Abstract
BACKGROUND: Robotic-assisted percutaneous coronary intervention (R-PCI) is an efficacious and safe treatment option for coronary artery disease. However, predictors of manual support during R-PCI are unknown, which we aimed to investigate in a multi-center study. METHODS: We utilized patient-level data from R-PCIs carried out from 2020 to 2022 at four sites in Germany. Manual support was defined as the combination of partial manual assistance, where the procedure is ultimately completed using robotic techniques, and manual conversion. A two-step selection process based on akaike information criteria was used to identify the ideal multivariable model predicting manual support. RESULTS: In 210 patients (median age 69.0 years; 25.7% female), a total of 231 coronary lesions were treated by R-PCI. Manual support was needed in 46 lesions (19.9%). Procedures requiring manual support were associated with significantly longer procedural times, greater total contrast fluid volumes, longer fluoroscopy times, and higher dose-area products. Amongst the predictors of manual support were lesions in the left anterior descending artery [OR: 1.09 (95%-CI: 0.99-1.20)], aorto-ostial lesions [OR: 1.35 (95%-CI: 1.11-1.64)], chronic total occlusions [OR: 1.78 (95%-CI: 1.38-2.31)], true bifurcations [OR: 1.37 (95%-CI: 1.17-1.59)], and severe calcification [OR: 1.13 (95%-CI: 1.00-1.27)]. CONCLUSION: Our findings reveal that nearly one out five of patients undergoing R-PCI required manual support, which was linked to longer procedure durations. Predictors of manual support reflected characteristics of more complex coronary lesions. These results highlight the limitations of current R-PCI platforms and underscore the need for technical advancements to address different clinical scenarios.