Novel Mathematical Model to Predict Repeated Ablation for Recurrent Atrial Fibrillation

Overview

Atrial fibrillation (AF) is the most common heart rhythm abnormality and a leading cause of stroke, costing $26 billion/year. Radiofrequency Catheter Ablation is used to treat AF. However, recurrence of AF can occur requiring repeated ablation procedures. Currently, there are no accurate models to predict probability of repeated ablation. I developed a novel mathematical model combining Electrocardiographic, Echocardiographic and clinical parameters to predict recurrence of AF after ablation. Procedure: In a retrospective review (n=46), 23 patients who underwent repeated ablation for recurrent AF were compared to 23 controls that underwent ablation only once. Data: Of the analyzed parameters, age was not predictive. P wave duration (PWD) by Electrocardiogram, Left atrial enlargement (LAE) by echocardiogram, Gender and Obstructive Sleep Apnea (OSA) were significant predictors of repeated ablation. Highly significant predictors were, PWD (p-value of < 0.0001, Chi- square 7.64) and OSA (p-value = 0.0004, Chi-square 2.40). Less significant predictors: Age, Gender and LAE were removed by backward selection procedure. Final simplified model was developed using the highly significant predictors: PWD and OSA, and a risk score was developed assigning a weighted integer to each, based on the predictor’s coefficient in the final regression model. A score of ≥ 4 predicted increased risk of repeated ablation. When scores were compared with outcome, the final model had an overall accuracy of 91.3%. Conclusions: A summed risk score predicted probability of repeated ablation with overall accuracy of 91.3%. The developed model will help doctors select proper patients, avoid repeated procedures and reduce costs.

Competition history

• AJAS 2017 Category not listed

Related projects

Closest projects by meaning, across every fair and year in the corpus.

Browse more like this

• All Biomedical and Health Sciences projects
• AJAS Biomedical and Health Sciences projects
• Every AJAS 2017 project