A Risk Prediction Model For Pre-discharge Major Residual Lesions Or Unplanned Reinterventions Following Congenital Mitral Valve Repair
Aditya Sengupta, MD, MPH, Kimberlee Gauvreau, ScD, Ji M. Lee, BS, Christopher W. Baird, MD, Sitaram Emani, MD, Pedro J. del Nido, MD, Meena Nathan, MD, MPH
Boston Children's Hospital, Boston, MA, USA.
Objective(s): We sought to develop a novel risk prediction model for pre-discharge major left atrioventricular valve (LAVV) residua or unplanned LAVV reintervention following congenital LAVV repair.
Methods: This was a single-center review of patients that underwent congenital LAVV repair (excluding primary repair of canal-type defects) from 01/2000-12/2020 and survived to discharge. The primary outcome was major LAVV residua (mean gradient >6 mmHg or ≥moderate regurgitation on the discharge echocardiogram) or pre-discharge unplanned LAVV reintervention. Risk factors of interest included age, preoperative LAVV stenosis and regurgitation severity, mixed LAVV disease, single ventricle palliation, concomitant left heart procedure, unfavorable LAVV anatomy, repair technique, and intraoperative post-repair LAVV stenosis and regurgitation severity. Logistic regression was used to develop a weighted risk score for the primary outcome. Internal validation using a bootstrap-resampling approach was performed.
Results: Of 866 patients meeting entry criteria, 202 (23.3%) developed the primary outcome. The final model had a C-statistic of 0.82. A weighted risk score was formulated based on the coefficients in the final model (Table). Patients were categorized as low (score 0-5), medium (score 6-10), high (score 11-15), or very high (score ≥16) risk. The probability of the primary outcome was 5.0±1.7%, 15.2±6.7%, 45.9±12.6%, and 76.7±8.8% for low, medium, high, and very high risk patients, respectively.
Conclusions: Our prediction model may guide prognostication of high-risk patients following congenital LAVV repair. 
Back to 2023 Abstracts