This was a nonprespecified retrospective analysis of prospectively collected data. Information regarding patients’ baseline characteristics and procedural details was collected prospectively. Patients underwent regular follow-ups after the procedure, and data on adverse events and readmissions were recorded in each local database.

From an initial cohort of 1653 patients who underwent percutaneous LAAC, we excluded those who died during the procedure or before discharge (9 [0.5%]) and those whose procedure was performed less than 1 year before the analysis (225 [14%]). Thus, a total of 1419 patients undergoing LAAC in 9 university hospitals in Canada and Europe from 2009 to 2022 were included in the analysis.

The indications for LAAC for each patient were assessed by a multidisciplinary team, including interventional cardiologists, electrophysiologists, and specialists in cardiac image at each center. LAAC procedures were performed using commercially approved devices. A LAAC procedure was considered successful if the patient was discharged alive, without any in-hospital embolic events, and with the device implanted in the left atrial appendage. Complete occlusion was defined as patients with no significant peri-device leak.

Baseline, procedural, and follow-up data were prospectively collected in a dedicated database. If present, contraindication for OAC was classified as relative (eg, previous gastrointestinal [GI] bleeding, anemia) or absolute (eg, history of intracranial [IC] bleeding, major/life threatening bleeding within the first 24hours following LAAC) (table 1 of the supplementary data).

Clinical follow-up was performed via clinical visits or telephone contacts at 1, 6, and 12 months after LAAC and yearly thereafter. Clinical outcomes were defined according to Valve Academic Research Consortium-2 criteria.7 The study was performed in accordance with the local ethics committee of each participating center and all patients provided written informed consent before the procedure.

Readmission was defined as the need for hospitalization in a hospital ward or intensive care unit. Visits to the emergency room lasting less than 24hours were not counted as hospital readmissions. Readmission date, length of stay, primary and secondary reasons for hospitalization, and in-hospital death were recorded after a detailed medical records review. The time to readmission was calculated from the date of procedure discharge to the first hospital readmission day. Readmissions were classified as early (≤ 30 days) or late (31 to 365 days).

The causes of hospital readmission were classified as either cardiac or noncardiac. Cardiac causes included heart failure, acute coronary syndrome, arrhythmia, LAAC-related hospitalization (such as endocarditis, device-related thrombosis [DRT], need for reintervention, and procedure-related pericardial effusion), or planned cardiac interventions. Noncardiac causes included acute respiratory disease, bleeding, stroke, peripheral vascular events, infectious diseases, trauma, gastrointestinal events, neurological events, any noncardiac elective intervention, or other causes.

Qualitative variables were reported as percentages, and continuous data as mean±standard deviation or median [interquartile range], depending on their distribution. Continuous variables were compared using the Student t test or Mann-Whitney U rank-sum tests, as appropriate. Qualitative variables were compared using the chi-square or Fisher exact tests.

The predictors of early readmission were determined using a logistic regression analysis. Variables with P values<.20 in the univariable analysis were included in the multivariable model, and the Akaike information criterion were used to select the best predictive Cox model. Cox multivariable regression analysis was performed to identify independent predictors of late readmission and predictors of mortality. Survival curves were summarized using Kaplan-Meier estimates, and log-rank tests were used to compare groups. A 2-sided alpha level of 0.05 was used for all statistical testing. All statistical analyses were performed using SAS version 9.4 (SAS Institute, United States).

A total of 257 patients (18.1%) were readmitted within the first year after the LAAC procedure. Early readmission (≤ 30 days) occurred in 46 patients (17.9%). From 31 days to 1 year after LAAC, 211patients (82.1%) were readmitted (figure 1).

Figure 1.

Rate early and late readmissions after LAAC procedure. LAAC, left atrial appendage closure.

(0.1MB).

The causes of readmission are summarized in table 2 and figure 2. Noncardiac origin accounted for most rehospitalizations (182 hospitalizations [70.8%]), with bleeding events being the most common cause (24.5% of all readmissions). A total of 75 patients (29.2%) were admitted due to cardiac causes, with heart failure being the primary cardiac cause and the second most common cause overall (20.6% of total hospitalizations). The causes of admission were similar for both early and late rehospitalizations.

Figure 2.

Causes of readmission after LAAC. CHF, congestive heart failure; LAAC, left atrial appendage closure; TIA, transient ischemic attack.

(0.13MB).

Patients requiring readmission were older (75.7±8.3 years vs 77.3±7.9 years; P=.005) and more often had a history of diabetes mellitus (32.2% vs 42.0%, P=.003), peripheral artery disease (22.3% vs 31.5%; P=.002), and heart failure (22.9% vs 37.4%; P <.001). Consequently, patients readmitted after LAAC had a higher ischemic risk as assessed by the CHA2DS2-VASc score (4.33±1.63 vs 4.54±1.38; P=.007). In addition, those readmitted within the first year after LAAC had a lower body mass index (27.4±4.9kg/m2 vs 26.6±4.7kg/m2; P=.008) and more often had a previous history of impaired renal function (35.0% vs 50.9%; P <.001), and GI bleeding (32.2% vs 49.8%, P <.001). The HAS-BLED score was higher in patients with readmission (3.49±1.09 vs 3.94±0.99; P <.001). Procedural characteristics did not differ between groups, and the rate of procedural success was also equivalent (99.3% vs 100%; P=.364) (table 1).

The predictors for early and late readmission after LAAC are presented in table 3. In the multivariable analysis, the only factor associated with early readmission was a previous history of GI bleeding (odds ratio [OR], 2.65; 95% confidence interval [95%CI], 1.23-5.71; P=.013). A lower BMI (hazard ratio [HR], 0.96; 95%CI, 0.93-0.99; P=.008), history of diabetes mellitus (HR, 1.38; 95%CI, 1.02-1.86; P=.036), chronic kidney disease (CKD) (HR, 1.60; 95%CI, 1.21-2.13; P=.001) and previous history of heart failure (H,: 1.69; 95%CI, 1.26-2.27; P <.001) were associated with a higher risk of late readmission.

Last, a lower BMI (HR, 0.96; 95%CI, 0.93-0.98; P=.004), history of diabetes mellitus (HR, 1.34; 95%CI, 1.02-1.76; P=.039), CKD (HR, 1.40; 95%CI, 1.07-1.82; P=.013), heart failure (HR, 1.51; 95%CI, 1.16-1.97; P=.002), previous GI bleeding (HR, 1.68; 95%CI, 1.30-2.20; P <.001) and dual antiplatelet therapy (DAPT) at hospital discharge (HR, 1.40; 95%CI, 1.08-1.80; P=.009) were associated with a higher risk of any rehospitalization (early or late) (table 4). DAPT was not independently associated with a higher risk of early readmission. The median time to readmission in patients on DAPT was 116 [36-216] days, and 19% of them were readmitted due to bleeding.

Patients were followed up for a median time of 33 months (IQR, 17 to 55 months), and follow-up was complete in all but 54 (3.8%) patients who were lost to follow-up. A total of 274 patients (19.3%) died within the first 2 years after LAAC. The predictors of mortality within the first 2 years are presented in table 5.

Early readmission (HR, 2.12; 95%CI, 1.22-3.70; P=.008) and late readmission (HR, 1.75; 95%CI, 1.41-2.17; P <.001) were independently associated with a higher risk of 2-year mortality from any cause. Factors associated with a higher risk of cardiovascular mortality are detailed in table 6. The Kaplan-Meier curves for mortality within the first 2 years after LAAC according to the time or the readmission status (early, late or no readmission) are shown in figure 3.

Figure 3.

Kaplan-Meier survival curves for 2-year mortality. LAAC, left atrial appendage closure.

(0.17MB).

Several studies have evaluated short-term (30-day) readmission rates after percutaneous interventions. After transcatheter aortic valve replacement (TAVR) and percutaneous coronary intervention (PCI), these rates ranged from approximately 3% to 15%.8,9 In the LAAC population, the 30-day readmission rate after the procedure has been ∼9%, which is higher than the 3.2% rate observed in our study.10 This reduced rate of early readmissions may be partially related to the inclusion of more contemporary LAAC cases (40% of our study cohort was treated in the last 5 years), along with potential differences in post-procedural management (eg, more aggressive antithrombotic management, with a combination of anticoagulation and antiplatelet has been commonly used in US centers, compared with dual or single antiplatelet therapy in Canada and Europe).11

Information on late readmissions after LAAC is scarce. Barsoum et al.12 reported a 15% readmission rate within the first 45 days after LAAC, while Murthi et al.13 found a 90-day rehospitalization rate of 17%.12,14 In our cohort, 18.1% of patients were readmitted within the first year after LAAC. This rate is lower than that reported for other percutaneous procedures, such as TAVR (∼44%) and PCI (∼35%).8,9 The difference between the rates reported in this study versus those performed in other percutaneous interventions may be explained by several factors: LAAC is a preventive procedure, while PCI are TAVR are often performed in unstable and often more vulnerable patients, which likely leads to a higher risk of readmission. Furthermore, because LAAC is a prophylactic intervention, patients at a higher risk of reintervention might have been excluded to avoid futility.

We found that noncardiac causes accounted for 70.8% of all readmissions, while cardiac causes contributed 29.2%. Bleeding events represented approximately 25% of all rehospitalizations, which agrees with previous reports in which bleeding-related readmissions ranged from 9% to 19% .13,15 The main indication for LAAC is relative or absolute contraindication for OAC, reflecting the high bleeding risk in these patients.1,4 Bleeding events remain a significant concern after LAAC, particularly during the phase of intensive pharmacotherapy after implantation to prevent DRT.16,17

In our population, 3 patients (1.2%) had a DRT leading to rehospitalization. This rate is lower than the 3.8% reported in a previous meta-analysis.18 However, our study includes only DRT patients requiring hospitalization. Therefore, we cannot directly compare our findings with the aforementioned meta-analysis, as our study did not take into account DRT patients who were managed in the outpatient clinic, which likely represents the majority of cases. In addition, 1 patient (0.4%) had a device embolization, leading to late readmission. A systematic review evaluating device embolization after LAAC reported a 0.6% rate (when centers did report the overall numbers of LAAC performed), which is similar to the rate in our study.19 Although most device embolizations occur within 24hours of the procedure, almost 20% occured more than 45 days later.19

The stroke rate in our study was 2.2% in the first 30 days and 5.2% from 31 days to 1 year, which is similar to other studies reporting rates ranging from 0.5% to 6.6% (30 days) and up to 1.2% (180 days).10,11,15 The rate of IC bleeding was 2.2% at 30 days and 2.4% at 1 year. These findings, together with previous experiences, highlight that in patients undergoing LAAC, the risk of bleeding may outweigh the risk of thrombotic or ischemic events.

To date, no definitive evidence exists regarding the optimal antithrombotic regimen following LAAC, and different regimens have been used in real-world practice.20,21 In our study, DAPT at discharge was associated with a higher risk of overall readmission. DAPT is commonly stopped 3 to 6 months after the procedure (in our population 86% underwent de-escalation of their antithrombotic therapy in the first 6 months after the procedure). This association between DAPT and readmission might be driven by the combination of early rehospitalizations and those occurring between 1 and 6 months after the procedure. While our study design prevents us from formulating a causal inference, this association would suggest that requiring a more intense antithrombotic therapy could increase the risk of readmission in patients undergoing LAAC. However, it is crucial to balance this with the risk of device thrombosis when choosing the antithrombotic regimen. Dedicated clinical trials are needed to identify the optimal antithrombotic strategy, considering both ischemic and bleeding risks, to better guide therapy after LAAC.

Several ongoing trials are currently evaluating the safety and efficacy of different antithrombotic regimens after LAAC, such as the ANDES trial (NCT03568890), the APPENDAGE trial (NCT04796714) and the ADALA study.22 Furthermore, an observational study conducted by Mesnier et al.23 supported the interruption of all antithrombotic medication once the device is endothelized and reported no association with increased risk of mortality or thromboembolic events (2 years follow-up).

Heart failure was the second most common cause of rehospitalization, accounting for approximately 20% of readmissions. This observation is consistent with those of previous studies in which heart failure was either the primary or secondary cause of rehospitalization.13,15 The left atrial appendage has been demonstrated to play a role in hemodynamic regulation through its endocrine function.24–26 Removing this structure might have a negative impact on hemodynamic regulation.26 The LAA is also known to serve as a reservoir when atrial pressure and volume increase, which contributes to atrial compliance during ventricular systole.26 Whether LAAC is associated with a poorer fluid management leading to an increased risk of heart failure should be evaluated in dedicated studies. Additionally, the volume overload during the procedure, intended to improve device sizing, might exacerbate congestive status, increasing the risk of heart failure-related readmissions in the first few days following the procedure.27 Therefore, in patients undergoing LAAC, a comprehensive assessment of congestion status—through physical exams and NT-proBNP levels—and optimization of diuretic therapy could help reduce heart failure-related rehospitalizations.

In the present study, previous GI bleeding was associated with a higher risk of readmission within the first 30 days after the procedure, which is consistent with previous experiences.15,28 We also identified several predictors of late readmissions, including lower BMI, diabetes mellitus, CKD, and heart failure. Interestingly, Mesnier et al.29 identified similar items as predictors of early mortality after LAAC, such as older age, lower BMI, reduced estimated glomerular filtration rate, prior diabetes, and prior heart failure. It is well established that patients with a significant comorbidity burden are at higher risk of complications following catheter-based procedures, leading to an increased risk of readmission.8 In addition, both early and late readmissions were independently associated with a higher risk of mortality after 2 years.

These findings highlight that comorbidity burden is associated with poorer results after LAAC including an increased risk of rehospitalizations and a reduced survival. Thus, an accurate preprocedural evaluation is paramount to make a proper assessment of the benefit/risk ratio, avoiding unnecessary interventions in patients who may not obtain a favorable net benefit due to a high risk of adverse events or early mortality.

Our study has several limitations that should be acknowledged. First, although data were collected prospectively, this analysis was not prespecified, which may have introduced unmeasured confounding factors, such as patient frailty.

Second, while some patients experienced multiple readmissions after LAAC, only the first readmission for each patient was analyzed. Thus, the impact of subsequent hospitalizations could not be assessed in this study.

Third, the level of experience of the operators and participating centers was not considered, which might have influenced the procedure outcomes. Due to these limitations, the conclusions drawn from this analysis and should be validated through future dedicated studies.