Prediction of sinus rhythm in patients undergoing concomitant Cox maze procedure through a median sternotomy

Abstract

One of the challenges that exists when discussing the Cox maze procedure for atrial fibrillation (AF) with patients is predicting the success for a given patient. The purpose of the present study was to develop a scoring system using well-established clinical factors to predict the probability of sinus rhythm (SR) after surgery. The data from patients 1 and 2 years postoperatively were analyzed using logistic regression to predict SR, including the most accepted variables associated with failure (age, left atrium size, AF duration, AF type). Regression models were applied using hypothetical patients to examine the predicted probability of SR. The predictors of 1-year SR were a shorter AF duration and greater surgeon experience performing surgical ablation. The predictors at 2 years were a shorter AF duration and smaller left atrium. The 1-year prediction model applied to hypothetical data found a 1-cm increase in left atrial size associated with a 0.4% reduction in SR probability, a 5-year increase in AF duration associated with a 0.8% reduction, and a reduction by 50 cases of surgeon experience associated with a 1.0% reduction. The 2-year model found a 1-cm increase in left atrial size associated with a 1.0% reduction in SR probability, a 5-year increase in AF duration associated with a 0.8% reduction, and a reduction by 50 cases of experience associated with a 0.2% reduction. Our findings are the first step in establishing a risk scoring system to better predict the outcomes after surgical ablation for AF and improve the ability to discuss the risk and benefits with patients.