Association between venous access-site closure device use and postoperative nausea and vomiting after atrial fibrillation ablation under propofol sedation: a single-centre retrospective observational study in Okayama, Japan

Abstract

Objectives

To examine whether the use of a venous access-site closure device is associated with the occurrence of postoperative nausea and vomiting (PONV) after atrial fibrillation (AF) ablation under propofol sedation.

Design

Observational study.

Setting

A single-centre retrospective observational study in Okayama, Japan.

Participants

We retrospectively analysed consecutive patients who underwent AF ablation under deep propofol sedation with adaptive servo-ventilation. A total of 686 patients were included. Patients were managed using a standardised sedation protocol with or without low-dose pentazocine. Patients treated with conventional manual compression for haemostasis (n=383) were compared with those treated using a venous access-site closure device (n=303).

Interventions

Postprocedural bed rest duration and the incidence and timing of PONV were compared between groups. Associations between closure device use and PONV were evaluated using logistic regression analysis.

Primary outcome measure

The primary outcome was the occurrence of PONV following AF ablation.

Results

All procedures were completed under propofol sedation without conversion to general anaesthesia. The duration of postprocedural bed rest was shorter in the device group than in the conventional-compression group (mean difference −14.7 hours, 95% CI −15.2 to −14.0).

PONV occurred in 6/303 patients (2.0%) in the device group and 20/383 patients (5.2%) in the conventional-compression group, corresponding to a relative risk of 0.38 (95% CI 0.15 to 0.93), an OR 0.25 (95% CI 0.10 to 0.62) and a risk difference of −3.2% (95% CI −6.0% to −0.5%).

In multivariable analysis, use of a venous closure device was associated with a lower likelihood of PONV.

Conclusions

In this single-centre observational study, use of a venous access-site closure device was associated with a lower occurrence of PONV after AF ablation under propofol sedation. These findings suggest that postprocedural management strategies enabling earlier mobilisation may be associated with improved patient comfort; however, causal inference is limited by the observational design.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• This study used a single-centre retrospective observational design with a consistent ablation strategy, sedation protocol and clinical workflow throughout the study period, minimising procedural heterogeneity.

• Consecutive patients were included before and after the introduction of a venous access-site closure device, allowing pragmatic comparison under real-world clinical conditions.

• Postprocedural management protocols, including bed-rest duration and mobilisation timing, were predefined and uniformly applied within each study period.

• The retrospective design limits causal inference, and unmeasured confounding factors related to clinical practice changes over time cannot be completely excluded.

• As all patients underwent propofol sedation with adaptive servo-ventilation at a single institution, the generalisability of the findings to other centres or anaesthetic strategies may be limited.

Introduction

Postoperative nausea and vomiting (PONV) remain clinically relevant adverse events following invasive cardiovascular procedures, including catheter ablation for atrial fibrillation (AF).¹ Even when not life-threatening, PONV substantially impairs patient comfort, delays recovery and negatively influences patients’ willingness to undergo repeat procedures. This issue is particularly important in AF ablation, where multiple sessions are often required to achieve durable rhythm control.²

AF ablation procedures frequently involve prolonged procedural times and the use of deep sedation or anaesthesia to ensure patient immobility and procedural stability. Propofol-based sedation has been widely adopted in electrophysiology laboratories because of its rapid onset, short recovery time and relatively low incidence of PONV compared with inhalational anaesthetic agents.³,⁵ Nevertheless, PONV continues to occur in a subset of patients even under propofol sedation, suggesting that factors beyond the choice of sedative agent may contribute to its development.

In AF ablation, vascular access is typically obtained through multiple femoral venous sheaths, necessitating secure post-procedural haemostasis. Traditionally, haemostasis has been achieved using prolonged manual compression followed by rigid and extended bed rest, often lasting overnight.⁶ Such enforced immobility is known to induce a variety of adverse physiological effects, including discomfort, autonomic imbalance and gastrointestinal dysregulation.^{7 8} Although bed rest has not been classically considered a direct cause of PONV, prolonged and rigid immobilisation may exacerbate nausea through altered autonomic tone and impaired gastrointestinal motility.

Recent advances in venous access-site closure technology have enabled faster haemostasis after femoral venous catheterisation, allowing earlier ambulation and less restrictive post-procedural management.⁹ These devices have primarily been evaluated in terms of vascular safety, bleeding complications and feasibility of same-day discharge. However, their potential impact on patient-centred outcomes, such as PONV, has not been systematically examined.

Given the multifactorial nature of PONV, we hypothesised that post-procedural management—specifically the duration and rigidity of bed rest—plays a meaningful role in its occurrence following AF ablation. We further hypothesised that the use of a venous closure device, by enabling earlier and less restrictive mobilisation, would be associated with a lower incidence of PONV, even under standardised deep propofol sedation.

Accordingly, the present study aimed to compare the incidence and temporal pattern of PONV between patients undergoing AF ablation managed with conventional manual compression and prolonged bed rest and those managed with a venous access-site closure device permitting early ambulation. By focusing on postprocedural care rather than sedative technique alone, this study seeks to provide new insight into modifiable factors that may improve patient comfort and perioperative quality of care after AF ablation.

Methods

Patients

This single-centre retrospective observational study was conducted at the Okayama Heart Clinic in Okayama, Japan. We analysed 686 consecutive patients with AF who underwent AF ablation between January 2023 and October 2024. Initially, 383 patients (mean age 66±10 years) who underwent AF ablation with conventional haemostasis using manual compression and bandaging between January and December 2023 were included in the conventional-compression group. After an initial learning period of five cases, a venous access-site closure device was introduced, and 303 patients (mean age 68±10 years) treated between January and October 2024 were included in the device-used group, allowing shorter and less rigid bed rest.

The ablation and haemostasis procedures were performed by a multidisciplinary team consisting of operators, specialised nurses and medical engineers. The composition of the ablation team did not change substantially during the study period.

Target trial emulation framework

A flow diagram of the study population is shown in figure 1.

Figure 1. Flow diagram of the study population. AF, atrial fibrillation.

Eligibility criteria

Patients were excluded if they met any of the following criteria: (1) unsuccessful AF ablation; (2) major complications; (3) local or systemic infection at or near the vascular access site; (4) history of venous thromboembolism or bleeding tendency; (5) platelet count <100×10⁹/L; (6) femoral catheterisation within the preceding 30 days or (7) failure to achieve acute haemostasis using the venous closure device.

Time zero

Time zero was defined as the start of AF ablation.

Treatment strategies

Haemostasis at the vascular access site was achieved using either manual compression or a venous access-site closure device.

Follow-up period

Patients were followed from the start of ablation until the morning after the procedure, when ambulation was permitted.

Outcome definition

The outcome was the occurrence of PONV, defined as a binary variable (yes/no).

Estimand

In patients undergoing AF ablation under deep propofol sedation, the estimand of interest was the occurrence of PONV after the procedure, comparing the venous closure device group with the conventional compression group, regardless of subsequent rescue antiemetic use.

Primary endpoint

The primary endpoint was the occurrence of PONV after AF ablation.

Main statistical measure

The occurrence of PONV was summarised using ORs and risk differences with 95% CIs. Multivariable logistic regression analysis was additionally performed to adjust for potential confounding factors.

Patient and public involvement

Patients and the public were not involved in the design, conduct, reporting, or dissemination of this research.

AF ablation

AF ablation was performed by an experienced multidisciplinary team following an established institutional protocol. Because the ablation strategy and catheter techniques were identical to those described in our previous reports, only a concise overview is provided.^{6 10} Pulmonary vein isolation was achieved using an electroanatomical mapping system with standard radiofrequency or temperature-controlled ablation catheters. Additional ablation procedures were performed at the operator’s discretion based on individual clinical indications.

During the procedure, pentazocine and/or etilefrine hydrochloride were administered at the operator’s discretion when pain occurred and/or arterial blood pressure decreased.

Haemostasis using a venous access-site closure device

Haemostasis at femoral venous access sites was achieved either by conventional manual compression or by use of a venous access-site closure device. In the device group, femoral venous haemostasis was achieved using a venous closure device (VASCADE MVP; Cardiva Medical Inc., Santa Clara, California, USA) according to the manufacturer’s instructions. Detailed technical aspects and vascular safety outcomes of this device have been reported previously and are therefore not reiterated here.^{9 11 12} Three catheters were inserted via the right femoral vein and one via the left femoral vein, and a 6-Fr catheter was placed in the coronary sinus via the right internal jugular vein. A closure device was applied to each femoral access site, resulting in the deployment of four devices after sheath removal; no device was used at the jugular venous site. In the conventional-compression group, haemostasis was obtained by manual compression followed by suturing and elastic bandaging. All patients were monitored for bleeding or access-site complications as part of routine postprocedural care.

Acute haemostasis was defined as cessation of bleeding without haematoma formation or expansion. Sustained closure success was defined as the absence of vascular complications during the follow-up period.

Post-ablation management protocol

Figure 2 illustrates the post-AF ablation management protocol with time-dependent occurrence of PONV. Patients in the conventional-compression group underwent rigid and prolonged overnight bed rest and were allowed to ambulate the following morning. In contrast, patients in the device-used group experienced shorter and less rigid bed rest and were permitted to ambulate 2–4 hours after AF ablation if no bleeding or access-site complications were observed.

Figure 2. Post-atrial fibrillation-ablation management protocol and time sequence of postoperative nausea and vomiting (PONV) occurrence. Numbers shown in the blue bars indicate cumulative PONV occurrence.

Postablation nausea and vomiting follow-up

Assessment of PONV was conducted using a standardised institutional protocol.¹ Patients were routinely evaluated for gastrointestinal symptoms during scheduled nursing assessments within the first 12 hours after ablation. In addition to these fixed assessments, patient-reported symptoms were continuously captured via the nurse call system.

Episodes of nausea and/or vomiting were documented as binary outcomes. Antiemetic therapy with metoclopramide was administered at the discretion of the attending physician when clinically indicated. No standardised protocol for the treatment of PONV was implemented in this study. Because the primary aim of the present study was to compare the incidence of PONV between the conventional compression group and the device-used group, antiemetic treatment after the onset of PONV was not included in the study protocol as an outcome measure. Metoclopramide administration was performed at the discretion of the attending physician after the occurrence of PONV and therefore was not considered to affect the incidence of PONV itself.

Complications

Major thromboembolic events included cerebral embolism, transient ischaemic attack after exclusion of intracranial haemorrhage, pulmonary embolism and deep vein thrombosis. Major bleeding events included bleeding requiring transfusion, haematomas necessitating surgical intervention and cardiac tamponade requiring drainage.

Statistical analyses

All statistical analyses were performed using R software, V.3.2.2 (R Foundation for Statistical Computing, Vienna, Austria).¹³

The incidence of PONV was compared between the conventional-compression and device-used groups using 2×2 contingency tables. Crude effect estimates were expressed as risk ratios, ORs and risk differences (RDs), each with 95% CIs.

Continuous variables are presented as means with SDs or medians with IQRs, as appropriate.

To identify factors associated with PONV, multivariable logistic regression analysis was performed. PONV was treated as the dependent variable, and independent variables included age, sex, Apfel score, use of analgesics and sedatives and use of a venous closure device. Adjusted associations are reported as ORs with 95% CIs.

A propensity score for the use of a venous closure device was estimated using logistic regression including baseline covariates. The propensity score was then included as an additional covariate in the logistic regression model to adjust for potential confounding.

In accordance with current reporting recommendations, results are presented as effect estimates with 95% CIs.

Results

Patient characteristics

Baseline characteristics, including hypertension, diabetes mellitus, smoking status and pentazocine use, were broadly comparable between groups, with no clinically meaningful differences (table 1 and online supplemental Table S1). Overall, the two groups were well balanced with respect to demographic variables, comorbidities and ablation-related factors. Creatinine clearance was higher and left atrial diameter was lower in the device-used group than in the conventional-compression group; however, the absolute differences were small and not considered clinically meaningful.

Table 1. Patient characteristics.

	Conventional-compression group	Device group	Difference (device - conventional)	95% CI	Standardised difference
Number of patients	383	303
Patient background characteristics
Age	66±10	68±10	−0.89	−2.45 to 0.66	−0.087
Female sex n (%)	112/383 (29.2)	82/303 (27.1)			−0.048
Body mass index	24.17±4.25	24.12±3.64	−0.06	−0.65 to 0.54	−0.014
Creatinine clearance (mL/min)	75.72±27.97	80.00±25.38	4.28	0.27 to 8.29	0.160
Left atrial diameter (mm)	42.16±7.34	41.14±6.29	−1.03	−2.05 to 0.00	−0.150
Left ventricular ejection fraction (%)	64.66±9.19	65.11±8.01	0.54	−0.84 to 1.75	0.053
Hypertension	239/383 (62.4)	196/303 (65.8)			0.070
DM	62/383 (15.9)	41/303 (13.5)			−0.057
Smoking	50/383 (13.1)	41/303 (13.6)			−0.014
Type of AF
paroxysmal AF	202 (53)	172 (53)
Persistent AF	159 (42)	121 (40)
Long-standing persistent AF	22 (6)	10 (5)
Procedural factors
Procedure time (min)	103±33	102±28	−1.0	−5.6 to 3.6	−0.057
Use of etilefrine hydrochloride n (%)	48 (13)	35 (12)			−0.030
PONV related factors
Propofol administered (mL/kg/hour)	0.93±0.26	0.89±0.53	−0.043	−0.107 to 0.023	−0.102
Use of pentazocine n (%)	186/383 (48.6)	166/303 (54.8)			0.125
Total pentazocine administered (mg)	7.6±8.9	9.1±9.5	1.5	0.10 to 2.90	0.125
Apfel score	1.31±0.76	1.22±0.70	−0.09	−0.20 to 0.02	−0.123
Complications
Major thromboembolic complications, n (%)	0 (0)	0 (0)
Major bleeding complications, n (%)	0 (0)	0 (0)

AF, atrial fibrillation; DM, diabetes mellitus; PONV, postoperative nausea and vomiting.

Factors related to PONV

Sedative management was comparable between the two groups (table 1). The total amount of propofol administered and the proportion of patients receiving pentazocine were broadly comparable. Although the mean dose of pentazocine was slightly higher in the device-used group, this difference was modest. The Apfel score, a commonly used predictor of PONV, showed similar distributions in both groups. In contrast, the duration of postprocedural bed rest was 18.1±2.1 hours in the conventional-compression group and 3.4±0.7 hours in the device group, with a mean difference of −14.7 hours (95% CI −14.93 to −14.47).

Procedural course under propofol sedation with adaptive servo-ventilation

All AF ablation procedures were completed successfully under deep propofol sedation with adaptive servo-ventilation (ASV). No patient required conversion to inhalational anaesthesia with endotracheal intubation, and no procedure was interrupted because of sedation-related issues. Hypotension requiring etilefrine hydrochloride occurred infrequently and was broadly comparable between the two groups (table 1). Procedural duration was within the expected range for AF ablation and was comparable between the conventional-compression and device-used groups.¹⁴ Likewise, the total volume of propofol administered was broadly comparable between groups.

Haemostasis outcomes

Haemostasis at venous access sites was achieved successfully in all patients. No cases of clinically relevant re-bleeding or access-site haematoma were observed. In the device-used group, both acute and sustained haemostasis were achieved in all patients. Importantly, ambulation after the prescribed bed-rest period did not necessitate any additional haemostatic intervention.

Incidence and temporal pattern of PONV

The incidence of PONV was 2.0% in the device group and 5.2% in the conventional-compression group (RR 0.38, 95% CI 0.15 to 0.93; OR 0.25, 95% CI 0.10 to 0.62; RD −3.2%, 95% CI −6.0% to −0.5%) (figure 3).

Figure 3. Occurrence of postoperative nausea and vomiting (PONV) after atrial fibrillation ablation in the conventional-compression and device groups.

Figure 2 shows the temporal distribution of PONV events. During the initial postprocedural bed-rest period (within 2–4 hours after ablation), fewer patients in the device-used group experienced PONV than in the conventional-compression group (5 vs 13 patients). After ambulation was permitted, only one patient in the device-used group developed PONV, whereas PONV events in the conventional-compression group continued to occur mainly during prolonged overnight bed rest.

Multivariate analysis

Results of the multivariable logistic regression analysis are presented in table 2. Use of a venous closure device was associated with a lower likelihood of PONV after AF ablation. Female sex was associated with a higher likelihood of PONV. In contrast, the Apfel score was not associated with PONV occurrence in this cohort.

Table 2. Multivariable logistic regression analysis for PONV.

Dependent variable: PONV occurrence (no=0, yes=1)
Independent variables	OR	95% CI	P value
Age (years)	0.965	0.926 to 1.000	0.080
Sex (male vs female)	0.184	0.064 to 0.534	0.002
Administered pentazocine (mg)	1.240	0.517 to 2.960	0.631
Propofol infusion (mL/kg/hour)	0.183	0.023 to 1.470	0.110
Apfel scores	1.020	0.535 to 1.930	0.963
Device use (yes vs no)	0.332	0.129 to 0.855	0.022

PONV, postoperative nausea and vomiting.

After adjustment using the propensity score, the association was attenuated, possibly reflecting model instability due to the limited number of events. When adjusted for age, sex, pentazocine dose, propofol infusion and Apfel score, the adjusted OR was similar to the crude odds ratio (table 3). The consistency between crude and adjusted estimates supports the robustness of the findings.

Table 3. Crude and adjusted comparison of PONV between groups.

Outcome	Device group	Conventional-compression	Crude OR	Adjusted OR
		Group	(95% CI)	(95% CI)
PONV occurrence	6/303 (2.0%)	20/383 (5.2%)	0.25	0.33
			(0.10 to 0.62)	(0.13 to 0.86)

Adjusted for age, sex, pentazocine dose, propofol infusion and Apfel score.

PONV, postoperative nausea and vomiting.

Discussion

This retrospective observational study suggests that differences in postprocedural management may be associated with a lower incidence of PONV after AF ablation. Specifically, the use of a venous access-site closure device, which enabled earlier and less restrictive mobilisation, was associated with a PONV incidence of 2% in patients undergoing AF ablation under deep propofol sedation with ASV.

A notable strength of the present study is the consistency of procedural and institutional conditions across the two comparison groups. Patients treated with conventional manual compression and those treated with a venous closure device were enrolled during consecutive periods before and after device introduction, while ablation techniques, sedation protocols, and team composition remained unchanged. This design minimised procedural heterogeneity and allowed a pragmatic comparison of outcomes under routine clinical practice. Although the mean dose of pentazocine was slightly higher in the device-used group, this difference was not associated with a higher incidence of PONV. Furthermore, clinical characteristics and procedural parameters were largely comparable between groups, and multivariable analyses supported the robustness of the observed association between closure-device use and lower PONV incidence.

Deep propofol sedation combined with ASV was successfully maintained in all patients, without the need for conversion to inhalational anaesthesia or procedure interruption. Both inhalational anaesthesia and propofol-based sedation are widely used during AF ablation, each with distinct advantages and limitations. Inhalational anaesthesia provides stable anaesthetic depth and eliminates patient movement but requires anaesthesiology resources and specialised equipment. In contrast, propofol-based sedation necessitates careful titration and close monitoring but offers logistical simplicity and has been associated with lower rates of PONV. Previous studies have reported acceptable procedural times and low complication rates with deep propofol sedation using ASV, supporting its feasibility in the electrophysiology laboratory.^{6 10 15} In the present study, the sedation strategy was standardised across groups, allowing the effects of post-procedural management to be evaluated independently of anaesthetic technique.

The incidence of PONV observed in the conventional-compression group (approximately 5%) was already lower than that reported in many surgical and interventional settings. Prior studies have described PONV rates of 20%–40% in patients receiving propofol sedation for various procedures, although these rates vary widely depending on sedative combinations and clinical context.^{16 17} Propofol has been shown to exert intrinsic antiemetic effects, which may partially explain the relatively low baseline incidence observed in this cohort.¹⁸,²⁰ Large database analyses of catheter ablation procedures have also demonstrated substantial variability in PONV incidence, reflecting heterogeneity in sedative regimens, perioperative management and patient populations.²¹ Within this context, the association of the further reduction in PONV incidence observed in the device-used group appears clinically meaningful.

Importantly, the temporal distribution of PONV events provides insight into potential contributing factors. In the conventional-compression group, PONV events occurred predominantly during prolonged and rigid bed rest, whereas PONV was infrequent after ambulation in the device-used group. These findings suggest that not only the duration but also the rigidity of enforced bed rest may be associated with PONV occurrence. To our knowledge, no previous studies have specifically examined the relationship between venous access-site closure devices, post-procedural mobilisation and PONV after AF ablation.

Established risk factors for PONV include inhalational anaesthesia, opioid use, female sex, nonsmoking status and a history of motion sickness or PONV, which are incorporated into the Apfel score.^{22 23} In the present cohort, inhalational anaesthesia was not used, opioid exposure was limited and the proportion of smokers was small. Under these conditions, the Apfel score did not effectively discriminate PONV risk. In contrast, use of a venous closure device—reflecting shorter and less restrictive bed rest—was associated with a lower likelihood of PONV in multivariable analysis. These findings highlight post-procedural management as a potentially modifiable factor influencing PONV risk after AF ablation.

The mechanisms linking enforced bed rest and PONV were not directly investigated in this study. PONV is mediated by complex interactions involving central emetic pathways, peripheral neurotransmitter systems and the enteric nervous system.²⁴ Prolonged immobility has been associated with autonomic alterations, which may influence gastrointestinal function and nausea perception.²⁵ Earlier and less restrictive mobilisation may mitigate these effects and contribute to improved gastrointestinal function, although this hypothesis remains speculative and warrants further investigation.

Limitations

Several limitations should be considered. First, this was a retrospective, single-centre observational study rather than a prospective randomised trial, which limits causal inference. In addition, the comparison between groups across different time periods may introduce temporal confounding. Second, all patients underwent propofol sedation with ASV, precluding comparison with other anaesthetic strategies. Third, mechanistic data explaining the association between bed rest and PONV were not available. Despite these limitations, the relatively large sample size and consistent procedural conditions strengthen the validity of the findings.

Conclusions

In patients undergoing AF ablation under deep propofol sedation with ASV, the use of a venous access-site closure device was associated with a lower incidence of PONV. These findings suggest that optimisation of post-procedural management—particularly through shorter and less restrictive bed rest—may be associated with improved perioperative patient comfort beyond the effects of sedation strategy alone.

Data availability statement

Data are available on reasonable request.