfect, this model is going to be one of the most high-fidelity, reliable, valid models for public reporting. In the data shown, one can see that there still is variation. But to me, at least it’s reassuring that roughly 90% of TAVR programs are performing as expected or slightly better, and it will give the 11% of TAVR programs the ability to continue to improve. This is the first step that will now start to allow for patients to be represented in the quality outcomes data world and be able to see these data.” – by Scott Buzby

Reference:

Desai N, et al. Joint American College of Cardiology/New England Journal of Medicine Late-Breaking Clinical Trials. Presented at: American College of Cardiology Scientific Session; March 28-30, 2020 (virtual meeting).

Disclosures: Desai reports he received consultant fees and/or honoraria from Edwards, Medtronic, St. Jude Medical and W.L. Gore and Associates; and research grants from Cook. Herrmann reports he received consultant fees and/or honoraria from Abbott Vascular, Edwards, Leerink Swann, Medtronic, Siemens and Wells Fargo; has a fiduciary role at Holistick; other financial disclosures from Mass Medical Society; has ownership interest in Micro-Interventional Devices; and research grants from Abbott Vascular, Ancora, Bayer, Boston Scientific, Cardiovascular Research Foundation, Edwards, Medtronic, Shockwave, St. Jude Medical and W.L. Gore and Associates. Cleveland reports no relevant financial disclosures.

Nimesh Desai

A novel analysis of transcatheter aortic valve replacement in the U.S. determined that 1 in 10 facilities that perform the procedure confer worse than expected mortality and quality of life outcomes.

According to findings from the Society of Thoracic Surgeons/American College of Cardiology Transcatheter Valve Therapies (TVT) registry presented at the virtual American College of Cardiology Scientific Session, researchers identified 30-day outcomes that most affect patients’ quality of life at 1 year, including stroke, severe bleeding, stage 3 acute kidney injury, moderate to severe paravalvular leak and death.

Using these outcomes, researchers assessed 54,217 patients at 301 U.S. facilities that performed TAVR between 2015 and 2017, and they found that 80% of sites had an as-expected rate of 30-day complications while 11% were worse than expected and 8% were better than expected.

Among the patients in the cohort, at 30 days, 3.2% died, 2% had a stroke, 5.8% had a severe bleeding event, 0.6% had stage 3 acute kidney injury, 2.5% had moderate or severe paravalular leak and 85.9% had none of those outcomes.

In comparison, between 2018 and June 30, 2019, researchers observed that 9% of sites had worse-than-expected 30-day outcomes while 88% were as expected and 3% had better-than-expected outcomes.

“We developed a novel patient-centered composite outcome for TAVR based on 30-day outcomes, and their ranked association with 1-year mortality and quality of life,” Nimesh Desai, MD, PhD, cardiac surgeon and associate professor of surgery at the Hospital of the University of Pennsylvania, said during his presentation. “We have identified significant site-level variation, mortality and major complications after TAVR procedures in the United States. The model demonstrated excellent performance, including internal validity and moderate to high reliability, even when including lower volume programs, and this 30-day composite metric is appropriate for high-stakes applications such as public reporting.”

“There’s clearly an opportunity to improve processes and try to better standardize care to decrease variation between different sites,” Desai said in a press release. “The overarching goal of this work is to provide transparency to the public and also to provide feedback to sites so that they can review their practices and develop ways to improve the results in their patients.”

 

Model design

For this novel analysis, researchers analyzed data from the STS/ACC TVT registry, a CMS-mandated registry that includes patient data from virtually every TAVR performed in the U.S., to create a model to assess complications that patients may want to consider when choosing TAVR. After researchers applied this model to STS/ACC TVT registry data and identified average in-hospital complication rates across all TAVR sites, they categorized sites as performing better or worse than expected if their outcomes were outside the 95% CI of that average.

PAGE BREAK

 

Confidence for public reporting

Howard C. Herrmann

In a follow-up discussion of this analysis of the STS/ACC TVT registry, Howard C. Herrmann, MD, FACC, director of the cardiac catheterization laboratories at the Hospital of the University of Pennsylvania, said he questioned the researchers’ confidence in their findings, given missing data.

“It’s clear that we all, both public and payers, want more transparency related to TAVR outcomes, and that we, as operators, want to be able to improve our quality outcomes and prioritize quality over just volume,” Herrmann said during the discussion. “How confident are you in these data, given that it was derived initially from only about 50% of the hospitals that participated in the registry, due to missing data? Are you confident that we can use this for public reporting and that the outcomes are going to be useful?”

“In the end, payers like insurance companies and commercial entities, which perform hospital rankings, already penalize hospitals, either in their reimbursement or in their public rankings for not publicly reporting their STS data for surgical AVR,” Desai said in response to Herrmann. “We’ve already heard that similar incentives likely being performed for TAVR. So not having publicly reportable data for TAVR will likely end up being a significant disincentive for sites for reasons outside of the registry.”

 

Allowing the 11% to improve

“The significance of this is very important in the way that we are now beginning to really see, if you will, the fruits from a high-quality mandated stated registry, which was part of the rollout of TAVR,” Joseph C. Cleveland Jr., MD, professor of surgery-cardiothoracic at University of Colorado School of Medicine in Aurora, said during a press conference. “Particularly, the strength of this model, as Dr. Desai emphasized, we’re really beginning to drill down to patient-centered planning. What do our patients care about? They want to be alive and well.

“Dr. Desai did a very good job of showing that this model has been tested, internally validated, is reliable and is very robust. While no model can be perfect, this model is going to be one of the most high-fidelity, reliable, valid models for public reporting. In the data shown, one can see that there still is variation. But to me, at least it’s reassuring that roughly 90% of TAVR programs are performing as expected or slightly better, and it will give the 11% of TAVR programs the ability to continue to improve. This is the first step that will now start to allow for patients to be represented in the quality outcomes data world and be able to see these data.” – by Scott Buzby

Reference:

Desai N, et al. Joint American College of Cardiology/New England Journal of Medicine Late-Breaking Clinical Trials. Presented at: American College of Cardiology Scientific Session; March 28-30, 2020 (virtual meeting).

Disclosures: Desai reports he received consultant fees and/or honoraria from Edwards, Medtronic, St. Jude Medical and W.L. Gore and Associates; and research grants from Cook. Herrmann reports he received consultant fees and/or honoraria from Abbott Vascular, Edwards, Leerink Swann, Medtronic, Siemens and Wells Fargo; has a fiduciary role at Holistick; other financial disclosures from Mass Medical Society; has ownership interest in Micro-Interventional Devices; and research grants from Abbott Vascular, Ancora, Bayer, Boston Scientific, Cardiovascular Research Foundation, Edwards, Medtronic, Shockwave, St. Jude Medical and W.L. Gore and Associates. Cleveland reports no relevant financial disclosures.

    See more from American College of Cardiology