Avoiding Common Technical Errors in Subclavian Central Venous Catheter Placement

Michael J Kilbourne, MD, Grant V Bochicchio, MD, MPH, FACS, Thomas Scalea, MD, FACS, Yan Xiao, PhD
Proficiency in placing infraclavicular subclavian venous catheters can be achieved through practice and repetition. But few data specifically document insertion technical errors, which mentors could teach novice operators to avoid. STUDY DESIGN: Surgical, medical, and anesthesia textbooks and procedural handbooks were reviewed. Subclavian catheter placement technical errors described were identified and consolidated. Video captures from 86 consecutive patients receiving subclavian central venous catheterizations at an urban trauma center were evaluated. In each video segment, the number of attempts at insertion, the number of failures at insertion, and the technical error observed during failed attempts were recorded and tabulated. RESULTS: Of the 86 subclavian line placements attempted, 77 were successful (89.5%), with a total of 357 subclavian venipuncture attempts and 279 failures (78% attempt failure rate). There was a mean of 3.2 failed attempts per line (left side, 2.1 attempts; right side, 5.5 attempts). Junior residents (PGY 1 to 2) had more failures per line than senior residents (PGY 3 to 5): 4.1 versus 3.6. The most common technical errors observed were improper site for needle insertion relative to the clavicle; insertion of the needle through the clavicular periosteum; too shallow of a trajectory for the needle; improper or inadequate anatomic landmark identification; aiming the needle too cephalad; and inadvertent movement of the needle out of the vein before or during wire placement. CONCLUSIONS: In subclavian central venous access attempts, there are six common technical errors. Mentors can improve novice operators proficiency by teaching them to avoid these errors. (J Am Coll Surg 2009;208:104109. 2008 by the American College of Surgeons)
BACKGROUND:

About 5 million central venous catheters (CVCs) are placed annually by physicians in the United States.1 The infraclavicular subclavian vein is the most frequently used access site.2 Current evidence supports the fact that subclavian vein catheter infection rates are lower (4 per 1,000 catheter-days) than rates for both internal jugular (8.6 per 1,000 catheter-days) and femoral catheters (15.3 per 1,000 catheter-days).3 The subclavian vein is more accessible to the operator in trauma patients with cervical collars than the internal jugular. In addition, the subclavian catheter can be placed without disrupting airway management during the initial stage of resuscitation.4 The femoral vein can be cannulated without disrupting airway management, but, as stated earlier, there
Disclosure Information: Nothing to disclose. Received June 11, 2008; Revised September 3, 2008; Accepted September 5, 2008. From the Departments of Surgery, Walter Reed Army Medical Center, Washington, DC (Kilbourne), and R Adams Cowley Shock Trauma Center, Baltimore, MD (Bochicchio, Scalea, Xiao). Correspondence address: Michael J Kilbourne, R Adams Cowley Shock Trauma Center, 22 South Greene St, T1R60, Baltimore, MD 21201.

is a higher rate of femoral catheter infection. So, a femoral catheter frequently needs be changed out to a subclavian or internal jugular catheter once the patient arrives at the ICU, subjecting the patient to two separate line procedures. The failure rate for subclavian vein catheterization using the landmark method and ultrasonographic assistance in first- and second-year residents (relative novices) was studied by Gualtieri and colleagues in 1995.5 The failure rates (failed venipunctures to total attempts) were 82% and 65% for the landmark group and ultrasound-assisted group, respectively. These numbers are remarkably high. Atul Gawande6 encapsulated most residents lack of technical understanding about CVC placement in his book, Complications: A Surgeons Notes on an Imperfect Science. He wrote, I still have no idea what I did differently that day . . . for days and days, you make out only the fragments of what to do. And then one day youve got the thing whole . . . and you cannot say precisely how. Gawande was not alone in his observation. Most physicians experience the same frustrations during their introduction to CVC placement.

2008 by the American College of Surgeons Published by Elsevier Inc.

104

ISSN 1072-7515/09/$36.00 doi:10.1016/j.jamcollsurg.2008.09.025

Vol. 208, No. 1, January 2009

Kilbourne et al

Subclavian Central Venous Catheter Placement

105

Table 1. Denition of Technical Errors in Subclavian Line Placement

Technical error Denition

Improper or inadequate anatomic landmark identification Improper insertion position relative to the clavicle Insertion of the needle through the clavicular periosteum Taking too shallow of a trajectory of the needle Aiming the needle too cephalad Failure to keep the needle in place for wire passage

Failure to palpate two bony landmarks, the sternal notch and the middle to medial third of the clavicle, before and during each attempt.9 Failure to insert the needle at a recommended distance of about 1 cm inferior and lateral to the middle or medial third of the clavicle.10 In an effort to walk the clavicle down to locate the vein posteriorly, using significant force or aggressively pushing the needle can drive it through, instead of beneath, the periosteum.11 After the needle is passed posterior to the clavicle, the angle is dropped significantly, causing the needle to only nick the vein anteriorly.8 In order to avoid the pleural apex (and pneumothorax), the needle trajectory is superior to the sternal notch.12,13 Backward retraction of the needle with syringe removal can prematurely pull the needle out of the vein and cause inability to pass the wire.14

How can mentors improve their teaching strategy? There are many resources available that describe how to correctly place a subclavian central venous catheter. But few comprehensively address common technical errors during failed insertions. Our aim was to describe the technical aspects of CVC placement in a consolidated fashion to improve the process for teaching subclavian line placement.

METHODS After institutional review board approval, video captures (split-screen) from 86 consecutive subclavian central venous catheterizations were recorded from noon to midnight during a 4-month period in an urban trauma resuscitation unit. A team of video technicians trained in observing central venous catheter placements recorded each procedure. Video cameras were placed at the edge of the trauma bay, providing both wide angle and close-up angle views directly onto the procedural field. The video technicians were present during the line placements to ensure that the procedural field was thoroughly taped with appropriate lighting and zoom-in capability for each line placement. Each physician operator was made aware of the video process for CVC placement and gave written consent to be observed during routine orientation before clinical activity at the institution. The camera systems were small and were placed in unobtrusive locations in the trauma bay, rendering them nearly invisible to the working physicians. Operators were either surgical or emergency medicine residents rotating from more than nine different training programs in the United States and Canada. No formal subclavian line placement training program at our institution was given to the residents before beginning their rotation. The authors were blinded to all operators level of training and specialty. The number of insertion attempts and failures were recorded for each operator. An attempt was defined as a

puncture of the skin with the cannulating needle. An attempt was successful when the subclavian vein was entered and the guidewire successfully passed. An attempt was defined as unsuccessful if the needle was removed without having the guidewire in place (ie, the vein was not cannulated, the wire was in a fascial plane, or the wire would not pass even if the needle was in the vein). Ultrasonographic guidance was not used in any of the placement attempts. Synthesizing information from anatomy textbooks, procedural handbooks, and previous journal articles, we developed a list of potential technical errors. Table 1 lists 6 errors in technique that lead to failure when attempting subclavian central venous catheter placement. For each unsuccessful insertion, one of the six technical errors was assigned. A team of two authors reviewed each video segment and determined the main cause for failure of a line placement. If there was difficulty in determining a cause for the line failure, a third author was consulted for his evaluation of the video. If, after the second review, no cause could be specifically elicited, that video segment was not included in the study. All authors had substantial previous clinical experience instructing residents on how to place lines and developing interdepartmental primers for teaching subclavian cannulation to novices. To address interrater differences related to the subjectivity involved in determining a cause for failure, a subset of 20 line placements was reviewed independently by a different author. A kappa test was then performed to demonstrate the interrater variance and the subsequent degree of agreement. A Students t-test was used for all other statistical analyses, with a p value of less than 0.05 considered statistically significant. All values are reported as mean values with SEM.

106

Kilbourne et al

Subclavian Central Venous Catheter Placement

J Am Coll Surg

Table 2. Subclavian Catheter Placement Technical Errors and Percentage of Failures

Technical error Frequency of errors, (n 277) Failures, %

Improper landmark identification Improper insertion relative to the clavicle Insertion through periosteum Too shallow trajectory Aiming too cephalad Failure to keep needle in place for wire passage

41 90 61 45 21 21

14.7 32.3 21.9 16.1 7.5 7.5

RESULTS Ninety-two videos were evaluated; 86 of these were included in the study. The six excluded videos did not have clear tape evidence of any one specific technical error, or the picture quality was such that a reliable decision could not be made as to the cause of the line failure. Seventy-seven of the 86 patients (89%) had subclavian lines successfully placed. Nine patients had the procedure aborted. There were 357 venipuncture attempts, with 279 failures, for an overall needle insertion failure rate of 78.2%. Table 2 breaks down the failed venipunctures with their associated technical errors. There were 242 venipuncture attempts on the right side and 115 attempts on the left side. The mean number of failed attempts per line was 3.2 (right side 5.5 4.3 and left side 2.1 3.0, p 0.016). Figure 1 demonstrates the technical error percentages based on laterality of placement. There was no statistically significant difference between technical error rates based on the side attempted. The presence of a cervical collar did not change the number of failed venipunctures per line (with collar, 2.8; without collar, 2.8). Fifty-two of the 86 subclavian central lines were placed by residents. The remaining 34 lines were placed by nonresidents, meaning a fellow or an attending physician. Statistical comparisons were not conducted on this group of senior operators, because the focus of the study was to help teach relative novices to avoid technical errors, not to critique already seasoned physicians on their line practice. Among resident operators, 29 were considered junior (PGY 1 to 2) and 23 were considered senior (PGY 3 to 5). Table 3 shows the mean failures per line for each resident

Figure 1. Subclavian catheter placement technical errors by laterality.

group. Figure 2 shows the technical error rates for each resident group. The kappa test for interrater consistency was performed, demonstrating very good correlation (0.840) among multiple authors evaluating a subset of 20 videos.

DISCUSSION Multiple venipunctures can increase the complication rate of central line placement. Complication rates tend to decrease as an operators experience increases, with about 50% fewer complications after a person has placed 50 lines.7 Mechanical complications, such as pneumothorax and hematoma, are infrequent, with rates between 1.5% and 3.1% and 1.2% and 2.1%, respectively.3 Although no mechanical complications were observed in any of the line placements in this series, it stands to reason that more sticks translates into more complications. The subclavian CVC success rate for all lines observed in our series was near 90%. Aborted procedures had a mean of 8.0 failed venipunctures per line. Even in eventually successful line placements, the mean number of failed venipunctures per line was still 3.2. The right subclavian had significantly more failures per line than the left side. We found no obvious anatomic or patient-related reason for this. Instead, we postulated that right-handedness (as was

Table 3. Failed Venipunctures per Central Line Placement by Resident Group

Resident group Lines attempted, n Failed attempts, n Mean failed attempts per line, n

Junior resident (PGY 1 to 2) Senior resident (PGY 3 to5)

*p 0.074.

29 23

119 83

4.1 3.6

2.4* 2.1

Vol. 208, No. 1, January 2009

Kilbourne et al

Subclavian Central Venous Catheter Placement

107

Figure 2. Subclavian catheter placement technical errors by resident group.

the case with nearly all observed operators) plays a role. Because all right-handed operators use their dominant hand for venipuncture, their most comfortable position seemed to be on the patients left side. Physician body position on the patients left can be likened to holding a pool-stick while playing billiards (ie, the left hand outstretched on the sternal notch/clavicle and the right hand maneuvering the needle and syringe). So, we recommend attempts on the patients left side (if the operator is righthanded), as long as it is appropriate for the clinical situation. In some of the video segments, maximal sterile techniques were not used because of the critical nature of the clinical situation. But we would like to stress the importance of using full protective barriers, including hat, mask, sterile gown, and gloves, during all line procedures. Overall, the most common technical error encountered was improper needle insertion position relative to the clavicle (video segment 1, online). In all cases, the needle was inserted too closely to the bone itself (less than 1 cm inferior and lateral to the middle to medial third of the clavicle). Close proximity to the clavicle creates a steep angle for cannulating the vein beneath the clavicle. Usually this causes the needle to miss the vein in a caudal direction, because the needle will not advance in between the clavicle and the first rib. Other times, the operator would actually obtain a flash of blood but be unable to pass the guidewire distally. The significant opposing resistance results from the wire striking the side wall of the vein at such a steep angle that it could not advance. The second most common error noted was insertion of the needle through the periosteum of the clavicle (video segment 2, online). Many operators touch the clavicle with

the needle tip to help guide themselves posteriorly, ie, walking the clavicle down to locate the vein beneath the clavicle. Many residents do not realize that the periosteal layer is both sensitive and thin. It is relatively easy to drive the needle through the periosteal layer and miss the subclavian vein anteriorly. Using significant force or aggressively pushing the needle can drive it through, instead of beneath, the periosteum. In other cases, operators attempt to bend or curve the needle around the clavicle, using the opposite hand to push down. Unfortunately, the needle will often catch the periosteal layer during this maneuver, causing it to become bent or deformed. We recommend placing gentle backward retraction on the needle between walking steps to keep from pushing the needle into the periosteum. The third most common technical error was taking too shallow of a trajectory of the needle (video segment 3, online). Avoiding a pneumothorax is a consideration for any operator performing subclavian vein catheterization.8 As a result, many physicians are concerned about the angle of the needle once it is posterior to the clavicle. This concern frequently causes the operator to mistrust the normal anatomic position of the vein and subsequently drop the needle angle too much in the coronal axis as it is passed beneath the clavicle. We suggest dropping the angle of the needle only slightly to ensure it is directed into the vein between the clavicle and first rib. The fourth most common technical error was improper or inadequate anatomic landmark identification. The two bony landmarks that must be palpated before each attempt are the sternal notch and the middle to medial third of the clavicle. The course of the subclavian vein is basically parallel to these landmarks. The sternal notch serves as the reference point for needle directionality; the middle third of the clavicle provides the starting point for skin puncture. Retaining the proper orientation of the needle is important before each attempt and throughout the attempt, because it is relatively easy to lose track of ones line of insertion amid needle manipulation and obscurative sterile draping. The fifth most common error observed was aiming the needle too cephalad (video segment 4, online). Part of the motivation to do this probably lies in the fact that mechanical complications like pneumothorax, as mentioned previously, are a significant concern. Consequently, the urge to aim cephalad and away from the pleural apex can cause the operator to miss the vein superiorly. This video segment also illustrates the danger in sweeping the needle angle after approaching the vein too cephalad. Making an arc with the needle deep in the skin not only can lacerate the subclavian vein but can injure the nearby subclavian artery.

108

Kilbourne et al

Subclavian Central Venous Catheter Placement

J Am Coll Surg

The least frequent error was failure to keep the needle in place for wire passage despite successful cannulation of the vein (video segment 5, online). The syringe must be withdrawn while maintaining a secure position of the needle hub. Operators who do not have the wire on the field or have to turn their body to retrieve it from the catheter tray are most prone to this mistake. Their movement and shift of focus off the procedural field can disrupt the correctly placed needle. At this juncture, an assistant (for the novice) can be very helpful, because he can hand the wire to the operator, avoiding unnecessary needle movement. The mean number of failures per line was higher in junior-level residents (4.1) versus senior-level residents (3.6), although this difference was not statistically significant (p 0.074). We made several observations about the distribution of errors in the two PGY groups. Senior-level residents had, proportionally, twice as many landmark errors. We believe that it is common for a senior operator to be lulled into thinking that increased experience with line placement could negate a rigid step-wise placement approach. In this case, the most likely step for an experienced operator to skip would be the landmark identification. Senior residents were also more likely than junior residents to displace the needle after it was already in the vein. Again, as with landmark identification, more experienced operators are probably less deliberate and rigid in their approach. Placement of the CVC is more on the fly. With speed comes less attention to the little details, like having the wire at the ready instead of behind the operator or on the tray. Junior-level residents had proportionally more technical errors from improper needle placement relative to the clavicle than did senior-level residents. Almost universally, this meant that junior residents would enter the skin too close to the clavicle. Two thought processes probably are responsible. First, the clavicle is the main anatomic landmark around which the needle is maneuvered. Less experienced operators are more likely to insert the needle close to the clavicle, because they believe it is easier to control its advancement under the clavicle. In addition, keeping the insertion close to the clavicle is likely more reassuring to a novice operator fearful of creating a pneumothorax. Along the same lines, junior residents had proportionally higher rates of aiming too shallow with the needle. We believe that this is much like the clavicle errors. The novice does not want to create a pneumothorax, so he will attempt to be as shallow as possible and hug the posterior portion of the clavicle to keep a mechanical complication from occurring. The principal limitation of this study was the judgment involved in deciding which technical errors were observed

in the line placements. We strove to identify the most notable error observed. There were times when an operator may have had more than one error during a placement. Occasionally, a failed venipuncture happened very rapidly, or the camera angle made it difficult to determine decisively what error was made. In these cases, the tape was reviewed several times and the authors best impression of the events was recorded. There were some failed venipunctures for which there appeared to be no specific technical error present. For these attempts, the error was either too subtle or there was a different reason for failure (ie, anomalous patient anatomy, equipment failure, an uncooperative patient, and so on). We addressed these limitations by having different reviewers give their interpretations in unclear circumstances. We would exclude an attempt if the video capture was so difficult to interpret that the reason for failure would be nothing more than a guess. In conclusion, subclavian CVC placement is an important skill in the trauma bay and the intensive care unit. Mentor teaching to avoid the six common line technique errors can hopefully improve novice proficiency.
Author Contributions Study conception and design: Kilbourne, Bochicchio, Xiao Acquisition of data: Kilbourne, Bochicchio, Xiao Analysis and interpretation of data: Kilbourne, Bochicchio, Xiao Drafting of manuscript: Kilbourne, Bochicchio, Xiao Critical revision: Bochicchio, Scalea Acknowledgment: We thank Steve Seebodee for video technical support.

REFERENCES 1. Mansfield PF, Hohn DC, Fornage BD, et al. Complications and failures of subclavian-vein catheterization. N Engl J Med 1994; 331:17351738. 2. Roberts JR, Hedges JR, Chanmugam AS, et al. Clinical procedures in emergency medicine. 4th ed. Philadelphia: WB Saunders; 2004. 3. Graham AS, Ozment C, Tegtmeyer K, et al. Central venous catheterization. N Engl J Med 2007;356:21. 4. Taylor RW, Palgiri AV. Central venous catheterization. Crit Care Med 2007;35:13901396. 5. Gualtieri E, Deppe SA, Sipperly ME, Thompson DR. Subclavian venous catheterization: greater success rate for less experienced operators using ultrasound guidance. Crit Care Med 1995;23:692697. 6. Gawande A. Complications: a surgeons notes on an imperfect science. New York: Metropolitan Books; 2002.

Vol. 208, No. 1, January 2009

Kilbourne et al

Subclavian Central Venous Catheter Placement

109

7. Sznajder JI, Zveibil FR, Bitterman H, et al. Central vein catheterization: failure and complication rates by three percutaneous approaches. Arch Int Med 1986;146:259 261. 8. Thompson EC, Calver LE. Safe subclavian vein cannulation how I do it. Am Surg 2005;71:180183. 9. Miller RD, Fleisher LA, Johns RA, et al. Millers anesthesia. 6th ed. Philadephia: Churchill Livingstone; 2005. 10. Nilsson KR, Piccini JP. The Osler medical handbook. 2nd ed. Philadelphia: WB Saunders; 2006.

11. Von Goedecke A, Keller C, Moriggl B, et al. An anatomic landmark to simplify subclavian vein cannulation: the deltoid tuberosity. Anesth Analg 2005;100:623628. 12. Procter A. A radiological examination of the subclavian vein in vivo. Can J Anesth 2005;52:A110. 13. Mitchell SE, Clark RA. Complications of central venous catheterization. Am J Roentgenol 1979;133:467476. 14. Townsend CM, Beauchamp RD, Evers MB, Mattox KL. Sabiston textbook of surgery: the biological basis of modern surgical practice. 18th ed. Philadelphia: WB Saunders; 2008.

NOTICE TO AUTHORS
Surgeon at Work articles submitted after October 31, 2007 will no longer be published in the print Journal. They will be published ONLINE ONLY and referenced in the table of contents with an e-page number. We encourage multimedia submissions (see Instructions for Authors at www.journals.org)