DISCUSSION:
The main findings of our current investigation are: (1) Over the study period from 2015-2017, there has been a significant increase in the number of Watchman device implantation procedures in the United States (from 1,195 in year 2015 to 11,165 devices in year 2017, p < 0.01). (2) There has been a decline in the rate of complications over the study period primarily driven by lower rates of cardiovascular and neurological complications. (3) Overall mortality continues to be low during the study period, with an even lower trend towards reduced mortality over the study years (1.3% in year 2015 vs. 0.1% in year 2017, p < 0.01). (4) The total length of stay and hospitalization costs after Watchman implantation declined over the study period.
Percutaneous LAAO with the Watchman device provides a viable alternative to oral anticoagulation in select patients based on randomized trials that have shown efficacy and safety of utilizing this approach for stroke risk reduction (10,11). The FDA approved the device for commercial use in United States in March of 2015 and the Watchman implant procedure currently carries a class IIb recommendation in patients with non-valvular AF at risk for stroke per the latest American College of Cardiology guidelines (14). Our analysis of a contemporary, real-world, national database sampling US practice since FDA approval of Watchman showed consistent increased utilization of device procedures suggesting gradual assimilation of this device implantation procedure in clinical practice. Additionally, our analysis also showed that complications and in-patient mortality associated with implantation of Watchman devices continued to show a downward trend.
Our study showed a significant decline in the overall complication rate over the study period (26.4% in year 2015 vs. 7.9% in year 2017, p < 0.01). This downtrend was primarily driven by a reduction in cardiovascular and neurological complications over the study period. Cardiac perforation complications including cardiac tamponade were encountered in the landmark PROTECT trial where its prevalence was approximately 4.3% (10). Subsequently, with improved operator experience and better implant techniques, the incidence of this complication was lowered to 1.9% in the PREVAIL trial, 1.4% in CAP (Continued Access to PROTECT), 1.9% in CAP2 (Continued Access to PREVAIL), and 0.3% in EWOLUTION registries (11,15,16). The overall rate of cardiac tamponade in the current study of contemporary Watchman patients was 0.8%. This rate is similar to post-FDA approval study led by Reddy et al. in which the authors reported a nearly similar rate of cardiac tamponade at 1%; still higher than the European EWOLUTION registry which reported cardiac tamponade rate of 0.3% (17). It is also similar to a recently published contemporary registry of Watchman implantations from National Cardiovascular Data Registry which analyzed nearly 38,000 patients (18). It may be expected that with more widespread availability of the Watchman device in US practice that the rate of cardiac tamponade may continue to decline over the coming years. The rate of ischemic stroke/transient ischemic attack (TIA) was 1% in our study cohort with most cases reported in 2015 and a significant downtrend since that time (0.7% in year 2016 and 0.4% in year 2017, p < 0.01). Vuddanda et al. have shown nearly similar ischemic stroke/TIA rates of about 0.5% when analyzing Watchman implants from year 2016 (13). These strokes are presumed to be due to inadvertent air or clot embolization from the trans-septal sheath and enhanced physician training should continue to mitigate this risk.
The current study showed a downward trend in in-hospital mortality suggesting improved safety with the device with more operator experience (from 1.3% in year 2015 to 0.1% in year 2017, p < 0.01). Improvement in mortality rate trends over time seen in the current study of real-world patients was also seen in previous clinical studies, including the CAP2 registry which supplemented the PREVAIL trial and was designed to continue long-term accrual of data, and showed a mortality rate of 0.2% within 7 days of Watchman implant (15). A subsequent study utilizing National Inpatient Sample database by Vuddanda showed a mortality rate of 0.3% at discharge for combined endocardial and epicardial based approaches for LAA occlusion (13). In another post-FDA approval analysis of more than 3,800 patients undergoing Watchman implantation from March, 2015 to May, 2016, Reddy et al. demonstrated procedure related mortality of 0.078% (17). The prospective EWOLUTION registry that enrolled more than 1,000 consecutive patients undergoing Watchman implantation outside the US also showed low procedure related mortality of 0.1% (16). These earlier studies along with our more contemporary data suggest that overall implantation of the Watchman device is associated with low absolute rates of mortality that is decreasing over time.