Echocardiographic Surveillance of AS Doesn’t Measure Up to Guidelines

Patients with aortic stenosis (AS) who undergo regular echocardiographic surveillance according to current US guidelines have better clinical outcomes, observational data from a California health system confirm. Still, as the new study makes clear, such monitoring doesn’t yet occur at an optimal level.

The findings, published online recently in Circulation: Population Health and Outcomes, show that patients with severe AS were, counterintuitively, the least likely to have guideline-concordant monitoring. Fewer than half of these patients were treated according to recommendations.

American College of Cardiology and American Heart Association guidelines for the management of valvular heart disease, released in 2020, “recommend repeat echocardiography every 3 to 5 years for mild AS, 1 to 2 years for moderate AS, and 6 to 12 months for severe AS,” Jesse K. Fitzpatrick, MD (Kaiser Permanente San Francisco Medical Center, CA), and colleagues point out.

Yet, as Fitzpatrick told TCTMD, there are a few evidence gaps behind that advice. Current “recommendations for echocardiographic surveillance of aortic stenosis are based largely on data describing the natural progression of the disease,” he wrote in an email. “However, the extent to which adherence to these recommendations may improve clinical outcomes in usual care is unknown.”

This study represents the first large, real-world assessment of adherence patterns and outcomes in AS, Fitzpatrick said.

Guidelines were consistent, for the most part, across the decade-long study period. “Later enrollment in the study was very modestly associated with improved guideline adherence,” he noted, adding that researchers adjusted for calendar year of AS diagnosis “to account for temporal changes in practice patterns and therapies such as the introduction and expansion of TAVR.”

In particular, the monitoring of patients with severe aortic stenosis—or the lack thereof—came as a surprise, given that they’re “at the highest risk of adverse outcomes from their valve disease,” Fitzgerald pointed out.

Christopher B. Scoma, MD, and Nicole M. Bhave, MD (both from University of Michigan Health, Ann Arbor), in an editorial, also point to the lack of existing data on the relationship between clinical outcomes and guideline-directed surveillance in AS.  The new study, they say, helps “to corroborate existing surveillance window recommen­dations.”

Longitudinal management is becoming ever relevant in this area as TAVI has expanded to low-risk and even asymptomatic patients, they observe. “In this era of surgical and transcatheter aortic valve replacement options, [it’s already known that] prompt diagnosis and intervention for severe, symptomatic AS improves longevity and quality of life for patients in all age groups.”

Death Decreased, AVR Increased

The researchers used natural language processing algorithms to look at patterns in AS echo surveillance that took place in the Kaiser Permanente Northern California healthcare system between 2008 and 2017. They assessed the frequency of surveillance across categories of AS severity.

The dataset included 20,571 patients with AS (mean age 75.7 years; 48% women) who had not yet received aortic valve replacement. Guideline concordance for echo surveillance was 74% for mild AS, 51% for mild-moderate AS, 63% for moderate AS, 51% for moderate-severe AS, and 49% for severe AS.

Patients were more likely to undergo echo at the recommended intervals if they were male, younger, or already seeing a cardiologist. Certain comorbid conditions also impacted the odds, though these associations varied across AS severity and not all were significant.

Over a median follow-up of 5.2 years, outcomes were better when surveillance aligned with guidelines. Death rates were 28.8% and 42.2% with versus without concordance, while AVR rates were 26.6% versus 14.5%, respectively.

For all-cause death, the difference in risk was significant for patients with at least moderate AS. Across the AS spectrum, guideline concordance was linked to greater likelihood of AVR.

Outcomes With vs Without Guideline-Directed Echo Surveillance

“These findings highlight the need to improve longitudinal AS management,” the researchers conclude.

Scoma and Bhave, though, highlight a few ways the results might reflect bias. For example, the people who tend to undergo more frequent testing might also seek out healthcare more generally, they note. “Conversely, ongoing surveillance may have been deemed clinically inappropriate in some situations based on individual patient characteristics—for example, advanced dementia, metastatic cancer with limited life expectancy, or hospice care.”

The editorialists add: “Because the dataset did not capture clinicians’ thought processes or outcomes of shared decision-making conversations, it is unclear whether other factors such as patient refusal or fragmented follow-up may have played a role.”

Barriers and Bridges

Fitpatrick acknowledged that several barriers stand in the way of AS surveillance.

“Patients with aortic stenosis are largely asymptomatic and may not be aware of the need for routine longitudinal follow-up,” he said. “On average, patients with aortic stenosis are older and there can be barriers to communication and transportation as well as their having other medical problems that are prioritized clinically. Follow-up intervals can span years over which time care can transition between providers or the patient may relocate.”

These barriers are consistent across health systems, Fitzgerald added, “though integrated healthcare delivery models may be better positioned to implement system-wide interventions to improve surveillance.”

Notably, their analysis showed that being under the care of a cardiologist helped facilitate monitoring.

“Cardiologists are in many cases already doing the right things—documenting recommendations, setting clinical reminders, communicating with patients, and using clinic-based panel management tools,” he commented. That said, “many barriers to care are outside of the cardiology clinic and better addressed through system-level population management strategies.”

Artificial intelligence, such as the natural language-based algorithms used in this study, could be used to interpret echo reports obtained during routine care, Fitzgerald suggested. “Echocardiographic data extracted through these or similar methods could be leveraged to generate automated provider alerts or to initiate appropriate follow-up testing that could bridge fragmented care.”

He said that what’s needed going forward is implementation research to test practical interventions at a system level, pointing to the DETECT-AS trial as an example of this approach. “We need to extend this type of research to earlier stages of valve disease and leverage technologies that enable it to be scaled across entire health systems,” Fitzgerald concluded.

The editorialists take a similar system-level perspective, highlighting the American Heart Association’s Target: Aortic Stenosis quality-improvement program as a potential way to bridge the gap between diagnosis and treatment.