Successful Implementation of Standardized Longitudinal Follow-up after Surgery for Congenital Heart Disease

To understand and improve long-term congenital heart surgery outcomes, longitudinal data are essential but challenging to capture. We evaluated a novel program to facilitate routine longitudinal data collection. Patients undergoing Society of Thoracic Surgeons Benchmark Operations at two centers (2007-2019) were included. Annual follow-up data (vital status and age-specific burden of disease questionnaires) were collected using an automated email system along with medical record and local registry data queries, with coordinator follow-up if needed. Completeness of follow-up and associated factors were analyzed in multivariable logistic regression. Of 4,075 subjects, vital status was captured for 98% (median follow-up 6.0 years; IQR 3.1-9.2). Vital status at one year was known for 95%; overall follow-up completeness (observed-to-potential follow-up time) was 81%. At least one follow-up questionnaire was completed by 85%, with 58% utilizing the automated email system. Those with genetic/non-cardiac abnormalities (OR 2.28, 95%CI 1.28-4.28), single ventricle palliation (Norwood OR 11.2, 95%CI 3.01-72.7; Glenn/hemi-Fontan OR 7.79, 95%CI 2.21-49.5), and tetralogy of Fallot repair (OR 2.43, 95%CI 1.11-5.59) had higher odds of known vital status. Follow-up questionnaire completion had higher odds for non-Hispanic white patients (OR 1.54, 95%CI: 1.27-1.86) and higher median income households (OR 1.05/$10,000, 95%CI: 1.01-1.09). This program captured vital status and burden of disease, with one-year survival determined for nearly all patients. Automation reduced human resource needs by more than half. Future efforts will focus on expanding to other centers and improving follow-up for disadvantaged groups.