Understanding the learning curve in robotic-assisted cardiac surgery and its application on curriculum development - systematic narrative review.

Since robotic-assisted cardiac surgery was established, the adoption rate was slow despite proven its efficacy and safety. Limited knowledge about the learning curve (LC) has led to lacking recognised training programme which in turn jeopardises patient safety and reduces the uptake rate. This study aims to review current literature, identify any knowledge gaps and quantify the LC. In addition, the study focuses on the application of LC knowledge on the curriculum development which has not been covered in the literature before. This systematic narrative review was conducted according to PRISMA-20 guideline. Ovid MEDLINE, PubMed, EMBASE and SCOPUS were searched from database inception to the 5th of February 2024. Eligible articles were those assessing LC in robotic-assisted cardiac surgery and reported in English. Among 24 studies which met the eligibility criteria (Table 1), 12 studies are for robotic-assisted coronary artery bypass, 9 for robotic-assisted mitral valve repair and 3 for robotic-assisted atrial septal defect repair. All studies were observational. Reporting LC exhibited substantial heterogeneity in terms of outcome variables and statistical analysis. None of the studies have quantified the surgeons' previous experience. Finally, having structured training programme is the most recommended method to mitigate the steep LC while creating standardised reporting system has been advised to decrease heterogeneity in the future studies. Current literature shows high heterogeneity in defining LC which creates challenges in developing safe curriculum. Nonetheless, adopting structured programme with good exposure to simulation sessions are deemed effective approach to reduce LC and improve patient safety.