Introduction: We aimed to investigate whether a modified implantation method facilitating a fully open umbrella can reduce the pericardial effusion/pericardial tamponade (PE/PT) rate after left atrial appendage closure (LAAC) with the LAmbre device compared with the conventional method (CM) in patients with non-valvular atrial fibrillation (NVAF). Methods and results: Patients with NVAF who received either isolated LAAC or combined catheter ablation and LAAC using the LAmbre device at the First Affiliated Hospital of Wenzhou Medical University from January 2018 to December 2019 were enrolled. CM was used for device implantation in the initial 59 patients, while a modified method (MM) was used in the remaining 165 patients. Successful implantation was achieved in 98.3% of patients in the CM group and in 98.8% in the MM group. A higher rate of a fully open umbrella (98.8% vs 69%, P<0.001), less requirement for recapture (46% vs 62.1%, P=0.036), and a lower incidence of delayed PE/PT (1.2% vs 8.6%, P=0.005) were found in the MM group compared with in the CM group. All of the five delayed PT events occurred in patients with combined treatment. An umbrella that was not fully open was the only factor associated with delayed PE/PT events in a multivariable Cox model. Conclusions: LAAC with the LAmbre device using an MM significantly increases the rate of a fully open umbrella and decreases the requirement for recapture and the incidence of delayed PE/PT. This method is more effective in patients with combined treatment.