Risk score model for predicting complications in patients undergoing ventricular tachycardia ablation: insights from the National Inpatient Sample database

AIMS: Outcome data on ventricular tachycardia (VT) ablation has been limited to few experienced centres. We sought to identify complication rates, predictors, and create a risk score model for predicting complications in patients from real-world data. METHODS AND RESULTS: A total of 25 451 patients undergoing VT ablation from year 2006 to 2013 were identified from the National Inpatient Sample (NIS) database. The whole cohort was randomly divided into derivation cohort to derive the model and validation cohort to validate the model. Multivariate predictors of any complication were identified using regression model. Each predictor was assigned a risk score and each patient was assigned to one of the four groups (risk score in parenthesis) based on total combined risk score: Group 0 (0), Group 1 (1-5), Group 2 (6-10), and Group 3 (>11). The rate of 'any complication' and 'in-hospital mortality' in whole cohort was 14.7% and 2.8%, respectively. The predictors of any complication include chronic kidney disease, coagulopathy, chronic liver disease, stroke (cerebrovascular accident), emergency procedure, age ≥ 65 years, coronary artery disease, peripheral vascular disease, and female gender. There was a significant increase in complication rate in a linear fashion as the risk score increased. The incidence of any complications increased from 2.7% in Group 0 to 31% in Group 3. The risk score model performed well in predicting complications associated with VT ablation. CONCLUSION: Patients with higher risk scores have significant increase in any complication and in-hospital mortality from VT ablation. The simple risk score model can help to risk stratify patients prior to VT ablation.