Prospective, Randomized Trial of Doppler-Assisted Subclavian Vein Catheterization FREE

PERCUTANEOUS subclavian vein catheterization is one of the most common invasive procedures performed in the United States. This is a "blind" procedure and requires localization of a deep central vein using only superficial anatomical landmarks.¹ Because of the blind nature of the procedure, it can be extremely difficult in some patients to localize the vein; thus, the procedure frequently requires multiple needle passes or attempts. In reported series, the complication rates range from 3% to 12%, depending on the experience of the persons performing the procedure and the definition of a complication.^{2 - 4} We have previously shown that additional factors affecting the rate of complications include an abnormal weight-height ratio and prior surgery or radiotherapy in the region of the subclavian vein.² The potential problems and complications associated with attempted subclavian vein catheterization include failure to locate and successfully cannulate the vein, catheter malposition, pneumothorax, mediastinal hematoma, subclavian artery puncture, hemothorax, and brachial plexus or other nerve injury.² These complications may increase patient morbidity, prolong hospitalization, or delay therapy.

The most common complication is failure to cannulate the subclavian vein. A variety of positional techniques are available that may increase the rate of successful cannulation; however, a method of subclavian vein cannulation that guarantees success has yet to be identified.^{5 - 6} Surface and Doppler ultrasonography have been recommended to determine the exact location of the subclavian vein and artery and guide the physician in placement of a central venous catheter. Some investigators have found this to be of significant benefit, while others have found no effect on the rate of successful subclavian vein cannulation.^{7 - 11} These studies, however, were limited by the lack of real-time Doppler information to locate vascular structures. Recent advances in ultrasound microtechnology have allowed the development of a Doppler sensor that fits into the lumen of a standard 18-gauge needle (Smart Needle, Peripheral Systems Group, Mountain View, Calif). This device provides real-time information about the proximity of the needle tip to adjacent vascular structures and may permit the differentiation of venous and arterial signals.¹² Reports of Smart Needle use have been primarily anecdotal and based on the use of the technique in specialized clinical settings such as the intensive care unit or in pediatric patients.^{13 - 15} Doppler guidance for the routine, elective insertion of a subclavian venous catheter has not been subjected to a prospective, randomized study. A meta-analysis of preliminary studies has suggested that the use of real-time ultrasound or Doppler ultrasound may increase the success rate of central venous catheter placement as well as decrease complications.¹⁶

Patient characteristics may predict the rate of success for subclavian vein cannulation. It has previously been shown that prior surgery adjacent to the site of catheterization, an abnormal body mass index, and prior catheter insertion may place patients at a higher risk of failure.² In addition, patients who have had prior catheter insertion may have occult subclavian vein thrombosis, which prevents the successful cannulation of the vein.¹⁷ Real-time Doppler guidance could allow for the immediate identification of this difficulty through the observation of an absent venous flow signal. For these reasons, we conducted a prospective, randomized study to determine whether the use of the Smart Needle can decrease the failure or complication rate during routine, elective placement of a subclavian catheter.

A prospective, randomized trial was performed at the University of Texas M. D. Anderson Cancer Center, Houston, to compare Doppler-guided percutaneous subclavian vein catheterization using the Smart Needle with the standard Seldinger catheter insertion technique. This study was reviewed and approved by the institutional review board. All patients participating in the study provided written informed consent and all data were collected by a single research nurse. Smart Needles were provided by the manufacturer and patients were not charged for the use of the Smart Needle or for any services associated with placement of the subclavian catheter other than the general charge for central venous catheter insertion using the standard insertion technique. At our institution, 8 to 20 patients per day undergo subclavian vein catheter insertion in a dedicated procedure suite. The primary indication for the placement of long-term venous access catheters in our patients is for the administration of chemotherapy. All catheters were inserted under sterile, nonemergency, controlled conditions.

Two hundred forty patients were enrolled in the study. All patients had at least 1 factor that may be associated with an increased risk for failure or complication (body mass index [determined by the weight in kilograms divided by the square of the height in meters] of <20 kg/m² or >30 kg/m², prior surgery in the region adjacent to the subclavian vein, or prior radiotherapy at the attempted catheterization site). Patients were stratified with regard to prior surgery within the region of subclavian vein catheterization (eg, mastectomy, axillary node dissection, or thoracotomy) and body mass index to minimize any effect these confounding variables might have on the success rate. All procedures were performed in the outpatient infusion therapy clinic by highly experienced fellows from the Department of Surgical Oncology (postgraduate year 6-10). Eighteen physicians participated in the study, inserting from 1 to 27 catheters each (median, 13 catheters). All physicians underwent rigorous instruction in the use of the Smart Needle and had demonstrated competence in the use of the Doppler probe and appropriate recognition of vascular flow signal characteristics prior to initiation of the study.

The standard technique for insertion of a subclavian vein catheter has previously been described.² The Smart Needle is a 14-MHz continuous-wave Doppler instrument that operates in conjunction with the Smart Needle vascular access device. The continuous-wave Doppler instrument is contained within the lumen of an 18-gauge needle, which allows real-time transmission and reception of a sonographic signal to localize the subclavian vein. This technique allows the physician to identify a vascular structure, though some limitations exist to differentiate between arterial and venous flow signals. The Smart Needle is advanced with frequent reassessment of the signal quality and intensity until the subclavian vessel is punctured, at which time the Doppler probe is withdrawn from the lumen of the 18-gauge needle to allow catheterization of the subclavian vein by the standard Seldinger technique. All patients had a chest radiograph after the catheter was placed to evaluate for complications (pneumothorax, hemothorax, mediastinal hematoma, and misplacement of the catheter). Procedures were monitored for arterial punctures, evidence of nerve injury, and accurate recording of the number of passes.

The following factors were evaluated: age, sex, height, weight, body mass index, the number of needle passes (defined as separate skin punctures) attempted, the side on which the catheter insertion was attempted, prior catheterization or attempted catheterization on the same side, and prior surgery or radiotherapy in the same region. Routine evaluation prior to subclavian vein catheterization included a chest radiograph, platelet count, prothrombin time, and partial thomboplastin time. Abnormalities in the platelet count or coagulation times were normalized by the administration of blood products prior to the catheterization attempt. If catheterization of the subclavian vein was unsuccessful after 2 attempts, the patient was crossed over to the other catheterization technique. After 4 unsuccessful attempts (2 attempts using each technique), a chest radiograph was obtained to evaluate for complications and the patient was declared a failure of subclavian vein cannulation.

Randomization was performed using a computer-generated block randomization process. One hundred twenty patients were entered into each arm of the trial to permit detection of a decrease in patients without a catheter after 2 passes from 28.7% (based on the results of a prior trial of subclavian vein catheterization in a similar group of patients² ) to 10%, with a significance of .01 and a power of 80%. Futhermore, to examine the utility of Doppler guidance as a salvage technique, 25.7% of patients randomized to receive standard placement technique would require more than 2 passes with an overall 48% failure rate and a decrease of this, compared with historical controls, to less than 10% would be detected with a significance of less than .05 and a power of 80%.

The statistical analysis included logistic regression and likelihood-ratio χ² tests.¹⁸ For each variable, differences between the 2 groups were initially evaluated with a univariate logistic regression model that contained a single independent variable. Significance was determined according to the P value of the likelihood-ratio test. A forward stepwise method of selection was used to construct models that identified the most important risk factors and to predict the probability of a succesful procedure. The likelihood-ratio χ² test (performed with contingency tables) was used to determine the factors that affected success. This test is equivalent to the likelihood test for the significance of the coefficient for the variable in a univariate logistic regression model. A P value less than .05 (2 tailed) was considered to indicate statistical significance.

Of the 240 patients enrolled in the study, 119 were randomly assigned to undergo initial subclavian vein catheterization using the Smart Needle, while the remaining 121 were assigned to the standard technique (control group). There were no unevaluable patients. The patient characteristics were similar between the 2 groups (Table 1). One hundred eighty-one patients (75.4%) had subclavian catheters successfully inserted using the initial method of insertion. The initial success rate (1 or 2 attempts) was 81.0% for the standard technique and 69.7% for the Smart Needle technique (P=.04; Table 2). Two patients in the Smart Needle group suffered a complication (1 hematoma, 1 catheter malposition) for a complication rate of 1.7%; 1 patient in the standard technique group suffered a complication (hemothorax) for a complication rate of 0.8% (Table 2).

Attempt at subclavian vein catheterization using the initial technique failed in 59 patients. The failure rate was independent of body mass index, prior irradiation, prior surgery, prior catheterization, or prior attempt at catheterization. Fifty-five of these patients (34 patients in the Smart Needle group and 21 in the standard technique group) underwent crossover to the other technique; 4 patients refused further attempts at subclavian vein catheterization. Smart Needle subclavian catheter insertion was successful in 18 of 21 patients crossed over to this technique after the standard technique failed, for a secondary success rate of 86%. This was slightly higher than but not statistically different from the 79% success rate for the standard technique (27 of 34 patients underwent successful catheterization by standard technique after failing the Smart Needle method; P =.56 vs Smart Needle technique, Table 2). No clinical parameter was a significant predictor of failure on the initial attempt or secondary attempt (data not shown).

The Smart Needle is a novel technique of subclavian vein catheterization and, as with all novel techniques, there may be a learning curve associated with the development of proficiency. The physicians involved in the present study were required to have demonstrated proficiency with the Smart Needle on 10 to 15 patients prior to entering the study. In the first 120 patients enrolled in the study, the success rate for the Smart Needle technique was significantly lower than the success rate for the standard technique (Table 3). However, the difference between the success rates for the 2 techniques in the second half of the study was not statistically significant, suggesting a much longer learning curve. Furthermore, there was a small subgroup of physicians who had a significantly lower success rate throughout the study. These "poor performers" (5 of 18 physicians) had a mean overall success rate of only 64.5%, while the other 13 physicians had a mean overall success rate of 80.5% (P=.007). The performance of these 5 physicians was examined, and their low overall success rate was found to be due to a poor success rate using the Smart Needle (success rate for Smart Needle, 52.4%; success rate for standard technique, 79.4%; P =.01). The remaining 13 physicians exhibited no difference in the success rate using the Smart Needle compared with the standard technique (success rate for Smart Needle, 80.2%; success rate for standard technique, 81.6%; P=.70).

The overall success rate for placement of a new subclavian catheter was not influenced by a previous ipsilateral catheter (success rate, 75.0% compared with 75.9% without previous catheter) or by a failed previous attempt (success rate, 81.2% compared with 75.8% without previous attempt). Furthermore, use of the Smart Needle did not lead to a higher success rate in those patients in whom previous subclavian vein cannulation on the ipsilateral side had failed; the success rates were 83.3% for the Smart Needle and 78.6% for the standard technique (P=.73).

Finally, because the Smart Needle did not increase the initial or secondary success rate of subclavian vein catheterization, there was no associated cost benefit. In fact, use of the Smart Needle was associated with longer times for catheter insertion (data not shown), and the needle itself is significantly more costly than the needle used for standard percutaneous cannulation ($40-$70 for the Smart Needle and $3-$5 for the standard needle). Complications may also add to the medical and financial cost to the patient. In the present study, the rate of complications was extremely low and not significantly different between the 2 groups. Therefore, detailed cost analysis was not calculated as the Smart Needle could not be demonstrated to significantly improve success of catheterization or decrease the rate of complications.

We found no benefit associated with use of the Smart Needle to achieve Doppler guidance for the insertion of subclavian vein catheters. Furthermore, the Smart Needle was not significantly beneficial as a salvage procedure after a failed attempt using the standard technique based on surface anatomical landmarks. However, the salvage rate using the Smart Needle was 85.7%, the highest success rate of any group. Our results demonstrate, however, that the development of technical proficiency with the Smart Needle technique requires a learning curve that extends beyond the rigorous training period and that not all physicians are able to demonstrate technical aptitude.

Previous central venous catheters, previous attempts at an ipsilateral subclavian catheter, abnormal body mass index, and surgery or radiotherapy in the region adjacent to the site of cannulation may decrease the likelihood of successful catheter insertion.² In our study, no clinical parameter was identified that was associated with a higher rate of success using the Smart Needle. The overall success rate of approximately 94% (226 of 240 patients had catheters successfully inserted at the initial setting) in the patients included in the present study is fairly high, given the clinical factors that may have predisposed to failure. This may reflect the expertise of the physicians involved in subclavian vein catheterization and be one possible reason why the Smart Needle failed to demonstrate a significant benefit.

Various reports have been published on the use of either surface or Doppler ultrasound guidance for catheterization of the central veins.^{7 - 13} Doppler guidance is a well-established technique for the identification and localization of vascular structures, especially arteries. The addition of color flow imaging may allow for clear differentiation of arterial and venous signals; however, at present, this type of duplex probe is not available as an intraneedle device. Identification of arterial structures may be accomplished by Doppler ultrasound techniques, though identification of venous structures may be limited given variables such as the angle of the Doppler signal to the vessel, the flow of blood in the vessel, and other anatomical limitations. Although we did not observe a statistically significant benefit in the present study, there was a slight trend toward this result (86% success rate for the Smart Needle vs 79% success rate for the standard technique as salvage methods of catheterization). In a larger study, crossover to the Smart Needle may have demonstrated a statistically significant advantage over the standard technique and this might be considered a potential selective use of the device.

Some reports have documented a higher rate of successful cannulation with Doppler or ultrasound guidance, though only in the hands of less-experienced physicians or under technically complicated circumstances.^{7 ,15 ,19} Branger et al¹⁵ found Doppler guidance particularly useful in the salvage situation. Two prospective studies of central venous catheterization performed in the setting of the intensive care unit under an urgent clinical scenario demonstrated a trend toward a higher success rate using either ultrasound guidance for the subclavian vein⁷ or Doppler guidance for the internal jugular vein.¹⁹ Furthermore, a recent meta-analysis of both surface and Doppler ultrasound guidance suggests that this technology improves success rates and decreases complications, though the majority of these studies involved jugular vein catheter placement.¹⁶

Advances in microtechnology have introduced a variety of novel techniques into medicine, and it is hoped that these will translate into better patient care. The usefulness of these techniques must be examined carefully before widespread adoption based on anecdotal evidence occurs. The present study represents the largest prospective, randomized, crossover study of Doppler guidance in vein catheterization. This technique was associated with a much longer learning curve than anticipated, did not decrease complications, and therefore could not be cost-effective. Furthermore, the individual technical abilities of each physician may further influence success and complication rates. For these reasons, the Smart Needle cannot be recommended for routine use in subclavian vein cannulation, although it may have a role in the hands of some physicians or in the setting of an emergency central venous catheterization.

This study was sponsored, in part, by a grant from Peripheral Systems Group, Mountain View, Calif.

Reprints: Paul F. Mansfield, MD, Department of Surgical Oncology, Box 106, University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd, Houston, TX 77030.