BackgroundAtrial fibrillation (AF) is common among patients in the intensive care unit (ICU) and can be triggered by severe illness or preexisting conditions. It is debated if AF is an independent predictor of poor outcome. MethodsData derives from a single center retrospective registry including all patients with a stay on the medical ICU for >24 h. The primary endpoint was ICU survival. Secondary endpoints included receiving mechanical support (renal, respiratory or circulatory), hemodynamic parameters during AF, rate and rhythm control strategies, anticoagulation, and documentation. ResultsA total of 616 patients (male gender 62.3%, median age 75 years) were included in our analysis. New-onset AF was diagnosed in 87 patients (14.1%), 136 (22.1%) presented with preexisting AF, and 393 (63.8%) did not develop AF. Initial episodes of new-onset AF exhibited higher hemodynamic instability than episodes in preexisting cases, with elevated heart rates and increased catecholamine doses (both p < 0.001). ICU survival in new-onset AF was 80.5% (70/87) compared to 92.4% (363/393) in patients without AF (OR 0.340, CI 0.182–0.658, p < 0.001). Likewise, ICU survival in preexisting AF was 86.8% (118/136) was significantly lower compared to no AF (OR 0.542, CI 0.290–0.986, p = 0.050*). Independent predictors of ICU survival for patients were atrial fibrillation (p = 0.016), resuscitation before or during ICU stay (p < 0.001), and receiving acute dialysis on ICU (p = 0.002). ConclusionsICU survival is noticeably lower in patients with new-onset or preexisting atrial fibrillation compared to those without. Patients who develop new-onset AF during their ICU stay warrant special attention for both short-term and long-term care strategies.