New Patient-Focused TAVR Quality Score Rolling Out to US Hospitals

A new patient-focused composite metric, one that moves beyond periprocedural mortality, will soon be available for ranking the quality of transcatheter aortic valve replacement programs in the United States, according to researchers.

The metric, which includes in-hospital or 30-day mortality, stroke, VARC major/life-threatening/disabling bleeding, acute kidney injury (stage 3), and moderate/severe paravalvular leak, is designed to measure patient-centric outcomes that capture TAVR quality.

“When we consider quality of care in TAVR, we must address the emerging need to go beyond mortality endpoints and consider the patient more holistically,” said Nimesh Desai, MD, PhD (University of Pennsylvania Hospital, Philadelphia), during a late-breaking clinical trial session at the recent virtual American College of Cardiology 2020 Scientific Session. “Patients care about outcomes beyond periprocedural mortality. Importantly, they want to be alive and well with improved functional status and quality of life after their procedure.”

We let the data tell the story. Nimesh Desai

The steering committee of the Society of Thoracic Surgeons/American College of Cardiology (STS/ACC) TVT Registry commissioned the composite metric, said Desai, and it has already been approved for use. “We’re going to start sending out detailed confidential reports to sites imminently,” he predicted. “That will eventually lead to a voluntary public reporting effort, which we will expect in the next year or so.”

Speaking with the media, Desai said that TAVR has been safely rolled out in the US through careful collaboration between the Food and Drug Administration, Centers for Medicare & Medicaid Services, specialty societies, and industry, but there is an ongoing concern about hospital or site-level variations in mortality. Additionally, quality-improvement efforts are moving away from surrogate measures, such as volume, toward real outcomes. A composite metric, as opposed to mortality alone, is already well established in quality assessments of CABG and valve surgery programs, he observed.

“When we developed our composite-outcome measure, we wanted to understand what mattered most to patients,” said Desai. “In addition to mortality, we selected, using a data-driven methodology, four periprocedural outcomes that were highly associated with 1-year mortality and 1-year poor functional status as determined by the [Kansas City Cardiomyopathy Questionnaire (KCCQ)]. We did not use expert opinion. . . . We let the data tell the story.”

Roughly 90% Perform as Well as Expected or Better

In a logistic regression model adjusted for 46 measured variables, the researchers calculated the “site difference,” which is how well each center performed compared with the average performance of all the US programs. Using data from 301 hospitals and more than 52,000 patients who underwent TAVR between 2015 and 2017, the researchers found that 11% of centers performed worse than expected, 80% did as well as expected, and 8% exceeded the performance of the average US hospital. When the analysis was restricted to 2018 to 2019, a period that reflects more contemporary TAVR outcomes, 9% of hospitals performed worse than expected, 88% did as well as expected, and 3% fared better than expected.

For cardiothoracic surgeon Joseph Cleveland, MD (University of Colorado School of Medicine, Aurora), who wasn’t involved in the study, the strength of the model is that it moves beyond periprocedural or 30-day mortality as a surrogate for quality to a patient-focused model that includes disabling stroke, an event most patients fear more than death. “While no model can be perfect, this is going to be one of the most high-fidelity, valid, reliable models for public reporting,” said Cleveland.

While the STS/ACC TVT Registry data showed there are still differences in quality across TAVR centers in the US, roughly 90% are performing as expected or better than expected, he said.

Howard Herrmann, MD (University of Pennsylvania Perelman School of Medicine, Philadelphia), pointed out that physicians and payers want more transparency related to clinical outcomes and said operators want to improve their results by prioritizing quality over a surrogate outcome like volume. However, he questioned how this metric might fare if used in public reporting, noting that outcomes from roughly 50% of hospitals weren’t included in devising the composite measure because of missing data.

Desai said there were concerns about missing numbers, specifically missing KCCQ data, which limited the number of centers that could be included in the study. In a sensitivity analysis, the researchers excluded KCCQ scores and gait speed, a move that allowed them to include data from 447 eligible centers. In doing so, the results hardly changed, with only one hospital shifting from its initial performance category. Desai also noted there are ongoing education efforts to improve compliance and inclusion of these variables in the TVT Registry. More recent data, from 2019, have complete KCCQ information.

The researchers also tested how well the composite metric performed under different conditions, specifically when applied to TAVR programs with lower volumes. In estimating the reliability as a function of volume, they found the metric performed well even in centers that did as few as 25 TAVRs per year, and that the overall composite measure performed better than mortality alone.

For her part, Mayra Guerrero, MD (Mayo Clinic, Rochester, MN), praised the development of the patient-centric metric. She noted that the need for a new permanent pacemaker didn’t correlate with mortality or quality of life in their model, which is reassuring given that younger patients are now being treated with TAVR.

Desai said the jury is still out on the long-term implications of pacemaker implantation, adding that the data from 2015-2017 included high-risk and intermediate-risk patients. However, as TAVR moves into lower-risk patients, and if it proves to be correlated with mortality or quality of life, the composite metric can be adjusted to take it into account, he said.