Perioperative arterial catheterization: A prospective evaluation of ultrasound, infection, and patient-focused outcomes

Abstract

Background:

There is little information regarding complications of arterial catheterization in modern clinical care. We aimed to determine the incidence of abnormal duplex vascular ultrasound and catheter related infections following perioperative arterial catheterization.

Methods:

Patients requiring arterial catheterization for elective surgery were included and insertion details collected prospectively. Duplex ultrasound evaluation was performed 24 h after catheter removal. Symptomatic patients were identified by self-reported questionnaire. On Day 7, patients answered questions by telephone, related to the insertion site, pain, and function. Results of catheter tip and blood culture analyses were sought. Univariate associations of patient and surgical characteristics with abnormal ultrasound were assessed with p < 0.05 considered significant.

Results:

Of 339 catheterizations, 105 (40%) had ultrasound evaluation. Catheters were indwelling for median (IQR, range) duration of 6.0 h (4.4–8.2, 1.8–28) with no catheter-related infections. There were 16 (15.2%, 95% CI 9.0%–23.6%) abnormal results, including 14 radial artery thromboses, one radial artery dissection, and one radial vein thrombosis. Those with abnormal ultrasound results were more likely to have had Arrow catheters inserted (68.8% vs 27%, p = 0.023) and more than one skin puncture (37.5% vs 26.8%, p = 0.031). Two of the 16 (12.5%) patients with abnormal ultrasound results reported new symptoms related to the hand compared with nine of the 88 (10.2%) with normal results (p = 0.1). No patients required urgent referral for management.

Conclusions:

Thrombosis was the most common abnormality and was usually asymptomatic. There were no infections, few post-operative symptoms, and minimal functional impairment following arterial catheterization.

Introduction

Invasive arterial blood pressure monitoring allows continuous hemodynamic monitoring and blood sampling. The current literature suggests that asymptomatic arterial thrombosis occurs commonly, ¹ with serious complications occurring in ⩽0.9%.^2,3 With increasingly co-morbid patients requiring surgery, the requirement for arterial catheterization is likely to rise. Current knowledge is limited, relating to outdated practice and restricted patient populations.

Prospective studies published between 1977 and 2005 reported radial artery occlusion rates between 12% and 30% following insertion of 18 and 20-Gauge (G) catheters.^4,5,6–8 Most of these studies included patients undergoing cardiac surgery with extracorporeal circulation and systemic heparinization.^4,7,8 Catheter-over-needle percutaneous insertion techniques were used^5–8 and one study included surgical arterial cut-down. ⁴ Heparinized saline^4–6,8 or papaverine ⁷ solutions were used during catheter maintenance. The outcomes of thrombosis or reduced flow were detected by the application of Allen’s test,^4,5,7 Doppler flowmeter,^4,5,7 pulse oximetry, ⁶ and thumb skin temperature. ⁴

There has since been considerable change in clinical practice. Real-time ultrasound is increasingly used for the first attempt at arterial line insertion, or to rescue a difficult or failed insertion. ⁹ Seldinger technique is used more commonly, ¹⁰ while heparinized saline solutions for flushing are used less commonly. ¹ Modern publications refer to procedures using 14 or 16G catheters, which are used in interventional procedures and are not relevant to the perioperative context.^11,12

There is no information to guide anesthetists and patients regarding complications rates following short-term upper limb arterial catheterization in the modern perioperative context. In this prospective service evaluation, we aimed to determine the incidence of abnormal duplex vascular ultrasound and catheter-related infection following upper limb arterial catheterization in adults undergoing elective surgery. We also assessed the post-operative patient experience.

Methods

The Human Research Ethics Committee of the Royal Brisbane and Women’s Hospital (RBWH) classified this project as a service evaluation, approving waiver of consent (Protocol version 6, approved 9/8/21, HREC/2021/QRBW/77338). SQUIRE Guidelines were followed. ¹³ The RBWH is a quaternary facility in Queensland, Australia, providing anesthesia services for adult surgical sub-specialties, excluding cardiac and transplant. Doctors performing arterial catheterization are specialist anesthetists or supervised anesthesia trainees. Table 1 shows the products available for arterial catheterization. Ultrasound guidance and device selection were determined by the anesthetic team. For monitoring, an Edwards LifeSciences TruWave™ transducer (Edward Lifesciences Corp, Irvine, California, USA) was connected to a pressurized bag of 0.9% normal saline. Arterial catheters were inserted pre-operatively and removed in the post-anesthesia care unit unless the patient was admitted to the intensive care unit (ICU), high dependency unit (HDU), or coronary care unit (CCU). In those admitted to ICU/HDU/CCU, catheter removal was at the discretion of the treating clinicians.

Table 1.

Product details of the available arterial catheterization equipment.

Product name	Technique	Needle gauge (G)	Catheter gauge and length (cm)	Catheter material	Wire diameter, length, tip
Arrow® Standard Seldinger Catheter a	Seldinger technique	20 G	20 G, 8 cm	Radio-opaque polyurethane	0.53 mm diameter, 35 cm long, straight soft tip on both ends
Arrow® Quickflash® a	Seldinger technique (with integral wire)	21 G	20 G, 3.8 cm	Radio-opaque polyurethane	Integral, 0.46 mm diameter, extends 1.6 cm beyond the needle tip
Vygon Leadercath Arterial® b	Seldinger technique	20 G,	20 G, 8 cm	Radio-opaque polytetrafluoroethylene (Teflon)	0.53 mm diameter, 29 cm long, soft tip one end
BD Insyte® c	Catheter-over-needle	20 G	4.8 cm	Vialon™ biomaterial (an elastomer of polyurethane)	Not applicable
BD Arterial Cannula with Flowswitch® c	Catheter-over-needle	20 G	20 G, 4.5 cm	Polytetrafluoroethylene	Not applicable
Arrow® Spring-Wire Guides a	Use Seldinger technique with a catheter-over-needle kit	Not applicable	Not applicable	Not applicable	0.46 diameter, 25 cm long, soft tip both ends

^aTeleflex®, Morrisville, North Carolina, United States. During data collection we did not differentiate between these and while Standard Seldinger is used much more commonly in our institution, the two kits apply different insertion techniques.

^bVygonUSA®, Lansdale, Pennsylvania, United States.

^cBecton Dickinson, Franklin Lakes, New Jersey, United States.

Elective surgery patients were included if arterial catheterization was indicated and the evaluation team were available in the following 24 h. Patient characteristics were collected, including age, sex, body mass index (BMI), and surgical specialty. Insertion data were obtained prospectively from the treating anesthetist, including site, catheter details; date and time of insertion; location when removed; post-operative ward destination; anatomical sites attempted; number of operators required; Seldinger technique use initially or as rescue; point-of-care ultrasound use initially or rescue; level of consciousness during insertion; difficulty of insertion (“difficult” defined as greater than one site or operator required or any deviation from the original plan); number of skin punctures; difficulties such as positional or unrecoverable loss of waveform.

Post-operative data included: date and time of removal; ultrasound evaluation (normal or abnormal); patient-reported symptoms; and recommended course of action. Qualified vascular ultrasonographers performed the ultrasound within 24 h of catheter removal. The ultrasound protocol was established according to local expert opinion and included examination of the radial artery; ulnar artery; brachial artery; identification of anatomical variations, vessel wall calcification or vessel wall thickening; arterial and venous flow measured by Doppler shift at multiple sites; presence of abnormalities (i.e. thrombus or dissection) and if present, their location and extent. The evaluation was reported as “normal” or “abnormal.” Ultrasounds were performed by two sonographers, using the same machine (Sonosite EDGE II System, model number L21822, Fujifilm, Bothell, Washington, United States) to limit inter-rater variability.

At the time of the ultrasound evaluation, patients were asked if their hand felt abnormal, looked abnormal, if they experienced numbness, tingling, pain, or weakness, responding “yes” or “no.” The authors developed these screening questions, based on the early signs of peripheral ischemia. The presence of one or more of these symptoms was considered “symptomatic.” There remains a lack of consensus regarding management following the identification of a symptomatic arterial thrombosis. ¹ A flowchart was developed by the authorship team for the purposes of this project, incorporating the expertise of vascular surgeon (MO) and vascular sonographers (TD). This allowed the systematic and individual evaluation of patients with symptomatic thrombosis. The available courses of action in the flowchart were: No action; referral to general practitioner; non-urgent referral to vascular surgeon; urgent referral to vascular surgeon (Supplemental Figure 1).

Day 7 post-discharge data were obtained by telephone survey (Supplemental Figure 2). The questions were developed by the authors to detect symptoms of ischemia and infection. Questions related to the side of catheter insertion only and identified: if the insertion site was on the dominant hand; presence of lump or bruise over the insertion site; color of the insertion site (bruise, redness, discharge); tingling in hand or fingers; pain at the site on a scale of 0–10; color of the hand and fingers (normal, increased, pale); pain related to the hand or fingers (on a scale of 0–10). Functional assessment used questions adapted from the QuickDash validated questionnaire. ¹⁴ Patients responded on a 5-point scale to questions concerning their ability to perform daily activities both before and after the procedure (to open a jar or carry a bag, Supplemental Figure 2).

Microbiology results were searched for seven post-operative days to identify catheter tip and associated blood culture results within the same 24 h period. ¹⁵

Statistical analysis plan

The rate of partial or complete arterial thrombosis estimated from older studies and detected by Doppler only is approximately 20%.^4–6,8 We intended to evaluate the arteries and experiences of 500 patients over a 12–24 month period, to explore factors associated with abnormal ultrasound applying multivariable logistic regression. Staff shortages limited the availability of vascular sonographers to continue the evaluations. We continued to collect patient experience information if ultrasound evaluation was unavailable. The service evaluation was ceased after data on 339 patients was obtained, including 105 with ultrasound evaluations.

Study data were collected and managed using REDCap (Research Electronic Data Capture). ¹⁶ Patient characteristics were summarized using descriptive statistics. The proportion of patients with an abnormal ultrasound was calculated with 95% confidence intervals (CI). Univariate associations of patient and surgical characteristics with abnormal ultrasound were assessed using Fisher’s exact tests for categorical variables and Mann-Whitney U tests for continuous variables. Statistical significance was set at a p-value <0.05 (two-sided). Stata version 15 (StataCorp, College Station, TX, U.S.A.) was used for analyses.

Results

Data were collected for 339 patients between October 2021 and May 2023 and 282 (83%) provided Day 7 patient experience information. The entire cohort had a median (IQR) age of 64 (54–73) years, 206 (61%) were male and they had a median (IQR) body mass index (BMI) of 28 (24–32) kg m⁻². The most common surgical specialties providing care were neurosurgery (107, 32%), vascular surgery (60, 18%), and urology (46, 14%). In 336 patients, arterial catheterization was attempted at the radial artery and in one the ulnar artery was attempted. In 336 (99.1%) the radial artery was the ultimate site used, with the brachial artery used in 2 (0.6%) and ulnar artery in 1 (0.3%). The catheters inserted were Arrow (107, 32%), Vygon (100, 30%), Insyte 77 (23%), and Flow-switch 55 (16%). Seldinger technique was applied either as the initial attempt and/or as rescue in 223 (66%). Ultrasound was used as the initial technique in 230 (68%). The median (IQR, range) catheter dwell-time was 6.5 (4.5–11.3, 1.0–170.7) hours. Intravenous heparin was administered intraoperatively in 26 (8%). No catheters were sent for microbiological analysis. The full details pertaining to the entire cohort are shown in Supplemental Tables 1 to 3.

Ultrasound evaluations were performed in 105 (31%) of the total cohort and we obtained Day 7 patient experience information from 93 (89%). These 105 patients had a median (IQR) age of 64 (53–73), 63 (60%) were male and they had a median (IQR) BMI of 28.2 kg m⁻² (24–32). The catheters were indwelling for a median (IQR, range) duration of 6.0 h (4.4–8.2, 1.8–28). There were 16 (15.2%, 95% CI 9.0%–23.6%) abnormal results which are described in Figure 1. All abnormalities were identified on the side of catheter insertion. In 15 patients the catheter was inserted in the radial artery and one was inserted in the brachial artery. The abnormalities included radial artery thrombosis in 14, radial artery dissection and calcification in one and radial vein thrombosis at the puncture site in one patient. The patient with radial vein thrombosis had a history of deep vein thrombosis. In 13 patients with measurements, the mean (SD) length of the thrombus was 9.6 (3.0) cm, ranging from 5 cm to 14 cm.

Figure 1.

Vascular ultrasound evaluations of 16 patients with abnormal scans Day 1 following removal of arterial catheter: (a) symptomatic—hand feels abnormal, pain in hand or fingers, weakness; this patient was lost to follow-up and (b) asymptomatic at time of ultrasound, developed symptoms in first week, weakness, tingling, pain. Non-urgent review by vascular team by which time the clinical examination was normal and symptoms had resolved.

Table 2 shows the details of arterial catheter insertion for the 105 patients with ultrasound evaluation, comparing those with normal and abnormal ultrasound results. There was no difference between those with normal and abnormal ultrasound results, in terms of age, sex, or BMI. In one patient with radial artery thrombosis, there was an unrecoverable loss of waveform intraoperatively suggesting early arterial occlusion soon after catheterization. This patient had non-invasive blood pressure measurement for the remainder of the case. Another case of radial artery thrombosis was identified in a patient where the arterial waveform was damped in the post-anesthesia care unit and non-invasive measurements were commenced. Those with abnormal ultrasound results were significantly more likely to have had an Arrow catheter inserted (Standard or Quickflash) and significantly more likely to have two or more skin punctures (Table 2). In Table 3, information regarding the catheter brand (Arrow or Other) and use of real-time ultrasound for insertion, is summarized according to number of skin punctures and ultrasound evaluation result. Five of the 11 patients with abnormal ultrasound results as well as an Arrow catheter inserted, had two or more skin punctures.

Table 2.

Characteristics of 105 patients who had a post-operative Day 1 vascular ultrasound (US) evaluation following removal of arterial catheter, by US result. All had attempted radial artery catheterization.

Insertion detail	Normal US	Abnormal US	p-Value
	n (%)	n (%)
	N = 89	N = 16
Location of catheter—side of body			0.84
Left	58 (65.2)	10 (62.5)
Right	31 (34.8)	6 (37.5)
Size of catheter—gauge			0.67
20G	88 (98.9)	16 (100.0)
22G	1 (1.1)	0 (0.0)
Brand of catheter			0.023
Arrow	24 (27.0)	11 (68.8)
Vygon	28 (31.5)	2 (12.5)
Insyte	23 (25.8)	2 (12.5)
Flow-switch	14 (15.7)	1 (6.3)
Number of anatomical sites attempted			0.49
1	85 (96.6)	15 (93.8)
2	3 (3.4)	1 (6.3)
Site of catheter placement			0.15
Radial artery	89 (100)	15 (93.8)
Brachial artery	0 (0)	1 (6.3)
Seldinger technique (wire) used a			0.17
Seldinger used initial attempt	55 (61.8)	12 (75.0)
Seldinger not used	33(37.1)	3 (18.8)
Seldinger Rescue	1 (1.1)	1 (6.3)
Ultrasound used			0.23
Ultrasound as initial technique	56 (62.9)	10 (62.5)
Palpation/landmark insertion only	27 (30.3)	3 (18.8)
Ultrasound as rescue after palpation attempt	6 (6.7)	3 (18.8)
Depth of anesthesia during insertion b			0.54
Awake	68 (76.4)	14 (87.5)
Sedated	6 (6.7)	1 (6.3)
Anesthetized	15 (16.9)	1 (6.3)
Difficult insertion c	8 (9.0)	2 (12.5)	0.66
Number of skin punctures			0.031
1	74 (83.1)	10 (62.5)
2	14 (15.7)	4 (25)
3	1 (1.1)	2 (12.5)
Intravenous heparin administered intraoperatively	14 (15.7)	0 (0.0)	0.088
Catheter removed in PACU	72 (80.9)	11 (68.8)	0.27
Catheter dwell-time (hours) median (IQR) [range]	6.0 (4.4–8.0) [1.8–28.0]	6.6 (4.7–8.5) [3.0–24.3]	0.54

PACU: post-anesthesia care unit.

^aMay have been as part of the Arrow or Vygon kit or as an individually packaged wire via an Insyte cannula.

^bThis refers to initial attempts.

^c“Difficult” defined as greater than one site or operator required or any deviation from the original plan.

Table 3.

Catheter brand “Arrow” (Arrow® Standard Seldinger Catheter or Arrow® Quickflash®, Teleflex®, Morrisville, North Carolina, United States) or “Other” and use of real-time ultrasound (US) for insertion, according to number of punctures and US evaluation result, 105 patients with duplex ultrasound evaluation following radial arterial attempt/catheterization.

Number of punctures	Normal US evaluation			Abnormal US evaluation
	Other	Arrow	Total	Other	Arrow	Total
1	54	20	74	4	6	10
2	11	3	14	1	3	4
3	0	1	1	0	2	2
Total number of patients	65	24	89	5	11	16
Number of punctures	Normal US evaluation			Abnormal US evaluation
	No initial US	Initial US-guided	Total	No initial US	Initial US-guided	Total
1	25	49	74	2	8	10
2	8	6	14	3	1	4
3	0	1	1	1	1	2
Total number of patients	33	56	89	6	10	16
Difficulty a	Normal US evaluation			Abnormal US evaluation
	No initial US	Initial US-guided	Total	No initial US	Initial US-guided	Total
Easy	30	51	81	5	9	14
Difficult	3	5	8	1	1	2
Total number of patients	33	56	89	6	10	16

^a“Difficult” defined as greater than one site or operator required or any deviation from the initial plan.

One or more symptoms were reported by three of the 16 patients with abnormal ultrasound results. In one patient, the symptoms were pre-existing and unchanged, leaving 2 of the 16 (12.5%) symptomatic compared with nine (10.2%) of the 89 with normal results (p = 0.67). One patient with partial occlusion of the radial artery reported pain, weakness, and the hand feeling abnormal. This patient had a terminal diagnosis and was lost to follow-up. One patient reported tingling, pain, and weakness in the hand or fingers that developed in the first post-operative week. Non-urgent review was undertaken by the vascular team by which time the clinical examination was normal and symptoms had resolved (Figure 1). In the presence of a normal ultrasound result, the most common symptoms reported on Day 1 post-operatively were: the hand feeling abnormal (5, 4.8%), the hand looking abnormal (3, 3.4%), and pain in the hand or fingers (3, 3.4%). Based on the combination of abnormal ultrasound results and symptoms, no patients required urgent referral to the vascular surgery team and no patients required pharmacological or surgical intervention.

On Day 7 post-operatively, 13 of the 16 patients with abnormal ultrasound results and 80 of the 89 patients with normal ultrasound results completed the post-operative survey. A lump or bruise was reported at the site in 5 (38%) of those with abnormal ultrasound results and 36 (45%) of those with normal ultrasound results (p = 0.77). Tingling in the hand or fingers was reported in 3 (23%) of those with an abnormal result and 5 (6%) of those with a normal ultrasound result (p = 0.08). In the 13 patients with abnormal ultrasound results, the highest pain score at the insertion site was 5 (1, 8%), while in the 88 patients with normal results, the highest pain score was 7 (1, 1%). The median (IQR) pain score was 0 (0–0) in those with normal and abnormal ultrasound results (p = 0.17). Increased difficulty opening a new or tight jar was reported in 1 (8%) of those from the abnormal ultrasound group and 2 (3%) from the normal ultrasound group (p = 0.37). Increased difficulty carrying a shopping bag or briefcase was reported in none of those from the abnormal group and 2 (3%) from the normal group (p = 1.00). Of the total cohort of 339, 282 (83.2%) completed the Day 7 survey, with some missing responses. Complete results of the total cohort are provided in Supplemental Table 3.

Discussion

This study investigated complications following radial artery catheterization in patients undergoing elective surgery, reporting radial artery thrombosis rate of 14%, at the lower end of rates reported between 1980 ⁴ and 2005. ⁶ Seldinger technique and ultrasound guidance were used in over a half of procedures. Detailed vascular ultrasonography permitted quantification of the arterial thromboses identified and one venous thrombosis. Abnormal ultrasound results were associated with the Arrow catheter brand (applying two different insertion methodologies) and more skin punctures. Despite abnormal findings in 15% of ultrasound evaluations, symptoms occurred in two patients, with no surgical or medical interventions required. The post-operative patient experience was largely positive, with pain and dysfunction infrequently reported.

Previous studies have confirmed the relatively common occurrence of asymptomatic arterial thrombosis following radial arterial catheterization.^4,6 Quantifying the magnitude and incidence of these of complications is difficult due to the rarity of severe complications and the required resources. We have provided outcomes specific to the modern, non-cardiac surgery context, in which arterial catheters are inserted using different techniques and for short dwell times.

Previous studies have evaluated landmark/palpation-based insertion using catheter over needle techniques.^4–8 Seldinger techniques were less common prior to the 2000s, with over needle products (Medicut,^4,5 Angiocath, ⁵ Longdwell, ⁴ and Abbocath). ⁷ Our cohort is the first to report radial artery thrombosis rates in a general perioperative setting in which the Seldinger technique was commonly used. This is important given the wire is threaded some distance into the artery, risking endothelial damage. In our cohort, the Arrow brand (but not the Seldinger technique) was over-represented in cases with abnormal ultrasounds. Thrombosis rates may be influenced by catheter materials,^1,17 length, ¹ insertion method, ¹ and the size of the artery. ¹ The 8 cm Arrow Standard Seldinger catheter is made of polyurethane and is longer than other products, features previously associated with lower thrombosis rates.^18,19 The included wire is longer than that in the Vygon kit, a factor which has not been previously evaluated. Our sample is too small to draw conclusions about causation. Arrow brand catheters were also used when multiple skin punctures occurred. Clinicians may choose an Arrow product when predicting technical difficulty.

Ultrasound guidance was used in over two-thirds of patients. Ultrasound use to guide arterial catheterization in elective surgery is not yet a recognized standard of care, ⁹ despite a 2016 meta-analysis finding increased success rates with use.^20,21 Ultrasound-guidance may reduce the number of attempts and trauma to the artery,^21,22 however we identified no difference between the groups with and without an abnormal ultrasound result, according to use of point-of-care ultrasound.

In the presence of a normal ultrasound, some patients reported new symptoms related to the hand or fingers on the affected side, such as bruising of the site and tingling of the hand and fingers. These symptoms could be related to other perioperative interventions (e.g. intravenous cannulation or patient positioning). Ideally patients should be advised that thrombosis may occur in 15%–20% of cases but this is likely to be asymptomatic, even if the thrombosis is occlusive. Catheter-related infections are an important source of morbidity and mortality ²³ but no patients had evidence of local or systemic infection. This may reflect the short catheter dwell-time and short follow-up period.

Previously, complications of arterial catheterization utilized clinical assessments and Doppler flowmeters.^4–8 Our more sensitive evaluation allowed measurement of thromboses, the most extensive being 14 cm. Despite our sensitive outcome evaluation, the incidence was at the lower end of the range quoted in the literature. Modern techniques (point-of-care ultrasound and use of the Seldinger technique) may have reduced endothelial damage, with fewer reported thromboses despite the sensitive detection method.

Strengths and limitations

The strengths of our study are its prospective design and detailed, objective assessment of outcomes. We are the first to evaluate patient reported outcomes following arterial line catheterization, with the procedure well-tolerated. Our flow-chart to guide management requires further validation. There are also limitations to our project. The small sample size was inadequate for estimation of rare, serious complications and precluded the use of multivariable logistic regression to identify risk factors for thrombosis. The insertion technique was not standardized and our method of data collection did not discriminate between the two Arrow kits. This is important, given that the two kits require different insertion methods, have catheters of a different length and needles of a different gauge. ²⁴ We did not report the years of training completed by trainee anesthetists and their level of experience may have influenced the outcomes. Reproducing these results in greater numbers, applying current best practice ²⁰ and a standardized insertion technique, would contribute to our understanding of the pathogenesis and long-term consequences of catheter-related thrombosis. Further evidence is required to determine if the Seldinger technique and point-of-care ultrasound influence thrombosis rate and extent in this context. Future work should focus on catheter materials, wire length, and measures to increase first-pass insertion success.

Using a detailed and objective assessment method, we observed abnormal ultrasounds in 15% of patients following short-term elective radial arterial catheterization for perioperative use and no infectious complications. Despite significant thromboses, a small fraction reported symptoms and none required medical or surgical intervention. Overall patients reported few symptoms related to the hand and fingers and minimal functional impairment.