Predictors of New Onset Atrial Fibrillation Burden in the Critically Ill

Abstract

Introduction Atrial fibrillation (AF) is common in the intensive care unit (ICU) setting, and has been associated with adverse outcomes. In this context, there is increasing research interest in AF burden as a predictor of subsequent adverse events. However, the pathophysiology and drivers of AF burden in ICU are poorly understood. This study sought to evaluate the predictors of AF burden in critical illness associated new onset AF (CI-NOAF). Methods Out of 7,030 admissions a tertiary general intensive care unit between December 2015 and September 2018, 309 patients developed CI-NOAF. AF burden was defined as the percentage of monitored time in AF, as extracted from hourly interpretations of continuous ECG monitoring. Low and high AF burden groups were defined relative to the median AF burden. Clinical, laboratory and echocardiographic parameters were extracted, and multivariable modelling with binary logistic regression was performed to evaluate for independent associations with AF burden. Results The median AF burden was 7.0%. Factors associated with increased AF burden were age, dyslipidaemia, chronic kidney disease, increased creatinine, CHA2DS2-VASc score, ICU admission diagnosis category, amiodarone administration and left atrial area (LAA). Factors associated with lower AF burden were previous alcohol excess, burden of ventilation, the use of inotropes/vasopressors and beta blockers. On multivariate analysis, increased LAA, chronic kidney disease and amiodarone use were independently associated with increased AF burden, whereas beta blocker use was associated with lower AF burden. Conclusion Left atrial size and chronic cardiovascular comorbidities appear to be the primary drivers of CI-NOAF burden, whereas factors related to acute illness and critical care intervention paradoxically did not appear to be a substantial driver of arrhythmia burden. Further research is needed regarding drivers of AF and the efficacy of rhythm control intervention in this unique setting.

The Cockroach of Critical Care: Atrial Fibrillation and Predictors in the ICU Setting

“…In this issue of cardiology, Lancini and colleagues [2] investigated predictors of AF burden in critically illness-associated new-onset AF. They evaluated 7,030 admissions yielding a cohort of 309 patients (5%) with critically illness-associated new-onset AF.…”

“…This study is not without limitations, largely the challenges of defining “new onset” AF and accurately determining the observed burden in such a large cohort of patients. While Lancini and colleagues [2] excluded patients with previously documented AF, it is likely this cohort of proposed critical illness-associated new-onset AF patients is contaminated with those patients in whom preexisting paroxysmal AF was newly discovered but not truly “new” at all. This limitation unfortunately is a real-world clinical challenge in caring for and studying the AF population at large and may explain the lack of effect by illness severity on AF.…”

“…Despite these pitfalls, Lancini and colleagues’ [2] careful analysis adds to mounting evidence that the size of a patients left atrium may be the single most important variable in assessing risk for incident AF, analogous to the HbA1C and diabetic control. Left atrial size may be the summative, quantifiable aspect of a host of preexisting clinical factors that can easily be obtained in the ICU and certainly assist in uncovering and influencing treatment of our own version of a roach problem.…”