Deaths After Ablation of Atrial Fibrillation ‘Concerning’ in Real-World Analysis

(UPDATED) The risk of death following catheter ablation for atrial fibrillation is higher than rates observed in randomized controlled trials, with the majority of deaths occurring not during the procedure but rather in the 30 days after discharge, according to a large analysis of real-world data.

The early mortality rate, defined as death during the initial admission or 30-day readmission period, was 0.46% among more than 60,000 patients treated for A-fib ablation between 2010 and 2015. Moreover, quarterly rates of early mortality after ablation increased over time, increasing from 0.25% in 2010 to 1.35% in 2015 (P < 0.001).

“There are some year-to-year fluctuations, but there is a statistically significant trend,” said lead investigator Jim Cheung, MD (Weill Cornell Medicine-NewYork Presbyterian Hospital, New York, NY), in reference to the increase in early mortality. “While it’s certainly concerning, at least we were able to observe that there was a trend toward increasing comorbidity burden among those patients undergoing catheter ablation procedures. If you adjust for this, that trend [of increasing mortality] is attenuated. So, it suggests that it’s not so much an issue of hospitals or operators doing a worse job with the procedure, but that we’re ablating sicker and sicker patients.”

To TCTMD, Cheung said the overall mortality rate following A-fib ablation was surprising because it is significantly higher than what has been reported in randomized trials. For example, there were no deaths in the pivotal FIRE AND ICE, CASTLE-AF, and CABANA trials, and no deaths reported in other analyses from large, high-volume academic medical centers.

When I first read this paper, my initial reaction was that the findings must be incorrect. Hugh Calkins

In an editorial accompanying the study, Hugh Calkins, MD (Johns Hopkins Hospital, Baltimore, MD), expresses consternation at the real-world mortality rate. “When I first read this paper, my initial reaction was that the findings must be incorrect,” he writes. “I have performed thousands of AF ablation procedures over the past two decades, and I have never had a patient die as a result of the procedure.”

Oussama Wazni, MD (Cleveland Clinic, OH), who was not involved in the study but was part of the Cleveland Clinic group that published data on more than 10,000 patients who underwent A-fib ablation, noted that more and more centers with less experience are performing the procedure. This inexperience translates into worse clinical outcomes, he said. In their series, there were no reported deaths when they published their report in early 2019. Inexperienced, or low-volume, centers performing ablation is concerning, but Wazni noted that patients also have a responsibility, noting that they sometimes “choose convenience over excellence.”

“Patients aren’t doing their part in seeking out the best possible centers,” he told TCTMD.

Hospital Volume and Early Mortality

The analysis of the Nationwide Readmissions Database (NRD), which was led by Edward Cheng, MD, PhD (NewYork-Presbyterian/Weill Cornell Medicine), and published October 28, 2019, in the Journal of the American College of Cardiology, identified several predictors of early mortality, among them procedural complications (adjusted OR 4.06; 95% CI 2.40-6.85), congestive heart failure (adjusted OR 2.20; 95% CI 1.20-4.03), and low AF ablation hospital volume (adjusted OR 2.35; 95% CI 1.33-4.15).

Of the 276 patients who died early following catheter ablation of A-fib, 126 died during the index admission and 150 died during the 30-day readmission after ablation. The median time to death was 11.6 days, and those who died were older and had a higher burden of comorbidities, such as congestive heart failure, CAD, and previous placement of a pacemaker, among other conditions. Patients who died early were also less likely to be treated at a higher-volume hospital or academic medical center. Overall, 62.3% of the early deaths after ablation were patients treated at a low-volume center, defined as a hospital that performed less than 21 ablations per year.    

“It’s significant part of the picture,” said Cheung. “Hospital volume was a very potent predictor of outcomes. That lines up with what has been shown in the past, and it was nice to see it replicated in our analyses.”

The rate of procedural complications during the index admission was 6.7%, increasing from 4.8% to 7.4% between 2010 and 2015 (P < 0.001). Perforation and bleeding/vascular complications occurred in 1.9% and 4.3% of treated patients. Like early mortality, the trend toward increased complications over time attenuated after adjusting for age, hospital volume, and patient comorbidities (P = 0.332 for trend).

EP Competency and Standards

Speaking to the volume issue, Cheung said procedural experience is largely focused on fellowships, noting that the American College of Cardiology (ACC)/American Heart Association(AHA)/Heart Rhythm Society recommend a minimum of 50 A-fib ablations (among other procedures) to complete a fellowship in clinical cardiac electrophysiology. To maintain competency after the fellowship, the professional societies recommend doing “at least several AF ablation procedures per month.”

For Cheung, their research raises the issue about whether or not a firmer recommendation might be necessary, something akin to the clinical competency statement for PCI from the ACC, AHA, and Society for Cardiovascular Angiography and Interventions. He noted, however, they only had access to hospital-level data and not information on individual operators. For that reason, the investigators are unable to precisely say whether the early mortality rate is directly tied to the electrophysiologist. While the operator is a “huge component of outcomes,” if the patient has a complication there are entire systems of care involved in treatment, including other physicians and nurses.

“Our research raises the question, but there are complexities that need to be drilled down deeper so that we know exactly what the issue is,” said Cheung. “I feel like this is more than just the operator. Part of this is the institution as well.”

To TCTMD, Angelo Biviano, MD (Columbia University Irving Medical Center, New York, NY), said outcomes in the real world can differ from those in clinical trials, particularly since the patient populations differ. He agreed with the investigators’ assessment that the differences in mortality and complications between these two settings are likely the result of increasing comorbidity burden.

“As time goes on and you feel more comfortable with the procedure, the referring physicians, as well as those doing the ablation, may take on riskier patients,” he said. “AF ablation is becoming more mainstream and with that the complication rates can go up.”

Additionally, experience is also likely a factor. Unlike in trials, where patients are treated at high-volume academic medical centers, some hospitals and providers don’t have the same degree of experience. The question of minimum volume requirements to perform AF ablations is one that needs to be addressed, said Biviano, noting that select structural heart interventions, such as TAVR, have such minimum standards in place. One of the limitations of the present study is that the causes of death are unknown, so a direct link between the ablation and early mortality is not necessarily the case, he added.

Hospital volume was a very potent predictor of outcomes. Jim Cheung

Wazni advised that physicians need to perform at least 50 A-fib ablations annually to main clinical competency. In his opinion, ablation of A-fib has become almost too routine, with operators taking the procedure “for granted.” For example, catheter ablation is now offered to sicker and sicker patients, such as those with congestive heart failure. While studies have shown that ablation is superior to medical therapy in patients with congestive heart failure, those studies limited the severity of disease with cutoffs based on ejection fraction, he said. The present NRD analysis does not include data on ejection fraction, but congestive heart failure was a leading cause of hospital readmission and a predictor of mortality.

At the Cleveland Clinic, there have been two deaths following A-fib ablation since their series was published, but these two patients were extremely sick, said Wazni. Thus far, clinicians at their center have ablated more than 20,000 patients with A-fib. “We’re very proud of our outcomes,” he said. “Now, it’s hard for others to replicate this because we’re a very big center and we do a lot of ablations. It’s also a matter of [patient] selection. That’s very, very important. It’s something that should be taken into consideration.”

In his editorial, Calkins also stresses the importance of operator experience. He highlights a 2013 study showing that more than 80% of 93,801 ablations performed between 2000 and 2010 were done by operators doing less than 25 procedures per year. In that study, annual operator volume (less than 25 procedures) and hospital volume (less than 50 procedures) were significantly associated with adverse outcomes.

This latest study by Cheng and colleagues “should serve as a wake-up call to all electrophysiologists who perform AF ablation, all cardiologists who refer patients for this procedure, and all patients who are considering undergoing AF ablation. What is clear is that AF ablation is not a benign procedure, and mortality is a very real complication of the procedure,” Calkins concludes.