Po-02-200 circadian occurence patterns of atrial fibrillation after cardiac surgery during 2.5 years of continuous rhythm monitoring

Atrial fibrillation (AF) is a frequent complication after cardiac surgery and current theories suggest that autonomic nervous system plays a major role in early-POAF development, but its role in late-recurrences is still unknown. Previous studies in non-surgical patients have shown distinct preferential AF-occurrence patterns, such as adrenergic/daytime AF, or nocturnal AF. However, the circadian onset patterns for POAF episodes have never been studied before. To investigate POAF circadian onset patterns after cardiac surgery during 2.5 years of continuous rhythm monitoring. Patients with and without history of AF undergoing cardiac surgery were included. Continuous rhythm monitoring was performed over a period of 2.5 years. Circadian distribution of AF-episodes onset was presented in radial histograms for the whole FU period, early postoperative phase (first 90 days), and late postoperative phase (after first 90 days). In addition, Bayesian circular regression was performed to calculate the average episode onset, along with 95% credibility intervals, for different classes of episode duration. The study population consisted of 134 patients with a median follow-up of 2.7 years (IQR: 2.2-2.9 years). Eighty-one (60.4%) patients developed AF during the follow-up period. Early-POAF (first 90 days) occurred in 75 patients (56.0%) and late-POAF (after first 90 days) in 47 patients (35.1%). In patients without history of AF, there was a heterogenuous POAF-episode onset pattern with peak incidence at 4:18pm (95%CI: 6:53am to 2:16am) (Figure 1). Paroxysmal AF and persistent AF patients showed more homogenuous occurrence patterns with peak incidence at 10:07am (95%CI: 9:29am to 10:51am) for paroxysmal AF, and 9:16am (95%CI: 8:42am till 9:50am) for persistent AF (Figure 1). Peak incidence of early-POAF was at night and in the morning for all rhythm groups (Figure 1), whearas late-POAF also started during the afternoon in patients without history of AF (Figure 1). In patients without history of AF, shorter AF episodes started predominantly during the afternoon whereas longer episodes started at night and in the morning (Table 1). In paroxysmal and persistent AF patients, most episodes started in the morning hours (Table 1). In patients with history of AF, POAF starts mostly during the night and in the morning, wheras patients with new-onset AF show a more heterogeneuous daytime occurrence pattern.