Can Cardiologist Care Really Improve Atrial Fibrillation Survival?

Patrice Wendling

December 05, 2017

TORONTO, ON — Patients with newly diagnosed atrial fibrillation (AF) live longer when cared for by cardiologists, despite—or perhaps because of— having more frequent hospital admissions, researchers report[1].

In the retrospective study, seeing a cardiologist vs other physicians was associated with a 32% relative risk reduction in death at 1 year (5.3% vs 7.7%; hazard ratio [HR] 0.68, 95% CI 0.55–0.84).

After adjustment for the risk of death, however, patients who saw a cardiologist were at increased risk of hospitalizations for:

  • AF (17.9% vs 8.2%; HR 2.3, 95% CI 2.0–2.7).

  • Stroke syndromes (1.7% vs 0.5%; HR 3.4, 95% CI 1.8–6.1).

  • Bleeding (3.1% vs 2.0%; HR 1.5, 95% CI 1.1–2.1).

  • Heart failure (3.2% vs 1.4%; HR 2.2, 95% CI 1.5–3.1).

"One would have thought that excellent cardiovascular care would have reduced these hospitalizations, so this is a curious finding for us," lead author Dr Sheldon M Singh (Sunnybrook Health Sciences Centre, University of Toronto, Ontario) commented to | Medscape Cardiology.

"We wondered whether this in part may help mediate some of the survival benefit we noticed. In other words, maybe patients came in when they were less sick and had more intensive therapy provided to them when they were an inpatient to allow them to do better in the long term," he said.

In contrast, the recent TREAT-AF study found that cardiology care within the first 90 days did not affect heart-failure hospitalizations among veterans with AF but lowered their adjusted risk of stroke and death by 9% and 11%, respectively.

Commenting for | Medscape Cardiology, TREAT-AF senior author Dr Mintu Turakhia (Stanford University and VA Palo Alto Health Care System, CA) said in an email that "these results should be interpreted with caution" and that "looking at patients who did not see cardiologists through the observation period may bias results in unmeasured ways, and this could explain the findings."

He also points out that the mortality benefit in TREAT-AF was not statistically mediated by higher observed anticoagulation use in cardiology-treated patients and was included as hypothesis-generating due to the potential for confounding.

"I don't think you can walk away and say that cardiologists save lives without a randomized trial," Turakhia added.

For the present study, published in the December 2017 issue of the Canadian Journal of Cardiology, Singh and his colleagues analyzed data from 5804 propensity-matched patients drawn from an original cohort of 22,032 individuals presenting to the emergency department with a primary diagnosis of AF, 85% of whom were seen by a cardiologist at least once within 1 year of the index visit.

Among the 15% of patients not seen by a cardiologist, 91.8% visited a family physician and 52.1% an internal medicine specialist during follow-up. Patients seen by a cardiologist, however, were even more likely to visit these practitioners (95.4% and 57.9%, respectively).

Novel oral anticoagulant use was also significantly higher in the cardiologist group than the noncardiologist group (18.1% vs 14.2%), as was the use of echocardiography (51.9% vs 24.5%), stress testing (30.6% vs 13.7%), coronary angiography (11.2% vs 0.3%), and percutaneous revascularization (1.6% vs 0).

"We have to appreciate there is literature suggesting that many patients with atrial fibrillation have subclinical coronary disease, so one wonders if some of these tests may uncover other concomitant conditions such as heart failure, systolic or diastolic dysfunction, or coronary artery disease and that general care if attended to may result in improved survival," Singh said.

He continued, "I think what our work should do is stimulate researchers to understand what are the precise processes of care that are mediating this survival benefit, because it's clear that not everyone will have access to a cardiologist. There's some jurisdictions where it's even challenging to see an internal-medicine specialist."

Dr Stephen B Wilton (University of Calgary, AB) suggests the "most startling finding" of the study is the high 6.5% rate of death in the overall cohort, despite the group being relatively young (mean age <64 years) and low risk (median CHA2DS2-VASc score 3). By contrast, 1-year mortality was only 3.9% in a recent study of Canadians discharged after an uncomplicated ST-elevation MI.

"Therefore, a new diagnosis of AF, although not immediately life-threatening, should be regarded as an important marker of near-term risk of cardiovascular events. This observation alone provides a potential rationale for desiring early cardiovascular specialist evaluation of these patients," he writes in an editorial accompanying the study[2].

On the other hand, Wilton notes that "the results in their totality are not easily explained" and the data cannot explain how and why patients were referred and whether the 15% not seen were never referred or declined a referral. Current Canadian and international clinical guidelines for management of AF express no preference for specialist- vs primary-care delivery for AF patients.

"The results of this study will not shift these recommendations, and in my opinion that is just fine," Wilton writes. "Even if it were possible for every patient with AF to be cared for by a cardiologist, this should be unnecessary.

"Especially in the increasing proportion of those with AF who are asymptomatic, we can empower family physicians to perform initial investigations and initiate stroke-prevention therapy," he concluded.

Singh said the vast majority of those looking after AF patients think about AF and stroke, which is a reasonable association because no other therapies have been shown to improve survival in this population. "But what other registry studies have shown is that stroke deaths are only the minority of deaths."

He noted that neither they nor TREAT-AF had data on cause of death—a key limitation, but that the latter used last hospitalization as a surrogate. In the present analysis, hospitalizations for stroke-related events were less common (1.1%), for example, than those for heart failure (2.3%).

"It was important for us to publish this paper to kind of get the message out that people die from [causes] other than strokes and bleeds with atrial fibrillation. So providing general comprehensive care is helpful in this group of patients," Singh said.

The study was supported by a donation from the Tambakis family, with additional support from the Institute for Clinical Evaluative Sciences (ICES). Singh reports no relevant financial relationships. Disclosure for the coauthors are listed in the paper. Wilton reports no relevant financial relationships. Turakhia reports receiving research grants from Medtronic and Janssen Pharmaceuticals and having served as a consultant/advisory board member for Medtronic, St Jude Medical, and Abbott.

Follow Patrice Wendling on Twitter: @pwendl. For more from | Medscape Cardiology, follow us on Twitter and Facebook.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.