Atrial fibrillation (AF) is the most common acquired arrhythmia in the intensive care unit (ICU) and is associated with major cardio-vascular complications. However, the association between AF and in-hospital morbi-mortality in ICU remains debated. Furthermore, no prospective study previously assessed long-term outcomes of new-onset AF.

To determine the incidence, clinical features, in-hospital and long-term outcomes of new-onset AF in the ICU.

We performed a prospective mono-centric observational study. All consecutive patients admitted during a 5-month period were included. Three groups were identified: no AF (No-AF); prior AF (Pr-AF) in case of AF history; and new-onset AF (New-AF) when first AF diagnosis was made during ICU stay. Survivors were followed up during a 6-month period after discharge.

We included 110 patients: 23 New-AF, 14 Pr-AF, and 73 No-AF. New-AF were significantly older than No-AF patients. New-AF presented significantly more acute pulmonary oedema and cardiogenic shock than No-AF patients. New-AF was significantly associated with more inotropes or vasopressors use, mechanical ventilation and renal replacement, and had higher in-hospital mortality (43.5% vs. 12%, P=0.001) than No-AF patients. New-AF had significantly higher levels of troponin and BNP than No-AF patients. At the time of AF occurrence in New-AF group, plasmatic potassium levels were lower than upon admission (3.5 vs. 4.35mmol/L; P<0.001). New-AF patients presented significantly more left ventricular systolic and diastolic dysfunction, right ventricular systolic dysfunction and pulmonary arterial hypertension. There was no long-term difference in mortality, heart failure or embolic events between New-AF and No-AF patients.

In the ICU, patients with New-onset AF exhibited more cardiogenic and respiratory failure, and had increased in-hospital mortality. No difference in morbi-mortality was observed 6 months after discharge.