Limitations
The retrospective, nonrandomized, single-centre design of our study has some limitations. Because the CM group comprised the initial patients who underwent LAAC, there might have been a learning curve, which would be in favour of the MM group. However, PT events occurred in the 110th and 174th patients in the MM group. This finding suggests there was no significant correlation between the incidence of PE/PT and the learning curve. Additionally, the limited number of events limited the power for our ability to detect a significant association, as shown by the wide confidence intervals.