Venous Vascular Closure System Versus Manual Compression Following Multiple Access Electrophysiology Procedures: The AMBULATE Trial
This study compared the efficacy and safety of the VASCADE MVP Venous Vascular Closure System (VVCS) device (Cardiva Medical, Santa Clara, California) to manual compression (MC) for closing multiple access sites after catheter-based electrophysiology procedures.
The VASCADE MVP VVCS is designed to provide earlier ambulatory hemostasis than MC after catheter-based procedures.
The AMBULATE (A Randomized, Multi-center Trial to Compare Cardiva Mid-Bore [VASCADE MVP] VVCS to Manual Compression in Closure of Multiple Femoral Venous Access Sites in 6 - 12 Fr Sheath Sizes) trial was a multicenter, randomized trial of device closure versus MC in patients who underwent ablation. Outcomes included time to ambulation (TTA), total post-procedure time (TPPT), time to discharge eligibility (TTDe), time to hemostasis (TTH), 30-day major and minor complications, pain medication usage, and patient-reported outcomes.
A total of 204 patients at 13 sites were randomized to the device arm (n = 100; 369 access sites) or the MC arm (n = 104; 382 access sites). Baseline characteristics were similar between groups. Mean TTA, TPPT, TTDe, and TTH were substantially lower in the device arm (respective decreases of 54%, 54%, 52%, and 55%; all p < 0.0001). Opioid use was reduced by 58% (p = 0.001). There were no major access site complications. Incidence of minor complications was 1.0% for the device arm and 2.4% for the MC arm (p = 0.45). Patient satisfaction scores with duration of and comfort during bedrest were 63% and 36% higher in device group (both p < 0.0001). Satisfaction with bedrest pain was 25% higher (p = 0.001) for the device overall, and 40% higher (p = 0.002) for patients with a previous ablation.
Use of the closure device for multiple access ablation procedures resulted in significant reductions in TTA, TPPT, TTH, TTDe, and opioid use, with increased patient satisfaction and no increase in complications. (A Randomized, Multi-center Trial to Compare Cardiva Mid-Bore VVCS to Manual Compression in Closure of Multiple Femoral Venous Access Sites in 6 - 12 Fr Sheath Sizes [AMBULATE]; NCT03193021).
The volume of catheter ablation procedures for the treatment of atrial fibrillation and other arrhythmias continues to rise year after year, both in the United States and worldwide (1). Specifically, for atrial fibrillation, innovation and refinement in ablation tools, techniques, and strategies have led to greater efficacy, shorter procedure times, and a decline in complication rates (2). Despite these procedural advances, certain critical aspects of post-procedural management have remained relatively unchanged. For example, manual compression (MC) followed by post-procedure bedrest—often up to 8 h when multiple access sites are present—remains the standard of care for post-procedural closure of venous access sites. Prolonged bedrest may itself be associated with other interventions, including use or persistence of an in-dwelling bladder catheter and longer length of stay (3).
Device-based vascular closure has dramatically shortened bedrest and post-procedure care after diagnostic and interventional arterial procedures (4). It has facilitated early ambulation, which itself has been associated with favorable clinical and economic outcomes in a variety of medical procedures, greater patient satisfaction, same-day discharge, and lower risk of complications (5,6).
Closure of venous access sites with specialized devices could have similar potential to shorten bedrest and facilitate early ambulation and discharge after catheter-based procedures. The VASCADE MVP (Cardiva Medical, Santa Clara, California) Venous Vascular Closure System (VVCS) is a novel, investigational, single-use vascular closure device designed to achieve hemostasis at mid-bore femoral venous access sites (Online Figure 1). The device establishes hemostasis by delivering a resorbable, extravascular collagen plug into the tissue tract, created by a procedural introducer sheath. The closure device is designed for use with up to 12-F inner diameter sheaths, including the Flexcath Steerable Sheath (Medtronic, Minneapolis, Minnesota), for use with the Artic Front CyroAblation Balloon (Medtronic), and the WATCHMAN left atrial appendage occlusion device (Boston Scientific, Natick, Massachusetts). The device is based on the design of the VASCADE VCS (Cardiva Medical), which is currently approved for arterial and venous closure after 5-F to 7-F procedures. The results of the randomized controlled U.S. pivotal RESPECT (A Prospective, Randomized, Pivotal Trial of a Novel Extravascular Collagen-Based Closure Device Compared to Manual Compression in Diagnostic and Interventional Patients) trial (n = 420) demonstrated that the VASCADE VCS was safe and effective in sealing femoral arterial access sites after endovascular procedures performed through 6-F or 7-F introducer sheaths, with no major vascular complications, fewer minor vascular complications, and shorter time to hemostasis, ambulation, and discharge eligibility compared with MC (7,8).
We evaluated the efficacy and safety of the investigational venous closure device compared with that of MC, for closing multiple access sites after catheter-based electrophysiology procedures. In addition to clinical outcomes, we also conducted several analyses that evaluated patient satisfaction with recovery and patterns of pain medication use.
The AMBULATE (A Randomized, Multi-center Trial to Compare Cardiva Mid-Bore VVCS to Manual Compression in Closure of Multiple Femoral Venous Access Sites in 6 - 12 Fr Sheath Sizes) trial was a randomized, controlled, open-label investigational device exemption clinical trial conducted at multiple U.S. sites ( NCT03193021). The aim was to demonstrate the efficacy and safety of the closure device compared with that of MC. The trial was conducted in compliance with Food and Drug Administration and International Council for Harmonisation (ICH) regulations for Good Clinical Practice. The protocol and informed consent forms were approved by the institutional review board at each site.
Adults 18 years or older who underwent elective, nonemergent, catheter-based procedures via the common femoral vein(s) using a 6-F to 12-F inner diameter introducer sheath, a minimum of 3 and maximum of 4 femoral venous access sites, and a maximum of 2 access sites per leg were enrolled. Exclusion criteria included active systemic or cutaneous infection or inflammation in the vicinity of the groin; any pre-existing immunodeficiency disorder; long-term use of high-dose systemic steroids; history of bleeding diathesis, coagulopathy, hypercoagulability, or thromboembolic events; platelet count <100,000 cells/mm3; or severe comorbidities with life expectancy <12 months in the opinion of the site investigator.
After providing informed consent, patients underwent the appropriate ablation procedure. During the ablation procedure, patients could be further excluded from the study before randomization if any of the following occurred: attempted femoral arterial access or inadvertent arterial puncture; procedural complications that interfered with routine recovery, ambulation, or discharge times; difficulty with needle puncture or insertion of the introducer sheath; incorrect sheath placement; intraprocedural bleeding or thrombotic complications; final sheath size <6-F or >12-F inner diameter; or tissue tract <2.5 cm deep.
The follow-up period consisted of in-hospital assessments through 30 days after patient discharge.
Sample size and power
Sample size calculations assumed a normal distribution for time to ambulation (TTA) according to Food and Drug Administration requirements for a more conservative treatment of outlier data. Based on the results of the RESPECT study (7), mean TTA for MC was estimated at 7.2 h. The same mean for Cardiva Mid-Bore VVCS was estimated at 5 h including an outlier and at 4.3 h excluding the outlier. With an estimated joint SD of 5.4 h, the power for reaching a statistically significant difference at a 2-sided p < 0.05 level, with a total sample size of 204 patients (102 per group), was 0.83 including the outlier in the calculation; the same power was 0.97 excluding the outlier. Although the primary TTA analysis used an analysis of covariance (ANCOVA) model that accounted for the randomization stratification factor (number of access sites: 3 vs. 4), the distribution of access sites and their impact on TTA was unknown at the planning stage. Therefore, the sample size estimation was based on a simple 2-arm comparison as a reasonable approximation.
At the end of the catheter-based procedure, and with the procedural sheaths still in place, patients who met all pre-operative and intraoperative study criteria were randomly assigned in a 1:1 scheme to vascular closure with the closure device or MC. Variable block randomizations were performed separately, that is, stratified, by number of venotomy sites (3 vs. 4) at each investigational site.
Endpoints and definitions
The primary efficacy endpoint was time to ambulation, defined as the elapsed time between removal of the final closure device or removal of final sheath and when the patient stood and walked 20 ft without evidence of venous re-bleeding from the femoral access site. The primary safety endpoint was the incidence of major complications related to venous access site closure occurring within 30 ± 7 days after the procedure, by limb. A schema defining all primary and secondary outcomes is shown in Figure 1.
Secondary efficacy endpoints were total post-procedure time (TPPT; elapsed time between removal of the last procedural device for the index procedure and when the patient was able to successfully ambulate); time to hemostasis (TTH; elapsed time between removal of the closure device or the sheath and first observed and confirmed venous hemostasis, for each access site); time to discharge eligibility (TTDe) (elapsed time between removal of the final closure device or final sheath and when the patient was eligible for hospital discharge based solely on the assessment of the access site, as determined by the medical team); time to discharge (TTD; elapsed time between removal of the final closure device or final sheath and when the patient was discharged from the institution); and time to closure eligibility (TTCe; elapsed time between removal of the last procedural device for the index procedure and removal of the first closure device or first sheath).
The secondary safety endpoint was incidence of minor complications within 30 days after the procedure. According to a pre-specified substudy, a subgroup of 25 patients from each treatment group underwent an ultrasound examination of the bilateral access sites to further evaluate safety. The ultrasound was performed at the 30-day office visit.
Other endpoints were patient experience with duration of, comfort during, and pain associated with bedrest, and use of analgesic and antianxiety medication.
Patients randomized to intervention underwent attempted closure of all femoral venous sites using the previously described closure device. Deployment of the VASCADE MVP closure device was carried out after completion of each catheterization procedure through the 6-F to 12-F inner diameter introducer sheath(s) placed during the catheter ablation procedure. First, the introducer sheaths were exchanged for short sheaths of equivalent diameter at each access site. The closure device catheter was inserted through the short sheath and then the distal tip of the closure device catheter was deployed to open and expand a low-profile, 23-F polyurethane covered nickel-titanium disc within the lumen of the femoral vein (Online Figure 1). The introducer sheath was removed over the closure device catheter, and the low-profile disc was placed against the inner wall, located the vessel wall, temporarily sealed the venotomy, and positioned the device for collagen-patch deployment within the tissue tract immediately adjacent and extravascular to the venotomy site. Placement was verified using imaging before release of the collagen plug, and once the plug was deployed, the disc collapsed and the device removed. The total deployment procedure took approximately 1 min per access site. Gentle MC was applied for up to 5 min. Hemostasis time was recorded and reconfirmed 5 min later, and at 15-min intervals for the next hour. Intraprocedural device failures were converted to MC. Protocol guidelines recommended bedrest for 2 to 2.5 h, then ambulation if stable, with groin checks immediately and 15 min after ambulation.
In case of repeat procedures, the same femoral vein could be accessed after 30 days of the previous procedure using the VASCADE MVP device for venous closure.
For the MC arm, sheaths were exchanged or left in place according to institutional protocol. After final sheath removal, sustained compression was applied according to local practice of the institution, operating physician, and care team. Hemostasis time was recorded and re-confirmed 5 min following sheath pull and at 15-min intervals for the next hour. Protocol guidelines advised investigators to follow institutional guidelines for standard of care for ambulation, with groin checks immediately and 15 min after ambulation. Periprocedural anticoagulation management (choice or drug and dose, interruption, resumption) and protamine reversal were not pre-specified in the protocol and were done at the discretion of the operator and in accordance with operator or institutional usual care or practice.
All participants completed a 3-part patient experience survey (Online Figure 2) after successful ambulation and discharge eligibility determination, but before discharge. The survey was developed according to guidance from the Food and Drug Administration (9) and previous studies of pain and discomfort after other types of interventional cardiology procedures (10–19). Questions were designed to characterize patients’ level of satisfaction with their current experience and their hypothetical preferences for duration, discomfort, and pain associated with supine bedrest post-procedure. Patients who underwent a repeat ablation procedure were asked to compare the current episode with past experience. Answers were scored on an anchored rating scale of 0 (very dissatisfied) to 10 (very satisfied).
Pain medication use
Analgesic or antianxiety medication for pain management during bedrest was prescribed according to local institutional protocols and recorded. When patients received multiple medications, use was categorized according to the strongest and/or most potent therapeutic class prescribed (i.e., analgesic/antipyretics less than nonsteroidal anti-inflammatory drugs less than opioids).
Adjudication of endpoints
An independent combination of the Clinical Events Committee and Data Safety Monitoring Board (referred to as the Data Safety Monitoring Committee) was responsible for systematic review and adjudication of all reported major and minor access site closure−related endpoint complications. All other events reported to be potentially device- or procedure-related by the investigator were reviewed by the chairman of the Data Safety Monitoring Committee to determine whether it met criteria for adjudication. To enhance objectivity and reduce the potential for bias, the Data Safety Monitoring Committee members were independent of the sponsor as well as the investigational sites and/or investigators. Members consisted of at least 3 independent physicians with experience in catheter-based procedures and a statistician. The methodology for performing these responsibilities was developed and outlined in a pre-specified adjudication charter. Operational provisions were established to minimize potential bias.
Patient demographics, baseline characteristics, medical history, periprocedural anticoagulant regimen, experience, and use of pain medication were summarized descriptively by randomized treatment arm.
Efficacy analysis was conducted according to the intent-to-treat principle. The intent-to-treat population consisted of all randomized patients. The null hypothesis for the primary endpoint TTA was that mean TTADevice = mean TTAMC.
The primary TTA analysis was an ANCOVA with treatment as the independent variable of interest, as well as randomization stratification factor and number of access sites as the only covariates. In addition, Student’s t-test and Wilcoxon’s rank-sum test were also performed for unadjusted TTA comparison between the closure device and MC. A 2-sided p value <0.05 in favor of the device was considered statistically significant for demonstration of the superiority of the closure device efficacy over MC.
To limit the study-wide type I error rate to 0.05 (2-sided), a gated hierarchical testing approach was pre-specified, with endpoints ordered to reflect their clinical relevance. If TTA results were positive, secondary endpoints were analyzed in the order of TPPT, TTH, TTDe, and TTD between the 2 study arms; testing would stop whenever a p value >0.05 was encountered. For analysis, TPPT, TTDe, and TTD were compared between treatments using the same analyses as described for the TTA analysis. Because there were multiple access sites per patient (and per limb in many cases), TTH was analyzed by the bootstrap method with the patient as the re-sampling unit. A noninferiority comparison was conducted, with a noninferiority margin of 5 min for the closure device compared with MC.
Summary statistics for additional efficacy endpoints of time to closure eligibility, procedure success, and proportion of device success by access site (closure device only) were presented without hypothesis testing.
Safety analyses were carried out at the limb level. The primary safety endpoint was the combined limb incidence rate of major complications related to access site closure, with a noninferiority margin of an absolute 5% for the closure device. To account for potential correlation between the 2 limbs of the same patient, bootstrap analysis was performed with the patient as the re-sampling unit. All limbs that received any portion of the closure device treatment were analyzed as a closure device limb. Otherwise, all limbs that were randomized to the MC control treatment were analyzed as MC limbs.
For secondary safety measures, the combined and individual incidence rates for minor access site closure complications were reported by limb, by treatment received, without noninferiority testing. Other reported adverse events were analyzed by treatment group, regardless of relationship to access site closure. All index procedures and assessments were completed for all patients.
The study database was locked in April 2018. Data analyses were carried out using the SAS version 9.4 (SAS Institute Inc., Cary, North Carolina), R statistical Environment (R Foundation for Statistical Computing, Vienna, Austria), and Microsoft Excel (Microsoft, Redmond, Washington).
Between September 2017 and March 2018, 258 patients across 28 operators and 13 sites consented to enroll into the randomized trial (Figure 2, Central Illustration). Of those, 204 consecutive patients met pre-operative and intraoperative selection criteria and were randomized to receive the closure device (n = 100; 369 access sites) or MC (n = 104; 382 access sites). Of those, 87 of 100 (87%) patients who received the closure device and 91 of 104 (88%) patients who underwent MC completed a follow-up office visit per protocol. The remaining patients received the 30-day follow-up by phone and/or received follow-up outside of the protocol specified window. One patient in each arm was lost to follow-up.
Demographics and baseline clinical characteristics were similar between groups (Table 1). After randomization, there were 369 access sites in the closure device intent-to-treat group. Ultrasound guidance, either alone or in combination with other techniques, was used for access in 65% of closure device and 74% of MC access sites. Actual attempts to deploy the closure device occurred in 363 access sites, and 351 were successfully deployed (97% device success in the as-treated population). Procedure success was 98% for the device group and 99% for the MC group. Six closure device access sites across 3 patients were converted to MC and were reported as protocol deviations. In 1 of those patients (4 access sites), the intraluminal positioning disc pulled through during the procedure in the first access site. The operator deviated from the protocol and converted the remaining 3 sites (including 2 on the contralateral limb) to MC. Thus, according to the intent-to-treat protocol, the patient was included in the closure device group for efficacy analyses. However, each limb was considered separately for the safety analysis on an as-treated basis. Protamine was used in 78 of the 100 (78%) patients randomized to the device (78 of 85 patients; 92% as-treated with heparin) and in 86 of 104 (83%) patients randomized to MC (86 of 94 patients; 91% as-treated with heparin).
|Closure Device (n = 100)||Manual Compression (n = 104)||p Value∗|
|Age, yrs||61.5 ± 11.6||63.4 ± 11.1||0.24|
|Body mass index, kg/m2||29.0 ± 4.6||29.3 ± 5.2||0.69|
|Male||67 (67)||64 (62)||0.47|
|Hispanic or Latino||3 (3)||6 (6)||0.50|
|Not Hispanic or Latino||97 (97)||95 (91)|
|Unknown||0 (0)||3 (3)|
|White||86 (86)||93 (89)||0.52|
|Black or African American||6 (6)||4 (4)|
|American Indian or Alaska Native||3 (3)||0 (0)|
|Asian||2 (2)||1 (1)|
|Other||3 (3)||6 (6)|
|Hypercholesterolemia||57 (57)||58 (56)||0.88|
|Hypertension||58 (58)||69 (66)||0.25|
|Current or ex-smoker||42 (42)||42 (40)||0.89|
|Diabetes mellitus||14 (14)||18 (17)||0.57|
|Insulin||5 (5)||3 (3)||0.49|
|Diet||5 (5)||5 (5)||1.00|
|Oral hypoglycemics||11 (11)||15 (14)||0.53|
|Lower right limb neuropathy||1 (1)||2 (2)||1.00|
|Lower left limb neuropathy||1 (1)||2 (2)||1.00|
|Other relevant medical history||16 (16)||16 (15)||1.00|
|No. of access sites|
|3 sticks||31 (31)||34 (33)|
|4 sticks||69 (69)||70 (67)||0.88|
|AF||64 (64)||68 (65)|
|AF + AFL||14 (14)||19 (18)|
|AFL||6 (6)||7 (7)|
|SVT||14 (14)||8 (8)|
|VT||2 (2)||2 (2)||0.58|
|RF||77 (77)||79 (76)|
|Cryo||20 (20)||23 (22)|
|Both (RF + Cryo)||3 (3)||2 (2)||0.83|
|Aspirin||29 (29)||21 (20)||0.19|
|Warfarin||5 (5)||3 (3)||0.49|
|Clopidogrel||0 (0)||2 (2)||0.50|
|Ticagrelor||1 (1)||0 (0)||0.49|
|Rivaroxaban||29 (29)||25 (24)||0.43|
|Dabigatran||3 (3)||4 (4)||1.00|
|Apixaban||37 (37)||46 (44)||0.32|
|Low molecular weight heparin||1 (1)||2 (1)||1.00|
|Any of the above||84 (84)||88 (85)||1.00|
|None||16 (16)||16 (15)||1.00|
Compared with MC, use of the closure device to achieve venous access site closure after catheterization resulted in an approximately 3.3-h (54%) decrease (2.8 ± 1.3 h vs. 6.1 ± 1.6 h; ANCOVA p < 0.0001) in the primary efficacy outcome, overall mean TTA per patient (Table 2, Figure 3A).
|Closure Device||Manual Compression||p Value∗|
|Total (N = 100)||3 Access Sites (n = 31)||4 Access Sites (n = 69)||Total (N = 104)||3 Access Sites (n = 34)||4 Access Sites (n = 70)|
|Mean ± SD||2.8 ± 1.3||2.5 ± 0.8||2.9 ± 1.5||6.1 ± 1.6||5.9 ± 1.2||6.2 ± 1.7||<0.0001|
|Median (min–max)||2.2 (2.0–11.5)||2.2 (2.0–5.6)||2.3 (2.0–11.5)||6.1 (3.4–15.7)||5.3 (4.2–9.1)||6.2 (3.4–15.7)||<0.0001|
|Mean ± SD||3.1 ± 1.3||2.7 ± 0.8||3.3 ± 1.5||6.8 ± 1.7||6.4 ± 1.3||6.9 ± 1.9||<0.0001|
|Median (min–max)||2.6 (2.2–11.8)||2.4 (2.2–5.9)||2.7 (2.2–11.8)||6.4 (4.2–15.9)||6.2 (4.5–9.8)||6.6 (4.2–15.9)||<0.0001|
|Mean ± SD||3.1 ± 1.3||2.7 ± 0.8||3.2 ± 1.5||6.5 ± 1.9||6.2 ± 1.3||6.6 ± 2.2||<0.0001|
|Median (min–max)||2.5 (2.3–11.7)||2.5 (2.3–5.9)||2.6 (2.3–11.7)||6.3 (4.3–21.3)||5.7 (4.6–9.4)||6.5 (4.3–21.3)||<0.0001|
|Mean ± SD||21.8 ± 13.4||20.5 ± 10.8||22.3 ± 14.5||21.8 ± 9.5||22.7 ± 10.6||21.4 ± 9.0||0.98|
|Median (min–max)||22.3 (2.3–96.1)||22.9 (2.3–48.2)||22.3 (3.5–96.1)||22.1 (5.7–72.9)||22.8 (5.7–71.5)||21.6 (5.8–72.9)||0.85|
|Mean ± SD||10.5 ± 6.0||9.0 ± 4.1||11.1 ± 6.6||37.6 ± 33.2||32.2 ± 27.6||40.3 ± 35.5||<0.0001|
|Median (min–max)||10.1 (1.7–47.5)||9.8 (1.7–17.5)||10.2 (2.0–47.5)||25.2 (1.8–132.3)||21.1 (2.0–108.9)||27.8 (1.8–132.3)||<0.0001|
The first secondary outcome was TPPT (Table 2, Figures 2 and 3B). There was a 3.7-h (54%) decrease (3.1 ± 1.3 h vs. 6.8 ± 1.7 h; ANCOVA p < 0.0001) in TPPT when the closure device was compared with MC.
According to the pre-specified hierarchical testing order, TTH was evaluated next by bootstrap analysis, at the level of individual access site (Table 3, Figure 3C). There was a mean decrease of 7.6 min (6.1 ± 3.7 min vs. 13.7 ± 6.5 min; 95% confidence interval: −8.8 to −6.3) in the closure versus MC groups. The result met pre-specified significance thresholds for noninferiority and superiority. The time course of hemostasis in Figure 3E shows that 90% of access sites closed with the closure device reached hemostasis within 7 min, whereas it took between 20 and 25 min for 90% of access sites to reach hemostasis in the MC group.
|Outcome||Closure Device Access Sites||Manual Compression Access Sites|
|Total (N = 369)||3 Access Sites (n = 93)||4 Access Sites (n = 276)||Total (N = 382)||3 Access Sites (n = 102)||4 Access Sites (n = 280)|
|Mean ± SD||6.1 ± 3.7||5.4 ± 2.0||6.3 ± 4.1||13.7 ± 6.5||11.4 ± 6.4||14.5 ± 6.4|
|Median (min–max)||5.1 (0.4–33.3)||5.1 (1.3–23.3)||5.1 (0.4–33.3)||11.7 (0.6–37.1)||10.0 (2.9–32.7)||12.5 (0.6–37.1)|
Consistent with other measures, there was a 3.4-h (52%) decrease in TTDe (3.1 ± 1.3 h vs 6.5 ± 1.9 h; ANCOVA p < 0.0001) (Table 2, Figure 3D). There was no statistically significant difference in overall TTD between the 2 groups (21.8 ± 13.4 h vs. 21.8 ± 9.5 h) (Table 2, Figure 3E). Most cases involved an overnight stay with next-day discharge.
Finally, TTCe was evaluated (Table 2, Figure 3F). Mean TTCe times were 27 min (72%) shorter for the closure device compared with MC (10.5 ± 6.0 min vs. 37.6 ± 33.2 min; nominal p < 0.0001). Formal hypothesis testing was not pre-specified for this endpoint in the protocol; therefore, nominal p values are presented in Table 2 for descriptive purposes only.
Variability in TTCe among study operators was observed in the MC groups. It was mainly driven by the location of the sheath pull as per the preference of the operators. The mean time to sheath removal in the MC group varied widely based on whether the sheath was removed in the laboratory (n = 39) or at recovery (n = 65) (14.76 ± 17.1 min vs. 51.33 ± 33.04 min, respectively).
At 30 days post-procedure, there were zero access site closure−related major complications in either group. As a result, the planned bootstrap analysis could not be performed. Instead, Newcombe’s Wilson score 95% confidence intervals were constructed both by limb and by patient to estimate confidence intervals accounting for potential limb correlation in both independent and completely correlated scenarios, respectively (7). The resulting 2-sided 95% confidence intervals for the VVCS-MC differences were −1.9% to 1.8% and −3.6% to 3.7%, respectively. In other words, an absolute 5% or higher VVCS closure device major complication incidence was ruled out (i.e., rejected with a 2-sided p < 0.05 in either case), and the noninferiority criterion was met.
There were 2 minor complications related to venous access site closure in the device treatment group and 5 in the MC group (Table 4). In the ultrasound substudy, minor complications per limb were 0.0% in the closure device group and 2.0% in the MC group.
|Closure Device (n = 199)||Manual Compression (n = 209)||p Value|
|Any access site closure–related minor complication||2 (1.0)||5 (2.4)||0.45|
|Access site–related hematoma >6 cm documented by ultrasound||0 (0.0)||2 (1.0)||0.50|
|Localized access site infection confirmed and treated with intramuscular or oral antibiotics||1 (0.5)||1 (0.5)||1.00|
|Arteriovenous fistula not requiring treatment||0 (0.0)||1 (0.5)||1.00|
|Pseudoaneurysm requiring thrombin/fibrin adhesive injection or ultrasound-guided compression||1 (0.5)||0 (0.0)||0.49|
|Transient access site–related nerve injury||0 (0.0)||1 (0.5)||1.00|
|Any access site closure–related adverse event, excluding major or minor access site closure–related complications||15 (7.5)||12 (5.7)||0.55|
|Access site venous re-bleeding after initial hemostasis confirmed for 5 min||12 (6.0)||6 (2.9)||0.15|
|Access site hematoma <6 cm||2 (1.0)||5 (2.4)||0.45|
|Pain at the access site||2 (1.0)||1 (0.5)||0.62|
|Superficial arterial bleed||1 (0.5)||0 (0.0)||0.49|
|Access site oozing||1 (0.5)||0 (0.0)||0.49|
|Right leg edema||0 (0.0)||1 (0.5)||1.00|
Adverse events (any) not related to the access closure sites were reported in 19 (19.0%) of the patients who received the closure device and in 17 (16.3%) of the patients who underwent MC (p = 0.71). Serious adverse events not related to the access closure sites were reported in 8 (8%) of the device group patients and 2 (2%) of patients who underwent MC (p = 0.06). There were no deaths in either group. A full listing of the adverse events is available in Online Table 1, and serious adverse events not related to access site closure are listed in Online Table 2.
Six sites elected to participate in the ultrasound substudy. A total of 25 patients from each treatment group were enrolled. Because 1 patient in the MC group declined to return for the ultrasound, 98 examinations were performed in 49 participants. The ultrasound substudy identified no clinically significant safety issues in either treatment group.
Patient survey responses indicated that patients in both groups exhibited a strong preference for the shorter duration of supine bedrest associated with the closure device group compared with the MC group (Table 5). Patients in the MC group expressed greater dissatisfaction with the duration of bedrest and level of pain than their counterparts in the closure device group. Mean satisfaction scores for the length of time spent on supine bedrest were 63% higher in the closure device group compared with the MC group, and satisfaction scores regarding discomfort from supine bedrest were 36% higher (p < 0.0001 for both comparisons). Patients in the closure device group also indicated 25% less pain during supine bedrest compared with patients in the MC group (p = 0.001).
|Satisfaction Measure||Closure Device (n = 100)||Manual Compression (n = 104)||p Value†|
|n||Mean Rating||n||Mean Rating|
|Level 1: all patients|
|Duration, current episode||100||8.3 ± 2.4||102||5.1 ± 3.4||<0.0001|
|Discomfort, current episode||100||7.2 ± 3.1||102||5.3 ± 3.1||<0.0001|
|Pain, current episode||100||7.5 ± 3.2||102||6.0 ± 3.4||0.001|
|Level 2: patients randomized to VVCS|
|Duration, if 2–3 h longer||98||2.6 ± 3.1||―||―||―|
|Discomfort, if 2–3 h longer||98||2.7 ± 2.9||―||―||―|
|Pain, if 2–3 h longer||98||3.2 ± 3.4||―||―||―|
|Level 2: patients randomized to MC|
|Duration, if 2–3 h shorter||―||―||102||9.1 ± 1.7||―|
|Discomfort, if 2–3 h shorter||―||―||101||8.4 ± 2.2||―|
|Pain, if 2–3 h shorter||―||―||100||8.2 ± 2.5||―|
|Level 3: patients with a previous cardiac ablation procedure|
|Duration, compared with previous experience||30||7.9 ± 2.3||39||5.6 ± 3.0||0.001|
|Discomfort, compared with previous experience||30||7.5 ± 2.1||39||5.4 ± 2.8||0.001|
|Pain, compared with previous experience||30||7.7 ± 2.8||38||5.5 ± 2.9||0.002|
When the 2 groups were queried separately about how they would feel if the duration of their bedrest was altered, patients in the closure device group indicated that their satisfaction would have decreased dramatically if they had experienced 2 to 3 additional hours of supine bedrest (Table 5). Conversely, patients who underwent MC believed that they would have preferred spending 2 to 3 fewer hours lying on their backs and would have experienced less discomfort and pain.
A total of 30 patients in the closure device group (30%) and 41 patients in the MC group (39%) reported that they had previously undergone previous cardiac ablation (Table 5). In the MC group, patients who had previous ablation rated their current experience as approximately equal or slightly better than previous (i.e., no significant change in satisfaction with the duration, discomfort, and pain from supine bedrest). However, in the closure device group, patients with previous ablation reported markedly higher levels of satisfaction with the duration, discomfort, and pain from supine bedrest compared with their previous experience.
Pain medication use
An evaluation of post-procedural use of pain medication showed lower use of pain medications in the closure device group (24% vs. 49%; p = 0.0003) (Online Table 3). When categorized according to the most potent drug class prescribed per patient, 36% of patients who underwent MC were treated with narcotic drugs compared with 15% of patients in the closure device group (58% relative decrease; p = 0.001).
The AMBULATE multicenter randomized trial evaluated the safety and efficacy of the investigational vascular closure device, VASCADE MVP, in electrophysiological procedures using venous access. Closure with the VASCADE MVP device had greater efficacy and safety compared with MC, including for the primary efficacy endpoint of TTA. Secondary efficacy endpoints were also improved in the closure device arm, including TPPT, TTH, and TTDe. TTD was the only outcome that was comparable between the 2 groups, although there were no instructions in the study protocol directing study sites to attempt earlier-than-standard discharge times after standard TTDe. There were no major access site complications reported in either group.
Cardiac catheter ablation procedures using either radiofrequency energy or cryogenic sources require multisite venous access (20,21). Absolute bedrest with minimal limb movement for a certain period is necessary to achieve effective MC hemostasis (22). MC protocols commonly require at least 10 min of sustained compression at the access site followed by 6 to 8 h of supine bedrest. Prolonged immobilization can lead to higher usage of indwelling bladder catheters and can also cause discomfort, back pain, and fatigue. In a randomized trial of patients who had a cardiac catheterization, intermittent position changing and earlier ambulation and undertaking of self-care activities were shown to be associated with an increase in satisfaction and comfort, and a decrease in the fatigue level (16). Early ambulation is also known to reduce the hospital staff use, thereby increasing the workflow efficiency (23). Furthermore, extended immobility is reported to be associated with high risk of venous thromboembolism, including deep vein thrombosis and pulmonary embolism (24). Moreover, although no formal analysis was undertaken in this study, selected patients were eligible to be discharged on the same day after being mobile without any pain or access site complications.
In our study population, the total post-procedure time was significantly lower in the closure device group. Removal of the femoral sheath and achieving hemostasis in a short time thereafter, along with early mobility, were reported to improve patient satisfaction, reduce groin complications, as well as post-procedural bedrest, hospital stay, and subsequent health care cost (25,26).
Operators worldwide aim for safe venous access and uncomplicated and adequate hemostasis (26). Achieving hemostasis may be more challenging when procedures are performed under uninterrupted anticoagulation. However, despite the large number of cases performed under continuous anticoagulation in our study, 90% of access sites closed with the device reached initial hemostasis within 7 min, whereas it took between 20 and 25 min for the same outcome in the MC group. Earlier randomized trials of arterial closure reported similar results (7,27).
An alternative approach, temporary subcutaneous Figure-of-eight suture to maintain hemostasis after the venous sheath removal was described in published reports (20,28,29). In the single-center retrospective study conducted by Issa et al. (28), the time to hemostasis was reported to be significantly shorter in the Figure-of-eight cohort compared with the MC group. However, other published trials also reported longer immobilization after the removal of the Figure-of-eight sutures (20,29). Lakshmanadoss et al. (29) reported the post-procedure bedrest time to be at least 6 h (29), and in the study by Aytemir et al. (20), the time to mobilization was similar in the Figure-of-eight group and the MC cohort. We found no published prospective trial data of a Figure-of-eight suture to define bedrest times, workflow, or safety. In the experience of some of the authors, the Figure-of-eight suturing technique is associated with pain at the groin site and some risk of bleeding, particularly with inadvertent vascular injury from deeply placed sutures. Still, the cost of the venous closure device may be a consideration, particularly if >1 is needed. An economic analysis factoring in cost differences in outcomes based on the AMBULATE trial is planned.
TTD was comparable between the groups. The overnight stay might reflect long-standing institutional practice. We were cautiously optimistic that shortened TTA might, in turn, facilitate shorter TTD, including obviating the need for an overnight stay in some cases.
The use of pain medications of all kinds was lower in the device group compared with patients treated with standard of care. This suggested that, as in other surgical fields, interventions and protocols that facilitated early ambulation after surgical procedures might help reduce the risk of pain medication or opioid requirements and the associated risks. In a recently published retrospective analysis of ablation, femoral venous closure using the 5-F to 7-F VASCADE VCS device was associated with higher comfort and satisfaction level and lower catheter-related urinary complications compared with MC (30). In that study, patients who had previous experience with the post-ablation recovery procedure expressed a notable increase in satisfaction with the shorter duration of bedrest and for the associated decrease in pain and discomfort.
First The complication rates were low, which could be due to the narrow patient inclusion criteria and stringent pre-screening as part of a controlled randomized trial, as well as high use of ultrasound-guided access. However, in a retrospective study, no access site hematoma, infection, or other complications were reported in real-world patients (31). Second, the protocol to assess the eligibility for ambulation was not standardized in the MC group. Rather, it was pre-specified that the sites would follow their standard of care with the rationale that the goal was to compare the device against usual operator or institutional care or practice. Third, this was an unblinded study, and all care providers even downstream of the intervention (e.g., nurses) knew the treatment assignment because it informed clinical care. Fourth, our patient-reported outcome questionnaire was not validated or applied in previous studies. The survey included questions to compare the current experience of patients with a hypothetical experience. In the absence of any validated measures, the questionnaire was developed by a global health outcome research firm with 3 levels of questions; level 2 questions were hypothetical in nature to understand patient preferences regarding pain and discomfort. Finally, post-anesthesia assessment data were not collected, and that could have affected TTA.
When used for closure of multiple, 6-F to 12-F venous access sites after cardiac catheter ablation, the VASCADE MVP closure device shortened TTA and other clinical measures of the post-procedure recovery course, increased patient satisfaction, and decreased the use of opioid narcotics, with no increase in complications compared with MC.
COMPETENCY IN MEDICAL KNOWLEDGE: MC followed by prolonged bedrest is the standard of care for post-procedural closure of venous access sites in patients undergoing catheter-based procedures. However, this approach may often lead to patient discomfort, pain, prolonged immobility, additional interventions, and delayed discharge. The findings of this randomized trial demonstrated that the strategy of venous vascular closure and short bedrest is efficacious, compared with usual care, across a number of primary and secondary patient-centered outcomes.
TRANSLATIONAL OUTLOOK: This was the first randomized trial to compare the safety and efficacy of an investigational venous access site closure device with standard MC in achieving hemostasis at multiple sites following catheter-based electrophysiology procedures (e.g., atrial fibrillation ablation). We also evaluated patient satisfaction with recovery and prevalence of pain medication usage in both groups. Our findings demonstrated the superiority of the vascular closure device in attaining effective hemostasis at a shorter time and thereby providing the patients the ability to ambulate and the satisfaction and freedom to perform simple self-care tasks. These results indicated that a strategy of vascular closure with shortened bedrest can improve a variety of outcomes that are important to patients, improve workflow, potentially minimize interventions, such as pain medications or indwelling bladder catheters, and facilitate earlier discharge.
Abbreviations and Acronyms
analysis of covariance
total post-procedure time
time to ambulation
time to closure eligibility
time to discharge
time to discharge eligibility
time to hemostasis
1. "Catheter ablation for cardiac arrhythmias: utilization and in-hospital complications, 2000 to 2013". J Am Coll Cardiol EP 2017;3:1240-1248.
2. "Temporal trends of in-hospital complications associated with catheter ablation of atrial fibrillation in the United States: an update from Nationwide Inpatient Sample database (2011-2014)". J Cardiovasc Electrophysiol 2018;29:715-724.
3. "Ambulation after femoral sheath removal in percutaneous coronary intervention: a prospective comparison of early vs. late ambulation". J Clin Nurs 2009;18:1862-1870.
4. "Systematic review of randomized controlled trials comparing manual compression to vascular closure devices for diagnostic and therapeutic arterial procedures". Surg Technol Int 2015;27:32-44.
5. "Early ambulation after diagnostic transfemoral catheterisation: a systematic review and meta-analysis". Int J Nurs Stud 2014;51:39-50.
6. "Enhanced recovery after surgery: a review". JAMA Surg 2017;152:292-298.
7. "A prospective, randomized, pivotal trial of a novel extravascular collagen-based closure device compared to manual compression in diagnostic and interventional patients". J Invasive Cardiol 2015;27:129-136.
8. "Interval estimation for the difference between independent proportions: comparison of eleven methods". Stat Med 1998;17:873-890.
9. "Femoral haemostasis after transcatheter therapeutic intervention: a prospective randomised study of the Angio-Seal device vs. the Femostop device". Int J Cardiol 2000;76:235-240.
10. "The effect of early ambulation on patient comfort and delayed bleeding after cardiac angiogram: a pilot study". Heart Lung 1994;23:112-117.
11. "Effect of transradial access on quality of life and cost of cardiac catheterization: a randomized comparison". Am Heart J 1999;138:430-436.
12. "Closure of the femoral artery after cardiac catheterization: a comparison of Angio-Seal, StarClose, and manual compression". Catheter Cardiovasc Interv 2008;71:518-523.
13. "A randomized trial comparing compression, Perclose Proglide and Angio-Seal VIP for arterial closure following percutaneous coronary intervention: the CAP trial". Catheter Cardiovasc Interv 2008;71:1-5.
14. "Treating specialty and outcomes in newly diagnosed atrial fibrillation: from the TREAT-AF study". J Am Coll Cardiol 2017;70:78-86.
15. "Patient satisfaction after femoral arterial access site closure using the ExoSeal® vascular closure device compared to manual compression: a prospective intra-individual comparative study". Cardiovasc Intervent Radiol 2016;39:21-27.
16. "The effect of changing position and early ambulation after cardiac catheterization on patients’ outcomes: a single-blind randomized controlled trial". Int J Nurs Stud 2009;46:1047-1053.
17. "Transradial access compared with femoral puncture closure devices in percutaneous coronary procedures". Int J Cardiol 2009;137:199-205.
18. "Safety of early ambulation after diagnostic and therapeutic neuroendovascular procedures without use of closure devices". Neurosurgery 2010;66:493-496.discussion 496.
19. "Comparison of comfort and local complications after cardiac catheterization". Clin Nurs Res 2001;10:29-39.
20. "Usefulness of ‘Figure-of-eight’ suture to achieve haemostasis after removal of 15-French calibre femoral venous sheath in patients undergoing cryoablation". Europace 2016;18:1545-1550.
21. "Importance of non-pulmonary vein triggers ablation to achieve long-term freedom from paroxysmal atrial fibrillation in patients with low ejection fraction Heart". Rhythm 2016;13:141-149.
22. "Venous access site closures using the VASCADE vascular closure system". J Vasc Interv Radiol 2016;27:1885-1888.
23. "Access site management with vascular closure devices for percutaneous transarterial procedures". J Vasc Surg 2010;52:1682-1696.
24. "Prolonged work- and computer-related seated immobility and risk of venous thromboembolism". J R Soc Med 2010;103:447-454.
25. "Periprocedural outcomes of prophylactic protamine administration for reversal of heparin after cryoballoon ablation of atrial fibrillation". J Interv Card Electrophysiol 2014;41:129-134.
26. "Femoral venous hemostasis after atrial fibrillation ablation: is Figure-of-eight suture the way to go". Indian Pacing Electrophysiol J 2017;17:132-133.
27. "A randomized comparison of a novel bioabsorbable vascular closure device versus manual compression in the achievement of hemostasis after percutaneous femoral procedures: the ECLIPSE (Ensure’s Vascular Closure Device Speeds Hemostasis Trial)". J Am Coll Cardiol Intv 2009;2:785-793.
28. "Venous hemostasis postcatheter ablation of atrial fibrillation while under therapeutic levels of oral and intravenous anticoagulation". J Interv Card Electrophysiol 2015;44:97-104.
29. "Figure-of-eight suture for venous hemostasis in fully anticoagulated patients after atrial fibrillation catheter ablation". Indian Pacing Electrophysiol 2017;17:134-139.
30. "Venous access-site closure with vascular closure device vs. manual compression in patients undergoing catheter ablation or left atrial appendage occlusion under uninterrupted anticoagulation: a multicentre experience on efficacy and complications". Europace 2019;21:1048-1054.
31. "Use of VASCADE Vascular Closure System for venous hemostasis". Vasc Dis Mgmt 2017;14:E171-E173.
For a complete list of the AMBULATE trial investigators as well as supplemental tables and figures, please see the online version of this paper.
This study was sponsored by Cardiva Medical, Inc. Dr. Natale has been a consultant for Biosense Webster, Boston Scientific, Medtronic, Biotronik, Abbott, and Baylis Medical. Dr. DeLurgio has been a consultant for BSCI, Medtronic, and Attricure. Dr. Spear has received fees from Cardiva Medical; and has been a member of the Speakers Bureau for Biosense Webster, Janssen Pharmaceuticals, Bristol-Myers Squibb, and Stereotaxis. Dr. Bunch has received research grants from Boston Scientific and Boehringer Ingelheim. Dr. O'Neill is a member of the Speakers Bureau for Pfizer; and has received research grants from Medtronic and St. Jude Medical. Dr. Turakhia has received compensation for services from Akebia, St. Jude Medical, Precision Health Economics, Cardiva Medical, and iRhythm Technologies; has received Speakers Bureau fees from Medscape; holds equity interest/stocks options (no-public) in AliveCor, Zipline Medical, and iBeat, Inc.; and has received research grants from AstraZeneca Pharmaceuticals, Janssen Pharmaceuticals, Medtronic, Inc., Veterans Administration, and Boehringer Ingelheim. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the JACC: Clinical Electrophysiology author instructions page.