Previous studies have shown the safety of early discharge pathways in selected patients and using selected transcatheter heart valves. Hence, we sought to evaluate the safety of next-day discharge (NDD) in patients who underwent transfemoral transcatheter aortic valve implantation (TF-TAVI) with the ACURATE neo/neo2 (Boston Scientific, Marlborough, Massachusetts) self-expanding aortic bioprosthesis. Patients who underwent TF-TAVI between January 2018 and April 2023 were prospectively included. Patients were stratified into 3 groups according to discharge times within 24 hours (NDD), between 24 and 48 hours, and those discharged >48 hours after TAVI. The primary outcome was the first unplanned readmission at 30 days after TAVI. Log-rank test was used to assess the differences in the outcome of interest between the groups. A total of 368 all-comers were included in this study. According to discharge times, 204 patients followed NDD, 69 patients 24 to 48 hours discharge, and 95 patients >48 hours discharge after TAVI. The mean age was 84 ± 6.3 years and 61% were women, without differences between the groups. The mean Society of Thoracic Surgeons score was lower in those with NDD versus 24 to 48 hours and >48 hours (2.9 ± 1.0, 3.2 ± 1.2, and 3.4 ± 1.4, respectively, p = 0.014). There were no differences between the groups in terms of preprocedural right bundle branch block or pacemaker. The need for new permanent pacemaker implantation was the leading postprocedural complication; it occurred more frequently in the >48 hours group than the 24 to 48 hours, and <24 hours groups (24% vs 8.6% and 2.2%, p <0.001). There were 5 strokes (1.4%) and all of them occurred in the >48 hours group (p = 0.005). At 30 days after discharge, there were no deaths and no differences in all-cause readmissions (9.3% in <24 hours, 8.6% in 24 to 48 hours, and 19% in >48 hours, log-rank p = 0.087). The readmission rates for new permanent pacemaker implantation requirement were 3.3% (n = 6) in NDD, 0% in 24 to 48 hours, and 1.6% (n = 5) in the >48 hours groups (p = 0.27). In conclusion, in unselected patients who underwent TF-TAVI with the ACURATE neo/neo2 self-expanding bioprosthesis, the NDD pathway is feasible and appears to be safe, without an increased risk of death or all-cause rehospitalization through 30 days after hospital discharge.
A major step forward in the management of patients who underwent transcatheter aortic valve implantation (TAVI) has been the establishment of institutional pathways for early discharge (ED). , However, these ED protocols were for selected patients and using selected transcatheter heart valves (THVs). The ACURATE neo/neo2 (Boston Scientific, Marlborough, Massachusetts) is a self-expanding supra-annular functioning THV with relatively low rates of new conduction disorders needing permanent pacemaker implantation (PPI) and, owing to the impact of these procedural-related complications on length of stay (LOS), the use of this valve may facilitate ED strategies. Furthermore, because the ACURATE neo/neo2 THV has been underrepresented in previous ED studies, we sought to evaluate the safety of next-day discharge (NDD) in patients who underwent TAVI with the ACURATE neo/neo2 THV.
Methods
The authors declare that all supporting data are available within the article and its online supplementary files. Institutional review board and ethics committee approval was obtained from the Western University Health Science Research Ethics Board.
Data from consecutive all-comers patients who underwent outpatient transfemoral TAVI (TF-TAVI) with the ACURATE neo/neo2 THV at the University Hospital, London Health Sciences Centre, Western University between January 2018 and April 2023 were prospectively collected in a dedicated local database and retrospectively analyzed. Patients who died during the index admission for TAVI were excluded from the analysis.
After TAVI completion, patients recovered as per our local policies in the postanesthesia care unit for 1 hour, then transferred to a regular cardiology/cardiac surgery ward unit bed with telemetry capabilities. A 12-lead electrocardiogram (ECG) was performed while in postanesthesia care unit and repeated 4 hours after TAVI. Patient mobilization was recommended within 4 hours, and a transthoracic echocardiogram was performed either the same day (afternoon) or the next morning. All patients had a repeat ECG before discharge.
Patients were deemed suitable for a NDD pathway in the absence of new-onset conduction delays on ECG or uncontrolled arrhythmia (i.e., rapid atrial fibrillation) on telemetry, chest pain secondary to cardiac ischemia, signs of congestive heart failure (CHF), fever, or ongoing infection, maintained urine output, and any signs of hemodynamic instability or major adverse events (i.e., stroke).
The ED population was further divided into 2 groups: those discharged within the first 24 hours (NDD) and those discharged between 24 and 48 hours after TF-TAVI. We hypothesize that NDD is a safe strategy compared with ED after 24 to 48 hours. In addition, to provide a better understanding of the interpretation of the results, a third group comprising patients who were discharged later than 48 hours mainly because of complications is provided for comparison purposes.
The primary outcome of interest was unplanned readmission at 30 days after TF-TAVI. The secondary end point was an exploratory analysis looking at the causes for readmissions, further divided into cardiovascular and noncardiovascular causes. All outcomes were reported according to the Valve Academic Research Consortium 3 definitions. Procedural aspects have been previously described. , ,
Continuous variables are reported as mean ± SD or median (interquartile range), whereas dichotomous parameters are reported as frequencies and percentages. Comparisons were carried out using a chi-square test for categorical data, and the Student’s t test or analysis of variance for continuous variables, as appropriate. To assess the occurrence of the primary outcome, survival analysis was conducted using the Kaplan–Meier method, and the distribution of time-to-event occurrence was compared with the log-rank test. A 2-step analysis was used to assess factors associated with the outcome of interest. First, a single logistic regression was performed. Variables judged, a priori, clinically relevant along with those with a p <0.10 were entered into a multivariable logistic regression analysis. The results are reported as odds ratio (OR) with a 95% confidence interval (CI). All statistical analyses are 2-tailed, and a p value of <0.05 was considered statistically significant. The statistical software STATA, Version 14.1 (StataCorp, College Station, Texas) was used for all the statistical analyses.
Results
Of 372 patients who underwent TF-TAVI with ACURATE neo/neo2 THV, 4 patients died during the index hospitalization and were excluded from the analysis. The study population comprised 368 patients who were discharged at home after TF-TAVI. The mean age was 84 ± 6.3 years, 61% were women, and the mean preprocedural risk profile, as assessed by the Society of Thoracic Surgeons Predicted Risk of Mortality and European System for Cardiac Operative Risk Evaluation II, were 3.1 ± 1.1 and 2.7 ± 1.8, respectively.
A total of 204 patients (55.4%) followed an NDD pathway, 69 patients (18.8%) were discharged between 24 and 48 hours and 95 patients (25.8%) after 48 hours. The mean preprocedural Society of Thoracic Surgeons Predicted Risk of Mortality score was similar in those who followed NDD versus 24 to 48 hours but higher in those discharged >48 hours (2.9 ± 1.0, 3.2 ± 1.2, and 3.4 ± 1.4, p = 0.28 and p = 0.014, respectively). There were no differences between the groups in terms of preexisting right bundle branch block (RBBB), preprocedural pacemaker, or the proportion of patients presenting with left ventricular ejection fraction (LVEF) <35% (p = no significant [NS] comparing the NDD versus 24 to 48 hours groups and between the 3 groups) ( Table 1 ).
| Variables | Overall n=368 | <24h n=204 | 24-48h n=69 | >48h n=95 | P-value * | P-value † |
|---|---|---|---|---|---|---|
| Baseline characteristics | ||||||
| Age, years | 84±6.3 | 84±5.8 | 83±6.1 | 84±7.5 | 0.57 | 0.58 |
| Female sex | 225 (61) | 124 (61) | 36 (52) | 65 (68) | 0.20 | 0.11 |
| Hypertension | 326 (89) | 179 (88) | 60 (87) | 87 (92) | 0.86 | 0.72 |
| Diabetes | 89 (24) | 45 (22) | 17 (25) | 27 (28) | 0.65 | 0.18 |
| Coronary artery disease | 109 (30) | 59 (29) | 20 (29) | 30 (32) | 0.99 | 0.23 |
| Previous PCI | 59 (16) | 32 (16) | 10 (14) | 17 (18) | 0.81 | 0.71 |
| Previous CABG | 50 (14) | 27 (13) | 10 (14) | 13 (14) | 0.79 | 0.32 |
| Peripheral artery disease | 49 (13) | 25 (12) | 9 (13) | 15 (16) | 0.86 | 0.89 |
| Porcelain aorta | 21 (5.6) | 11 (5.4) | 5 (7.2) | 5 (5.2) | 0.57 | 0.85 |
| Atrial fibrillation | 110 (30) | 50 (25) | 22 (32) | 36 (38) | 0.22 | 0.25 |
| Any conduction delay | 64 (17) | 32 (16) | 13 (19) | 19 (20) | 0.54 | 0.51 |
| Right bundle branch block | 42 (11) | 19 (9.3) | 9 (13) | 14 (15) | 0.37 | 0.60 |
| Previous pacemaker | 41 (11) | 19 (9.3) | 11 (16) | 11 (12) | 0.12 | 0.25 |
| STS-PROM score | 3.1±1.1 | 2.9±1.0 | 3.2±1.2 | 3.4±1.4 | 0.28 | 0.014 |
| EuroSCORE II score | 2.7±1.8 | 2.5±1.2 | 2.9±2.0 | 3.2±2.6 | 0.39 | 0.007 |
| Echocardiographic data | ||||||
| Aortic valve area (cm 2 ) | 0.60±0.15 | 0.60±0.16 | 0.60±0.16 | 0.60±0.14 | 0.99 | 0.97 |
| Mean gradient (mmHg) | 46±13 | 45±11 | 47±13 | 48±15 | 0.26 | 0.16 |
| Ejection Fraction (%) | 57±13 | 58±12 | 58±12 | 56±15 | 0.99 | 0.37 |
| Ejection Fraction <35% | 31 (8.4) | 14 (6.8) | 5 (7.2) | 12 (13) | 0.91 | 0.30 |
| Procedural data | ||||||
| Single access | 178 (48) | 144 (70) | 14 (20) | 20 (21) | <0.001 | <0.001 |
| General anesthesia | 110 (30) | 18 (8.8) | 31 (45) | 61 (64) | <0.001 | <0.001 |
| Surgical cutdown access | 45 (12) | 8 (3.9) | 6 (8.6) | 31 (33) | 0.12 | <0.001 |
| Valve size | ||||||
| Small (23 mm) | 72 (20) | 40 (20) | 17 (25) | 15 (16) | 0.46 | 0.35 |
| Medium (25 mm) | 136 (37) | 73 (36) | 21 (30) | 42 (44) | ||
| Large (27 mm) | 153 (42) | 82 (40) | 30 (43) | 41 (43) | ||
| Postdilatation | 120 (33) | 55 (29) | 24 (37) | 41 (43) | 0.21 | 0.05 |
| Right ventricular pacing ‡ | 183 (50) | 53 (26) | 57 (83) | 73 (77) | <0.001 | <0.001 |
| ProGlide/Style closure device | 260 (71) | 138 (67) | 61 (88) | 61 (64) | 0.001 | 0.001 |
| AngioSeal closure devise | 31 (8.4) | 22 (11) | 1 (1.4) | 8 (8.4) | 0.01 | 0.09 |
| Manta vascular closure devise | 63 (17) | 58 (28) | 2 (2.8) | 3 (3.1) | <0.001 | <0.001 |
| Procedure duration | 46±21 | 42±10 | 46±13 | 62±26 | 0.18 | <0.001 |
⁎ For differences between the <24h versus 24-48h groups.
† For overall differences between the three groups.
‡ Right ventricular as opposed to left ventricular stimulation for rapid pacing. Data are presented as mean±SD or n (%). Some percentages may not add up to 100% owing to rounding.
Higher rates of general anesthesia were observed comparing the NDD and 24 to 48 hours groups and the 3 groups (8.8%, 45%, and 64%, p <0.001 for both comparisons). Surgical cutdown for vascular access tended to be higher in the 24 to 48 hours group versus NDD but did not reach statistical significance, whereas it did when comparing the 3 groups, being higher in those discharged >48 hours (3.9%, 8.6%, and 33%, p = 0.12 and p <0.001, respectively). Patients who followed a NDD pathway had higher proportion of single- vascular access (pigtail through the same large-bore sheath, 70% vs 20% in the 24 to 48 hours and 21% in the >48 hours groups, p <0.001 for both comparisons), use of the TAVI wire for left ventricular pacing (as opposed to venous access for right ventricular pacing, 74% vs 17% in the 24 to 48 hours and 13% in >48 hours groups, p <0.001 for both comparisons), and Manta (Teleflex, Teleflex Inc., Malvern, Pennsylvania) vascular closure device (28% vs 2.8% in the 24 to 48 hours and 3.1% in >48 hours groups, p <0.001 for both comparisons). Procedural duration time was comparable between the NDD and 24 to 48 hours groups but significantly longer in those discharged >48 hours (42 ± 10 minutes vs 46 ± 13 minutes in the 24 to 48 hours and 62 ± 26 minutes in the >48 hours groups, p = 0.18 and p <0.001, respectively). This is most likely the translation of the previously mentioned higher use of surgical cutdown in this group. The remaining baseline clinical, electrocardiographic, echocardiographic, and procedural characteristics of the study population are listed in Table 1 .
Postprocedural echocardiographic data were similar between groups, except for a slightly but significantly higher mean gradient in 24 to 48 hours group (9.7 ± 5.3 vs 8.3 ± 3.2 mm Hg in the NDD p = 0.03 and 8.5 ± 4.3 mm Hg in the 48 hours groups, p = 0.04) and this may be explained by a numerically higher proportion of small-sized THVs that were implanted in this group.
As expected, patients who had procedural-related complications were more often discharged >48 hours (p <0.001) but no difference was found between the NDD and 24 to 48 hours groups (p = 0.58). The need for new PPI was the leading postprocedural complication (n = 29, 8.9%) and occurred more frequently in those discharged >48 hours than the 24 to 48 hours and NDD groups (24%, 8.6%, and 2.2%, respectively, p <0.001 for the overall comparison and p = 0.02 comparing the NDD and 24 to 48 hours groups). In patients who required PPI, baseline RBBB was present in 11 patients, 3 in the NDD group, 3 in the 24 to 48 hour groups, and 5 in the >48 hours group. There were a total of 5 strokes (1.4%) and all of them occurred in the >48 hours group (p = 0.01). No significant differences were found in terms of vascular complications, bleeding, and infections (p = NS for all comparisons) ( Table 2 ).
