New Model Looking at TAVR Outcomes, Site-Level Variations Could Inform Decision-Making

A new model that assesses mortality and quality of life following transfemoral TAVR procedures and site-level variations in TAVR outcomes could help inform decision-making for hospitals and patients, according to results of a study presented March 29 at ACC.20/WCC during a Late-Breaking Clinical Trial session.

Nimesh Desai, MD, PhD, et al., used data from the STS/ACC TVT Registry to assess outcomes in 54,217 TAVR patients treated at 301 sites from 2015 through 2017 to determine whether there were site-level variations in quality of care. The researchers created a model evaluating a composite of serious 30-day complications – including mortality, stroke, major bleeding, stage 3 acute kidney injury, and moderate to severe paravalvular leak – that patients likely would want to consider when making decisions regarding TAVR procedures.

The researchers then applied the composite metric across all sites, categorizing sites with outcomes outside the 95% confidence interval average as performing better than or worse than expected. Of the 301 participating sites, 34 (11%) experienced worse composite outcomes than expected, while the rate of complications was as-expected at 242 sites (80%) and better-than-expected at 25 sites (8%).

Researchers reported a substantial difference in complication rates among sites with worse-than-expected performance vs. those with better-than-expected performance. For example, sites with worse-than-expected outcomes had a risk-adjusted expected mortality ratio of 1.25, compared with 0.71 for sites with better-than-expected outcomes and 1.01 for sites with as-expected outcomes. For major bleeding, the ratio was 2.13 for worse-than-expected sites, 1.02 for as-expected sites and 0.45 for better-than-expected sites.

The model can assess mortality and quality of life following TAVR, the researchers conclude. Because the model incorporates multiple outcomes on a ranked basis, the researchers report it is reliable even when including sites with low TAVR volumes. The researchers plan to further analyze the data to identify any features or factors that may be associated with worse-than-expected performance. In addition, the model can help establish a platform for public reporting that patients and hospitals could use to inform decision-making, the researchers note.

"There's clearly an opportunity to improve processes and try to better standardize care to decrease variation between different sites," says Desai, the study's lead author. "The overarching goal of this work is to provide transparency to the public and to provide feedback to sites so that they can review their practices and develop ways to improve the results in their patients."

Keywords: ACC Annual Scientific Session, acc20, Transcatheter Aortic Valve Replacement, Aortic Valve Stenosis, Heart Valve Diseases, Heart Failure, National Cardiovascular Data Registries, STS/ACC TVT Registry


< Back to Listings