NDT Advance Access first published online on November 5, 2007
This version published online on January 24, 2008
Nephrology Dialysis Transplantation, doi:10.1093/ndt/gfm767
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Prospective follow-up of a novel design haemodialysis catheter; lower infection rates and improved survival
Department of Nephrology, Vrije Universiteit Medical Center, Amsterdam, The Netherlands
Correspondence and offprint requests to: M. C. Weijmer, Department of Nephrology, Sint Lucas Andreas Hospital, PO Box 9243, 1006 AE Amsterdam, The Netherlands. Tel: +31-205108911; Fax: +31-206837720; E-mail: mc.weijmer{at}weijmer.nl
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and methodsResultsCatheter outcomesDiscussionReferences |
|---|
Background. Untunnelled straight jugular catheters (USC) are uncomfortable for patients and cannot be well fixated. This could be a reason for the high incidence of catheter-related complications.
Methods. We prospectively analysed the outcome of a novel designed untunnelled precurved catheter (UPC) with better fixation properties and compared it with the outcome of USC. The outcome was also related to data on tunnelled cuffed catheters (TCC).
Results. The outcome of USC was documented over a 32-month period. Thereafter, we switched to an UPC. The same catheter care protocol was used and not changed over time. A total of 104 USC and 65 UPC were inserted. Compared to USC, less UPC had to be removed for a complication (53 versus 15%; P < 0.001) and less periods of catheter-related bacteraemia were observed in UPC compared to USC [0 versus 5.6 per 1000 catheter days (cd); P < 0.01]. Removal for flow problems was similar. Compared to 64 TCC, inserted in the same period, UPC had more flow problems. Other outcomes and complication rates were similar. Complication rates for TCC inserted before and after the switch from USC to UPC were similar.
Conclusions. UPC have better patency rates and a lower risk for bacteraemia and exit-site infection compared to USC.
Keywords: Bacteraemia; catheter; clinical study; comparison; haemodialysis; infection; outcome; patency; precurved; temporary; tunnelled cuffed; untunnelled; vascular access
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsCatheter outcomesDiscussionReferences |
|---|
Temporary untunnelled haemodialysis catheters for vascular access are regularly used in haemodialysis practice. Today, the preferred site for a temporary catheter is the jugular vein. Untunnelled catheters are often left in place for a prolonged period of time, despite the recommendation in the recently updated NKF-DOQI guidelines to use tunnelled cuffed catheters (TCC) whenever it can be anticipated that a catheter will be needed for >3 weeks [1]. This is probably because inserting TCC requires more experience, prolonged procedure time and special skills of the operator.
In incident haemodialysis patients, 48% of catheters in the United States and 75% of catheters in Europe are untunnelled and even in prevalent patients over a third of all catheters are untunnelled [2]. It has been shown that catheter-related complications are higher for temporary untunnelled jugular catheters compared to tunnelled catheters [3,4]. The reasons for these increased rates are not clear. It is suggested that poor fixation could be largely responsible. Current models of untunnelled straight jugular vein catheters (USC) are placed in the external or internal jugular vein pointing upward from the place of insertion. They are fixated cranial from the point of insertion and either have curved extensions (Figure 1A) or are curved laterally. These catheters are uncomfortable for the patient and can easily dislocate. In addition, on connection to the haemodialysis machine, pulling lines can kink the catheter or cause laceration of the exit site, a known risk factor for infection [5,6].
|
Recently, a novel designed untunnelled precurved temporary jugular catheter (UPC) model has become available (Duoflow®, Medcomp, Harleysville, USA). Inserted close to the upper border of the clavicle with the catheter bending over the clavicle and fixated to the chest wall, this catheter is more comfortable for patients (Figure 1B). During haemodialysis treatment, movement of this catheter is minimal.
The primary aim of this study was to compare the patency and catheter-related complications of this UPC with USC. We also related the results to the outcome of TCC during this study.
|
Subjects and methods |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsCatheter outcomesDiscussionReferences |
|---|
Patients and data collection
We analysed all untunnelled temporary jugular and tunnelled cuffed haemodialysis catheters inserted in the jugular vein at the dialysis department of an academical teaching hospital over a 4-year period. On average, 60 haemodialysis patients are treated in our unit, 10% are patients with acute renal failure and 25% of patients depend on a catheter for vascular access. Data on all inserted catheters and catheter-related complications at follow-up were entered prospectively in a computerized patient data system by the attendant nephrologist and dialysis nursing team.
The decision on the type of catheter and place of insertion was left to the physician responsible for the patients’ care at the time of insertion. In general, TCC were more likely to be chosen whenever it could be foreseen that a catheter was needed for >4 to 6 weeks. When this could not be determined or the need for a catheter was expected to be shorter, an untunnelled jugular catheter could be inserted.
Reasons for catheter removal were entered in the database: elective removal, exit-site infection, catheter-related bacteraemia (CRB), flow problems, accidental removal or catheter fracture. Catheters were evaluated for the type, place of insertion, time of insertion, duration of use and reason for removal. Femoral catheters were excluded. All cultures of blood and exit sites taken from the patient during the period when a catheter was in place were collected from the computerized data system of the department of microbiology. Baseline patient characteristics and demographical data were collected at the time of insertion. In addition, hospitalization for non-catheter-related reasons at any time when a catheter was in place and nasal staphylococcal aureus carrier was documented. Data on two catheters (both USC) were lost to follow-up shortly after insertion, because of transferral of the patient to another dialysis centre. These were excluded from the analysis.
Catheters, catheter care protocol and catheter outcome assessment
From 1 January 1997 to 1 September 1999, we used an untunnelled dual lumen 11 Fr polyurethane catheter with curved extensions (Gamcath®, Hechingen, Germany; n = 89) and dual lumen 11.5 Fr polyurethane catheter with curved extensions (Mahurkar, Tyco, Mansfield, MA, USA; n = 15) for the jugular site (Figure 1A). The outcome of these catheters has been described in our previous study [3]. After this period, we switched to a novel design untunnelled dual lumen polyurethane catheter (11.5 Fr; Duo-Flow® IJ, Medcomp, Harleysville, USA) (Figure 1B). For TCC we used Neostar Circle-C® (13.5 Fr; cuffed silicon, Horizon Medical Products, Atlanta, GA, USA; n = 48), PermCath® cuffed silicon (16 Fr; Tyco, Mansfield, MA, USA; n = 5), Tesio twin-cath® cuffed silicon (10 Fr; n = 3) and Ash-split® cuffed polyurethane (14.5 Fr; both Medcomp; n = 8).
All catheters were inserted under local anaesthesia and strict asepsis, and sutured to the skin. All untunnelled and tunnelled cuffed catheters were inserted after cannulation of the internal jugular vein close to (about 1 cm above) the clavicle, low in the sternocleidomastoid triangle in accordance with the puncture technique described by Rao et al. [7]. A standard Seldinger guidewire procedure was used to introduce the catheter. Care was taken to insert the catheter including the first half of the curve. After insertion, the exit site should point forward to downward. Ultrasound guidance was used whenever considered necessary.
Catheters were only handled by experienced dialysis nurses or nephrology staff, using sterile gloves and masks. The semi-occlusive dressings we used for exit-site care were inspected at every dialysis session and changed at least once a week, the exit site being cleansed with a povidone–iodine solution. Before removing the caps, the catheter hubs were disinfected with a chlorhexidine solution (2.5%). After dialysis treatment, catheters were locked with unfractionated heparin (5000 U/ml) with a volume equivalent to the internal volume of the lumen noted on the catheter. Catheters were used for haemodialysis exclusively.
CRB was defined as fever (temperature >38°C) or cold chills not during a dialysis treatment, with at least one positive blood culture and no other obvious cause of infection. In patients who developed signs of bacteraemia without symptoms of an alternative source other than the catheter, at least two blood cultures were taken either from the catheter or from a peripheral vein. Subsequently, antibiotics for suspected CRB were given. When a CRB was established, the catheter was left in place in stable patients in whom fever disappeared after initiation of antibiotic treatment. In patients not improving within 48 h or with recurrent bacteraemia within 3 weeks after stopping antibiotic treatment, the catheter was removed. The policy was not different between untunnelled catheters and TCC at our institution in the study period. In the case of recurrent bacteraemia, only the first period was counted for the analysis.
Exit-site infection was defined as the development of a purulent exudate or redness around the site not resulting from residual stitches. After culturing, antibiotic treatment was recommended for at least 2 weeks. In the case of no improvement, the catheter had to be removed.
Flow problems
In accordance with national guidelines, the minimal acceptable Qb was 200 ml/min and the target was 250 ml/min. More important, the minimal acceptable dialysis dose was a urea Kt/V of at least 1.2 per treatment. When this could not be reached, the flow or dialysis duration could be increased. When a persistent inability to run a blood flow of >200 ml/min occurred despite positional changes of the patient and/or additional flushing, urokinase 10 000 IU/ml was installed in both pools with a volume equivalent to the internal volume of the lumen. After 15 min the urokinase was withdrawn. If a blood flow of >200 ml/min was not achieved after this procedure, 100 000–250 000 IU of urokinase could be infused in 3 h during dialysis according to the protocol of Twardowski [8]. When this was not successful, the catheter was removed or exchanged.
No patient started coumarines because of flow problems.
Statistical analysis
Statistical analysis was performed with SPSS software 11.2 (SPSS Inc., Chicago, IL, USA). Non-parametric tests for two (Mann–Whitney U test) and multiple continuous variables (Kruskal–Wallis test) were used. For comparing binary and categorical variables, Chi-square and Fisher's exact tests were applied where appropriate. ANOVA was used to compare age and time on dialysis between multiple groups. Kaplan–Meier survival curves were constructed to analyse the patency rates and infectious complications. Functional catheters at the end of the observation time and catheters removed because they were no longer needed were analysed as censored values. The log rank test was used to compare groups and identify individual risk factors associated with a premature removal or catheter-related infection. At an individual two-sided P value of <0.1, the factor was included to fit a Cox-regression model. We used a forward stepwise conditional technique to identify the factors independently associated with catheter failure and infection. Differences were considered statistically significant for P < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsCatheter outcomesDiscussionReferences |
|---|
A total of 233 haemodialysis catheters [14 434 catheter days (cd)] were included in the final analysis. There were 104 USC, 65 UPC and 64 TCC inserted.
Patient and catheter characteristics are given in Table 1. There were no statistical differences between USC and UPC. Patients who had a TCC inserted differed importantly from UC; patients with a TCC were more likely to be on chronic haemodialysis treatment, hospitalized, on coumarines and have diabetes diagnosed as primary cause for end-stage renal disease.
|
Patient and catheter characteristics of TCC inserted before (n = 31) and after (n = 33) the switch from USC to UPC were similar. Most catheters were right sided (87%). Since there was no significant difference in the outcome between catheters inserted in the right or left jugular vein, they were not separated in the definite analysis.
|
Catheter outcomes |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsCatheter outcomesDiscussionReferences |
|---|
Of 65 UPC inserted, 55 (85%) could be used until they were no longer needed compared to only 49 (47%) of 104 USC (relative risk for premature removal 0.22; 95% CI 0.11–0.44; P < 0.001). We experienced no episodes of pneumothorax on insertion. In 11 USC and 5 UPC insertions the carotid artery was accidentally punctured; no major bleeding was experienced. The rate of premature removal was reduced from 17.1 to 4.3 per 1000 cd after switching from USC to UPC.
Characteristics of premature removals are shown in Table 2. There was a reduction of 11.5 to 0 removals for catheter-related infections per 1000 cd after switching from USC to UPC (P < 0.01). However, removals for flow problems were similar (5.0 versus 4.3 per 1000 cd; P = 0.27). Log rank analysis of the risk for premature removal showed better rates for UPC compared to the USC group (P < 0.0001, Figure 2)
|
|
TCC were characterized by less premature removal rates compared to both untunnelled catheter groups. The risk for premature removal of TCC inserted before and after the switch from USC to UPC was not statistically different (relative risk 1.33; 95% CI 0.55–3.26; P = 0.52, Table 3, Figure 4).
|
|
Infectious complications
In patients with a UPC inserted, no episodes of CRB occurred. One catheter was instantly removed in an immunocompromised patient who presented with fever. However, an alternative diagnosis was made during the follow-up.
There was a reduction of CRB episodes after switching from USC to UPC (5.6 to 0 per 1000 cd; P < 0.05 by log rank testing) (Figure 3). Differences in CRB and exit-site infection rates within the catheter groups are presented in Table 2. The risk for CRB and exit-site infection in the TCC group during both periods was equal (relative risk 1.01; 95% CI 0.47–2.17; P = 0.97)
|
In 42% of cases, cultures yielded gram-positive micro-organisms, predominantly Staphylococcus aureus or Staphylococcus epidermidis, in 34% it concerned gram-negative micro-organisms. The remaining cultures revealed multiple micro-organisms or yeasts.
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsCatheter outcomesDiscussionReferences |
|---|
Our study demonstrates that a novel design forward-bended precurved haemodialysis catheter inserted in the low jugular site has a better survival and lower risk of infection than a straight jugular catheter.
There are some possible explanations for these findings. In the present study the reduction of exit-site infections for UPC (0 per 1000 cd) compared to USC (10 per 1000 cd) was remarkable and this probably partially explained the reduction in CRB. An important problem of USC is that an adequate fixation is difficult and that they are uncomfortable for the patient because neck and head movements are limited. Inadequate fixation and discomfort leads to more manipulation, sliding of the catheter through its port and can easily give laceration and secondary infection of the exit site. These are well-known risk factors for subsequent CRB [9,10]. In addition, straight jugular catheters have an upward directed exit site. In peritoneal dialysis catheters, this is a well-established risk factor for exit-site infections and peritonitis [5,6]. Probably, with an upward directed exit site, adequate drainage of debris is prohibited and colonization of the catheter exit site promoted, which can cause local infection and subsequent systemic infection. Therefore, in peritoneal dialysis catheter management, a downward directed exit site is recommended.
Exit-site colonization can also lead to catheter hub colonization and subsequent bacteraemia. This could also be an explanation for the differences found in this study, but as regular cultures of the hub were not performed, this remains speculative. The same reduction of colonization and infections in exit sites that permit an adequate drainage could also be the reason that in recent studies, USC had higher rates of infectious complications compared to untunnelled subclavian catheters [3,4]. The UPC used in this study has the advantage of a forward to downward directed exit site that makes adequate drainage of debris possible. The improved fixation properties and exit-site direction are probably the most important explanation for the reduction of infection rates in UPC compared to USC. As overall outcome between UPC and TCC was not different, UPC can be an attractive alternative for the time period studied. The low rates of infectious complications in UPC compared to TCC might have been influenced by the shorter period UPC were left in place compared to TCC.
There are limitations to our study. The prospective sequential cohort analysis of the USC and UPC respectively, instead of a randomized design, could have led to bias due to unnoticed differences in time, not accounted for by baseline characteristics. Furthermore, the propensity for inserting a TCC instead of an untunnelled catheter might have changed over time. However, since outcome in TCC in the two time periods did not change, this probably did not influence the results. Also, the analysis between untunnelled catheters on one hand and TCC on the other might be biased due to lack of randomization and differences in indication.
We observed differences in baseline characteristics between TCC and UPC, like the presence of diabetes and the incidence of acute renal failure. Despite the fact that we corrected for these differences in the outcome analysis, patients with a TCC are different to patients with a UC and a direct comparison is hazardous.
Another confounding factor in catheter studies influencing the outcome of catheters is the catheter care protocol. It has been shown in randomized studies that the risk for CRB can be reduced with a thorough protocol [15,16]. In our study, catheter care protocol was not changed over the entire observation period and the incidence of catheter-related infections of TCC in the time periods when USC (first period) or UPC (second period) were used, was not statistically different making a time bias and differences in catheter care as explanation for our findings unlikely.
Previous studies in tunnelled catheters have suggested that catheter-related complications, especially infections, are constant over time or tend to decrease [3,11,12]. Furthermore, as shown in the choice study, few patients will need a catheter for >3 months [13]. Considering catheter-related infections as a major drawback for catheter use, our results show that an UPC for the period of 3 months could be a safe option, more convenient for the patient and physician during insertion and probably cost-saving. This is in contrast with the NKF-DOQI guidelines, which state that a tunnellized catheter should be used whenever it can be foreseen that a catheter is needed for >2 to 4 weeks [1]. However, these guidelines are predominantly expert based or supported by studies with USC including our own [3,4].
Compared to TCC, more UPC had to be removed for flow problems. This was probably caused by the fact that TCC have a 2–3 Fr larger diameter and the flow resistance is proportional to the diameter of a catheter to the fourth power. Therefore, it is clear that flow characteristics are better in TCC and that there is more flow reserved in the case of partial obstruction. The function of UPC can probably be improved by increasing its diameter and tip construction [14].
In conclusion, our study demonstrates that in addition to improved patient comfort, an untunnelled precurved jugular haemodialysis catheter has better patency rates and a lower risk for infection compared to a straight jugular catheter with curved extensions. This novel design precurved catheter placed in the lower jugular position should be preferred as the untunnelled jugular catheter model over straight models with curved extensions and can be used safely when it can be foreseen that a catheter is needed for up to 3 months. Our results make a prospective randomized trial comparing TCC to UPC with a wider inner diameter an important issue for the future.
Conflict of interest statement. None declared.
|
References |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsCatheter outcomesDiscussionReferences |
|---|
- NKF-K/DOQI Clinical practice guidelines for vascular access: update. 2000. Am J Kidney Dis (2001) 37:S137–S181.[Medline]
- Pisoni RL, Young EW, Dykstra DM, et al. Vascular access use in Europe and the United States: results from the DOPPS. Kidney Int (2002) 61:305–316.[CrossRef][Web of Science][Medline]
- Weijmer MC, Vervloet MG, ter Wee PM. Compared to tunnelled cuffed haemodialysis catheters, temporary untunnelled catheters are associated with more complications already within 2 weeks of use. Nephrol Dial Transplant (2004) 19:670–677.
[Abstract/Free Full Text] - Kairaitis LK, Gottlieb T. Outcome and complications of temporary haemodialysis catheters. Nephrol Dial Transplant (1999) 14:1710–1714.
[Abstract/Free Full Text] - Warady BA, Sullivan EK, Alexander SR. Lessons from the peritoneal dialysis patient database: a report of the north American pediatric renal transplant cooperative study. Kidney Int Suppl (1996) 53:S68–S71.[Medline]
- Golper TA, Brier ME, Bunke M, et al. Risk factors for peritonitis in long-term peritoneal dialysis: the network 9 peritonitis and catheter survival studies. Academic subcommittee of the steering committee of the network 9 peritonitis and catheter survival studies. Am J Kidney Dis (1996) 28:428–436.[Web of Science][Medline]
- Rao TL, Wong AY, Salem MR. A new approach to percutaneous catheterization of the internal jugular vein. Anesthesiology (1977) 46:362–364.[Web of Science][Medline]
- Twardowski ZJ. High-dose intradialytic urokinase to restore the patency of permanent central vein hemodialysis catheters. Am J Kidney Dis (1998) 31:841–847.[Web of Science][Medline]
- Mermel LA. Prevention of intravascular catheter-related infections. Ann Intern Med (2000) 132:391–402.
[Abstract/Free Full Text] - Oliver MJ, Callery SM, Thorpe KE, et al. Risk of bacteremia from temporary hemodialysis catheters by site of insertion and duration of use: a prospective study. Kidney Int (2000) 58:2543–2545.[CrossRef][Web of Science][Medline]
- 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.
[Abstract/Free Full Text] - Lund GB, Trerotola SO, Scheel PFJ, et al. Outcome of tunneled hemodialysis catheters placed by radiologists. Radiology (1996) 198:467–472.
[Abstract/Free Full Text] - Astor BC, Eustace JA, Powe NR, et al. Timing of nephrologist referral and arteriovenous access use: the choice study. Am J Kidney Dis (2001) 38:494–501.[Web of Science][Medline]
- Weijmer MC, ter Wee PM. Temporary vascular access for hemodialysis treatment. Current guidelines and future directions. Contrib Nephrol (2004) 142:94–111.[Web of Science][Medline]
- Sesso R, Barbosa D, Leme IL, et al. Staphylococcus aureus prophylaxis in hemodialysis patients using central venous catheter: effect of mupirocin ointment. J Am Soc Nephrol (1998) 9:1085–1092.[Abstract]
- Vanherweghem JL, Dhaene M, Goldman M, et al. Infections associated with subclavian dialysis catheters: the key role of nurse training. Nephron (1986) 42:116–119.[Web of Science][Medline]
Accepted in revised form: 1.10.07
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||