Incidence and predictors of new-onset atrial fibrillation in noncardiac intensive care unit patients.

Background: Atrial fibrillation (AF) is thought to be a relatively common arrhythmia in the setting of noncardiac intensive care unit (ICU). However, data concerning AF deriving from such populations are scarce. In addition, it is unclear which of the wide spectrum of AF predictors are relevant to the ICU setting.

Objectives: The aim of our study was to evaluate the incidence of new-onset AF and investigate the factors that contribute to its occurrence in ICU patients.

Methods: We prospectively studied all patients admitted to our ICU during a 1-year period. Patients admitted for brief postoperative monitoring and patients with chronic or intermittent AF and AF present upon admission were excluded. A number of conditions incriminated as AF risk factors or "triggers" from demographics, medical history, present disease, and cardiac echocardiography as well as circumstances of AF onset were recorded.

Results: The study population consisted of 133 patients (90 males). Atrial fibrillation was observed in 15% of them. Age older than 65 years (P=.001), arterial hypertension (P=.03), systemic inflammatory response syndrome (P<.001), sepsis (P=.001), left atrial dilatation (P=.01), and diastolic dysfunction (P=.04) were significantly associated with the occurrence of AF. By multivariate analysis, it was demonstrated that only older than 65 years (odds ratio, 7.0; 95% confidence interval, 2.0-24.6; P=.003) and sepsis (odds ratio, 6.5; 95% confidence interval, 2.0-21.1; P=.002) independently predict new-onset AF. Patients manifesting AF were frequently hypovolemic (30%) and had electrolyte disorders (40%) as well as elevated and rising serum C-reactive protein (70%).

Conclusion: A significant fraction of ICU patients manifest AF. The predictors of interest for the ICU patients might be considerably different than those of the general population and other subgroups with systemic inflammation possibly having a pivotal role.