PhD defence X. (Xu) Wang

Brief summary on the aim of the doctoral thesis:

Many congenital heart defects (CHD) involve abnormalities in the right ventricular outflow tract (RVOT), necessitating surgical RVOT reconstructions, in which homograft is commonly used. Typically, lifelong follow-up is required to detect RVOT dysfunctions and support clinical decision-making to optimize outcomes for the patient. With more patients data obtained through long-term follow-up, advanced statistical methods provide powerful tools to gain new insights and thus optimize outcomes for the patient. The primary objective of this PhD project is to integrate complex statistical methodologies into clinical research focused on patients who have undergone RVOT reconstruction with a homograft, making analytical processes more practical, understandable, and reproducible. During this project, we firstly performed a meta-analysis systematic review about outcomes after RVOT reconstruction with valve substitutes and found most procedures can be performed with an acceptable complications rate. However, certain CHD groups exhibit higher mortality and reintervention rates, indicating the need to tailor follow-up protocols, accounting for factors, e.g., etiologies, ages, and implanted valve substitutes. A statistical prime about clinical applications of linear mixed-effects models was introduced afterwards. Survival analysis, mixed-effects models, and joint models were used to analyse longitudinal data obtained from patients of Tetralogy-of-Fallot (TOF) and Dextro-Transposition-of-the-Great-Arteries (d-TGA), focusing on male-female differences in long-term clinical and homograft outcomes. Given the importance of mortality of the-matched-general-population in survival analysis, methodologies to calculate it were illustrated thoroughly with a web-based shiny App available. The findings and outputs of this PhD project can be used to guide clinical-decision making and facilitate qualified clinical research.