- Email: [email protected]
Tunneled Internal Jugular Hemodialysis Catheters: Impact of Laterality and Tip Position on Catheter Dysfunction and Infection Rates
CLINICAL STUDY
Tunneled Internal Jugular Hemodialysis Catheters: Impact of Laterality and Tip Position on Catheter Dysfunction and Infection Rates Bjorn I. Engstrom, MD, Jeffrey J. Horvath, MD, Jessica K. Stewart, MD, Ryan H. Sydnor, MD, Michael J. Miller, MD, Tony P. Smith, MD, and Charles Y. Kim, MD
ABSTRACT Purpose: To determine rates of dysfunction and infection for tunneled internal jugular vein hemodialysis catheters based on laterality of insertion and catheter tip position. Materials and Methods: Retrospective review of a procedural database for tunneled internal jugular vein hemodialysis catheter placements between January 2008 and December 2009 revealed 532 catheter insertions in 409 patients (234 male; mean age, 54.9 y). Of these, 398 catheters were placed on the right and 134 on the left. The catheter tip location was categorized as superior vena cava (SVC), pericavoatrial junction, or mid- to deep right atrium based on review of the final intraprocedural radiograph. The rates of catheter dysfunction and catheter-related infection (reported as events per 100 catheter-days) were analyzed. Results: Catheters terminating in the SVC or pericavoatrial junction inserted from the left showed significantly higher rates of infection (0.50 vs 0.27; P ¼ .005) and dysfunction (0.25 vs 0.11; P ¼ .036) compared with those inserted from the right. No difference was identified based on laterality for catheter tip position in the mid- to deep right atrium. Left-sided catheters terminating in the SVC or pericavoatrial junction had significantly more episodes of catheter dysfunction or infection than catheters terminating in the mid- to deep right atrium (0.84 vs 0.35; P ¼ .006), whereas no significant difference was identified for right-sided catheters based on tip position. Conclusions: When inserted from the left internal jugular vein, catheter tip position demonstrated a significant impact on catheterrelated dysfunction and infection; this relationship was not demonstrated for right-sided catheters.
ABBREVIATIONS NKF = National Kidney Foundation, SVC = superior vena cava, tPA = tissue plasminogen activator
The optimal location for the long-term tunneled central venous hemodialysis catheter tip is controversial (1,2). The National Kidney Foundation (NKF) recommends that tunneled hemodialysis catheters be placed with the tip in the mid-right atrium to maximize flow rates (3), a recommendation suggested by data from previous studies (4,5). Insertion of catheters from a right-sided approach has been advocated on the basis of a straighter
From the Division of Vascular and Interventional Radiology, Duke University Medical Center, Box 3808, Durham, NC 27710. Received March 21, 2012; final revision received May 13, 2013; accepted May 14, 2013. Address correspondence to C.Y.K.; E-mail: [email protected] None of the authors have identified a conflict of interest. & SIR, 2013 J Vasc Interv Radiol 2013; 24:1295–1302 http://dx.doi.org/10.1016/j.jvir.2013.05.035
catheter course compared with the left-sided approach (3). However, the impact of catheter tip position and laterality on catheter dysfunction and infection has not been fully elucidated. Earlier data have suggested that catheter longevity is decreased from a left-sided access compared with the right side (6). Similarly, data from nonhemodialysis tunneled catheters (7,8), implantable ports (9), and nontunneled hemodialysis catheters (10,11) suggest that left-sided central venous catheters are associated with higher rates of dysfunction, requiring exchange. Tunneled hemodialysis catheters are typically of larger caliber, used more frequently over an extended period of time, and have higher flow requirements than other types of central venous catheters; therefore, these preceding data may not be applicable. In the latest report (3), the NKF guidelines emphasized a need for additional research to determine the ideal catheter tip position for long-term hemodialysis catheters. Therefore,
1296
’
Impact of Catheter Laterality and Tip Position on Dysfunction
the purpose of the present study was to determine whether access site laterality and catheter tip position correlate with catheter dysfunction and infection in tunneled internal jugular hemodialysis catheters.
MATERIALS AND METHODS Subjects Institutional review board approval and waiver of informed consent were obtained for the present retrospective study. Review of the procedural database for all tunneled hemodialysis catheter placements via the internal jugular vein revealed 746 insertions in 425 patients between January 2008 and December 2009. Catheter exchanges were excluded from the initial patient acquisition. The majority of catheters (n ¼ 532) were Decathlon catheters (manufactured by Spire Biomedical [Bedford, Massachusetts] during the catheter insertion time period; currently manufactured by Bard Access Systems [Salt Lake City, Utah]), with the remainder being of one of three other catheter types. To avoid potential bias resulting from inclusion of multiple different catheter types, catheters other than the Decathlon catheter were excluded. The Decathlon long-term hemodialysis catheter is 16 F in caliber with a floating split-tip configuration. Therefore, the study cohort consisted of 532 tunneled catheter insertions in 409 patients (234 male, 175 female). The disposition of each catheter was reviewed until catheter removal, patient death, or October 2011. The mean patient age was 54.9 years (range, 18–91 y).
Procedural Technique All procedures were performed under moderate sedation in our interventional radiology angiography suites by an attending interventional radiologist or a fellow under the supervision of an attending physician. A standardized surgical hat, mask, and sterile gown and gloves were worn. Antibiotic prophylaxis was used in all procedures per our routine clinical practice. One gram of cefazolin was typically administered intravenously just before initiation of the procedure. For patients with an allergy to cefazolin, 600 mg of clindamycin or 1 g of vancomycin was administered. Insertion of tunneled hemodialysis catheters was not performed if systemic infection was suspected.
Tunneled Hemodialysis Catheter Insertion Technique Tunneled hemodialysis catheter insertion was performed by using real-time sonographic guidance for venous access and fluoroscopic guidance for catheter positioning. The right internal jugular vein was the preferred for access when both sides were available, based on the NKF guidelines (3). The left internal jugular vein was selected as the access site at the discretion of the operator in cases such as right internal jugular vein thrombus or
Engstrom et al
’
JVIR
occlusion; known right brachiocephalic vein pathologic condition; presence of a wound, recently removed infected catheter, or pacemaker/defibrillator in the right chest wall; patient preference; multiple existing central catheters in the right upper extremity; or permanent access planned for the right arm or while awaiting maturation of a right-sided fistula. The planned catheter tip position was the cavoatrial junction or mid-right atrium per the catheter instructions for use (12). Immediately after insertion and after each hemodialysis session, both ports were locked with a heparinized saline solution at a concentration of 1,000 U/mL. Catheters were secured to the skin with 2–0 nylon monofilament sutures.
Data Collection The final intraprocedural fluoroscopic images obtained during initial catheter insertion were reviewed to determine the catheter tip position by one of two boardcertified radiologists (B.I.E. and J.J.H.). The radiographic concave angulation formed by the intersection of the right mediastinal border and the convex right heart border was used as a reference point (1,3,13). Although this landmark is often used as a marker for the cavoatrial junction, a study based on cross-sectional imaging data (13) reported that the actual cavoatrial junction is located 0.5–4.5 cm inferior to this point, at a median of 1.0 cm away. Therefore, we defined the pericavoatrial junction as a region extending inferiorly from this radiographic landmark a distance equal to the height of a vertebral body. Based on these definitions, the location of the distal tip of the catheter was categorized into three anatomic regions: (i) above the mediastinal–right heart border angle, (ii) within one vertebral body below this point, or (iii) greater than one vertebral body below this point (Fig 1). This approximates a distal catheter tip in the superior vena cava (SVC), the pericavoatrial region, and the mid- to deep right atrium, respectively (13). The medical records for these patients were then reviewed for demographic data, procedural details, complications, and follow-up. If the catheter was later exchanged, the number of exchanges, the dates, and reason for the exchange was recorded to determine the cumulative catheter dysfunction event rate. Patients referred for catheter exchange as a result of inadequate flow rates were considered to have catheter dysfunction. Catheter removal as a result of local venous thrombosis was not considered actual catheter dysfunction for the purposes of this study. Catheter lock therapy with tissue plasminogen activator (tPA) was variably performed at the dialysis centers but not reliably recorded; therefore, these therapies were not included for our analysis. Catheter infection was diagnosed by the nephrology service at our institution based on the presence of fever, bacteremia, leukocytosis, and/or signs and symptoms of
Volume 24
’
Number 9
’
September
’
2013
Figure 1. Fluoroscopic image of the chest in the supine position demonstrates zones used to categorize catheter tip location into (1) SVC, (2) pericavoatrial junction, or (3) mid- to deep right atrium. The concave angulation formed by the border of the right atrium convexity and the right mediastinal border defined the border between zones 1 and 2 (arrow). The inferior border of zone 2 was a distance equal to one vertebral body height. The superior and inferior borders of zone 2 are at superior endplates of consecutive vertebral bodies.
tunnel infection, such as pericatheter erythema, tenderness, and/or drainage. All catheters diagnosed with infection were uniformly treated with catheter removal. Catheter tip cultures were not routinely performed. However, if a low level of suspicion for infection was present—eg, bacteremia with an identified noncatheter source in the setting of an asymptomatic catheter— empiric catheter exchange over a guide wire was sometimes performed; this was not considered to represent catheter dysfunction or infection. The date and reasons for catheter removal were also determined. Major immediate (o 24 h), early (o 30 d), and late (4 30 d) complications requiring major therapy were recorded according to Society of Interventional Radiology (SIR) reporting standards for central venous access (14).
1297
was divided by the total number of days from catheter insertion until the first catheter exchange or removal for dysfunction, or until the catheter endpoint if no catheter exchanges were performed. The catheter exchange rate for dysfunction per access site was calculated by dividing the total number of catheter exchanges performed as a result of dysfunction by the total number of days from catheter insertion until the catheter endpoint. The catheter infection rate was calculated in the same fashion by using infectionrelated catheter removal as the endpoint. These values were then multiplied by 100 to achieve the dysfunction and infection rate per 100 catheter-days. To compare catheter dysfunction and infection rates based on tip position and/ or laterality, the Kaplan–Meier technique was used to determine the catheter freedom-from-dysfunction and freedom-from-infection rates based on the time until first catheter exchange or infection-related catheter removal, respectively. The resulting curves were compared by using the log rank test. Catheters were censored for the following reasons: death, loss to follow up, functional catheter at the end of the study period, catheter exchange as a result of questionable infection or dislodgment, catheter removal as a result of venous thrombosis, excessive exit site bleeding, catheter no longer needed, or inadvertent catheter retraction. For the purposes of Kaplan–Meier analysis of catheter dysfunction, catheter removal for any reason, including infection, resulted in the catheter being censored in that patient. Likewise, for analysis of catheter infection, an exchange for any reason, including dysfunction, resulted in catheter censoring. Because of the marginal numbers of catheter tips in the SVC, statistical analysis was performed on the SVC and pericavoatrial junction positions separately and combined. Statistical analysis was performed by using SPSS software (version 19.0.0; IBM, Armonk, New York). A P value less than or equal to .05 was considered statistically significant.
RESULTS The majority of tunneled hemodialysis catheters were inserted in the right internal jugular vein with a catheter tip position in the mid- to deep right atrium (Table 1). No differences in demographic variables were identified between patients who underwent right- versus left-sided catheter insertion (Table 2). Of 532 tunneled hemodialysis catheter insertions, 81 catheters were exchanged during the present study period Table 1 . Catheter Insertion Laterality and Tip Position Tip Position
Right
Left
Total
33 (8.3)
4 (3.0)
37 (7.0)
Statistical Analysis
SVC
Demographic variables among patients with right- versus left-sided catheters were compared by using the χ2 test for categoric variables and Student t test for continuous variables. For purposes of calculating catheter dysfunction rates, the number of catheters with first-time dysfunction
Pericavoatrial junction
129 (32.4)
41 (30.6)
170 (32.0)
Mid-/deep right atrium Total
236 (59.3) 398 (74.8)
89 (66.4) 134 (25.2)
325 (61.1) 532 (100)
Values in parentheses are percentages. SVC ¼ superior vena cava.
1298
’
Impact of Catheter Laterality and Tip Position on Dysfunction
Engstrom et al
’
JVIR
Table 2 . Patient Demographics Characteristic Female sex Mean age (y) BMI Diabetes
Right (n ¼ 398) Left (n ¼ 134) P Value 170 (42.7)
63 (47.0)
.42
55.4 29.1
53.2 30.8
.17 .23
196 (49.2)
70 (52.2)
.62
HIV Steroid therapy
18 (4.5) 67 (16.8)
8 (6.0) 19 (14.2)
.49 .59
Chemotherapy
18 (4.8)
2 (1.5)
.19
Malignancy Hypercoagulability
36 (9.0) 13 (3.3)
8 (6.0) 2 (1.5)
.36 .38
Values in parentheses are percentages. BMI ¼ body mass index.
(Table 2). Of these, 66 were exchanged because of dysfunction during hemodialysis, 10 because of a low level of suspicion of catheter infection, and five because of inadvertent catheter retraction. The overall dysfunction rate requiring catheter exchange was 0.09 episodes per 100 catheter-days. During this study period, 213 catheters were removed because of catheter infection, at a rate of 0.26 episodes per 100 catheter-days. The mean follow up was 15.7 months. Only one major complication was encountered occurring in the immediate (o 24 h) period after placement, which occurred during left internal jugular hemodialysis catheter placement. In this procedure, the SVC was perforated during the insertion of the peel-away sheath over the guide wire, resulting in a large hemothorax with cardiopulmonary compromise, requiring emergent surgical decompression.
Figure 2. Kaplan–Meier curves demonstrating dysfunction-free survival based on catheter laterality.
Catheter Laterality The right internal jugular vein was used in 74.8% of catheters, and the left in 25.2% (Table 2). A left internal jugular vein access site was associated with higher catheter dysfunction rates (0.13 vs 0.08 catheter exchanges for dysfunction per 100 catheter-days; P ¼ .089) and catheterrelated infection rates (0.33 vs 0.24 catheter removals for infection per 100 catheter-days; P ¼ .012) compared with right internal jugular vein access; however, statistical significance was reached only for the difference in infection rates (Figs 2, 3, Table 3). The rate of either event (ie, dysfunction or infection) was significantly higher with left-sided access (0.45 vs 0.31; P ¼ .003).
Figure 3. Kaplan–Meier curves demonstrating infection-free survival based on catheter laterality.
Catheter Tip Position Of the 532 catheters analyzed, 7.0% of catheter tips were positioned in the SVC, 32.0% were at the pericavoatrial junction, and 61.1% were in the mid- to deep right atrium (Table 1). Analysis of the effect of catheter tip position revealed no statistical difference in catheter dysfunction requiring catheter exchange (P ¼ .197) or removal as a result of concern for catheter infection (P ¼ .054; Table 3). However, catheters placed in the SVC or pericavoatrial junction exhibited a significantly higher rate of either event (ie, infection or dysfunction)
compared with catheters placed in the mid- to deep right atrium (P ¼ .028). Notably, relatively few catheter tips were in the SVC (6.7% on the right, 0.5% on the left).
Combination of Catheter Tip Position and Laterality Rates of dysfunction and infection were analyzed for each catheter tip position based on laterality (Figs 4, 5). Catheters with the tip in the SVC or pericavoatrial
Volume 24
’
Number 9
’
September
’
2013
1299
Table 3 . Rates of Catheter Dysfunction and Infection Expressed per 100 Catheter-days Rate per 100 Catheter-days No of. Pts.
Dysfunction
Infection
Either
Laterality Right
Characteristic
398
0.08
0.24
0.31
Left
134
0.13
0.33
0.45
P ¼ .089
P ¼ .012n
P ¼ .003n
Right vs left Tip position SVC
37
0.15
0.24
0.40
PCJ SVC or PCJ
170 207
0.12 0.12
0.32 0.30
0.45 0.44
RA
325
SVC/PCJ vs RA Right-sided SVC
0.08
0.24
0.30
P ¼ .197
P ¼ .054
P ¼ .028n
33
0.13
0.23
0.34
PCJ SVC or PCJ
129 162
0.10 0.11
0.28 0.27
0.38 0.37
RA
236
SVC vs PCJ SVC vs RA SVC/PCJ vs RA PCJ vs RA Left-sided SVC
0.07
0.23
0.28
P ¼ .571 P ¼ .362
P ¼ .654 P ¼ .842
P ¼ .713 P ¼ .708
P ¼ .321
P ¼ .251
P ¼ .170
P ¼ .454
P ¼ .207
P ¼ .148
4
0.56
0.40
1.69
PCJ SVC or PCJ
41 45
0.23 0.25
0.51 0.50
0.79 0.84
RA
89
0.10
0.27
0.35
SVC vs PCJ SVC vs RA
P ¼ .468 P ¼ .369
P ¼ .665 P ¼ .650
P ¼ .431 P ¼ .058
SVC/PCJ vs RA
P ¼ .285
P ¼ .032n
P ¼ .006n
P ¼ .371
n
P ¼ .018
P ¼ .013n
P ¼ .279
P ¼ .503
P ¼ .032n
PCJ vs RA Right vs left SVC
n
SVC or PCJ PCJ RA Total
532
n
P ¼ .036 P ¼ .043n
P ¼ .005 P ¼ .005n
P o .001n P ¼ .001n
P ¼ .272
P ¼ .184
P ¼ .125
0.09
0.26
0.34
Comparison of event-free survival rates was performed by Kaplan–Meier technique and log-rank test. PCJ ¼ pericavoatrial junction, RA ¼ mid-/deep right atrium. n Significant at P ≤ .05.
junction had a significantly higher rate of catheter dysfunction if the catheter was placed from the left side compared with the right (0.25 vs 0.11 events per 100 catheter-days; P ¼ .036; Table 3). Left-sided catheters with tips in the SVC or pericavoatrial junction were also similarly prone to infection (0.50 vs 0.27; P ¼ .005). In contrast, catheters with tip in the mid- to deep right atrium had similar dysfunction and infection rates whether inserted from the left or right side (P ¼ .272 and P ¼ .184, respectively). Analysis of right-sided catheters revealed no significant difference in rate of dysfunction or infection based on tip position in SVC/ pericavoatrial junction versus mid- to deep right atrium (0.37 vs 0.28 events per 100 catheter-days; P ¼ .170). However, left-sided catheters terminating in the SVC or
pericavoatrial junction exhibited a significantly higher rate of dysfunction or infection compared with left-sided catheters with tips in the mid- to deep right atrium (0.84 vs 0.35 events per 100 catheter-days; P ¼ .006).
DISCUSSION Tunneled hemodialysis catheters are associated with a number of short-term and long-term complications, including catheter-related infection, catheter dysfunction, malpositioning, venous thrombosis, and venous stenosis (1,15–18). Differing guidelines have been issued by various organizations regarding the optimal catheter tip position, including SIR, NKF, and United States
1300
’
Impact of Catheter Laterality and Tip Position on Dysfunction
Figure 4. Kaplan–Meier curves demonstrating dysfunction-free survival based on combined catheter laterality and tip position.
Figure 5. Kaplan–Meier curves demonstrating infection-free survival based on combined catheter laterality and tip position.
Food and Drug Administration (1,3,19). The Food and Drug Administration, based on reports before the common use of fluoroscopy, recommends catheter tip positioning in the SVC as a result of potential cardiacrelated complications occurring at the time of catheter placement if the catheter tip is in the right atrium (19–22). Current NKF Kidney Disease Outcomes Quality Initiative recommendations for hemodialysis catheters are to place short-term catheter tips in the SVC and long-term catheter tips in the right atrium, primarily to
Engstrom et al
’
JVIR
maximize flow rates for hemodialysis (3). The package insert for the tunneled hemodialysis catheter used in the present study recommends tip positioning at “the junction of the superior vena cava/right atrium or in the mid right atrium” (12). The SIR guidelines (14) state that the optimal tip position has yet to be determined. In terms of laterality of the catheter, the Kidney Disease Outcomes Quality Initiative recommends the right internal jugular vein approach as a result of the “more direct” route to the right atrium (12). Notably, none of these recommendations are based on rates of infection, dysfunction, or catheter longevity, all of which are critical considerations in this population. Catheter dysfunction has been shown to be a significant contributor to the high costs associated with maintenance of catheter-dependent hemodialysis, occurring with a reported rate of 0.05–0.30 events per 100 catheter-days (15–18). The present catheter dysfunction rate of 0.09 events per 100 catheter-days fell within the lower end of this range. Catheter dysfunction can result from multiple reasons and is commonly caused by fibrin sheath formation around the tip of the catheter and/or intraluminal thrombus (23). Although tPA can resolve catheter dysfunction caused by fibrin sheath and/or thrombus (24), more invasive procedures such as catheter exchange, fibrin sheath disruption, catheter stripping, or catheter insertion at a de novo site may be necessary for full resolution (1,25,26). Although other studies based on different catheter types have demonstrated a general correlation with catheter laterality and longevity (6) as well as catheter dysfunction (7–9), the present study showed that laterality impacted rates of tip dysfunction only when the catheter terminated in the SVC or pericavoatrial junction region. The exact reason why left-sided catheters terminating in the SVC or pericavoatrial junction have higher dysfunction rates is unclear. It has been postulated that the tip of left-sided catheters terminating in this region likely abuts the right lateral wall of the SVC as a result of the geometric relationship, which may result in endothelial irritation, intimal hyperplasia, accelerated fibrin sheath formation, and possibly thrombosis formation (1). Notably, catheter tips identified in the high right atrium (included in the pericavoatrial region in the present study) with the patient supine on the fluoroscopy table may migrate into the SVC when the patient is in the upright position (27). Also notable is that the hemodialysis catheter analyzed in the present study has a staggered split tip, with the tip position analyzed based on the location of the distalmost tip. Therefore, even though the distal tip of a catheter may be in the right atrium just beyond the cavoatrial junction, the proximal tip may be in the SVC, where it may be prone to dysfunction. Hemodialysis catheter–related infection is an important source of morbidity and mortality, with a reported incidence of 0.04–0.55 episodes per 100 catheter-days in tunneled hemodialysis catheters (18,28). The present catheter infection rate of 0.26 is within the middle of
Volume 24
’
Number 9
’
September
’
2013
this range. The cost implications of catheter infection are considerable, reported to be approximately $23,451 per hospitalization for catheter-related bloodstream infection (29). In addition, the Centers for Medicare and Medicaid Services have ceased reimbursements for treatment of catheter-related bloodstream infections (30). Catheter infection can occur from bacterial migration along the outer surface of the catheter from the exit site, luminal contamination from handling of the hub, contamination of infusates, or hematogenous contamination from another source (31). In the present study, the incidence of catheter infection was highest for catheter tips in the SVC or pericavoatrial junction when inserted from left-sided access. Although left-sided catheters were more prone overall to infection than rightsided catheters, this difference was not present when the catheter tip was in the mid- to deep right atrium. In addition, for left-sided catheters, those with the tip in the mid- to deep right atrium were significantly less likely to be infected, whereas, for right-sided catheters, no such difference was identified. Although it is not clear how laterality of insertion or catheter tip position would directly influence rates of infection we speculate that an indirect effect may be responsible. For example, if left-sided catheters are more prone to dysfunction, they may be prone to higher rates of catheter manipulation during dialysis, which may increase chances for hub and intraluminal contamination. Alternatively, fibrin sheath formation may potentially be related, as bacterial adherence has been shown to be greater with a fibrin sheath than the smooth surface of the catheter (32–34). Catheters treated with tPA lock solution instead of a heparin lock solution have lower rates of catheter dysfunction, presumably as a result of inhibition of fibrin sheath and thrombus formation (17). Interestingly, these patients treated with tPA lock solution also experienced a significantly lower catheter infection rate, which again raises the possibility of a direct correlation between fibrin sheath presence and susceptibility to catheter-related bloodstream infection. Dissolution or prevention of intracatheter thrombus that may potentially harbor bacteria is also a possible mechanism. Given that dysfunctional catheters may be more prone to infection (or that infection may predispose a catheter to dysfunction), our analysis included a combined analysis of dysfunction or infection, which demonstrated accentuated correlations. The present study is substantially limited by its retrospective nature, which results in variable availability of pertinent clinical and technical information and potential for bias. One of the primary limitations is the difficulty in anatomic determination of the catheter tip position based on a frontal radiograph (1). Although correlation with cross-sectional imaging may be the only way to truly correlate catheter outcomes based on the true anatomic position of the catheter tip, this would not be helpful to the interventionalist at the time of catheter insertion with the use of basic fluoroscopy. Therefore, an
1301
easily identifiable reference point with the use of fluoroscopy was considered to be the most practical for this study. A position greater than one vertebral body inferior to the right mediastinal–to–right heart border intersection has been shown to be highly likely to be within the right atrium (14,35). The actual catheter tip position may not be accurately categorized based on the final intraprocedural fluoroscopic image, for example, if subsequent catheter retraction occurs or if the patient was in an exaggerated phase of inspiration or expiration during the final intraprocedural image acquisition (36). Another limitation is our definition of catheter infection (ie, catheter removal for suspected catheter infection), which was based on clinical and laboratory parameters used at our institution based on clinical practice guidelines. Routine catheter tip culture was not performed at our institution given the lack of impact on management (37). Because the right internal jugular vein was generally preferred versus the left side in the absence of any other concerns, selection bias may exist. Finally, relatively few catheters in this study were inserted with the tip in the SVC, particularly from a left-sided approach (n ¼ 4). Although the rates of catheter dysfunction and infection were high for this group, the small sample size limits the statistical analysis of catheters tips in the SVC. In addition, based on current recommendations, tip positioning in the SVC is accepted to be suboptimal because of flow rates, and therefore tip positioning in the mid- to deep right atrium versus the pericavoatrial junction is a more critical decision point. Given the morbidity and costs associated with dysfunction and infection of dialysis catheters, identification of factors associated with such complications are important. Our results suggest that the catheter tip position is an important determinant of episodes of dysfunction or infection when inserted from the left internal jugular vein, but less important when the catheter is inserted from the right. Specifically, when the catheter tip is positioned at the SVC or pericavoatrial junction from a left-sided approach, the incidence of infection or dysfunction is significantly higher than when the tip is positioned in the mid- to deep right atrium. Therefore, if a tunneled hemodialysis catheter is inserted via the left internal jugular vein, positioning of the catheter tip within the mid- to deep right atrium is recommended to minimize the risk of catheter dysfunction or infection.
REFERENCES 1. Vesely TM. Central venous catheter tip position: a continuing controversy. J Vasc Interv Radiol 2003; 14:527–537. 2. Schutz JC, Patel AA, Clark TW. Relationship between chest port catheter tip position and port malfunction after interventional radiologic placement. J Vasc Interv Radiol 2004; 15:581–587. 3. National Kidney Foundation. KDOQI clinical practice guidelines and clinical practice recommendations for 2006 updates: Hemodialysis adequacy, peritoneal dialysis adequacy and vascular access. Am J Kidney Dis 2006; 48(suppl 1):S1–S322.
1302
’
Impact of Catheter Laterality and Tip Position on Dysfunction
4. Mandolfo S, Galli F, Costa S, Ravani P, Gaggia P, Imbasciati E. Factors influencing permanent catheter performance. J Vasc Access Devices 2001; 2:106–109. 5. Jean G, Chazot C, Vanel T, et al. Central venous catheters for haemodialysis: looking for optimal blood flow. Nephrol Dial Transplant 1997; 12:1689–1691. 6. Fry AC, Stratton J, Farrington K, et al. Factors affecting long-term survival of tunneled haemodialysis catheters—a prospective audit of 812 tunnelled catheters. Nephrol Dial Transplant 2008; 23:275–281. 7. Craft PS, May J, Dorigo A, Hoy C, Plant A. Hickman catheters: left-sided insertion, male gender, and obesity are associated with an increased risk of complications. Aust N Z J Med 1996; 26:33–39. 8. Brown-Smith JK, Stoner MH, Barley ZA. Tunneled catheter thrombosis: factors related to incidence. Oncol Nurs Forum 1990; 17: 543–549. 9. Puel V, Caudry M, Le Métayer P, et al. Superior vena cava thrombosis related to catheter malposition in cancer chemotherapy given through implanted ports. Cancer 1993; 72:2248–2252. 10. Salgado OJ, Urdaneta B, Colmenares B, García R, Flores C. Right versus left internal jugular vein catheterization for hemodialysis: complications and impact on ipsilateral access creation. Artif Organs 2004; 28:728–733. 11. Parienti JJ, Mégarbane B, Fischer MO. Catheter dysfunction and dialysis performance according to vascular access among 736 critically ill adults requiring renal replacement therapy: a randomized controlled study. Crit Care Med 2010; 38:1118–1125. 12. Decathlon DF. Long-term hemodialysis catheter [instructions for use]. Salt Lake City: Bard Access Systems, August 2011. 13. Aslamy Z, Dewald CL, Heffner JE. MRI of central venous anatomy. Implications for central venous catheter insertion. Chest 1998; 114: 820–826. 14. Silberzweig JE, Sacks D, Khorsandi AS, Bakal CW. Reporting standards for central venous access. J Vasc Interv Radiol 2003; 14(suppl):S443–S452. 15. Lee H, Manns B, Taub K, et al. Cost analysis of ongoing care of patients with end-stage renal disease: The impact of dialysis modality and dialysis access. Am J Kidney Dis 2002; 40:611–622. 16. Manns B, Tonelli M, Yilmaz S, et al. Establishment and maintenance of vascular access in incident hemodialysis patients: a prospective cost analysis. J Am Soc Nephrol 2005; 16:201–209. 17. Hemmelgarn BR, Moist LM, Charmaine EL, et al. Prevention of dialysis catheter malfunction with recombinant tissue plasminogen activator. N Engl J Med 2011; 364:303–312. 18. Little MA, O’Riordan A, Lucey B, et al. A prospective study of complications associated with cuffed, tunnelled haemodialysis catheters. Nephrol Dial Transplant 2001; 16:2194–2200. 19. Food and Drug Administration Task Force. Precautions necessary with central venous catheters. FDA Drug Bulletin July 1989;15–16. 20. Huyghens L, Sennesael J, Verbeelen D, Deraemaecker R, Corne L. Cardiothoracic complications of centrally inserted catheters. Acute Care 1985; 11:53–56.
Engstrom et al
’
JVIR
21. Robinson JF, Robinson WA, Cohn A, Garg K, Armstrong JD. Perforation of the great vessels during central venous line placement. Arch Intern Med 1995; 155:1225–1228. 22. Collier PE, Goodman GB. Cardiac tamponade caused by central venous catheter perforation of the heart: a preventable complication. J Am Coll Surg 1995; 181:459–463. 23. Rasmussen RL. The catheter-challenged patient and the need to recognize the recurrently dysfunctional tunneled dialysis catheter. Semin Dial 2010;23: 648–646. 24. Haymond J, Shalansky K, Jastrzebski J. Efficacy of low-dose alteplase for treatment of hemodialysis catheter occlusions. J Vasc Access 2005; 6:76–82. 25. Merport M, Murphy TP, Egglin TK, Dubel GJ. Fibrin sheath stripping versus catheter exchange for the treatment of failed tunneled hemodialysis catheters: randomized clinical trial. J Vasc Interv Radiol 2000; 11: 1115–1120. 26. Beathard GA. Dysfunction of new catheters by old fibrin sheaths. Semin Dial 2004; 17:243–244. 27. Nazarian GK, Bjarnason H, Dietz CA, Bernadas CA, Hunter DW. Changes in tunneled catheter tip position when a patient is upright. J Vasc Interv Radiol 1997; 8:437–441. 28. Saad TF. Central venous dialysis catheters: catheter-associated infection. Semin Dial 2001; 14:446–451. 29. Ramanathan V, Chiu EJ, Thomas JT, Khan A, Dolson GM, Darouiche RO. Healthcare costs associated with hemodialysis catheter-related infections: a single-center experience. Infect Control Hosp Epidemiol 2007; 28:606–609. 30. Mattie AS, Webster BL. Centers for Medicare and Medicaid Services’ “never events”: an analysis and recommendations to hospitals. Health Care Manag 2008; 27:338–349. 31. Miller DL, O’Grady NP. Guidelines for the prevention of intravascular catheter-related infections: recommendations relevant to interventional radiology for venous catheter placement and maintenance. J Vasc Interv Radiol 2012; 14(suppl):S355–S358. 32. Passerini L, Lam K, Costerton JW, Kin EG. Biofilms on indwelling vascular catheters. Crit Care Med 1992; 20:665–673. 33. Trerotola SO. Hemodialysis catheter placement and management. Radiology 2000; 215:651–658. 34. Mehall JR, Saltzman DA, Jackson RJ, Smith SD. Fibrin sheath enhances central venous catheter infection. Crit Care Med 2002; 30:908–912. 35. Baskin KM, Jimenez RM, Cahill AM, Jawad AF, Towbin RB. Cavoatrial junction and central venous anatomy: implications for central venous access tip position. J Vasc Interv Radiol 2008; 19:359–365. 36. Pan PP, Engstrom BI, Lungren MP, Seaman DM, Lessne ML, Kim CY. Impact of phase of respiration on central venous catheter tip position. J Vasc Access 2013, http://dx.doi.org/10:5301/jva.5000135. 37. Cooper ET, Cohen RM, Berns JS, Kornfield ZN, Trerotola SO. Impact of tip culture on the management of infected tunneled hemodialysis catheters. J Vasc Interv Radiol 2007; 18:1227–1231.
Tunneled Internal Jugular Hemodialysis Catheters: Impact of Laterality and Tip Position on Catheter Dysfunction and Infection Rates Bjorn I. Engstrom, MD, Jeffrey J. Horvath, MD, Jessica K. Stewart, MD, Ryan H. Sydnor, MD, Michael J. Miller, MD, Tony P. Smith, MD, and Charles Y. Kim, MD
ABSTRACT Purpose: To determine rates of dysfunction and infection for tunneled internal jugular vein hemodialysis catheters based on laterality of insertion and catheter tip position. Materials and Methods: Retrospective review of a procedural database for tunneled internal jugular vein hemodialysis catheter placements between January 2008 and December 2009 revealed 532 catheter insertions in 409 patients (234 male; mean age, 54.9 y). Of these, 398 catheters were placed on the right and 134 on the left. The catheter tip location was categorized as superior vena cava (SVC), pericavoatrial junction, or mid- to deep right atrium based on review of the final intraprocedural radiograph. The rates of catheter dysfunction and catheter-related infection (reported as events per 100 catheter-days) were analyzed. Results: Catheters terminating in the SVC or pericavoatrial junction inserted from the left showed significantly higher rates of infection (0.50 vs 0.27; P ¼ .005) and dysfunction (0.25 vs 0.11; P ¼ .036) compared with those inserted from the right. No difference was identified based on laterality for catheter tip position in the mid- to deep right atrium. Left-sided catheters terminating in the SVC or pericavoatrial junction had significantly more episodes of catheter dysfunction or infection than catheters terminating in the mid- to deep right atrium (0.84 vs 0.35; P ¼ .006), whereas no significant difference was identified for right-sided catheters based on tip position. Conclusions: When inserted from the left internal jugular vein, catheter tip position demonstrated a significant impact on catheterrelated dysfunction and infection; this relationship was not demonstrated for right-sided catheters.
ABBREVIATIONS NKF = National Kidney Foundation, SVC = superior vena cava, tPA = tissue plasminogen activator
The optimal location for the long-term tunneled central venous hemodialysis catheter tip is controversial (1,2). The National Kidney Foundation (NKF) recommends that tunneled hemodialysis catheters be placed with the tip in the mid-right atrium to maximize flow rates (3), a recommendation suggested by data from previous studies (4,5). Insertion of catheters from a right-sided approach has been advocated on the basis of a straighter
From the Division of Vascular and Interventional Radiology, Duke University Medical Center, Box 3808, Durham, NC 27710. Received March 21, 2012; final revision received May 13, 2013; accepted May 14, 2013. Address correspondence to C.Y.K.; E-mail: [email protected] None of the authors have identified a conflict of interest. & SIR, 2013 J Vasc Interv Radiol 2013; 24:1295–1302 http://dx.doi.org/10.1016/j.jvir.2013.05.035
catheter course compared with the left-sided approach (3). However, the impact of catheter tip position and laterality on catheter dysfunction and infection has not been fully elucidated. Earlier data have suggested that catheter longevity is decreased from a left-sided access compared with the right side (6). Similarly, data from nonhemodialysis tunneled catheters (7,8), implantable ports (9), and nontunneled hemodialysis catheters (10,11) suggest that left-sided central venous catheters are associated with higher rates of dysfunction, requiring exchange. Tunneled hemodialysis catheters are typically of larger caliber, used more frequently over an extended period of time, and have higher flow requirements than other types of central venous catheters; therefore, these preceding data may not be applicable. In the latest report (3), the NKF guidelines emphasized a need for additional research to determine the ideal catheter tip position for long-term hemodialysis catheters. Therefore,
1296
’
Impact of Catheter Laterality and Tip Position on Dysfunction
the purpose of the present study was to determine whether access site laterality and catheter tip position correlate with catheter dysfunction and infection in tunneled internal jugular hemodialysis catheters.
MATERIALS AND METHODS Subjects Institutional review board approval and waiver of informed consent were obtained for the present retrospective study. Review of the procedural database for all tunneled hemodialysis catheter placements via the internal jugular vein revealed 746 insertions in 425 patients between January 2008 and December 2009. Catheter exchanges were excluded from the initial patient acquisition. The majority of catheters (n ¼ 532) were Decathlon catheters (manufactured by Spire Biomedical [Bedford, Massachusetts] during the catheter insertion time period; currently manufactured by Bard Access Systems [Salt Lake City, Utah]), with the remainder being of one of three other catheter types. To avoid potential bias resulting from inclusion of multiple different catheter types, catheters other than the Decathlon catheter were excluded. The Decathlon long-term hemodialysis catheter is 16 F in caliber with a floating split-tip configuration. Therefore, the study cohort consisted of 532 tunneled catheter insertions in 409 patients (234 male, 175 female). The disposition of each catheter was reviewed until catheter removal, patient death, or October 2011. The mean patient age was 54.9 years (range, 18–91 y).
Procedural Technique All procedures were performed under moderate sedation in our interventional radiology angiography suites by an attending interventional radiologist or a fellow under the supervision of an attending physician. A standardized surgical hat, mask, and sterile gown and gloves were worn. Antibiotic prophylaxis was used in all procedures per our routine clinical practice. One gram of cefazolin was typically administered intravenously just before initiation of the procedure. For patients with an allergy to cefazolin, 600 mg of clindamycin or 1 g of vancomycin was administered. Insertion of tunneled hemodialysis catheters was not performed if systemic infection was suspected.
Tunneled Hemodialysis Catheter Insertion Technique Tunneled hemodialysis catheter insertion was performed by using real-time sonographic guidance for venous access and fluoroscopic guidance for catheter positioning. The right internal jugular vein was the preferred for access when both sides were available, based on the NKF guidelines (3). The left internal jugular vein was selected as the access site at the discretion of the operator in cases such as right internal jugular vein thrombus or
Engstrom et al
’
JVIR
occlusion; known right brachiocephalic vein pathologic condition; presence of a wound, recently removed infected catheter, or pacemaker/defibrillator in the right chest wall; patient preference; multiple existing central catheters in the right upper extremity; or permanent access planned for the right arm or while awaiting maturation of a right-sided fistula. The planned catheter tip position was the cavoatrial junction or mid-right atrium per the catheter instructions for use (12). Immediately after insertion and after each hemodialysis session, both ports were locked with a heparinized saline solution at a concentration of 1,000 U/mL. Catheters were secured to the skin with 2–0 nylon monofilament sutures.
Data Collection The final intraprocedural fluoroscopic images obtained during initial catheter insertion were reviewed to determine the catheter tip position by one of two boardcertified radiologists (B.I.E. and J.J.H.). The radiographic concave angulation formed by the intersection of the right mediastinal border and the convex right heart border was used as a reference point (1,3,13). Although this landmark is often used as a marker for the cavoatrial junction, a study based on cross-sectional imaging data (13) reported that the actual cavoatrial junction is located 0.5–4.5 cm inferior to this point, at a median of 1.0 cm away. Therefore, we defined the pericavoatrial junction as a region extending inferiorly from this radiographic landmark a distance equal to the height of a vertebral body. Based on these definitions, the location of the distal tip of the catheter was categorized into three anatomic regions: (i) above the mediastinal–right heart border angle, (ii) within one vertebral body below this point, or (iii) greater than one vertebral body below this point (Fig 1). This approximates a distal catheter tip in the superior vena cava (SVC), the pericavoatrial region, and the mid- to deep right atrium, respectively (13). The medical records for these patients were then reviewed for demographic data, procedural details, complications, and follow-up. If the catheter was later exchanged, the number of exchanges, the dates, and reason for the exchange was recorded to determine the cumulative catheter dysfunction event rate. Patients referred for catheter exchange as a result of inadequate flow rates were considered to have catheter dysfunction. Catheter removal as a result of local venous thrombosis was not considered actual catheter dysfunction for the purposes of this study. Catheter lock therapy with tissue plasminogen activator (tPA) was variably performed at the dialysis centers but not reliably recorded; therefore, these therapies were not included for our analysis. Catheter infection was diagnosed by the nephrology service at our institution based on the presence of fever, bacteremia, leukocytosis, and/or signs and symptoms of
Volume 24
’
Number 9
’
September
’
2013
Figure 1. Fluoroscopic image of the chest in the supine position demonstrates zones used to categorize catheter tip location into (1) SVC, (2) pericavoatrial junction, or (3) mid- to deep right atrium. The concave angulation formed by the border of the right atrium convexity and the right mediastinal border defined the border between zones 1 and 2 (arrow). The inferior border of zone 2 was a distance equal to one vertebral body height. The superior and inferior borders of zone 2 are at superior endplates of consecutive vertebral bodies.
tunnel infection, such as pericatheter erythema, tenderness, and/or drainage. All catheters diagnosed with infection were uniformly treated with catheter removal. Catheter tip cultures were not routinely performed. However, if a low level of suspicion for infection was present—eg, bacteremia with an identified noncatheter source in the setting of an asymptomatic catheter— empiric catheter exchange over a guide wire was sometimes performed; this was not considered to represent catheter dysfunction or infection. The date and reasons for catheter removal were also determined. Major immediate (o 24 h), early (o 30 d), and late (4 30 d) complications requiring major therapy were recorded according to Society of Interventional Radiology (SIR) reporting standards for central venous access (14).
1297
was divided by the total number of days from catheter insertion until the first catheter exchange or removal for dysfunction, or until the catheter endpoint if no catheter exchanges were performed. The catheter exchange rate for dysfunction per access site was calculated by dividing the total number of catheter exchanges performed as a result of dysfunction by the total number of days from catheter insertion until the catheter endpoint. The catheter infection rate was calculated in the same fashion by using infectionrelated catheter removal as the endpoint. These values were then multiplied by 100 to achieve the dysfunction and infection rate per 100 catheter-days. To compare catheter dysfunction and infection rates based on tip position and/ or laterality, the Kaplan–Meier technique was used to determine the catheter freedom-from-dysfunction and freedom-from-infection rates based on the time until first catheter exchange or infection-related catheter removal, respectively. The resulting curves were compared by using the log rank test. Catheters were censored for the following reasons: death, loss to follow up, functional catheter at the end of the study period, catheter exchange as a result of questionable infection or dislodgment, catheter removal as a result of venous thrombosis, excessive exit site bleeding, catheter no longer needed, or inadvertent catheter retraction. For the purposes of Kaplan–Meier analysis of catheter dysfunction, catheter removal for any reason, including infection, resulted in the catheter being censored in that patient. Likewise, for analysis of catheter infection, an exchange for any reason, including dysfunction, resulted in catheter censoring. Because of the marginal numbers of catheter tips in the SVC, statistical analysis was performed on the SVC and pericavoatrial junction positions separately and combined. Statistical analysis was performed by using SPSS software (version 19.0.0; IBM, Armonk, New York). A P value less than or equal to .05 was considered statistically significant.
RESULTS The majority of tunneled hemodialysis catheters were inserted in the right internal jugular vein with a catheter tip position in the mid- to deep right atrium (Table 1). No differences in demographic variables were identified between patients who underwent right- versus left-sided catheter insertion (Table 2). Of 532 tunneled hemodialysis catheter insertions, 81 catheters were exchanged during the present study period Table 1 . Catheter Insertion Laterality and Tip Position Tip Position
Right
Left
Total
33 (8.3)
4 (3.0)
37 (7.0)
Statistical Analysis
SVC
Demographic variables among patients with right- versus left-sided catheters were compared by using the χ2 test for categoric variables and Student t test for continuous variables. For purposes of calculating catheter dysfunction rates, the number of catheters with first-time dysfunction
Pericavoatrial junction
129 (32.4)
41 (30.6)
170 (32.0)
Mid-/deep right atrium Total
236 (59.3) 398 (74.8)
89 (66.4) 134 (25.2)
325 (61.1) 532 (100)
Values in parentheses are percentages. SVC ¼ superior vena cava.
1298
’
Impact of Catheter Laterality and Tip Position on Dysfunction
Engstrom et al
’
JVIR
Table 2 . Patient Demographics Characteristic Female sex Mean age (y) BMI Diabetes
Right (n ¼ 398) Left (n ¼ 134) P Value 170 (42.7)
63 (47.0)
.42
55.4 29.1
53.2 30.8
.17 .23
196 (49.2)
70 (52.2)
.62
HIV Steroid therapy
18 (4.5) 67 (16.8)
8 (6.0) 19 (14.2)
.49 .59
Chemotherapy
18 (4.8)
2 (1.5)
.19
Malignancy Hypercoagulability
36 (9.0) 13 (3.3)
8 (6.0) 2 (1.5)
.36 .38
Values in parentheses are percentages. BMI ¼ body mass index.
(Table 2). Of these, 66 were exchanged because of dysfunction during hemodialysis, 10 because of a low level of suspicion of catheter infection, and five because of inadvertent catheter retraction. The overall dysfunction rate requiring catheter exchange was 0.09 episodes per 100 catheter-days. During this study period, 213 catheters were removed because of catheter infection, at a rate of 0.26 episodes per 100 catheter-days. The mean follow up was 15.7 months. Only one major complication was encountered occurring in the immediate (o 24 h) period after placement, which occurred during left internal jugular hemodialysis catheter placement. In this procedure, the SVC was perforated during the insertion of the peel-away sheath over the guide wire, resulting in a large hemothorax with cardiopulmonary compromise, requiring emergent surgical decompression.
Figure 2. Kaplan–Meier curves demonstrating dysfunction-free survival based on catheter laterality.
Catheter Laterality The right internal jugular vein was used in 74.8% of catheters, and the left in 25.2% (Table 2). A left internal jugular vein access site was associated with higher catheter dysfunction rates (0.13 vs 0.08 catheter exchanges for dysfunction per 100 catheter-days; P ¼ .089) and catheterrelated infection rates (0.33 vs 0.24 catheter removals for infection per 100 catheter-days; P ¼ .012) compared with right internal jugular vein access; however, statistical significance was reached only for the difference in infection rates (Figs 2, 3, Table 3). The rate of either event (ie, dysfunction or infection) was significantly higher with left-sided access (0.45 vs 0.31; P ¼ .003).
Figure 3. Kaplan–Meier curves demonstrating infection-free survival based on catheter laterality.
Catheter Tip Position Of the 532 catheters analyzed, 7.0% of catheter tips were positioned in the SVC, 32.0% were at the pericavoatrial junction, and 61.1% were in the mid- to deep right atrium (Table 1). Analysis of the effect of catheter tip position revealed no statistical difference in catheter dysfunction requiring catheter exchange (P ¼ .197) or removal as a result of concern for catheter infection (P ¼ .054; Table 3). However, catheters placed in the SVC or pericavoatrial junction exhibited a significantly higher rate of either event (ie, infection or dysfunction)
compared with catheters placed in the mid- to deep right atrium (P ¼ .028). Notably, relatively few catheter tips were in the SVC (6.7% on the right, 0.5% on the left).
Combination of Catheter Tip Position and Laterality Rates of dysfunction and infection were analyzed for each catheter tip position based on laterality (Figs 4, 5). Catheters with the tip in the SVC or pericavoatrial
Volume 24
’
Number 9
’
September
’
2013
1299
Table 3 . Rates of Catheter Dysfunction and Infection Expressed per 100 Catheter-days Rate per 100 Catheter-days No of. Pts.
Dysfunction
Infection
Either
Laterality Right
Characteristic
398
0.08
0.24
0.31
Left
134
0.13
0.33
0.45
P ¼ .089
P ¼ .012n
P ¼ .003n
Right vs left Tip position SVC
37
0.15
0.24
0.40
PCJ SVC or PCJ
170 207
0.12 0.12
0.32 0.30
0.45 0.44
RA
325
SVC/PCJ vs RA Right-sided SVC
0.08
0.24
0.30
P ¼ .197
P ¼ .054
P ¼ .028n
33
0.13
0.23
0.34
PCJ SVC or PCJ
129 162
0.10 0.11
0.28 0.27
0.38 0.37
RA
236
SVC vs PCJ SVC vs RA SVC/PCJ vs RA PCJ vs RA Left-sided SVC
0.07
0.23
0.28
P ¼ .571 P ¼ .362
P ¼ .654 P ¼ .842
P ¼ .713 P ¼ .708
P ¼ .321
P ¼ .251
P ¼ .170
P ¼ .454
P ¼ .207
P ¼ .148
4
0.56
0.40
1.69
PCJ SVC or PCJ
41 45
0.23 0.25
0.51 0.50
0.79 0.84
RA
89
0.10
0.27
0.35
SVC vs PCJ SVC vs RA
P ¼ .468 P ¼ .369
P ¼ .665 P ¼ .650
P ¼ .431 P ¼ .058
SVC/PCJ vs RA
P ¼ .285
P ¼ .032n
P ¼ .006n
P ¼ .371
n
P ¼ .018
P ¼ .013n
P ¼ .279
P ¼ .503
P ¼ .032n
PCJ vs RA Right vs left SVC
n
SVC or PCJ PCJ RA Total
532
n
P ¼ .036 P ¼ .043n
P ¼ .005 P ¼ .005n
P o .001n P ¼ .001n
P ¼ .272
P ¼ .184
P ¼ .125
0.09
0.26
0.34
Comparison of event-free survival rates was performed by Kaplan–Meier technique and log-rank test. PCJ ¼ pericavoatrial junction, RA ¼ mid-/deep right atrium. n Significant at P ≤ .05.
junction had a significantly higher rate of catheter dysfunction if the catheter was placed from the left side compared with the right (0.25 vs 0.11 events per 100 catheter-days; P ¼ .036; Table 3). Left-sided catheters with tips in the SVC or pericavoatrial junction were also similarly prone to infection (0.50 vs 0.27; P ¼ .005). In contrast, catheters with tip in the mid- to deep right atrium had similar dysfunction and infection rates whether inserted from the left or right side (P ¼ .272 and P ¼ .184, respectively). Analysis of right-sided catheters revealed no significant difference in rate of dysfunction or infection based on tip position in SVC/ pericavoatrial junction versus mid- to deep right atrium (0.37 vs 0.28 events per 100 catheter-days; P ¼ .170). However, left-sided catheters terminating in the SVC or
pericavoatrial junction exhibited a significantly higher rate of dysfunction or infection compared with left-sided catheters with tips in the mid- to deep right atrium (0.84 vs 0.35 events per 100 catheter-days; P ¼ .006).
DISCUSSION Tunneled hemodialysis catheters are associated with a number of short-term and long-term complications, including catheter-related infection, catheter dysfunction, malpositioning, venous thrombosis, and venous stenosis (1,15–18). Differing guidelines have been issued by various organizations regarding the optimal catheter tip position, including SIR, NKF, and United States
1300
’
Impact of Catheter Laterality and Tip Position on Dysfunction
Figure 4. Kaplan–Meier curves demonstrating dysfunction-free survival based on combined catheter laterality and tip position.
Figure 5. Kaplan–Meier curves demonstrating infection-free survival based on combined catheter laterality and tip position.
Food and Drug Administration (1,3,19). The Food and Drug Administration, based on reports before the common use of fluoroscopy, recommends catheter tip positioning in the SVC as a result of potential cardiacrelated complications occurring at the time of catheter placement if the catheter tip is in the right atrium (19–22). Current NKF Kidney Disease Outcomes Quality Initiative recommendations for hemodialysis catheters are to place short-term catheter tips in the SVC and long-term catheter tips in the right atrium, primarily to
Engstrom et al
’
JVIR
maximize flow rates for hemodialysis (3). The package insert for the tunneled hemodialysis catheter used in the present study recommends tip positioning at “the junction of the superior vena cava/right atrium or in the mid right atrium” (12). The SIR guidelines (14) state that the optimal tip position has yet to be determined. In terms of laterality of the catheter, the Kidney Disease Outcomes Quality Initiative recommends the right internal jugular vein approach as a result of the “more direct” route to the right atrium (12). Notably, none of these recommendations are based on rates of infection, dysfunction, or catheter longevity, all of which are critical considerations in this population. Catheter dysfunction has been shown to be a significant contributor to the high costs associated with maintenance of catheter-dependent hemodialysis, occurring with a reported rate of 0.05–0.30 events per 100 catheter-days (15–18). The present catheter dysfunction rate of 0.09 events per 100 catheter-days fell within the lower end of this range. Catheter dysfunction can result from multiple reasons and is commonly caused by fibrin sheath formation around the tip of the catheter and/or intraluminal thrombus (23). Although tPA can resolve catheter dysfunction caused by fibrin sheath and/or thrombus (24), more invasive procedures such as catheter exchange, fibrin sheath disruption, catheter stripping, or catheter insertion at a de novo site may be necessary for full resolution (1,25,26). Although other studies based on different catheter types have demonstrated a general correlation with catheter laterality and longevity (6) as well as catheter dysfunction (7–9), the present study showed that laterality impacted rates of tip dysfunction only when the catheter terminated in the SVC or pericavoatrial junction region. The exact reason why left-sided catheters terminating in the SVC or pericavoatrial junction have higher dysfunction rates is unclear. It has been postulated that the tip of left-sided catheters terminating in this region likely abuts the right lateral wall of the SVC as a result of the geometric relationship, which may result in endothelial irritation, intimal hyperplasia, accelerated fibrin sheath formation, and possibly thrombosis formation (1). Notably, catheter tips identified in the high right atrium (included in the pericavoatrial region in the present study) with the patient supine on the fluoroscopy table may migrate into the SVC when the patient is in the upright position (27). Also notable is that the hemodialysis catheter analyzed in the present study has a staggered split tip, with the tip position analyzed based on the location of the distalmost tip. Therefore, even though the distal tip of a catheter may be in the right atrium just beyond the cavoatrial junction, the proximal tip may be in the SVC, where it may be prone to dysfunction. Hemodialysis catheter–related infection is an important source of morbidity and mortality, with a reported incidence of 0.04–0.55 episodes per 100 catheter-days in tunneled hemodialysis catheters (18,28). The present catheter infection rate of 0.26 is within the middle of
Volume 24
’
Number 9
’
September
’
2013
this range. The cost implications of catheter infection are considerable, reported to be approximately $23,451 per hospitalization for catheter-related bloodstream infection (29). In addition, the Centers for Medicare and Medicaid Services have ceased reimbursements for treatment of catheter-related bloodstream infections (30). Catheter infection can occur from bacterial migration along the outer surface of the catheter from the exit site, luminal contamination from handling of the hub, contamination of infusates, or hematogenous contamination from another source (31). In the present study, the incidence of catheter infection was highest for catheter tips in the SVC or pericavoatrial junction when inserted from left-sided access. Although left-sided catheters were more prone overall to infection than rightsided catheters, this difference was not present when the catheter tip was in the mid- to deep right atrium. In addition, for left-sided catheters, those with the tip in the mid- to deep right atrium were significantly less likely to be infected, whereas, for right-sided catheters, no such difference was identified. Although it is not clear how laterality of insertion or catheter tip position would directly influence rates of infection we speculate that an indirect effect may be responsible. For example, if left-sided catheters are more prone to dysfunction, they may be prone to higher rates of catheter manipulation during dialysis, which may increase chances for hub and intraluminal contamination. Alternatively, fibrin sheath formation may potentially be related, as bacterial adherence has been shown to be greater with a fibrin sheath than the smooth surface of the catheter (32–34). Catheters treated with tPA lock solution instead of a heparin lock solution have lower rates of catheter dysfunction, presumably as a result of inhibition of fibrin sheath and thrombus formation (17). Interestingly, these patients treated with tPA lock solution also experienced a significantly lower catheter infection rate, which again raises the possibility of a direct correlation between fibrin sheath presence and susceptibility to catheter-related bloodstream infection. Dissolution or prevention of intracatheter thrombus that may potentially harbor bacteria is also a possible mechanism. Given that dysfunctional catheters may be more prone to infection (or that infection may predispose a catheter to dysfunction), our analysis included a combined analysis of dysfunction or infection, which demonstrated accentuated correlations. The present study is substantially limited by its retrospective nature, which results in variable availability of pertinent clinical and technical information and potential for bias. One of the primary limitations is the difficulty in anatomic determination of the catheter tip position based on a frontal radiograph (1). Although correlation with cross-sectional imaging may be the only way to truly correlate catheter outcomes based on the true anatomic position of the catheter tip, this would not be helpful to the interventionalist at the time of catheter insertion with the use of basic fluoroscopy. Therefore, an
1301
easily identifiable reference point with the use of fluoroscopy was considered to be the most practical for this study. A position greater than one vertebral body inferior to the right mediastinal–to–right heart border intersection has been shown to be highly likely to be within the right atrium (14,35). The actual catheter tip position may not be accurately categorized based on the final intraprocedural fluoroscopic image, for example, if subsequent catheter retraction occurs or if the patient was in an exaggerated phase of inspiration or expiration during the final intraprocedural image acquisition (36). Another limitation is our definition of catheter infection (ie, catheter removal for suspected catheter infection), which was based on clinical and laboratory parameters used at our institution based on clinical practice guidelines. Routine catheter tip culture was not performed at our institution given the lack of impact on management (37). Because the right internal jugular vein was generally preferred versus the left side in the absence of any other concerns, selection bias may exist. Finally, relatively few catheters in this study were inserted with the tip in the SVC, particularly from a left-sided approach (n ¼ 4). Although the rates of catheter dysfunction and infection were high for this group, the small sample size limits the statistical analysis of catheters tips in the SVC. In addition, based on current recommendations, tip positioning in the SVC is accepted to be suboptimal because of flow rates, and therefore tip positioning in the mid- to deep right atrium versus the pericavoatrial junction is a more critical decision point. Given the morbidity and costs associated with dysfunction and infection of dialysis catheters, identification of factors associated with such complications are important. Our results suggest that the catheter tip position is an important determinant of episodes of dysfunction or infection when inserted from the left internal jugular vein, but less important when the catheter is inserted from the right. Specifically, when the catheter tip is positioned at the SVC or pericavoatrial junction from a left-sided approach, the incidence of infection or dysfunction is significantly higher than when the tip is positioned in the mid- to deep right atrium. Therefore, if a tunneled hemodialysis catheter is inserted via the left internal jugular vein, positioning of the catheter tip within the mid- to deep right atrium is recommended to minimize the risk of catheter dysfunction or infection.
REFERENCES 1. Vesely TM. Central venous catheter tip position: a continuing controversy. J Vasc Interv Radiol 2003; 14:527–537. 2. Schutz JC, Patel AA, Clark TW. Relationship between chest port catheter tip position and port malfunction after interventional radiologic placement. J Vasc Interv Radiol 2004; 15:581–587. 3. National Kidney Foundation. KDOQI clinical practice guidelines and clinical practice recommendations for 2006 updates: Hemodialysis adequacy, peritoneal dialysis adequacy and vascular access. Am J Kidney Dis 2006; 48(suppl 1):S1–S322.
1302
’
Impact of Catheter Laterality and Tip Position on Dysfunction
4. Mandolfo S, Galli F, Costa S, Ravani P, Gaggia P, Imbasciati E. Factors influencing permanent catheter performance. J Vasc Access Devices 2001; 2:106–109. 5. Jean G, Chazot C, Vanel T, et al. Central venous catheters for haemodialysis: looking for optimal blood flow. Nephrol Dial Transplant 1997; 12:1689–1691. 6. Fry AC, Stratton J, Farrington K, et al. Factors affecting long-term survival of tunneled haemodialysis catheters—a prospective audit of 812 tunnelled catheters. Nephrol Dial Transplant 2008; 23:275–281. 7. Craft PS, May J, Dorigo A, Hoy C, Plant A. Hickman catheters: left-sided insertion, male gender, and obesity are associated with an increased risk of complications. Aust N Z J Med 1996; 26:33–39. 8. Brown-Smith JK, Stoner MH, Barley ZA. Tunneled catheter thrombosis: factors related to incidence. Oncol Nurs Forum 1990; 17: 543–549. 9. Puel V, Caudry M, Le Métayer P, et al. Superior vena cava thrombosis related to catheter malposition in cancer chemotherapy given through implanted ports. Cancer 1993; 72:2248–2252. 10. Salgado OJ, Urdaneta B, Colmenares B, García R, Flores C. Right versus left internal jugular vein catheterization for hemodialysis: complications and impact on ipsilateral access creation. Artif Organs 2004; 28:728–733. 11. Parienti JJ, Mégarbane B, Fischer MO. Catheter dysfunction and dialysis performance according to vascular access among 736 critically ill adults requiring renal replacement therapy: a randomized controlled study. Crit Care Med 2010; 38:1118–1125. 12. Decathlon DF. Long-term hemodialysis catheter [instructions for use]. Salt Lake City: Bard Access Systems, August 2011. 13. Aslamy Z, Dewald CL, Heffner JE. MRI of central venous anatomy. Implications for central venous catheter insertion. Chest 1998; 114: 820–826. 14. Silberzweig JE, Sacks D, Khorsandi AS, Bakal CW. Reporting standards for central venous access. J Vasc Interv Radiol 2003; 14(suppl):S443–S452. 15. Lee H, Manns B, Taub K, et al. Cost analysis of ongoing care of patients with end-stage renal disease: The impact of dialysis modality and dialysis access. Am J Kidney Dis 2002; 40:611–622. 16. Manns B, Tonelli M, Yilmaz S, et al. Establishment and maintenance of vascular access in incident hemodialysis patients: a prospective cost analysis. J Am Soc Nephrol 2005; 16:201–209. 17. Hemmelgarn BR, Moist LM, Charmaine EL, et al. Prevention of dialysis catheter malfunction with recombinant tissue plasminogen activator. N Engl J Med 2011; 364:303–312. 18. Little MA, O’Riordan A, Lucey B, et al. A prospective study of complications associated with cuffed, tunnelled haemodialysis catheters. Nephrol Dial Transplant 2001; 16:2194–2200. 19. Food and Drug Administration Task Force. Precautions necessary with central venous catheters. FDA Drug Bulletin July 1989;15–16. 20. Huyghens L, Sennesael J, Verbeelen D, Deraemaecker R, Corne L. Cardiothoracic complications of centrally inserted catheters. Acute Care 1985; 11:53–56.
Engstrom et al
’
JVIR
21. Robinson JF, Robinson WA, Cohn A, Garg K, Armstrong JD. Perforation of the great vessels during central venous line placement. Arch Intern Med 1995; 155:1225–1228. 22. Collier PE, Goodman GB. Cardiac tamponade caused by central venous catheter perforation of the heart: a preventable complication. J Am Coll Surg 1995; 181:459–463. 23. Rasmussen RL. The catheter-challenged patient and the need to recognize the recurrently dysfunctional tunneled dialysis catheter. Semin Dial 2010;23: 648–646. 24. Haymond J, Shalansky K, Jastrzebski J. Efficacy of low-dose alteplase for treatment of hemodialysis catheter occlusions. J Vasc Access 2005; 6:76–82. 25. Merport M, Murphy TP, Egglin TK, Dubel GJ. Fibrin sheath stripping versus catheter exchange for the treatment of failed tunneled hemodialysis catheters: randomized clinical trial. J Vasc Interv Radiol 2000; 11: 1115–1120. 26. Beathard GA. Dysfunction of new catheters by old fibrin sheaths. Semin Dial 2004; 17:243–244. 27. Nazarian GK, Bjarnason H, Dietz CA, Bernadas CA, Hunter DW. Changes in tunneled catheter tip position when a patient is upright. J Vasc Interv Radiol 1997; 8:437–441. 28. Saad TF. Central venous dialysis catheters: catheter-associated infection. Semin Dial 2001; 14:446–451. 29. Ramanathan V, Chiu EJ, Thomas JT, Khan A, Dolson GM, Darouiche RO. Healthcare costs associated with hemodialysis catheter-related infections: a single-center experience. Infect Control Hosp Epidemiol 2007; 28:606–609. 30. Mattie AS, Webster BL. Centers for Medicare and Medicaid Services’ “never events”: an analysis and recommendations to hospitals. Health Care Manag 2008; 27:338–349. 31. Miller DL, O’Grady NP. Guidelines for the prevention of intravascular catheter-related infections: recommendations relevant to interventional radiology for venous catheter placement and maintenance. J Vasc Interv Radiol 2012; 14(suppl):S355–S358. 32. Passerini L, Lam K, Costerton JW, Kin EG. Biofilms on indwelling vascular catheters. Crit Care Med 1992; 20:665–673. 33. Trerotola SO. Hemodialysis catheter placement and management. Radiology 2000; 215:651–658. 34. Mehall JR, Saltzman DA, Jackson RJ, Smith SD. Fibrin sheath enhances central venous catheter infection. Crit Care Med 2002; 30:908–912. 35. Baskin KM, Jimenez RM, Cahill AM, Jawad AF, Towbin RB. Cavoatrial junction and central venous anatomy: implications for central venous access tip position. J Vasc Interv Radiol 2008; 19:359–365. 36. Pan PP, Engstrom BI, Lungren MP, Seaman DM, Lessne ML, Kim CY. Impact of phase of respiration on central venous catheter tip position. J Vasc Access 2013, http://dx.doi.org/10:5301/jva.5000135. 37. Cooper ET, Cohen RM, Berns JS, Kornfield ZN, Trerotola SO. Impact of tip culture on the management of infected tunneled hemodialysis catheters. J Vasc Interv Radiol 2007; 18:1227–1231.