| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Vascular and Interventional Radiology |
1 From the Departments of Radiology (W.W., M.A.B., B.B., D.R.L.) and Medicine, Section of Nephrology (B.R., M.C.), Dartmouth-Hitchcock Medical Center, One Medical Center Dr, Lebanon, NH 03756. Received November 15, 1999; revision requested December 28; revision received May 11, 2001; accepted June 5. Address correspondence to M.A.B. (e-mail: michael.bettmann@hitchcock.org).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: The authors retrospectively reviewed the medical records, hemodialysis records, and clinical information system data from 132 consecutive patients who were referred for placement of a tunneled catheter for hemodialysis access. A commercially available twin catheter system was placed in all patients. Outcomes evaluated included infection rate, complication rate, and catheter malfunction and failure rates. Performance parameters evaluated included blood flow rates, urea reduction percentages, and recirculation percentages.
RESULTS: One hundred eighty-four twin catheter systems were placed in 132 patients from January 11, 1996, to October 23, 1997. The initial technical success rate was 100%. There were four immediate procedural complications: Air emboli occurred in two patients, and prolonged bleeding necessitating intervention occurred in two. The total number of days a catheter was in place was 13,200 (mean, 74.6 days). Thirty-one infections occurred in 20 patients (total infection rate, 0.23 episodes per 100 catheter days). Sixty-five catheters malfunctioned during the study period, 19 of which necessitated removal, for a rate of 0.14 episodes per 100 catheter days. The average blood flow rate was 281.4 mL/min (range, 117.1–405.6 mL/min; median, 295.2 mL/min). Mean and median urea reductions were both 61%. Mean and median recirculation was 6.1% and 3.5%, respectively (range, 0%–31%).
CONCLUSION: Percutaneous placement of the tunneled twin catheter system can be performed with excellent technical success and safety and acceptable catheter performance and outcomes for effective intermediate- to long-term hemodialysis.
Index terms: Catheters and catheterization, central venous access, 566.129, 569.129, 907.1269, 946.1269 • Catheters and catheterization, complications, 566.458, 569.458, 907.458, 946.458 • Dialysis
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Catheter Insertion
The twin catheter system available during the study period consisted of two 10-F, 40-cm silicone catheters, each with six side-holes in a spiral configuration over the distal 4 cm. Each catheter was color coded (blue or red), had a luer lock configuration, and had a Dacron subcutaneous cuff fixed 22 cm from the intravascular tip. All catheters were inserted by interventional radiology staff or fellows. Before catheter placement, the platelet count and prothrombin time were obtained. The prothrombin time was considered acceptable if the international normalized ratio was 1.7 or less, and an acceptable platelet count was greater than 50,000 x 103/L. Prophylactic antibiotics were not routinely given.
All patients routinely underwent ultrasonography (US) of the internal jugular vein, the preferred site of placement, in the angiography suite to confirm the vein’s patency and location. Then, the catheter was placed with fluoroscopic and US guidance. All operators wore caps, masks, gloves, and gowns but did not perform a surgical scrub. The skin site was scrubbed well and draped in a sterile fashion. A standard angiography suite was used for all procedures. The internal jugular vein was punctured with a 21-gauge micropuncture needle (Cook, Bloomington, Ind). The coaxial introducer set with the micropuncture needle was then used to advance an 0.018-inch guide wire, followed by a 4-F dilator, to the superior vena cava. An 0.038-inch guide wire was then placed through the dilator and the dilator was removed. A second puncture of the internal jugular vein was made 1 cm inferior to the first, followed by placement of a second guide wire with the same technique. Two subcutaneous tunnels were made over the anterior chest wall by using a custom-made angled stainless steel tunneling device. The tunnels were first mapped by using fluoroscopy with the catheters lying on the patient’s chest so that both intravascular tips would lie between the mid-right atrium and the low superior vena cava, and the Dacron cuffs would be 1 cm from the skin exit sites on the anterior chest wall. Once both catheters were pulled through their respective tunnels, a 12-F peel-away sheath was placed over each 0.038-inch guide wire. The intravascular portion of each catheter was then placed in the peel-away sheath. Kinks were removed by gently tugging on the catheters. Fluoroscopy was used to position the intravascular tips in optimal position, with the blue venous catheter placed more centrally in the high right atrium and the red arterial catheter placed 4 cm proximal in the low superior vena cava. Once the position was confirmed, each catheter was filled with a mixture of urokinase (Abbokinase; Abbott Laboratories, North Chicago, Ill) and heparin, as directed by the manufacturer. Sutures were then placed at the venotomy sites as needed. We defined technical success as the ability to place the catheters.
To prevent the introduction of air emboli while inserting the catheters into the 12-F peel-away sheath, several maneuvers were used. Because our angiographic tables cannot be placed in the Trendelenburg position, a technologist raised the patient’s legs. During catheter insertion, we pinched the peel-away sheath while instructing the patient to hum to increase intrathoracic pressure. The use of a smaller, nontunneled catheter may lessen the likelihood of air embolism in certain patients, such as those who are uncooperative, those with tachypnea, and those with a very limited pulmonary reserve.
Catheter Care
Catheter evaluation and care were performed with an aseptic technique by dialysis nurses before each dialysis session. Patients suspected of having exit site and tunnel infections had cultures taken, and treatment with appropriate antibiotics was begun. If results of repeat cultures remained positive, the catheter(s) was either removed or exchanged over a stiff hydrophilic guide wire (Terumo wire; BSC Vascular, Natick, Mass), depending on the individual clinical scenario. Bacteremia related to the indwelling catheters was treated in a similar fashion, either with prolonged antibiotic coverage alone or, if persistent or symptomatic, with subsequent catheter removal or exchange. At the conclusion of each dialysis session, the catheters were dressed with nonadhesive dressing and povidone-iodine ointment and instilled with 5,000 U of heparin in an appropriate volume (1.8 mL per catheter).
Catheter malfunction (inability to maintain blood flow rates of at least 200 mL/min) or occlusion was treated initially with 5,000 U of urokinase in a total volume of 1.8 mL of normal saline, and subsequently with 1 mg of recombinant tissue plasminogen activator (Alteplase; Genentech, South San Francisco, Calif). If after 1 hour of treatment the catheter still did not function, an additional 5,000 U of urokinase or 1 mg of recombinant tissue plasminogen activator was instilled into the catheter lumen. If after 2 hours of therapy the catheter still did not function, the patient was sent to the interventional radiology suite for evaluation. During the initial part of the study, we routinely performed catheter stripping via a common femoral vein approach with the presumption that a fibrin sheath was responsible for the malfunction. We subsequently abandoned this approach and adopted the technique of exchanging the malfunctioning catheters over a stiff hydrophilic guide wire (Terumo wire) through the preexisting tunnel. Such catheter exchange avoids the risks and inconveniences of transfemoral catheterization and decreases the postprocedural observation time in the cardiovascular and interventional radiology recovery area (7).
Statistical Analysis
All catheter survival curves were generated by using the Kaplan-Meier method with a statistical software application (STATA, College Station, Tex). Survival curve differences were tested with the log-rank test.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The catheters were in place a total of 13,200 days during the study period (mean, 74.6 days; range, 1–541 days; median, 48 days). Indications for catheter placement were multiple and are listed in Table 1. At final follow-up, 10 (5%) catheters were in place and functioning, 32 (17%) were functioning when patients died, and seven (4%) were functioning but were in patients who were lost to follow-up. Catheters had been removed in 83 patients for various reasons (although they were subsequently replaced in some). Overall, 94 (51%) catheter sets were removed in relation to resolution of the need for these catheters for dialysis (Table 2). In addition, 22 (12%) catheter sets were removed secondary to infection and 19 (10%) were removed because of malfunction or dislodgment (Table 2).
|
|
|
|
Thirty-one infections were noted in 20 patients during the study period, for a total infection rate of 0.23 episodes per 100 catheter days. Twenty-two of these infections (18 episodes of bacteremia and four exit site infections) necessitated catheter removal in 15 patients, for a total infection rate necessitating catheter removal of 0.17 episodes per 100 catheter days. The bacteremia rate was 0.14 episodes per 100 catheter days. Of the 22 infections that necessitated catheter removal, the mean time until infection developed was 60.5 days (range, 3–266 days). Three of these infections occurred in the 1st week after placement. Of the nine infections that did not necessitate catheter removal, the mean time until infection developed was 72.3 days (range, 11–187 days). Three patients had multiple catheters removed owing to infection. Two patients had an episode of bacteremia and an exit site infection that necessitated catheter removal. One patient had two episodes of bacteremia that necessitated catheter removal.
Overall, there were 151 episodes of catheter malfunction (1.14 episodes per 100 catheter days) that necessitated intervention but not catheter removal. These episodes involved 63 (34%) of the 184 catheters placed. Interventions for the malfunctioning catheters included 129 urokinase infusions at dialysis, 15 catheter strippings, five guide wire manipulations, one catheter repositioning, and one urokinase infusion in the angiography suite. Of the 129 urokinase treatments administered at dialysis, 34 (26%) were performed in 30 catheters within the first 24 hours after catheter placement. We eventually recognized that, for reasons that are not apparent, the Tesio catheters do not provide blood flow rates adequate for dialysis until at least 24 hours after placement, a problem that is noted on the package insert. If those catheter malfunctions that necessitated urokinase infusion within 24 hours of catheter placement are excluded from analysis (probably a realistic exclusion, on the basis of our early recognition of this limitation), then the total number of catheter malfunctions necessitating intervention was 117, or 0.88 episodes per 100 catheter days.
Many catheters had multiple episodes of malfunction (inability to maintain blood flow rates greater than 200 mL/min). The episodes of malfunction per catheter are listed in Table 3. Of the 63 catheters necessitating interventional procedures to maintain patency, 13 (21%) accounted for 77 of the 151 interventions (51%). If urokinase infusions within the first 24 hours after catheter placement are eliminated from consideration (34 interventions), then these 13 catheters (12 patients) account for 66% of the interventions needed to maintain patency.
|
The recorded blood flow rates, although not as high as initially expected, were certainly adequate to achieve effective dialysis. The mean and median blood flow rates were 281.4 mL/min and 295.2 mL/min, respectively (range, 117.1–405.6 mL/min). Both the mean and median urea reduction percentages were 61%. The mean and median recirculation percentages, calculated by using the two-needle urea-based method, were 6.1% and 3.5%, respectively (range, 0%–31%).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Many tunneled catheter systems have been developed to achieve optimal longevity, low complication rates, and adequate blood flow rates. A two-catheter system for central venous catheterization was first described by Canaud et al (9) in 1986. Tesio et al (5) in 1994 and Milner et al (6) in 1995 described their clinical experience with the catheter system used in the current study. Their initial encouraging results, coupled with suboptimal results with the available dual-lumen hemodialysis catheters, prompted our study.
Because a two-catheter system requires two cannulations of the internal jugular vein and the creation of two subcutaneous tunnels, one might expect the technical success rate to be lower and the complication rate to be higher than that with a conventional dual-lumen hemodialysis catheter. Our technical success rate, however, was 100%, and the complication rate was only 2.2% (two air emboli and two episodes of bleeding necessitating intervention). There were no episodes of hemothorax, pneumothorax, or hemomediastinum. Our complication rate is in the same range as that of other published reports for hemodialysis catheters placed by interventional radiologists (0.1%–3.6%) (7,10,11). The low figure (0.1%) is from a study of 250 patients (11), all with right internal jugular vein placement, and included one Tesio catheter system. The high figure (3.6%) is from a study in which only the Tesio twin catheter system was used (10). We attribute our good technical success and low complication rate to our use of US guidance to establish venous access. The one death in our study on the day after catheter placement was related to an air embolus during catheter placement. The patient, however, was clinically unstable at catheter placement, and the prudent course of action might have been to delay placement or use a smaller, nontunneled catheter.
Our technical success and complication rates compare favorably with those in published surgical series, where the overall procedural complication rates are as high as 5.9%, and technical success is often not indicated (7,12,13). As noted by Trerotola et al (7), surgical complications have included pneumothorax (0%–1.8%), hemothorax (0%–0.6%), hemomediastinum (0%–1.2%), recurrent laryngeal nerve palsy (0%–1.6%), and bleeding necessitating repeat exploration and/or transfusion (0%–4.7%). In addition, surgical placement usually relies on a cutdown for insertion, and this is associated with a high incidence of occlusion of the internal jugular vein. Agraharkar et al (14) found a 33% thrombosis rate in right internal jugular veins cannulated with surgical cutdown; the thrombosis rate with percutaneous insertion was only 2%.
According to Trerotola et al (7), catheter infection and failure due to thrombosis remain the most important drawbacks to catheter hemodialysis. Our total infection rate of 0.23 episodes per 100 catheter days, rate of infections necessitating catheter removal of 0.17 episodes per 100 catheter days, and bacteremia rate of 0.14 episodes per 100 catheter days compare favorably with those of other published reports for hemodialysis catheters placed by interventional radiologists (Table 4), although most other reports (7,11,15) deal solely or primarily with single catheter systems.
|
During the study period, we routinely removed all catheters in patients with two positive blood cultures and then placed new catheters at two different access sites following several days of antibiotics. More recently, on the basis of results by Carlisle et al (16) and Schaffer (17), as well as our desire to preserve venous access sites, we have begun reviewing episodes of bacteremia on a case-by-case basis. Now in cases of positive blood cultures (except for certain gram-negative infections, at the discretion of attending nephrologists), the venous access site is preserved by guide wire exchange of the catheters, and a new subcutaneous tunnel is created. Frank tunnel infections, however, are still treated with catheter removal and catheter insertion at a different access site (usually the contralateral internal jugular vein) with creation of a new subcutaneous tunnel.
Catheter malfunction, like infection, remains a persistent problem with percutaneously placed hemodialysis catheters. Our overall probability of catheter survival of 66% at 6 months and 45% at 1 year compares favorably with those in recent reports by interventional radiologists: 44%–62% at 6 months and 25%–48% at 1 year (7,11).
The frequency of catheter malfunction in our study (1.14 episodes per 100 catheter days) was higher than that reported in other interventional series (0.16 and 0.81, Table 4). This is likely owing to the failure, recognized by the manufacturer but of unknown cause, of the Tesio twin catheter system to maintain high blood flow rates in the 24 hours immediately following placement. It is not clear why Prahbu et al (10) did not encounter similar difficulties in their use of the same catheter system. After recognizing this limitation, we attempted to delay initial dialysis for 24 hours after percutaneous placement. If catheter malfunctions within 24 hours of placement are excluded, our overall catheter malfunction rate becomes 0.88 episodes per 100 catheter days. This rate is comparable to the only other reported overall catheter malfunction rate in a large interventional radiology series (0.81 episodes per 100 catheter days) (11).
The subset of patients who chronically rely on percutaneous hemodialysis catheters for their sole dialysis access accounted for a large percentage of the interventions needed to maintain catheter patency. If we exclude interventional procedures performed within 24 hours of catheter placement, 13 catheters in 12 patients accounted for 66% of the interventions needed to maintain patency. All of these patients had undergone multiple failed attempts at surgical placement of dialysis access fistulae. It is likely that the larger the number of patients who rely solely on percutaneously placed hemodialysis catheters, the higher the catheter malfunction rate. It is also likely that, because of body habitus, personal habits, or comorbid conditions, access of all kinds will be difficult to maintain in certain patients.
The overall frequency of catheter malfunctions necessitating removal, 0.14 episodes per 100 catheter days, compares favorably with those reported in other interventional radiology series (Table 4). Only 19 (10%) of the 184 catheters placed were removed because of malfunction. Thirteen (7%) of the 184 catheters were removed because of consistently poor blood flow rates not responsive to urokinase infusion at dialysis, five (3%) were dislodged, and one (1%) had eroded through its subcutaneous tunnel. It is difficult to evaluate the true thrombosis rate as distinct from other causes of malfunction. If a patient was referred to us with repeatedly poor blood flow rates not responsive to urokinase infusion in the dialysis unit, we routinely exchanged the catheter(s) over a stiff guide wire by using the same subcutaneous tunnel rather than attempt high dose urokinase infusion or fibrin sheath stripping. We abandoned fibrin sheath stripping early in our experience with this catheter system because of associated costs, risks, and inconveniences of transfemoral catheterization and because, empirically, this approach rarely improved catheter function in the long term (7,18).
Blood flow rates, although not as high as initially expected and lower than reported in other interventional radiology series (10,15,19), were clearly adequate to maintain efficient long-term dialysis (Table 4). Our rates were nearly identical to those reported by Tesio et al (5) (284 mL/min at 1 year). These flow rates were substantially lower than those reported by Prahbu et al (10), perhaps because we calculated actual flow over time instead of using pump flow settings of Doppler integration. Flow rates may differ substantially if calculated from pump settings, from single or multiple US assessments, or by dividing the total liters processed by duration of dialysis.
Our mean blood flow rate of 281.4 mL/min was also less than the 300 mL/min that the Dialysis Outcomes Quality Initiative work group considered to be sufficient extracorporeal blood flow for hemodialysis catheters (1). This recommendation is based on opinion, rather than on data, presumably because there are few recent published reports in which the blood flow rates in tunneled hemodialysis catheters were evaluated. It is clear that additional studies about this topic, concerning both flow rates and long-term outcomes, are warranted.
Our mean and median recirculation percentages are similar to those previously reported with the same Tesio catheter system (10,19). Despite the paucity of other published data, our recirculation percentages are clearly acceptable and less than the recommended limit of 10% for the two-needle urea-based method suggested by the Dialysis Outcomes Quality Initiative work group (1). Both our nephrologists and dialysis personnel were satisfied with our mean and median urea reduction rate of 61%. This ratio is dependent on multiple variables, and the relative paucity of reported data about urea reduction with tunneled hemodialysis catheters makes comparison to other studies difficult.
In conclusion, overall outcomes were superior with right, versus left, internal jugular vein placement. A disproportionate number of catheter failures occurred in a small number of patients, specifically those with multiple failed access procedures at other sites. Although we recognize that hemodialysis catheters are currently not intended to replace surgically placed fistulas and grafts, the steady performance of the Tesio catheter makes it a suitable alternative for prolonged, efficient dialysis. These data, which address infection rate, malfunction rate, catheter survival, and flow, may serve as a basis for comparison with other tunneled dialysis access catheter systems.
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. Wang, J. M. LaBerge, G. M. Chertow, R. K. Kerlan, M. W. Wilson, and R. L. Gordon Tesio Catheter Access for Long-term Maintenance Hemodialysis Radiology, October 1, 2006; 241(1): 284 - 290. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. J. Rosa-Diez, R. G. Lambertini, O. Peralta, L. H. Diodato, S. L. Algranati, and R. D. Garcia-Monaco Transhepatic venous access as an alternative for Tesio catheter in the case of a patient on haemodialysis with antiphospholipid syndrome Nephrol. Dial. Transplant., July 1, 2006; 21(7): 2031 - 2033. [Full Text] [PDF]
|
|
|
|
|
|
V. Mickley Central venous catheters: many questions, few answers Nephrol. Dial. Transplant., August 1, 2002; 17(8): 1368 - 1373. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
