Management and prevention of complications of subcutaneous intravenous infusion port
Article Outline
- Abstract
- Introduction
- Materials and methods
- Results
- Discussion
- Conclusion
- Conflict of Interest Statement
- Authorship
- References
- Copyright
Abstract
Subcutaneous intravenous infusion port (SIIP) has become an increasingly and widely adopted technique in the management of oncology patients. This route has been used not only for chemotherapy but also for parenteral nutrition provision, blood transfusion, medication administration, blood sample collection, hemodialysis, and so on. This system provides a safe vascular access with low complication rate which helps preventing patients from vascular infection and catheter associated thrombosis. In this study, we reviewed 1247 cases of breast cancer patients that had subcutaneous intravenous infusion port implanted for chemotherapy in our general surgery department from 1990 to 2008. The result indicates that complication decreases as our technique and experience mature. We hereby share our accrued experience and improved technique, hoping to be of help to young surgeons.
Keyword: Subcutaneous intravenous infusion port
Introduction
Requirements for long-term intravenous access are common among cancer patients. In several previous studies, the complication rate of implanted port devices varied from 6% to 21% [1], [2], [3], [5], [7], [11], [12], [13]. Most of the studies were limited by the small number of samples and the relatively short follow-up interval. In our study, the low complication rate was accrued from a data base of over 1000 patients, documented since 1990. With a bigger sampling size, more accurate conclusions can be drawn.
Materials and methods
This study examines 1247 breast cancer patients from our data base between January 1990 and December 2008. In total, 1247 venous ports were placed in 6 men and 1241 women aged between 20 and 103 years with a mean age of 49.2. Among all the 1341 cases of breast cancer treated within 18 years, the implantation rate of SIIP was 93%. There were 670 cases collected during 1990–2001, and 487 cases collected during 2002–2008. All implantation procedures were performed by experienced general surgeons, not trainees. Follow-up data base was obtained through clinical records, personal contacts, chart reviews and telephone contacts. All patients were followed up until SIIP was removed at the end of the treatment or the death of the patients.
All of the procedures, which were performed in the operation room in aseptic environment, can be done in either one of the following four methods: (1) under general anesthesia at the time of mastectomy; (2) with the open method with direct venous cut down; (3) under local anesthesia by percutaneous insertion; or (4) image guide percutaneous insertion.
At the time of mastectomy during axillary lymph node dissection, a branch of the axillary vein near the chest wall (lateral thoracic vein) was identified and isolated, prepared for catheter insertion later. If this vein is too small as in some rare cases, then, as a second choice, after completion of the axillary dissection, the thoracodorsal vein lying on the subscapularis muscle can be dissected from its accompanying artery and all the important thoracodorsal nerves. This vein is always big enough to allow insertion of the catheter into the axillary-subclavian vein and then into the superior vena cava. The level of SVC can be grossly estimated by palpating the sternal notch, then the manubrio-sternal angle. Placing the catheter from this angle to the point of desired axillary vein entrance, adding the depth of the pectoralis major muscle, the desired length of the catheter can be estimated and marked with a silk tie, usually about 20 ± 1 cm. This gross estimation is accurate enough in our hands to obviate the need for X-ray imaging to ascertain whether or not the catheter tip is in the SVC. Then the catheter tip is pulled through the muscle just medial to the junction of the axillary vein with the prepared vein branch, and inserted into the vein to the pre-marked level (about 20 cm mark). The vein is tied around the catheter with an absorbable suture and left long to be later sutured to the under surface of the muscle. Note that this suture should not be silk because we have seen excessive fibrotic reaction around it, causing difficulty in catheter removal later. After the spot suitable for the port is chosen, usually between 2 ribs anteriorly, the catheter is cut at the appropriate length and securely connected to the port. The port is aspirated to check for blood return, then flushed with 10 ml of Heparin 1:100 diluted solution to prevent clotting. The marking tie is removed from the catheter, and the catheter is fixed to the under surface of the pectoralis major muscle as above mentioned. The port base is then fixed to the anterior surface of the muscle with 2 non-absorbable sutures. The wound is irrigated with antibiotic containing solution.
For the open venous cut-down method, the left cephalic vein was initially preferred in all patients. Skin at the insertion site was prepared with alcoholic beta-iodine from the nipple up to the mandible cranially and from the sternum to the mid-axillary line laterally. The skin was incised 2–3 cm to expose the underlying subcutaneous tissue. With the proper dissection of the tissue, the cephalic vein was carefully identified and suspended with silk. An incision was made on the vessel proximal to the suspended silk, then a 90° PORT-A-CATH needle was inserted. The catheter was inserted beneath the groove of the 90° PORT-A-CATH needle passing the atrio-caval junction to the level of right atrium. Next, the catheter was tied to the vessel with absorbable suture to prevent bleeding, air embolism or tissue fibrosis.
Percutaneously landmark-base insertion technique guiding the localization of a No. 25 spinal needle was done first, and then the puncture needle was carefully inserted along the lower border of middle third of the clavicle. Then the needle was pushed toward about 2–3 cm above manubrium notch and 10 to 15° angle beneath the clavicle until it reached the subclavian vein. A guide wire was advanced to the subclavian vein through the puncture needle into the superior vena cava, and further through the right atrium into the inferior vana cava to assure venous puncture (see Fig. 1). The puncture tract was dilated and a peel away sheath was inserted. The catheter was introduced into the superior vana cava through the peel away sheath. Catheter tip placement was attempted at the level of the atrio-caval junction, average about 20 cm in length, which was controlled by intra-operative chest X-ray film or fluoroscopy (see Fig. 2). A 2 to 3 cm skin incision at about 2 cm below puncture site was made. A subcutaneous reservoir pocket was created by dissecting the Scarpa fascia until it became large enough to bury the pocket reservoir. Once the pocket was created, the catheter was tunneled from the vein access site to the port and was trimmed to proper length. Then it was connected to the port and the port was secured with a suture to the pectoralis major muscle to prevent rotation of the port. By using winged infusion set™ (Bard Access System), the port was checked by aspiration of blood and then flushed with 100 u/ml of heparin solution. The incision was carefully closed with absorbable 4-0 coated vicryl interrupted inverted dermal sutures. 3M steri-strips were applied to keep the edges of the incision smooth. A chest X-ray was routinely performed at the end of the percutaneous insertion procedure. Then the patient was discharged home. Before starting chemotherapy, all patients had routine chest film to ensure the position of the catheter is acceptable for safety during chemotherapy.
Figure 1
The guide wire was advanced to the left subclavian vein under the guidance of surgical landmark. I: manubrium notch. II: coracoid process.
Figure 2
The subclavian venous catheter was inserted at the middle part of clavicle over the delto-pectoral groove. I: manubrium notch. II: coracoid process.
For image guide percutaneous insertion method, we use the SonoSite’s portable duplex ultrasound 180 plus to screen the patency and size of both jugular veins and subclavian veins to rule out partial or complete thrombosis. Internal jugular veins were our standard catheter access site. Venipuncture was performed under ultrasound guiding and catheter was inserted through puncture needle. Fluoroscopy was used to check the route and tip of catheter. A skin incision about 2 cm below clavicle was made and a subcutaneous reservoir pocket was created. A long subcutaneous tunnel crossing the clavicle was created. The catheter was tunneled from the vein access site to the pocket wound. Fluoroscopy was used to check the position and catheter tip again, and catheter was trimmed to proper length. Then, it was connected to the port. The wound was closed with 4-0 vicryl plus inverted suture.
Port-catheter irrigation with 1:100 diluted heparin solution was performed every 3 months after chemotherapy to prevent catheter occlusion. Routine removal of SIIP was mostly 2 to 3 years after implantation, in those who had implantation during the period from January 1990 to December 2000. However, to decrease worrying about possible complications, we changed our practice policy, so most patients who had implantation after January 2001 to December 2008 had their SIIP removed soon after their annual breast cancer follow up, very few needed reinsertion for another chemotherapy course for recurrent disease.
Results
Between January 1990 and December 2008, 1247 venous ports were carried out on adult patients with breast malignancies. Two cases of pneumothorax were caused by the percutaneous insertion method. One of the patients had chest tube inserted immediately after the procedure whereas another patient received conservative treatment including analgesics and oxygen inhalation. The most common complication during 1990–2001 was extravasation. However, between 2002 and 2008, the most common complication was catheter-related infection. Breakage of the catheter had markedly decreased from 0.79% to 0.20%. No venous thrombosis or migration of the port occurred during this period. The total complication rate had markedly decreased from 3.42% to 1.23%. Of the total 1247 implantations, no severe systemic infection was seen. Minor hematoma formation was observed in a few patients (data not included).
| Delayed Complications of Subcutaneous Intravenous Infusion port (SIIP) | ||||
|---|---|---|---|---|
| From January 1990 to December 2001, 760 cases | From January 2002 to December 2008, 487 cases | |||
| Extravasation | 7 | 0.92% | 1 | 0.20% |
| Catheter breakage/separation from port | 6 | 0.79% | 1 | 0.20% |
| Symptomatic venous thrombosis | 5 | 0.66% | 0 | 0% |
| Catheter occlusion | 2 | 0.26% | 1 | 0.20% |
| Catheter-related infection | 2 | 0.26% | 2 | 0.41% |
| Port migration | 2 | 0.26% | 0 | 0% |
| Difficulty at removal of the catheter | 2 | 0.26% | 1 | 0.20% |
| total | 26 | 3.42% | 6 | 1.23% |
Discussion
In this study, infectious complications account for the majority of device-related morbidity from 2002 to 2008. On the other hand, some studies by other groups show that catheter thrombosis is the most common complication. Each time after chemotherapy, we use 100 u/ml heparin solution 10 ml to flush the port and catheter to prevent catheter occlusion. Once catheter occlusion happened, removing it and replacing it with a new one is the policy. Symptomatic venous thrombosis is a rare complication now. Catheter tip position above the T3 level, especially with left subclavian vein cannulation, has been implicated as a major risk factor for thrombosis [9]. Catheter mechanical irritation and erosion of vessel wall may be the cause of subclavian vein and superior vena cava thrombosis. Another study indicated that a malpositioned catheter tip located outside the vana cava or right atrium may increase the rate of thrombosis [10], [14], [23], [24]. In our study, the low incidence of thrombosis might be attributed to the mature technique of surgeons, who could place the catheter perfectly even without the aid of X-ray guidance. Besides manual factors, improvement of catheter materials also contributed to the low thrombosis rate. Asymptomatic venous thrombosis may be much higher than what we can see. Now we use infusion port side upper limb venography and venous Doppler ultrasound before and after chemotherapy and six months later to evaluate the catheter-induced venous thrombosis. We hope to get detailed data in the near future. If we found patient’s arm swelling, we should distinguish lymphedema from venous thrombosis. Engorged upper arm veins or bluish discoloration of skin indicated venous thrombosis. Once venography or ultrasound detected venous thrombosis, at least six months of antithrombotic therapy should be given. The lower rate of symptomatic thrombotic complications in our series is attributed to the relatively lower platelet aggregation function in our race, correct position of the catheter tip, and the frequent port infusion with heparin solution.
In this study, one of the patients who had extravasation was a polio victim. The use of crutch compressed the axilla and distorted the pectoralis muscle tendon. With the traction force on the connection site of the port, the catheter was dislodged from the reservoir and caused extravasation. The resolution for this condition was choosing another place that will cause no compression effect such as internal jugular vein of the neck or contralateral side of crutch use. Other studies point out that separation of the catheter and the port was caused by forced flushing of the port, so gentle flushing of the port is necessary. Besides the chest film taken to check the integrity and position of the port-catheter system, blood-withdraw test and/or normal saline dripping via the port for 10 min should routinely be done before infusion of anti-neoplastic regimens. Once leakage was noted such as swelling of the infusion site or reporting of severe pain, chemotherapy should be stopped immediately. If massive chemotherapy drug was noted to be leaking in the chest wall, immediate open wound irrigation to reduce drug induced chemical burn must be done and broad spectrum antibodies should be given for infection prevention.
Normally, the port base is secured to the p. major muscle surface with 1–2 non-absorbable sutures. In 2 incidences, the port migrated to a much lower position. Perhaps the elastic bandage around the chest over the port was the culprit. We no longer cover the port with the elastic bandage, nor use absorbable suture to fix the port to the muscle.
As a safety measure, routine chest X-ray was taken before removal of the port-catheter system. If the catheter was noticed to be broken or disconnected from the port, immediate operation was undertaken to remove the catheter before it can migrate away. In cases where the broken catheter dropped into the subclavian vein or superior vena cava or heart chamber, cardiovascular consult should be immediately available, and the broken catheter can be removed via cardiac-catheterization technique. In one incidence, the broken catheter migrated into the pulmonary artery and could not be retrieved (see Figure 3, Figure 4).
Figure 3
The pocket was created and the catheter was tunneled from the vein access site to the port and was trimmed to proper length. I: manubrium notch. II: coracoid process.
Figure 4
Proper position of the port-A catheter in order to prevent pinch-off syndrome.
Catheter breakage or separation from the port occurred more in our series before 2002 (see Figure 5, Figure 6, Figure 7). Another theory was the phenomenon of pinch-off syndrome, which was caused by pinching the catheter between the clavicle and the first rib, leading to catheter fracture [6]. In our study only one case whose catheter breakage seemed to correlate to this phenomenon. To prevent pinch-off syndrome from happening, the tip of the puncture needle should be slightly moved down vertically so that it could punch the vessel more laterally to the intersection of these two bones (see Fig. 4). This procedure should be done more carefully to prevent pneumothorax. Another way to prevent this condition was to insert the catheter 1 cm lateral to the common insertion site. For patients with heavy use of upper limbs, we should avoid insertion of catheter to the subclavian vein, but into the internal jugular vein instead. At first, suspicion was raised about the quality of the catheter. So we promptly notified the manufacturer. Their microscopic inspection of the broken ends was reported as not likely due to wear-and-tear of the catheter, but rather a clean-break. But they have subsequently made improvements in their catheter material, and we have also improved our technique, such as not fixing the catheter to the upper surface of the muscle with sulk sutures, avoiding any fulcrum of shear during patients arm movement.
Figure 5
Migration of the breakage catheter through the heart chamber to the right pulmonary artery which can’t be removed.
Figure 6
Breakage of the catheter between the clavicle and the second rib due to pinch-off syndrome. Migration of the breakage catheter to the right ventricle which was removed by the cardiologist.
Figure 7
The broken fragment of the catheter and the port.
The infection of the port may have come from several ways. One might be the contamination of the equipment or the surgical devices during the insertion procedure. But this condition seldom happened in the hospital setting because of modern sterilization technique and well-trained staff. Another way of infection was at the time of chemotherapy. The insertion of the needle into the port might take some skin microbes to the port. Tenderness is a prominent sign of the port infection. If the skin of the infection site turns red and becomes swollen, the port site soft tissue infection should be highly suspected. Immediate wound care with antiseptic agent and intravenous antibiotic treatment should be done to eradicate the infection. For more severe infection, even with abscess formation, incision and drainage might be needed in addition to the antibiotic treatment. If port site infection persisted or the port aspiration blood culture revealed bacterial infection, the port should be removed as soon as possible. In our hospital, the infection rate was low because insertion site of the needle to the port was always well prepared by well-experienced nurses.
There were three patients who had difficult removal of the catheter due to strong tissue reaction to the silk sutures. One of these three patients even needed the cardiovascular surgeon for axillary vein repair. Tearing of the vessel was caused by forceful pulling of the catheter. Avoidance of using silk for this suture is the best way to prevent this from happening. Once difficulty in removal of the catheter is encountered, careful dissection of the tissue to avoid forceful pulling of the catheter can reduce the chance of this disaster.
In this study, the complication rate had markedly decreased in the latter period. The key point of this phenomenon may be attributed to early removal of the device, properly selected insertion technique, accurate catheter tip position and well-trained surgeons [15], [16], [19]. Previously, the device would be left in a patient’s body for years; however, in the latter part of this study, the device was removed in 6–12 months after the treatment had been completed. Even totally embedded, the device is still a foreign body to the human body. The longer it stays in the body, the more complications would appear.
There were several studies about image-guided port placement [4], [8], [17], [18], [20], [21], [22]. According to the papers published, the success rate almost reaches 100%. The advantage of the image-guided port placement counts on clear localization of the vessel, visible catheter tip placement, and decreasing possible injury to the other part of the body.
Recently, we used SonoSite’s portable duplex ultrasound 180 plus to screen the patency and size of both jugular veins and rule out partial or complete thrombosis. Internal jugular veins were our standard catheter access site, Ultrasound guidance have relatively less advantage in subclavian access because of deep position and bony structure. Ultrasound is used to guide venipuncture and fluoroscopy, which enables visualization of both the course and the tip position of the catheter. Image guidance virtually eliminates the risk of several complications with unguided placement such as pneumothorax, hemothorax, hematoma attributed to arterial puncture, pericardial tamponade, air embolism, chylothorax, hydrothorax, nerve injury, arrthythmia, and catheter malposition. In addition, image-guided venipuncture catheterization can be placed safely in patients who have undergone ipsilateral axillary lymph node dissection for breast cancer. Ultrasound may help to locate the vein and decrease some complications, but it will not decrease the technical difficulties such as long subclavian tunnel for the catheter, malpositioning of guide wire or catheter.
We have not used fluoroscopy X-ray for guidance after the initial period of learning because fluoroscopy does not help the implanting surgeon find the vein. Before chemotherapy, a PA chest X-ray was routinely taken to check the position of the SIIP. Any serious malposition can be corrected before infusion. Occasionally, the catheter may puncture cephalad in the jugular vein, which requires reimplantation immediately. In our hospital, there was one colon cancer patient, who was not included in this study, received port infusion chemotherapy with catheter tip in the internal jugular vein. She developed cerebral infarct after three times of chemotherapy. Anatomically, the right subclavian vein makes a more acute angle than the left subclavian vein when it joins the internal jugular vein to form brachycephalic vein, and thus the guide wire preferentially may go up to the right internal jugular vein. In addition, introducer sheaths can kink secondary to this angulation and prevent the placement of the catheter. This phenomenon may be decreased by the left subclavian vein catheterizations, but at that time, the thoracic duct can be at a high risk for injury by making use of the left side venous puncture [20], [21].
Mature technique plus careful implantation was the key to success. The operators in our hospital were all well-experienced general surgeons. That’s why our success rate almost reached 100% too. For a well-experienced surgeon, the landmark-base insertion technique for the subclavian vein is a simple and easy method. The average operation time is estimated to be around 20 min. The advantage of the open cut-down method and the intra-operative method during mastectomy is accurate identification of the vessel. In both methods, the catheter is inserted with direct visual control into the desired position. Intra-operative method is a one-step surgery that has best cost-effectiveness for the patient and less surgical requirements for the whole management process. However, if patients who have advanced breast cancer or metastatic breast cancer will receive neo-adjuvent or palliative chemotherapy, the percutaneous method would be more suitable. The percutaneous method is convenient and quite safe for a well-trained doctor to perform in an operation room, best under brief intravenous anesthesia. The fluoroscopy was used to confirm the correct position of the catheter tip at atrio-caval junction, a well-known condition being able to minimize the thrombosis rate. Upright chest X-ray was taken post-op to rule out pneumothorax.
The disadvantage of the surgical cut-down method is the need for a large and deep wound to expose the vessel. This procedure costs more and is time-consuming during the whole course of management and relatively higher infection rate. Although intra-operative method is a more convenient way for implantation, it may be technically difficult in case there is an extensive pathological change to the tissue, or, occasionally, no suitable branch of the axillary vein could be found.
In the complications of the implantable venous access device, pneumothorax still occurs with the percutaneous method. But, percutaneously placed catheters are associated with a lower infection rate than surgically implanted ones. We suggest that patients with tall, slim figure and high risk of bullae formation in their lungs should be treated with the ultrasound and fluoroscopy-guided insertion method.
Conclusion
Since 1980, the SIP (subcutaneous infusion port) implantable venous access device has been increasingly adopted and widely accepted by both patients and medical staff alike because of its many benefits and few disadvantages. As revealed by many clinicians in their recently published data, the complication rate since its initial introduction has decreased progressively due to the improvement of technique and product material. However, there is still room for improvement with accumulated experience and improved catheter materials. There were fewer complications in this series mainly because more attention to details was paid by the surgeons and the care team. The controversy of safety between image-guided approach and non-X-ray-guide direct approach is irrelevant because we no longer depend on X-ray guide for implantation after the first few cases. The use of Doppler ultrasound to assist with catheter placement was first reported in 1984 [20]. At least three meta-analysis paper [17], [25], [26] proved that ultrasound guidance for placement of internal jugular and subclavian central venous catheter in adult can reduce the risk of complications and cost-effectiveness. For a breast cancer patient during mastectomy, port implantation is a good and safe method. Direct venous cut-down method was not recommended now. For percutaneous insertion method, we prefer landmark-base insertion technique of subclavian vein first. If fine guiding needle cannot localize the position of subclavian vein easily, converting to image-guided internal jugular vein or subclavian vein catheterization is the best choice. Chest X-ray is still taken prior to chemotherapy and prior to the removal of the port. To our knowledge, SIIP is still very useful for chemotherapy patients. Even complications can be forestalled if they are forewarned, and the worst complication can be adequately managed without panic if surgeons are well prepared ahead of time, by own experience, or through others’ experiences.
Conflict of Interest Statement
All the authors of this paper claim herein that this research do not have any financial sponsor, hence it is not inappropriately influenced by any affiliated organization or person.
Authorship
Guarantor of the integrity of the study: Tze-Kai Chen, Mary Ann Lou.
Study concepts: Hsiang-Chun Jan
Study design: Hsiang-Chun Jan
Definition of intellectual content: Hsiang-Chun Jan
Literature research: Hsiang-Chun Jan
Clinical studies: Hsiang-Chun Jan, Shao-Jiun Chou, Tzu-Hung Chen
Experimental studies:Clinical studies: Hsiang-Chun Jan, Shao-Jiun Chou, Tzu-Hung Chen
Data acquisition: Hsiang-Chun Jan
Data analysis: Hsiang-Chun Jan
Statistical analysis: Hsiang-Chun Jan
Manuscript preparation: Hsiang-Chun Jan, Chuin-I Lee
Manuscript editing: Mary Ann Lou
Manuscript review: Mary Ann Lou
References
- Subcutaneous venous port implantation in adult patients: a single center experience. Diagn Interv Radiol. 2006;12:93–98
- . Infection assessment of totally implanted long-term venous access devices. J Vasc Access. 2006;7(1):24–28
- . Long-term infusional systems: complications in cancer patients. Tumori. 2001;87(5):308–311
- . Ultrasonography and fluoroscopy-guided insertion of chest ports. Eur J Surg Oncol. 2008;34(12):1340–1343
- . Rupture of totally implantable central venous access devices (Intraports) in patients with cancer: report of four cases. World J Surg Oncol. 2004;2(36):
- . The “pinch off” syndrome: a complication of implantable catheter systems in the subclavian vein. Schweiz Med Wochenschr. 1997;127(29–30):1247–1250
- . Catheter fracture and cardiac migration of a totally implantable venous device. Indian J Cancer. 2005;42(3):155–157
- . Central venous access ports placed by interventional radiologists: experience with 125 consecutive patients. Cardiovasc Intervent Radiol. 2001;24(3):180–184
- Superior vena cava thrombosis related to catheter malposition in cancer chemotherapy given through implanted ports. Cancer. 1993;72(7):2248–2252
- . Subcutaneously implanted central venous access devices in cancer patients: a prospective analysis. Cancer. 1997;79(8):1635–1640
- . Spontaneous rupture and embolization: a rare complication of port catheter implantation. Dtsch Med Wochenschr. 2000;125(20):628–630
- . Intravenous port catheter embolization: mechanisms, clinical features, and management. Angiology. 2008;59(1):90–97
- . Complications and management of long-term central venous access catheters and ports. J Vasc Access. 2004;5(4):174–178
- . Mechanical complications of subclavian vein catheterization: a prospective study. Int Surg. 1991;76:18–22
- . Complication of central venous catheterization. J Am Coll Surg. 2007;204:681–696
- . Surgical technique for totally implantable access port (TAIP) needs improvement: a multivariable analysis of 400 patients. J Surg Oncol. 2006;93(1):24–29
- . Ultrasound guidance for placement of central venous catheters: a meta-analysis of the literature. Crit Care Med. 1996;24(12):2053–2058
- Ultrasound guided implantation of chest port system via the lateral subclavian vein. Fortschr Röntgenster. 2006;178(3):324–329
- . Preventing complications of central venous catheterization. N Engl J Med. 2003;348(12):1123–1133
- . Doppler localization of the internal jugular rein facilitates central venous cannulation. Anesthesiology. 1984;60:481–482
- . Combined ultrasound and fluoroscopy guided port catheter implantation- high success and low complication rate. Eur J Radiol. 2009;69:517–522
- . Port-A-Cath placement without the aid of fluoroscopy or localizing devices: a community hospital series. Cancer J. 2005;11:157–159
- . Catheter-induced thrombosis in the superior vana cava diagnosed by transesophageal echocardiography. Acta Anesthesiol Scand. 1998;42:1227–1230
- . Central venous catheter vascular erosion: diagnosis and clinical course. Ann Surg. 1989;209:475–478
- Ultrasonic locating devices for central venous cannulation: meta analysis. BMJ. 2003;327–361
- . Ultrasound for central venous cannulation: economic evaluation of cost-effectiveness. Anaesthesia. 2004;59:1116–1120
PII: S0960-7404(10)00062-9
doi:10.1016/j.suronc.2010.07.001
© 2010 Elsevier Ltd. All rights reserved.
