Does Atrial Fibrillation Portend a Worse Prognosis in the ICU?

By Betty Tran, MD, MSc, Editor

SYNOPSIS: In this prospective, observational cohort study, both new-onset and recurrent atrial fibrillation were associated with increased hospital mortality, especially in patients without sepsis.

SOURCE: Shaver CM, et al. Atrial fibrillation is an independent predictor of mortality in critically ill patients. Crit Care Med 2015;43:2104-2111.

In this prospective, observational study from Vanderbilt University Medical Center, 1770 critically ill adults (1275 medical ICU and 495 surgical ICU patients) were followed to test the hypothesis that atrial fibrillation (AF) during critical illness was associated with increased hospital mortality independent of severity of illness and comorbidities. The authors also examined whether there were differences in risk factors and outcomes in patients with new-onset AF vs recurrent AF and in patients with sepsis and non-sepsis, and whether variables such as a more positive fluid balance, higher vasopressor requirement, and prior cardiac disease were associated with an increased risk of AF.

Overall, 236 patients developed AF during the 4-day study period in the ICU, with 123 of these categorized as new-onset AF and 113 experiencing recurrent AF. Compared to recurrent AF patients, those with new-onset AF tended to have more organ failure and shock, whereas recurrent AF patients were more likely to have a history of congestive heart failure, hypertension, and hyperlipidemia. After controlling for potential confounders such as age, cardiac history, APACHE II score, shock, and sepsis, development of any AF was associated with a 62% increased risk of hospital mortality (95% confidence interval [CI], 1.14-2.29; P = 0.007). This association occurred for both new-onset and recurrent AF, separately. Patients with any AF also had significantly longer ICU and hospital lengths of stay. Compared to patients with no AF or recurrent AF, new-onset AF patients had significantly greater net positive cumulative fluid balance (median 6.1 L vs 5.4 L; P = 0.034), more frequently had diastolic dysfunction, were more likely to be treated with vasopressors, and were on vasopressors longer. Although an occurrence of any AF was associated with increased hospital mortality regardless of the presence of sepsis, the association was magnified in patients without sepsis (odds ratio, 2.92; 95% CI, 1.52-5.60; P = 0.001).

COMMENTARY

AF commonly occurs in the ICU and is often difficult to manage, especially in the setting of hypotension. The study by Shaver et al suggests that its presence may have more significant implications. There are several strengths to this study, including the large sample size of mixed medical and surgical ICU patients and the ability to distinguish between new and recurrent AF. Although the authors included APACHE II score and the presence of shock and sepsis in their multivariable regression models in an attempt to control for AF being a mere marker of severe illness, it is possible there is still residual confounding present. Intuitively, it makes sense that patients with new-onset AF are usually sicker than non-AF patients; in the study, AF patients had higher APACHE II scores, more organ failures, higher cumulative fluid balance, and more frequent and/or longer need for vasopressors (implying more severe hypotension or hypotension of longer duration). Additionally, recurrent AF patients often have traditional cardiac comorbidities such as congestive heart failure, hypertension, and hyperlipidemia.

Although severity of illness is likely a contributor, the exact relationship between AF and mortality is still unclear. As suggested by the authors in citing a “two-hit” model for development of AF, patients who develop AF may have a genetic propensity to develop it in response to triggers commonly seen in critical illness. Given the interest in exploring genetics in sepsis, could there be an overlap in genetic risk for development of AF and sepsis, especially a more severe systemic inflammatory response? Or is development of AF in critically ill patients just less well-tolerated? Additional studies aimed toward treating AF and preventing it with a focus on patient outcomes will be enlightening.