Timing of outpatient stenting is associated with stent size in nonthrombotic iliac vein lesions

Abstract

Objective

Stent placement is commonly used to treat nonthrombotic iliac vein lesions (NIVLs). On the day of surgery, patients typically fast after midnight and may often present with volume deficits and various degrees of dehydration. This study aims to determine if the time of day stenting is performed predicts the size of the iliac vein stent used to treat NIVLs.

Methods

A retrospective study of patients who underwent outpatient stenting and first-time treatment for NIVLs at a single institution from January 2013 to March 2022 was performed. The exposure variable was the time of day the outpatient stenting was performed, serving as a proxy for fasting duration. Stent timing was dichotomized into an early cohort (procedures before 11 am) and a late cohort (procedures after 11 am). Small stent size was defined as 12 to 16 mm, and large stent size was defined as 18 to 22 mm. A multivariable logistic regression analysis evaluated the independent association between the timing of stenting and iliac vein stent size.

Results

Sixty-four patients met the inclusion criteria, with 53% (n = 34) in the late cohort. The median age of the study population was 53 years (interquartile range, 39-64 years). Patients were 75% female, 30% non-White, and 69% non-Hispanic. Although there were no significant differences in baseline characteristics or iliac vein measurements on intravascular ultrasound examination before stenting between the early and late cohorts, patients in the early cohort were significantly more likely to have a large stent size used compared with the late cohort (70% vs 44%; P = .037). On multivariable logistic regression, each additional hour later in the day that stenting was performed was associated with a 28% higher odds of using a smaller venous stent (P = .036). Older age also independently predicted the implantation of a smaller stent (odds ratio, 0.95; 95% confidence interval, 0.91-0.99; P = .022).

Conclusions

Outpatient stenting performed later in the day was associated with a smaller stent size in patients treated for NIVLs. A 9-hour delay in stent placement was associated with a 2-mm reduction in stent diameter (ie, one stent size smaller). Future studies should investigate the role of a preoperative hydration protocol in optimizing stent sizing for patients with iliac vein lesions.

Article Highlights.

• •Type of Research: Single-center retrospective cohort study
• •Key Findings: A small stent size (12-16 mm) was more likely to be used for left-sided nonthrombotic iliac vein lesions in patients undergoing first-time iliac venous stenting later in the day (n = 34) than earlier in the day (n = 30; P = .037).
• •Take Home Message: Outpatient stenting performed later in the day was associated with a smaller stent size used to treat nonthrombotic iliac vein lesions.

In the United States, approximately 25 million people suffer from chronic venous insufficiency (CVI), and 6 to 7 million people have advanced, disabling disease.¹ Nonthrombotic iliac vein lesions (NIVLs) are an increasingly recognized mechanism involved in the pathophysiology of CVI.² Iliac vein stenosis or compression ≥50% can cause symptoms of CVI in some patients, including chronic edema, pain, and venous ulceration.³ Venous stenting is the primary method for treating symptomatic NIVLs.⁴ Stenting can achieve an ulcer healing rate between 62% and 76% at 2 years post procedure.^5,6

The use of iliac vein stenting to treat CVI is increasing. However, there has been a concomitant increase in reports of venous stent migration to the heart, and the incidence of this devastating complication might be underreported.^7,8 In 2021, the US Food and Drug Administration recalled a venous stent system owing to reports of stent migration after implantation.9, 10, 11, 12, 13 Changes in intraluminal pressure and intravascular volume alter vein size, and these factors can vary based on the patient's position, respiration, and hydration status.^14,15 In cases of NIVLs, veins are uniquely affected by intraluminal pressure and intravascular volume, because these lesions tend to be short, elliptical, and less fibrotic than other CVI pathologies.¹⁶ Because vein size dictates the size of the venous stent implanted, accurate measurements are critical to prevent stent migration.

In 1964, Atkinson and Lee¹⁷ recommended that patients not consume anything by mouth (ie, nil per os, NPO) on the day of their procedure. Since then, hospitals have implemented a rule of absolute fasting starting at midnight.^18,19 With the accumulation of evidence, this traditional practice is now viewed as dogmatic.²⁰ Nevertheless, institutions vary in their NPO practices, and on average, patients forgo solids for 13.5 hours and liquids for 9.6 hours before procedures.^19,21

Vein size is affected by hydration status,²² yet the impact of fasting duration on the size of iliac venous stents has not been investigated. In 2024, a consensus statement from the VIVA Foundation, the American Venous Forum, and the American Vein and Lymphatic Society encouraged research into the effects of hydration on the endovascular management of NIVLs.¹⁶ The present study aimed to determine if fasting duration, operationalized as the time of day of an outpatient procedure, is associated with the size of the iliac vein stent used to treat NIVLs. We hypothesized that patients undergoing iliac vein stenting later in the day would have smaller venous stents implanted.

Methods

Study design and eligibility criteria

This retrospective cohort study included patients with CVI who underwent outpatient venography at a single institution from November 2013 to March 2022. Patients with a first-time iliac vein stenting procedure for left-sided NIVLs were eligible. Stenting of the left common iliac vein and the left external iliac vein were included. Outpatient procedures were defined as those for which the patient was admitted from outside the hospital and thus did not receive any intravenous hydration before the procedure and was discharged on the same day after the procedure. Patients with a history of venous stenting procedures were excluded, because prior stenting could affect vein size. Patients who underwent only right-sided iliac stenting procedures were excluded to strengthen the homogeneity of the study population. The study was conducted per the principles of the Declaration of Helsinki. The Yale University Institutional Review Board deemed the study exempt (#2000031926), and informed consent was waived because this was a retrospective study posing minimal risk to included patients.

Timing of outpatient stenting

The exposure of interest was the time of day the outpatient stenting was performed, serving as a proxy for fasting duration. At the study's institution, all patients scheduled for outpatient stenting are instructed to be NPO starting at midnight on the day of the procedure. Thus, fasting duration was operationalized as the time of day of the procedure, so patients who underwent procedures later in the day were theoretically fasting longer. The average time of day of the procedures in the overall cohort was 11 am. Therefore, for ease of interpretation, the timing of stenting was dichotomized into patients who had procedures before 11 am (early cohort) and patients who had procedures after 11 am (late cohort). On regression analyses, the timing of stenting was treated as a continuous variable and calculated as minutes after midnight. This approach preserved the variable's granularity and statistical power.

Patient and procedural characteristics

The baseline characteristics were age, sex, race, ethnicity, body mass index (BMI), smoking status, and medical comorbidities. Age and BMI were treated as continuous variables, and the other baseline characteristics were treated as categorical variables. Sex was ascertained from the electronic medical record. The procedural characteristics captured were procedure start time, anesthesia type, patient positioning, and intravascular ultrasound (IVUS) measurements. At our institution, vein measurements are obtained by manually tracing the lumen. IVUS measurements included the minimum and maximum cross-sectional areas, in square millimeters, of the left common and external iliac veins. The degree of stenosis was calculated as the maximum cross-sectional area minus the minimum cross-sectional area divided by the maximum cross-sectional area of the vein. Procedure start time and IVUS measurements were treated as continuous variables. The veins that were stented, the types of stents used, and the sizes of the stents were compared between the two groups.

Primary outcome

The primary outcome was the size of the iliac vein stent. Iliac vein stent size was reported as the diameter of the stent and treated as a binary variable of small stent size and large stent size and as a continuous variable in mm. Small stent size was defined as 12 to 16 mm and large stent size was defined as 18 to 22 mm.

Statistical analyses

Medians and interquartile ranges (IQRs) were calculated for continuous variables, and frequencies and percentages were calculated for categorical variables. The Wilcoxon rank-sum test was used for between-group comparisons for continuous variables. Pearson's χ² or Fisher's exact test were used for categorical variables. Multivariable logistic regression evaluated the association of stent timing with the binary outcome of iliac vein stent size, adjusting for age, sex, race, ethnicity, BMI, and patient positioning. These covariates were selected based on previous literature and clinical relevance. Significance was set at a P value of <.05. All statistical analyses were performed in R (version 2023.06.2, R Foundation for Statistical Computing).

Subgroup analysis

Cases performed under general anesthesia are more likely to involve intraoperative fluid administration and are subject to the vasodilatory effects of most general anesthetic agents. To account for anesthesia type as a potential confounder, a subgroup analysis was conducted, limited to patients who received local anesthesia, excluding those who underwent general anesthesia. This multivariable logistic model incorporated the same covariates as the primary logistic regression, except for ethnicity owing to insufficient sample size for reliable estimation.

Sensitivity analysis

To assess the robustness of the results, a sensitivity analysis using multiple linear regression was performed, treating iliac vein stent size as a continuous variable. This model included the same covariates as the primary logistic regression.

Results

Baseline demographics and clinical characteristics of the overall cohort

Of the 149 patients with CVI who underwent lower extremity venography from 2013 to 2022 in our single institution database, 64 (43%) met the eligibility criteria, namely, first-time outpatient iliac vein stenting for a left-sided NIVL (Fig 1). The cohort's median age was 53 years (IQR, 39-64 years). The cohort was 75% female (n = 48), 30% non-White (n = 19), and 69% non-Hispanic (n = 47). The median BMI was 28 kg/m² (IQR, 25-33 kg/m²). Regarding comorbidities, 8% (n = 5) were current smokers and 30% (n = 19) were former smokers. There were 2 patients (3%) with chronic kidney disease, 12 patients (19%) with coronary artery disease, 5 patients (8%) with a hypercoagulable disorder, 6 patients (9%) prescribed long-term anticoagulation, 9 patients (14%) with a history of deep venous thrombosis, and 2 patients (3%) with a history of pulmonary embolism. Seventeen patients (27%) had a history of an intervention for lower extremity superficial venous disease. No patients had a history of inferior vena cava filter placement.

Fig 1.

Flow diagram of the patient cohort. NIVL, nonthrombotic iliac vein lesion.

There were 34% of patients (n = 22) with advanced Clinical-Etiology-Anatomy-Pathophysiology classification of 4 to 6. Twenty-three patients (36%) received general anesthesia, and 9 procedures (14%) were performed with the patient in the prone position to obtain access via the popliteal vein. For most patients (83%), venous access was obtained via the common femoral vein. The most frequent iliac vein stent size placed was 18 mm (n = 22) and 16 mm (n = 21), and the least frequent iliac vein stent size placed was 12 mm (n = 2) (Fig 2).

Fig 2.

Histogram of iliac vein stent sizes for patients with left-sided nonthrombotic iliac vein lesions (NIVLs). Size was reported as the diameter, in millimeters, of the largest stent placed during the procedure.

Demographics and clinical characteristics of the early vs late cohorts

There were 30 patients (47%) in the early cohort and 34 in the late cohort (53%). The two cohorts had no significant differences in baseline characteristics (Table I). The median procedure start time for the early and late cohorts was 8:36 am (range, 7:45 am to 10:15 am) and 1:18 PM (range, 11:01 am to 5:36 PM), respectively. The median time from midnight to the start of the procedure was 516 minutes (8 hours 36 minutes) for the early cohort and 798 minutes (13 hours 18 minutes) for the late cohort (P < .01). There were no incidents of stent migration in either cohort.

Table I.

Demographic and clinical characteristics of patients with left-sided nonthrombotic iliac vein lesions (NIVLs) undergoing outpatient stent placement, stratified by timing of venography

Baseline characteristics	Timing of stenting		P valueb
	Early cohort (before 11 am) n = 30	Late cohorta (after 11 am) (n = 34)
Age, years	53 (40-66)	52 (38-62)	.77
Sex			.77
Female	22 (73)	26 (77)
Male	8 (27)	8 (24)
Race			.15
White	24 (80)	21 (62)
Black	1 (3)	6 (18)
Other	5 (17)	7 (21)
Ethnicity			.13
Hispanic	9 (30)	8 (24)
BMI (kg/m2)	28.0 (25.1-32.1)	30.3 (24.2-34.8)	.43
Smoking			.36
Current	1 (3)	4 (12)
Former	11 (37)	8 (24)
Never	18 (60)	22 (65)
Comorbidities
Chronic kidney disease	1 (3)	1 (3)	1.00
Coronary artery disease	7 (23)	5 (15)	.52
History of hypercoagulable state	2 (7)	3 (9)	.64
Chronic anticoagulation	2 (7)	4 (12)	.68
Previous history of DVT	5 (17)	4 (12)	.72
Anesthesia type			.91
Local	19 (63)	22 (65)
General	11 (37)	12 (35)
Patient positioning			1.00
Supine	26 (87)	29 (85)
Prone	4 (13)	5 (15)

CIV, Common iliac vein; DVT, deep vein thrombosis; EIV, external iliac vein; IVUS, intravascular ultrasound.

Percentages may not total 100 owing to rounding.

Values are median (interquartile range) or number (%).

^aPatients with procedure times at 11 am were included in the late group.

^bSignificance was set at P < .05.

IVUS measurements, stent characteristics, and primary patency

There was a trend toward larger minimum (64 mm²vs 55 mm²; P = .62) and maximum cross-sectional areas (205 mm²vs 166 mm²; P = .28) in the patients treated in the early group compared with the late group. These differences did not attain statistical significance (Table II). The median degree of stenosis was 60% (IQR, 54%-74%) in the early cohort vs 65% (IQR, 56%-74%) in the late cohort (P = .58). The type of iliac vein stent used did not vary between the two cohorts (P = .74). There was a trend toward a larger median stent size in the early group compared with the late group on unadjusted analysis (18 mm vs 16 mm; P = .06). After grouping into large and small stent categories, patients in the early cohort were significantly more likely to have a large stent size used compared with the late cohort (Fig 3).

Table II.

Intravascular ultrasound (IVUS) measurements and stent characteristics of patients with left-sided nonthrombotic iliac vein lesions (NIVLs) undergoing outpatient stent placement

Outcomes	Timing of stenting		P valueb
	Early cohort (before 11 am) (n = 30)	Late cohorta (after 11 am) (n = 34)
Left CIVc
Minimum cross-sectional area, mm2	64 (45-84)	55 (45-73)	.62
Missing	3 (10)	10 (29)
Maximum cross-sectional area, mm2	205 (140-239)	166 (141-182)	.28
Missing	5 (17)	10 (29)
Degree of stenosis, %	60 (54-74)	65 (56-74)	.58
Left EIVd	NR	NR	-
Left CIV stenting	30 (100)	31 (91)	.24
Left EIV stenting	11 (37)	11 (32)	.72
Type of stent			.74
Wallstent	20 (67)	21 (62)
Venovo	8 (27)	9 (27)
Abre	2 (7)	2 (6)
Zilver Vena	0 (0)	2 (6)
Iliac vein stent size (mm)	18 (16-20)	16 (15-17)	.06
Iliac vein stent size			.037
Large stent (18-22 mm)	21 (70)	15 (44)
Small stent (12-16 mm)	9 (30)	19 (56)

CIV, Common iliac vein; EIV, external iliac vein size; NR, not reported.

Percentages may not total 100 owing to rounding.

Values are median (interquartile range) or number (%).

Boldface entries indicate statistical significance.

^aPatients with procedure times at 11 am were included in the late group.

^bSignificance was set at P < .05.

^cPrestenting IVUS measurements.

^dIVUS measurements for the left EIV were not reported because >90% of these values were missing.

Fig 3.

Comparing the percentage of large stents used to treat left-sided nonthrombotic iliac vein lesions (NIVLs) between the early and late cohorts. Large stents were defined as 18 to 22 mm in size. The early cohort consisted of patients who had procedures before 11 am, and the late cohort consisted of those after 11 am. Pearson's χ² test was performed to compare the two cohorts.

Regression analysis of iliac vein stent size

On multivariable logistic regression, where iliac vein stent size was treated as a binary outcome, each additional hour later in the day that stenting was performed was associated with a 28% higher odds of using a smaller stent (odds ratio [OR], 0.78; 95% confidence interval [CI], 0.61-0.98; P = .045). Older age also independently predicted the implantation of a smaller stent (OR, 0.95; 95% CI, 0.91-0.99; P = .022). Sex, race, ethnicity, BMI, and prone positioning were not significantly associated with the stent size used in the adjusted analysis (Table III).

Table III.

Multivariable logistic regression of large iliac vein stent size

Characteristic	OR	95% CI	P valuea
Time from midnight, in hours	0.78	0.61-0.98	.045
Age, in years	0.95	0.91-0.99	.022
Sex (ref: male)	1.48	0.37-6.45	.587
Race (ref: White)	2.75	0.61-3.84	.200
Ethnicity (ref: Hispanic)	1.59	0.36-7.29	.540
BMI, kg/m2	0.97	0.87-1.08	.630
Prone (ref: supine)	0.23	0.03-1.23	.110

BMI, Body mass index; CI, confidence interval; OR, odds ratio.

Iliac vein stent size was treated as a binary outcome.

Boldface entries indicate statistical significance.

^aSignificance was set at P < .05.

Subgroup analysis excluding general anesthesia cases

For patients who received only local anesthesia, the findings remained unchanged (Supplementary Table I, online only). The timing of stenting, in hours after midnight (OR, 0.72, 95% CI, 0.50-0.97, P = .045) and older age (OR, 0.93, 95% CI, 0.88-0.98, P = .016) were associated with an increased odds of using a smaller stent.

Sensitivity analysis

A multiple linear regression was performed, treating iliac vein stent size as a continuous variable. The overall model was statistically significant, F(7, 56) = 2.62; P = .021. Stenting procedures performed later in the day were associated with a smaller stent size (beta, −0.258; t-value = −2.467; P = .017). Older age (beta, −0.035; t-value = −2.017; P = .049) and prone positioning (beta, −1.714; t-value = −2.121; P = .038) were also associated with smaller stent sizes (Supplementary Table II, online only).

Discussion

NIVLs are an increasingly recognized cause of CVI,²³ and iliac vein stenting is a safe and effective therapy.²⁴ The undersizing of iliac vein stents is a risk factor for stent migration, a rare, albeit potentially catastrophic, complication of iliac vein stenting.²⁵ Vein size is affected by hydration status,²² and vein size dictates the size of the stent. Thus, the primary aim of this study was to determine if outpatient venography performed later in the day is associated with smaller iliac vein stents placed to manage left-sided NIVLs. On adjusted analyses, patients who underwent iliac vein stenting later in the day, a surrogate for fasting duration, had 28% higher odds of having a smaller stent implanted than patients who underwent stenting earlier in the day.

Although the physiological impact of hydration on vein size is well-understood conceptually, limited research has quantified this impact in human subjects. Behzadi et al²² studied the effect of oral hydration on the volume and cross-sectional area of pelvic veins with cross-sectional imaging. After oral administration of 1 to 2 L of fluid, the common femoral vein cross-sectional area increased by 32% (59 mm² to 78 mm²; P < .01) for patients in the supine position.²² Similarly, the current study showed a trend toward a larger cross-sectional area of the iliac veins in the early cohort and, consequently, larger stents placed.

Iliac veins are elliptical, and their morphology is affected by intravascular volume.²⁶ This dynamic nature of the iliac veins can be more pronounced in the presence of NIVL than in fixed pathologies, such as post-thrombotic states.^16,27 The median stent diameter in our overall and early cohorts was 18 mm, equivalent to the diameter of the stents used in a study by Gagne et al²⁸ to treat nonthrombotic iliofemoral lesions. Iliac vein stents, particularly the Wallstent, are susceptible to underexpansion when used to treat NIVLs.²⁹ In both our study and that of Gagne et al, the majority of patients received a Wallstent. At our institution, it is standard practice to oversize iliac vein stents by approximately 2 mm relative to the mean reference diameter determined by IVUS. In contrast, the median stent diameter in our late cohort (ie, the more dehydrated patients) was smaller and similar to the diameter of the stents used in Gagne et al's study to treat post-thrombotic lesions.²⁸ These findings suggest that adequate hydration is critical when treating NIVLs to prevent the undersizing of stents, which might increase the risk of stent migration.^3,30

Our study examined the influence of age, patient positioning, and BMI on stent size and found that older age and prone positioning, but not BMI, were independent predictors of smaller venous stent size. Previous studies have not demonstrated an association between age and iliac vein stent size.²⁹ This may reflect case-specific intraoperative decision-making, because proceduralists may have exercised greater caution when selecting stent size in older patients, potentially owing to concerns around frailty, bleeding risk, or technical challenges related to decreased venous compliance.³¹ Given the retrospective nature of this study, it is challenging to determine the extent to which such clinical judgments could have influenced stent sizing beyond what was guided by IVUS measurements.

The inverse relationship between prone positioning and stent size emerged only in our sensitivity analysis, where iliac vein size was treated as a continuous variable. Nevertheless, this finding contrasts with a study of eight healthy subjects, in which prone positioning increased iliac vein size relative to the supine position.²² It is likely that popliteal access was obtained with the patient in the prone position when there are concerns about stenosis involving the common femoral vein or the distal external iliac vein, particularly near the inguinal ligament. In such cases, access to the common femoral vein can compromise visualization and treatment of the lesions. Even though this study excluded patients with thrombotic lesions, it is possible that some patients had silent thrombosis or a remote history of thrombosis that was not recognized. Thrombotic disease with recanalization and residual luminal narrowing, resulting in smaller vein size, might be a potential confounder that explains why those patients placed in the prone position had smaller stents.²⁸ However, post-thrombotic changes could restrict the compliance of the venous wall and possibly minimize the changes in diameter that are noted with change in volume status.

As for BMI, although the diameter of pelvic veins is larger in obese compared with nonobese individuals,³² our study, which included patients with BMIs ranging from 17.6 to 45.5 kg/m², did not find a relationship between obesity and iliac vein stent size. Most practices select a stent size of ≥2 mm greater than the diameter of the reference iliac vein on IVUS examination, aligning with expert recommendations.³³ Even with this practice in place, the dehydrated patients in our study were significantly more likely to have a smaller stent size implanted than their more hydrated counterparts. The 2023 American Society of Anesthesiologists Practice Guidelines for Preoperative Fasting recommend that patients fast for a minimum of 6 hours for solids and 2 hours for clear liquids before elective procedures to minimize the risk of pulmonary aspiration of gastric contents.³⁴ Yet, many hospitals, including this study's institution, have adopted an NPO after midnight policy for elective procedures to ensure efficiency and prevent delays and cancellations.

Based on our study's multiple linear regression, fasting for 9 additional hours predicted having one size smaller of a stent implanted (ie, 2 mm) for patients undergoing outpatient venography for NIVLs. The Society of Interventional Radiology Position Statement states that dehydration might result in overdiagnosis of NIVLs in cases where stenosis reflects a low intravascular volume status.³⁵ Next steps for our group are to evaluate preprocedural hydration protocols for patients with CVI to enhance the detection of pathologic iliac vein stenosis and prevent undersizing of iliac venous stents. This will be a prospective interventional study in which patients receive 1 L of normal saline in the preprocedure area. We will evaluate the size of iliac vein stents placed and compare them with historical controls.

Our study has several limitations. This was a retrospective cohort study with a small sample size, but it benefits from a relatively homogeneous cohort owing to its narrow inclusion criteria. This study was also limited by variability in whether and how proceduralists measured and reported iliac vein size. This resulted in missing data in the IVUS measurements, notably for the left external iliac vein and the diameter of the veins. Because there is no clearly defined algorithm at our institution for choosing stent diameter, there is also variability in how operators select the size of the venous stents. There were four types of stents used, and an interaction between stent type and diameter is possible. We restricted the number of covariates in our regression model because of small sample size, so we could not account for this interaction effect, which introduces the potential for residual confounding. A prospective study with a larger sample size and a standardized IVUS and stenting protocol is needed to better define the impact of hydration on iliac vein stent size.³⁶

Another limitation of this study is the inability to account for the timing of intraoperative fluids administered relative to the timing of stent placement. Because intravenous fluids might have been given before, during, or after stent sizing and placement, adjusting for the total procedural fluid volume could introduce misclassification bias. Cases with general anesthesia are more likely to involve intraoperative fluid administration than those with local anesthesia, and general anesthesia can cause vasodilatation. Still, the early and late cohorts were balanced in the types of anesthesia used. Nevertheless, we performed a subgroup analysis of only local anesthesia cases to address this limitation, and the timing of stenting was found to be an independent predictor of iliac vein stent size. An additional limitation is that, given its retrospective nature, our study could not account for respiratory variations and Valsalva maneuvers, which are known to affect vein size.³⁷ Also, we do not know precisely when patients started their preprocedure fast, and patients who had procedures later in the day might be more likely to begin fasting later. However, this potential would bias our results toward the null. Future studies could consider directly measuring hydration status and using it as the exposure variable, rather than relying on time of day as a proxy, as we did.

Conclusions

In patients treated for NIVLs, venous stent implantation performed later in the day was associated with a smaller stent size compared with procedures performed earlier in the day. This study has important implications, because a smaller stent size is a risk factor for stent migration. Future studies should investigate the role of a preoperative intravascular hydration protocol in determining appropriate stent sizing for patients with iliac vein lesions.

Author contributions

Conception and design: KS, PR, SA, CC

Analysis and interpretation: KS, PR, RA, EA, RG, CC

Data collection: PR, SA, RA, JL, SH, BT, EA, RG, CC

Writing the article: KS, CC

Critical revision of the article: KS, PR, SA, RA, JL, SH, BT, EA, RG, CC

Final approval of the article: KS, PR, SA, RA, JL, SH, BT, EA, RG, CC

Statistical analysis: KS, PR, CC

Obtained funding: RG, CC

Overall responsibility: KS, CC

Funding

None.

Disclosures

C.I.O.C. is a consultant for EnVVeno Medical, has IP of patent U.S.S.N. 10,524,89, and has received research support from Yale Department of Surgery, SVS, AVF, CT Innovation, VSGNE, NIH, Boston Scientific, Medtronic, EnVVeno Medical, and Inari Medical.

From the American Venous Forum