Guidelines on the Insertion and Management of Central Venous Access Devices in Adults L

TECHNICAL REPORT INTERNATIONAL JOURNAL OF LABORATORY HEMATOLOGY

Guidelines on the insertion and management of central venous access devices in adults L. BISHOP*, L. DOUGHERTY†,A.BODENHAM‡,J.MANSI*,P.CROWE§,C.KIBBLER–,M.SHANNON**, J. TRELEAVEN†

*Guys and St Thomas Hospital, SUMMARY London, UK †Royal Marsden Hospital, Sutton, Central venous access devices are used in many branched of medicine Surrey, UK where venous access is required for either long-term or a short-term ‡Consultant Anaesthetist, Leeds RoyalInﬁrmary,Leeds,UK care. These guidelines review the types of access devices available and §Birmingham Heartlands Hospital, make a number of major recommendations. Their respective advan- Birmingham, UK tages and disadvantages in various clinical settings are outlined. Patient – Royal Free Hospital, London, UK care prior to, and immediately following insertion is discussed in the **St George’s Hospital London, UK context of possible complications and how these are best avoided. Correspondence: There is a section addressing long-term care of in-dwelling devices. Jennifer Treleaven, Royal Marsden Techniques of insertion and removal are reviewed and management of Hospital, Sutton, Surrey SM2 5PT, the problems which are most likely to occur following insertion UK. Tel.: 0208 661 3782; Fax: 0208643 7958; E-mail: including infection, misplacement and thrombosis are discussed. Care [email protected] of patients with coagulopathies is addressed and there is a section addressing catheter-related problems. doi:10.1111/j.1751-553X.2007.00931.x

Received 20 December 2006; accepted for publication 6 February 2007

Keywords venous access devices, chemotherapy, coagulation, heparin, PBSC transplantation

for nursing staff (RCN, 2005) and are based on database INTRODUCTION searches using appropriate keywords, and a review of These guidelines are a review of basic principles and the existing published guidelines written by expert relevant research for medical and nursing staff involved groups [British Committee for Standards in Haematol- in the care of patients with central venous access ogy (BCSH Guidelines, 1997)]. The authors were selec- devices (CVADs). They complement existing guidelines ted to embrace a wide range of interest in central

Corresponding author belongs to British Committee for Standards in Haematology.

2007 The Authors Journal compilation 2007 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2007, 29, 261–278 261 262 L. BISHOP ET AL. GUIDELINES ON CENTRAL VENOUS ACCESS DEVICES FOR USE IN ADULTS

venous access devices, including clinical, haematologi- supervision and competence assessment pro- cal, radiological, anaesthetic, critical care and nursing. grammes in place. Paediatric specialists should insert The guidelines are not intended as a substitute for local catheters in children. policies and protocols but should provide a useful • Ultrasound guided insertion is recommended for all source of reference for those writing such documents. routes of central venous catheterization. The use of These guidelines relate to the insertion and manage- ultrasound is also recommended for the insertion of ment of nontunnelled and skin-tunnelled central PICC when the peripheral veins are not visible or venous catheters (CVC), apheresis catheters, implanted palpable. ports and peripherally inserted central catheters (PICC). • Imaging facilities (fluoroscopy, intravenous contrast studies and standard radiography) should be available for the insertion of skin-tunnelled CVCs and MAJOR RECOMMENDATIONS ports. Following a review of the current literature, which is • Catheter insertion should take place in an operating fully referenced in the body of the paper, the follow- theatre or similar clean environment. Bedside place- ing recommendations are made. ment should not be performed except in an emergency, apart from PICC placement. • Patients should receive clear and comprehensive • Rigorous skin cleansing with a chlorhexidine gluco- verbal and written information explaining the risks, nate 2% in alcohol or aqueous solution is recom- benefits and care of the catheter. Signed consent mended prior to catheter insertion. should be obtained prior to catheter insertion. • Antibiotic/antimicrobial impregnated catheters, for • Nontunnelled catheters are indicated for short-term example, chlorhexidine and silver sulfadiazine use when peripheral venous access is impractical. impregnated catheters should be considered for • Tunnelled central venous catheters are indicated for appropriate risk groups of patients to minimize the repeated administration of chemotherapy, anti- infection risk. These are becoming available for tun- biotics, parenteral feeding and blood products, and nelled devices. for frequent blood sampling. They are recommen- • Routine antibiotic prophylaxis is not recommended. ded for patients in whom long-term (>30 days) cen- • Flushing with heparin vs. normal saline remains tral venous access is anticipated. controversial. • Fully implanted catheters (ports) are more suitable • Routine replacement, for example, weekly change, for children and for less frequent accessing but of short-term catheters as a means to reduce infec- long-term use, whereas skin-tunnelled catheters are tion rates is not recommended. recommended for intensive access. • Guidewire-assisted catheter exchange to replace a • Peripherally inserted central catheters should be malfunctioning catheter is acceptable if there is no avoided for inpatient therapy because of limited evidence of infection. However, if infection is suspec- catheter longevity and increased incidence of ted the existing catheter should be removed and a thrombosis. They are more suited to ambulatory or new catheter inserted at a different site. This tech- outpatient-based therapy. nique is generally impractical for cuffed tunnelled • Polyurethane PICC allow easier infusion of blood catheters or ports when it may be technically easier products as greater flow rates are achieved because and safer to insert a new catheter into a clean site. the thinner walls provide a larger internal diameter • Dressings should be changed 24 h after catheter of the catheter. The decision to use polyurethane insertion and weekly thereafter. catheters should be balanced against the higher risk • Securing devices, for example, StatlokTM are prefer- of thrombosis with these catheters compared with able to stitches, and lines should not be sewn into silicone catheters. or around the vein. • The number of lumina and diameter of catheters • Needle-free connectors should be used to reduce risk should be kept to the minimum. of infection to patients and needle stick injury to • Experienced operators, regardless of speciality, staff. should perform catheter insertion with training,

2007 The Authors Journal compilation 2007 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2007, 29, 261–278 L. BISHOP ET AL. GUIDELINES ON CENTRAL VENOUS ACCESS DEVICES FOR USE IN ADULTS 263

• A positive pressure method of flushing (by protocol, have single or multiple lumina and can be open according to the type of catheter) is essential to ended or valved. Multiple lumina catheters are maintain catheter patency. advantageous in patients undergoing stem cell trans- • IV therapy giving sets should be changed every 24– plantation or chemotherapy where a number of 48 h if used for transfusing blood products, and agents and blood products require simultaneous infu- every 72–96 h otherwise. sion. Blood products may be administered concur- • Pre-existing haemorrhagic, thrombotic, or infective rently with another drug/infusion through a dual problems must be effectively managed before cath- bore catheter. Multiple lumina catheters are associ- eter insertion. ated with increased morbidity (Farkas et al., 1992; • Blood products may be administered concurrently Dezfulian et al., 2003), but in the haematology set- with another drug/infusion through a dual bore ting, the increased risk is likely to be offset by the catheter. convenience of multilumina catheters, thereby justi- • Low-dose warfarin prophylaxis is not recommen- fying their use. If TPN is being administered, a single ded, but therapeutic dosing may be required in lumen central venous catheter or lumen should be selected patients at risk of developing a thrombo- dedicated exclusively to this purpose (Pratt et al., sis. 2001). The smallest diameter catheter should be • Thrombosis and infection must be promptly diag- employed, to minimize the risk of catheter-related nosed and vigorously treated. Both complications thrombosis and/or subsequent venous stenosis may require removal of the catheter. (Knutstad, Hager & Hauser, 2003). However, it may • Tunnelled catheters can be pulled out if the cuff has be difficult to administer blood products via very nar- not anchored in the tissues. Otherwise, a cut-down row lumina. procedure is needed to free the cuff. Ports require surgical removal. All procedures should be underta- Nontunnelled central venous catheters ken by experienced personnel. • Units should audit complications associated with In an attempt to reduce catheter-related blood stream central venous catheterization and use the data to infection (CRBSI) rates, various materials have been develop preventative measures. Close liaison with investigated. These have been reviewed, and antimi- the local microbiology department is essential to crobial/antiseptic impregnated catheters, for example, monitor trends in infection. chlorhexidine and silver sulfadiazine short-term catheters have been shown to be effective in reducing catheter-related blood stream infections, but other Indications for catheter insertion types are commercially available. A large randomized These catheters are indicated (i) when venous access controlled trial showed that CVCs coated with chlor- is poor, (ii) when embarking on prolonged intraven- hexidine and silver sulfadiazine were associated with ous chemotherapy and/or total parenteral nutrition a 44 % reduction in colonization and a 79 % reduc- (TPN), or for repeated administration of blood tion in catheter-related blood stream infection (Maki products, (iii) when intravenous therapy involves et al., 1997), although the largest study to date, in drugs known to be venous sclerosants, (iv) when which the mean duration of catheterization was ambulatory chemotherapy is to be given as an outpa- 20 ± 12 days, failed to show any benefit (Logghe tient, (v) in the situation of repeated sampling, or et al., 1997). The use of antimicrobial/antiseptic venesection. impregnated catheters is recommended for adults who require short-term (<10 days) central venous catheterization and who are at high risk of infection (Pratt Choice of catheter et al., 2001; Pellowe et al., 2003). The debate contin- Catheters are categorized into (i) nontunnelled cathe- ues about such catheters and their propensity for ters, (ii) tunnelled catheters with anchoring cuff, (iii) inducing antibiotic resistance, and occasional severe implanted ports (iv) apheresis/dialysis catheters (tun- allergic reactions have been reported (Cicalini, Palmi- nelled and nontunnelled) and (v) PICC. They may eri & Petrosillo, 2004).

2007 The Authors Journal compilation 2007 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2007, 29, 261–278 264 L. BISHOP ET AL. GUIDELINES ON CENTRAL VENOUS ACCESS DEVICES FOR USE IN ADULTS

cy (e.g. 5000 U/ml of heparin). The volume of hep- Tunnelled catheters arin flush used should be equal to the volume of each Tunnelled catheters are recommended for patients in lumen to avoid systemic heparinization of the patient. whom long-term (>30 days) central venous access is It should be noted that for optimum flow rates, it necessary (Pratt et al., 2001). Devices exist with and may be necessary to position the tip of the catheter at without Dacron anchoring cuffs. Tunnelled catheters the junction of the right atrium and superior vena have been shown to be associated with lower infec- cava (SVC; Vesely, 2003) to avoid irritation/thrombus tion rates than nontunnelled catheters (Randolph formation when the catheter tip abuts onto the vein et al., 1998). The cuff induces an inflammatory reac- wall (Fletcher & Bodenham, 2000). tion within the subcutaneous tunnel leading to fibro- sis, with catheter fixation usually occurring within 3– PICC 4 weeks of insertion. Valved catheters have the advantage of not requiring heparin flushes but may Peripherally inserted central catheters represent a vas- need pressurized infusions to administer blood prod- cular access device (VAD) that can be considered to ucts. They also tend to be more costly. There is little have an intermediate role in central venous access. hard evidence to support one type of catheter over These catheters are usually inserted at the bedside via another. an antecubital vein and are available with single or multiple lumina. In the haematology setting, they are well suited for ambulatory or outpatient therapy Implanted ports (Whitman, 1996) as opposed to intensive inpatient Ports have been shown to have the lowest reported therapy but have been shown to be associated with a rates of catheter-related blood stream infections com- higher incidence of thombosis in patients with haema- pared with either tunnelled or nontunnelled CVC tological malignancies (Cortelezzi et al., 2005). This is (Pegues et al., 1992; Groeger et al., 1993). Most ports an important consideration in patients who have had are single lumen, which makes them more suited to previous thromboses, and in those who are receiving long-term intermittent therapy. They tend to be used therapy which may increase the thrombotic tendency more frequently in paediatrics, and in patients with like thalidomide. Peripherally inserted central cathe- solid tumours (Camp-Sorrell, 1992; Gabriel, 1999). In ters can be made of either silicon rubber or polyureth- the adult haematology setting, they may be of use in ane, the former being associated with a lower risk of sickle cell anaemia or thalassaemia, where patients thrombosis (Galloway & Bodenham, 2004). However, are receiving regular blood transfusions. Ports may polyurethane PICC are recommended because polyur- also be useful for oncology patients with poor periph- ethane is a tougher material, enabling thinner lumen eral venous access who are receiving less intensive walls and larger internal diameters of the lumina. This therapy unlikely to cause prolonged neutropenia. significantly increases flow rates and reduces the They allow less restricted bathing and swimming and potential for breakage and rupture of the catheter may appeal to patients concerned about the psycholo- (Hadaway, 1995; Mayer & Wong, 2002). This is an gical aspects of the presence of the external part of advantage because of the volume of blood and platelet the nonimplanted catheters. They are more expensive infusions required by haematology patients. to purchase, insert and remove, and they leave larger The various types of catheter available are sum- scars. marized in Table 1. An experienced member of the haemato-oncology team should make the decision regarding which type of catheter is most appropriate Apheresis/dialysis catheters at the outset of therapy, to avoid multiple catheteriza- These can be either nontunnelled (VascathsTM)or tions. The decision should be made based on diagno- subcutaneously tunnelled with a cuff, and a selection sis, length and type of therapy, patient preference, is commercially available for longer-term use. They clinical status, availability of patent veins, operator are larger bore catheters and usually require flushing experience and previous central venous access history with stronger solutions of heparin to maintain paten- (Hamilton & Fermo, 1998; Hamilton, 2000; Chernecky

2007 The Authors Journal compilation 2007 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2007, 29, 261–278 L. BISHOP ET AL. GUIDELINES ON CENTRAL VENOUS ACCESS DEVICES FOR USE IN ADULTS 265

Table 1. Types of commercially available catheter

Catheter type Advantages Disadvantages

Nontunnelled catheters Choice of sites Short-term use Easy to insert and remove Multiple lumina available Skin-tunnelled catheters* Lower infection rates than nontun- More complex insertion and nelled removal Long-term use Ports No external catheter Surgical insertion and removal Cosmetically attractive Less suitable for frequent repeated Patient can swim/bathe as normal access Low maintenance Long-term use Lower infection rates than skin- tunnelled catheters Apheresis/dialysis catheters Permit high blood flow rates Large bore Nontunnelled (e.g. VascathTM Kimal) Easier to insert and remove Require flushing with concentrated Skin-tunnelled Lower infection rates than nontun- heparin (for example 5000 U/ml, nelled devices according to manufacturer guide- Long-term use lines) solution to maintain paten- Good for patients with poor per- cy. Flush solution must be ipheral access who require both withdrawn prior to use PBSC harvest and transplant pro- Short-term use cedure Complex insertion and removal Best inserted via internal jugular or femoral routes PICC Easy to insert and remove Higher thrombosis rate particularly Do not require platelet support or with polyurethane variety. Polyur- correction of clotting prior to inser- ethane variety required to infuse tion/removal blood/platelets because has greater internal diameter than silicone variety Slower flow rates particularly in silicone/valved varieties Catheter longevity lower than with skin-tunnelled devices Incidence of malposition greater than in other types of CVC

*Valved skin-tunnelled catheters are not suitable for apheresis, whereas occasionally the other types of skin-tunnelled catheters can be used for returning blood during an apheresis procedure. et al., 2002). For example, if an allogeneic transplant take a physical assessment, vein assessment and his- is planned, a double or triple lumen skin-tunnelled tory of previous central venous catheterizations. catheter should be inserted at the outset to ensure Small, portable ultrasound imaging devices provide adequate access throughout chemotherapy and the quick confirmation of vein patency. The presence of subsequent transplant. venous collaterals on the chest wall/abdomen may signify deep venous obstruction. A history, or signs of SVC obstruction is highly significant, as is a history of Patient care prior to catheter insertion difficulties or failure of insertion by a competent oper- The procedure, including risks and benefits, should be ator using X-ray screening or ultrasound. If there is a explained to the patient. The ‘operator’ should under- history of a prior thrombus it is prudent to perform

2007 The Authors Journal compilation 2007 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2007, 29, 261–278 266 L. BISHOP ET AL. GUIDELINES ON CENTRAL VENOUS ACCESS DEVICES FOR USE IN ADULTS

formal ultrasound imaging studies with Doppler meas- • As treatment in response to a single blood culture urements, on both the affected and unaffected side, to positive for coagulase-negative staphylococcus, if exclude thrombus and ensure vein patency. Ultra- other blood cultures drawn in the same time frame sound gives useful information at most sites but can- are negative. not image the SVC. Other modalities including • For routine prophylaxis. conventional venography, CT with contrast, MRI ve- • When there is a catheter-related infection involving nography and transoesophageal ultrasound all have beta-lactam-sensitive organisms. advantages. Difficult, or potentially difficult cases • As continued empiric therapy for presumed infec- should be discussed with a vascular radiologist. Stent- tions in patients whose cultures are negative for ing techniques can be used to restore patency to ste- beta-lactam-resistant gram-positive organisms. nosed veins. Antibiotic cover may be appropriate if simulta- It is generally accepted that the platelet count neous insertion and removal procedures have to be should be >50 · 109/l prior to insertion of a catheter performed in a known case of catheter-related sepsis. other than a PICC (BCSH, 2003), and the INR <1.5. However, it is not routinely recommended. (Ansell et al., 2004; Douketis, Johnson & Turpie, 2004) Problems may arise when patients are refractory to platelets, have idiopathic thrombocytopenic Catheter insertion purpura (ITP) or thrombotic thrombocytopenic pur- As previously mentioned, it is essential that only pura (TTP), or in the presence of deranged clotting, experienced personnel insert central venous catheters, for example, in acute promyelocytic leukaemia. The to minimize infection and other complications, partic- risks and benefits of insertion in terms of type and site ularly in the presence of low platelets, deranged clot- of catheter must be assessed on an individual basis. ting and/or in the critically ill patient. This includes Where the risk of bleeding is increased, or when diffi- the short-term multiple lumina nontunnelled cath- culties with insertion are anticipated, use of experi- eter, which is often inserted by junior medical person- enced personnel and ultrasound guidance are nel who have undergone minimal training with little essential (Hatfield & Bodenham, 1999) to maximize clinical supervision in insertion techniques. the likelihood of an atraumatic, ‘first pass’ procedure. Where possible, the procedure should be performed Additionally, use of lidocaine with adrenaline in a clean area designated for CVC insertion such as 1 : 200 000 as local anaesthetic will reduce subcuta- the X-ray department, operating theatre, or a proce- neous bruising/bleeding. The increasing recognition of dure suite where a high standard of asepsis is prac- the risks associated with the use of blood products ticed. Units which currently insert tunnelled CVCs on mandates assessment of need on an individual basis the wards should audit their procedural complications rather than routine correction of all minor abnormal- and infection rates in order to support or refute con- ities of platelet count and coagulation studies. tinuation of this practice. Whatever the environment, maximum sterile barrier conditions (mask, cap, sterile Antibiotic prophylaxis gloves, gown and large drape) have been shown to lower the risk of acquiring catheter-related infections In a meta-analysis of published research, the Center (Mermel et al., 1991; Raad et al., 1994; O’Grady et al., for Disease Control in the United States of America 2002). Published evidence shows, however, that the (O’Grady et al., 2002) identified that the use of anti- risk of infection depends mainly on the presence of microbial prophylaxis routinely before insertion or bacteria on the skin (Fletcher, 2005). Hence, skin during use of an intravascular catheter does not pre- cleansing is the most important part of care before vent catheter colonization or BSI (evidence level 1A; catheter insertion. Two per cent aqueous chlorhexi- i.e. strongly supported by well-designed experimental, dine has been shown to be superior to either povi- clinical, or epidemiologic studies). The Department of done-iodine solution or 70% alcohol in reducing Health (Pratt et al., 2001) has issued guidelines limit- catheter-related blood stream infections (Maki, Ringer ing the use of vancomycin. They state that the agent & Alvarado, 1991). More recently, 2% chlorhexidine should not be used in the following circumstances.

2007 The Authors Journal compilation 2007 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2007, 29, 261–278 L. BISHOP ET AL. GUIDELINES ON CENTRAL VENOUS ACCESS DEVICES FOR USE IN ADULTS 267

in alcohol has been recommended (Pratt et al., 2001; cause secondary stenosis (Knutstad, Hager & Hauser, Pellowe et al., 2003) and this should be used for skin 2003). antisepsis prior to catheter insertion and for insertion A lateral approach when using the subclavian vein site dressing changes. avoids ‘pinch off’ of the catheter, which is a rare com- Anatomical surface markings can be unreliable for plication referring to entrapment of the catheter the initial puncture and ultrasound guided insertion between the clavicle and the first rib (Polderman & of catheters (excluding PICC) is strongly recommen- Girbes, 2002). Ultrasound-guided insertions in the ded, whenever possible, by operators experienced in subclavian area will tend to move the puncture site the use of ultrasound techniques (NICE, 2002; Hind laterally because of the presence of the probe and cla- et al., 2003). Ultrasound can be used for PICC inser- vicle (Sharma, Bodenham & Mallick, 2004). An alter- tion but is not usually necessary. It may, however, native is to use the internal jugular vein, which is be useful to check vein patency, for proximal approa- particularly useful in patients with abnormal coagula- ches or when the peripheral veins are not palpable tion, as it reduces the risk of inadvertent arterial because they lie deep within the subcutaneous tissue. puncture (Lameris et al., 1990), although tunnelling The use of ultrasound ensures patency of the vein can be more difficult. Debate continues based on ret- and has been shown to reduce insertion-related com- rospective audits concerning the relative merits of plications (NICE, 2002; Hind et al., 2003). An image right or left insertion sites and the perceived differ- intensifier can be used to image the target vein with ences in rates of thrombosis, infections and mechan- contrast and to check correct positioning of the ical problems. Other considerations such as cosmetic guidewire and catheter tip. Its use is recommended appearance may be important to patients and care when inserting tunnelled catheters and ports. Cathe- should be taken to avoid lumpy tunnelling in the ters are normally inserted under local anaesthesia neck and to arrange for the external portion of the with, or without sedation such as intravenous mida- line to be concealed under clothing. zolam or other combinations of sedative drugs, In patients in whom the internal jugular and both according to local policy. Occasional patients with subclavian veins are occluded or otherwise unavaila- severe needle phobia or who become dyspnoeic ble for puncture, or in the event of SVC obstruction, when lying flat/head down will require a general catheters may be inserted into the femoral vein, anaesthetic. The majority of procedures should be although infection rates are higher with this approach performed by Seldinger wire techniques, reserving (Goetz et al., 1998; O’Grady et al., 2002), as are rates open surgery for difficult cases or small children. of deep vein thrombosis (Trottier et al., 1995; Joynt Specialist vascular radiology input will be required et al., 2000). Tunnelling can take the exit site out of for the more difficult cases needing vein stenting or the groin onto the relatively clean anterior abdominal unusual routes of access. wall. The femoral approach may also be useful in No randomized controlled trials have satisfactorily emergency situations or in the presence of severe compared infection or thrombosis rates for catheters thrombocytopenia and/or coagulopathy, where it is placed via the jugular, subclavian and femoral sites easier to apply pressure or achieve haemostasis in the (Pellowe et al 2004), although in nonrandomized event of bleeding. The straight course of the vein comparisons, infection rates are higher with femoral makes it ideal for stiff dialysis-type catheters, which catheters (Goetz et al., 1998) [See below]. In adults, do not traverse tight angles easily. the subclavain approach may have an advantage With subcutaneous ports, subclavian or internal because there is some evidence to suggest that it has jugular venous access is achieved using aseptic condi- lower infection rates than the internal jugular tions, in the same way as for skin-tunnelled catheters. approach (Mermel et al., 1991; Richet et al., 1991), If patients are accessing their own port, then it should but this must be balanced against the risks of mechan- be located low on the rib cage or upper arm, for easy ical complications. For example, catheters placed via a access. When patients are not accessing their own subclavian approach are more likely to cause throm- port, then the port is usually located on the upper rib bosis than those placed via the internal jugular route cage near the clavicle. Ports near the sternum provide (Trerotola, 2000), and are therefore more likely to better needle stability and ease of access. Adequate

2007 The Authors Journal compilation 2007 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2007, 29, 261–278 268 L. BISHOP ET AL. GUIDELINES ON CENTRAL VENOUS ACCESS DEVICES FOR USE IN ADULTS

subcutaneous tissue will prevent erosion through the groups using gauze dressing. The risk for CRBSIs did skin. If the port is placed too deeply or there is excess not differ between the groups. The choice of dressing adipose tissue, it can make access difficult. Placement can be a matter of preference. If blood is oozing from under the arm, in the breast or the soft tissue of the the catheter insertion site, gauze dressing might be abdomen should be avoided (Goodman, 2000). There preferred (O’Grady et al., 2002). are both low and high profile ports available. For patients with a tunnelled catheter, the upper suture over the insertion site into the vein should be removed at 7–10 days and the lower one at the exit Immediate patient care after catheter insertion point should be removed after 3 weeks. Recent evi- Tip placement should be checked by X-ray prior to dence supports the use of securing devices, including use and the position of the catheter tip documented tapes, adhesives or staples (Motonaga, Lee & Kirsch, in the patient file. Optimum tip position is the distal 2004), particularly with nontunnelled CVCs and PICC. superior vena cava or the upper right atrium (Fletcher These obviate the need for sutures at the exit site or & Bodenham, 2000).The carina can be used as an around the vein, which can cause difficulties with approximate marker of the level of the pericardial subsequent line removal. Securing devices have also reflection. Catheters inserted from the left side tend to been shown to reduce infection rates when compared need the tip to lie at the junction of the SVC and RA with sutures (Crnich & Maki, 2002; Yamamoto et al., or within the upper RA, in order for the catheter tip 2002; Frey & Schears, 2006). Sutures over an to lie within the long axis of the vessel. Changes in implanted port insertion site are removed after 7– tip position on standing may be significant. Chest 10 days, although some operators may use dissolvable X-rays taken within 1–2 h of placement may not sutures to close the wound. Peripherally inserted cen- demonstrate a slowly developing pneumothorax or tral catheters and nontunnelled catheters should bleed. A further chest X-ray is required if the patient always be covered with a dressing, although skin-tun- becomes dyspnoeic or complains of laterat chest wall nelled catheters may not require a dressing once the discomfort/pain. Pneumothorax usually relates to sub- wound has healed (Morris et al., 1995; O’Grady et al., clavian vein catheterization but can also result from 2002). This can be reviewed on an individual basis. attempted internal jugular cannulation. Many pneu- Implanted ports do not require any dressing once the mothoraces do not require drainage but should be wound has healed. monitored by serial X-rays. Smaller bore Seldinger- type drains are suitable for draining slowly accumula- Long-term catheter care ting collections when treatment is required, and large bore traditional chest drains are rarely required. For skin-tunnelled devices, it is advisable to either A transparent semi-occlusive dressing such as Op- secure the ends to the chest wall with tape or to use a Site 1 V 3000TM is recommended (Reynolds, Tebbs & ‘neck-bag’ to take the weight of the free ends. Elliott, 1997; Treston-Aurand et al., 1997). Transpar- Showering is preferable to bathing, and swimming ent dressings reliably secure the device, permit con- must be avoided with any external catheter, in order tinuous visual inspection of the catheter site, permit to prevent colonization by Gram-negative organisms, patients to bathe and shower without saturating the especially Pseudomonas spp. dressing, and require less frequent changes than do Flushing with the correct solution and technique is standard gauze and tape dressings. The dressing essential to maintain catheter patency, and only sin- should be changed after the procedure if bleeding has gle-dose solutions should be used. The use of heparin occurred, but otherwise not until 24 h postoperative- flushes vs. normal saline intermittent flushes remains ly. It should then be changed weekly if there are no controversial. Many clinicians still recommend the use signs of bleeding and/or infection. However, a recent of heparin (10 U/ml) to prevent thrombus formation review has found no consistent benefit for any type of and ensure catheter patency, but the efficacy of this is dressing (Gillies et al., 2003). A meta-analysis has unproven (Pellowe et al. 2004). Exposure to heparin assessed studies that compared the risk for catheter- should be minimized to prevent the development of related BSIs for groups using transparent dressings vs. heparin-induced thrombocytopenia (HITS) and to

2007 The Authors Journal compilation 2007 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2007, 29, 261–278 L. BISHOP ET AL. GUIDELINES ON CENTRAL VENOUS ACCESS DEVICES FOR USE IN ADULTS 269

avoid development of bleeding complications because of inadvertent heparinization secondary to multiple heparin flushes (Passannante & Macik, 1998). A review of the current evidence concluded that heparin doses of 10 U/ml are no more beneficial than flushing with normal saline alone (Pellowe et al. 2004). How- ever, there are exceptions. For example, apheresis/ dialysis catheters require heparin flushes to maintain patency, and some manufacturers and clinicians recommend heparin flushes, particularly when catheters are infrequently accessed. The need for heparin may be a function of bore, as larger bore catheters 10 ml inrupture size, to avoidavoid catheter systemicusually printed on heparinization the CVC – lumen allow quicker back-tracking of blood up the lumen. Because thrombi and fibrin deposits on catheters might serve as a nidus for microbial colonization of intravascular catheters, the use of anticoagulants might have a role in the prevention of CRBSI. Because the majority of heparin solutions contain pre- servatives with antimicrobial activity, whether any decrease in the rate of CRBSI is a result of the reduced thrombus formation, the preservative, or both is unclear (O’Grady et al 2002). Flushing protocols for the main types of catheter are shown in Table 2 although the manufacturers’

recommendations should always be followed. Syringe intermittently, or weekly After each access or weeklyAfter each access, or at least weekly Calculate ‘dead space volume’ to After each access when used After each access or weekly After each access or monthly Do not use a syringe smaller than size can be important as the smaller syringes create greater pressure and may contribute to catheter rupture if excessive pressure is exerted (Conn, 1993; Primhak, 1998). Care must always be taken to maintain patency by using a pulsatile flush method and by maintaining positive pressure while removing the syringe at the end of flushing in order to avoid reflux of blood (Goodwin & Carlson, 1993; Dougherty, 2004). The use of needle-free connectors is recommended as these have been shown to reduce infection (Yebenes et al., 2004). Patients should be educated in the care of their catheters, and a recent randomized study has shown a reduction in catheter-related infections when patients are made responsible for care of the catheter 1000 U/ml or 5000volume U/ml only Catheter saline solution (10 U/ml) (Moller et al., 2005). saline solution (10 U/ml) heparinized saline solution (10 U/ml) heparinized saline solution (10 U/ml) Several prospective and randomized studies have shown that specialist teams can reduce infection rates (Miller et al., 1996; Fitzsimmons et al., 1997; Meier et al., 1998; Soifer et al., 1998; Solomon & Stoddard,

2001; Hamilton, 2004), and where possible the use of Catheter ﬂushing protocols these should be developed. Infections can be minimized by careful hand washing and catheter site care. Table 2. Apheresis/dialysisPICC* 10 ml saline 0.9% plus *Valved varieties require 10 ml saline 0.9% solution to ﬂush. No heparin required. Refer to manufacturer guidelines in each 10 case. ml saline 0.9% ± 5 ml heparinized The external surfaces of the access port should be Catheter typeNontunnelledSkin-tunnelled* SolutionPorts* 10 ml saline 0.9% ± 10 5 ml ml saline heparinized 0.9% ± 5 ml 10 ml saline 0.9% ± 5 ml Frequency Cautions

2007 The Authors Journal compilation 2007 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2007, 29, 261–278 270 L. BISHOP ET AL. GUIDELINES ON CENTRAL VENOUS ACCESS DEVICES FOR USE IN ADULTS

disinfected with a chlorhexidine gluconate solution Management of problem patients unless contraindicated by the manufacturers instruc- • If the patient is thrombocytopenic and there is evi- tions (Pellowe et al. 2004; Pratt et al., 2001). Either a dence of bleeding after catheter insertion, then the sterile (sterile gloves) or ‘nontouch’ (clean gloves) patient should receive further platelet transfusion(s) technique must be utilized when accessing any CVC to maintain the count in excess of 50 · 109/1 until lumen. A blood sampling protocol should be devel- bleeding stops, bearing in mind problems may exist oped locally, but must include instruction on the in patients refractory to random donor platelets or removal of the heparinized dead space (approximately in those suffering from ITP or TTP (BCSH, 2003). In 5 ml) prior to sampling, to avoid erroneous results. these situations the application of pressure dressings The volume to be removed before coagulation studies and topical tranexamic acid may help. Prolonged are performed is uncertain with central venous cathe- compression (15 min plus) will often stop bleeding. ters, but for APTT studies from arterial cannulae it is It should not be overlooked, however, that there recommended that six times the dead space volume is are no published prospective, randomized studies to removed (Laxson & Titler, 1994). Coagulation studies support or negate the theory that the level of plate- in such circumstances may produce erroneous results, let count at time of CVC insertion should be main- particularly with apheresis or dialysis catheters where tained above 50 · 109/l to reduce potential for stronger heparin solutions have been used, and, if significant bleeding problems. Barrera et al. (1996) there is any doubt, the sample should be taken from a concluded that thrombocytopenia is not the only peripheral vein. In the bone marrow transplant set- risk factor for bleeding and concludes that other ting, a lumen for taking blood for ciclosporin and variables such as insertion site, number of needle other drug levels should be identified and the drugs passes to cannulate the vein, and expertise are more administered through a different lumen. pertinent. Similarly, the study by Ray and Shenoy (1997) concluded that peri-procedural platelet Patient information transfusions have little effect on bleeding outcome. Other studies in patients with liver disease and A patient’s guide should include the following sections: complex haemostatic defects have found that a (i) What constitutes a central venous access device. (ii) platelet count of <50 · 109/l was an independent The advantages and disadvantages of having a central risk factor for bleeding in comparison with raised venous access device. (iii) Any risks involved in the INR or prolonged PT (Doerfler, Kaufman & Golden- insertion of a central venous access device. (iv) Care of berg, 1996; Fisher & Mutimer, 1999; Mumtaz et al., the device. (v) Removal of the device. 2000). Robust, randomized studies are necessary to Locally generated patient information leaflets are achieve evidence-base guidance in these complex recommended but should not be a substitute for care- situations. ful and detailed explanation by a nurse/doctor experi- • In patients with disseminated intravascular coagulation, enced in the care of central venous catheters. for example, in association with acute promyelocy- Generally speaking, the following information must tic leukaemia, there should be vigorous correction be provided, with a 24-h cover arrangement: of any abnormality of coagulation. The prothrombin • How to contact a nurse or doctor if the exit site is time should be <1.5 times normal and fibrinogen red, sore, oozing pus or in the event of a fever of >1.0 g/1. Patients taking oral anticoagulants should >38 C. stop their tablets to achieve an INR <1.5 before • How to contact a nurse or doctor if the catheter catheter insertion. If time is limited, factor concen- becomes damaged or leaks. trates, fresh frozen plasma (FFP), or vitamin K may • How to contact a nurse or doctor if the arm becomes be required, but the latter, in high doses, may inter- swollen or if any distended veins become apparent fere with subsequent anticoagulation (Baglin, Keel- on the chest or neck. ing & Watson, 2005). Reversal of therapeutic • How to contact a nurse or doctor if breathless or anticoagulation with vitamin K is usually achieved pain are experienced. within 4–6 h of intravenous administration of vita-

2007 The Authors Journal compilation 2007 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2007, 29, 261–278 L. BISHOP ET AL. GUIDELINES ON CENTRAL VENOUS ACCESS DEVICES FOR USE IN ADULTS 271

min K, and within 24 h of oral administration 2002). Guidewire-assisted catheter exchange to (Watson et al., 2001). Phytomenadione (Konak- replace a malfunctioning catheter is acceptable if there ion) doses of up to 2 mg intravenously, or 5 mg is no evidence of infection. However, if infection is orally are recommended. Complete and rapid rever- suspected, the existing catheter should be removed sal of over anticoagulation is more readily achieved and a new catheter inserted at a different site (Pratt with a factor concentrate rather than with FFP et al., 2001). This technique is generally impractical (Makris et al., 1997; Evans, Luddington & Baglin, for cuffed tunnelled catheters or ports, when it may 2001). Intravenous vitamin K should be given if be technically easier and safer to insert a new catheter reversal is to be sustained (Yasaka et al., 2002), and into a clean site. repeat administration may be required after 24 h. • A patient who has received previous radiotherapy to Intravenous unfractionated heparin (UFH) should be one side of the chest should have the catheter inser- stopped 3 h before catheter insertion and restarted ted on the opposite side. Patients who have under- when haemostasis is secured. In patients receiving gone mastectomy and lymph node dissection should prophylactic subcutaneous low molecular weight hep- have the catheter placed on the opposite side. Cath- arin (LMWH), catheter insertion can be undertaken eters can be tunnelled onto the arm or other sites if 12 h after the last injection; for patients receiving the chest wall is unsuitable. Catheters should be therapeutic subcutaneous LMWH, the time to catheter kept away from breast prostheses and pacemaker insertion should be extended to 18 h after the last boxes/wires. injection. Heparin can be recommenced once the operator has conﬁrmed haemostasis is secure, usually Prevention and management of catheter within 2 h of catheter insertion. Substitution with complications intravenously infused UFH or insertion of an IVC ﬁlter should be considered if there is a very high throm- The main complications are: (i) catheter-related infec- botic risk. Expert haematology advice should be tion; (ii) catheter malfunction; and (iii) catheter-rela- sought. ted thrombosis.

• Haemophiliac patients (with haemophilia A, B or C) will require appropriate factor replacement, as may Catheter-related infection patients with other inherited coagulopathies. Cor- Infection rates vary from 0.08 per 1000 days in rection should be maintained for >48 h. Clinicians oncology outpatients to 19/1000 catheter days in caring for such patients should seek advice from the critically ill (Fletcher, 2005). Haemato-oncology their local haemophilia reference centre. infection rates probably lie somewhere within this • Infection at the time of catheter insertion represents range but catheter-related blood stream infections a relative contra-indication to proceeding, and con- can be severe and life-threatening depending on the sideration should be given to temporary, nontun- micro-organism involved. The Department of Health nelled catheter placement or temporary use of (Pratt et al., 2001) has made recommendations for peripheral cannulae, but the risks and beneﬁts good practice regarding prevention, diagnosis and should be considered for each patient on an individ- treatment of infections (and other aspects of central ual basis. If the patient has a unilateral skin infec- venous catheterization), which have recently been tion on the anterior upper chest wall, the updated by Pellowe et al. (2004). There are three unaffected side should be used for catheter place- categories: ment. Targeted antibiotic prophylaxis may be war- ranted in these cases. • A catheter-related blood stream infection is deﬁned as at least two blood cultures positive with the same Routine replacement of nontunnelled CVCs should organism, obtained from at least two separate sites not be used as a method for preventing catheter-rela- at different times, in association with evidence ted infection, as this has not been shown to reduce of colonization of the catheter with the same infection rates (Cook et al., 1997; O’Grady et al.,

2007 The Authors Journal compilation 2007 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2007, 29, 261–278 272 L. BISHOP ET AL. GUIDELINES ON CENTRAL VENOUS ACCESS DEVICES FOR USE IN ADULTS

Table 3. Recommendations for the management of catheter related infections

Category of infection Non-neutropenic patient Neutropenic patient

Exit site infection Remove catheter if no longer needed Remove catheter if no longer needed Treat empirically with *ﬂucloxacillin Initial empirical therapy including glycopeptide Treat for 10–14 days or longer until infection resolved Modify according to isolates Remove catheter if evidence of progres- sion or if blood cultures are positive for Staph. aureus, Pseudomonas spp., Mycobac- terium spp., or fungi Tunnel infection Remove catheter if no longer needed Remove catheter if no longer needed Treat empirically with *ﬂucloxacillin Initial empirical therapy including glycopeptide Treat for 10–14 days or longer until reso- lution of soft tissue infection. Modify according to isolates If tracking continues to spread remove catheter Presumed catheter-related Remove catheter if no longer needed Remove catheter if no longer needed bloodstream infection Treat empirically with antibiotics Initial empirical antibiotic therapy. Mod- targeted against isolates ify according to isolates. Treat for at least 10–14 days Remove catheter if cultures remain positive after 48 h of therapy or if proven catheter-related infection with Staph. aureus, Pseudomonas spp., Mycobacterium spp., or fungi

*Unless known to be colonized with MRSA, when a glycopeptide should be used

organism. The latter part of the deﬁnition can only Table 3 summarizes current recommendations based be strictly fulﬁlled by removing the catheter. upon consensus and the literature, although the decis- • An exit site infection presents with erythema, tender- ion to salvage or remove a catheter should be made ness and occasionally a discharge at the insertion following discussion with the microbiologist and after site. consideration of the patient’s clinical status and his • A tunnel infection is characterized by pain and indu- position on the treatment pathway. Recent evidence ration along the track of the catheter. suggests that in situ use of glycopeptides may be effective for coagulase negative staphylococcal infec- The incidence of these infections varies in different tions (Ley et al., 1996; Giacometti et al., 2005). The centres with different groups of patients and different ‘antibiotic lock’ technique may also be effective in practices. reducing catheter-related bacteraemia (Carratala, The management of catheter infections remains 2002; Garland et al., 2005; Rijnders et al., 2005; Fer- controversial. Attempts should be made to make a nandez-Hidalgo et al., 2006; Kim et al., 2006). Certain microbiological diagnosis by culturing blood from all organisms carry a much greater risk of treatment fail- catheter lumina, a peripheral sample of blood and the ure and disseminated infection (see Table 3). A cath- exit site before commencing antibiotics. However, in eter lock solution containing citrate and taurolidine clinical practice, it is usual for broad-spectrum antibi- (TauroLockTM) has been used particularly in the otics to be initiated while awaiting culture results.

2007 The Authors Journal compilation 2007 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2007, 29, 261–278 L. BISHOP ET AL. GUIDELINES ON CENTRAL VENOUS ACCESS DEVICES FOR USE IN ADULTS 273

setting of haemodialysis catheters (Allon, 2003; Betjes Medac and is available on a named patient basis. The & van Agteren, 2004), and is designed to render the solution should be injected gently into the catheter internal flow passages resistant to clot formation and with a push-pull action to maximize mixing within hostile to bacterial and fungal growth. This may also the lumen. The lumen should then be clamped and prove useful in the haemato-oncology setting, but left for at least 2–3 h. The catheter should then be un- requires further evaluation. clamped and the solution containing disaggregated It is recommended that administration sets are clot aspirated (Gabriel, 1999; Dougherty, 2004). It has changed every 72 h when used for continuous infu- not been shown to be cost-effective or clinically sions of solutions, but these should be changed more necessary to leave the solution in the lumen for lon- frequently if used for intermittent therapies, blood, or ger periods, such as between episodes of haemodialy- lipids (NICE, 2003; RCN, 2005). Needle-free connec- sis. tors must be changed in accordance with manufac- Very recently, an alternative urokinase preparation turer’s recommendations (Medicines and Healthcare called Syner-KINASE has become available. This is Products Regulatory Agency (MHRA), 2005). manufactured by Syner-med and has been licensed in the UK for clearing blocked intravenous catheters. It is highly purified, extracted from human urine, and Catheter malfunction tested by PCR to exclude viral contamination. This is Partial and complete catheter blockage is evidenced by said to be free of the technical problems associated difficulty in aspirating blood or infusing fluid. Forcible with manufacture, which occurred with the previous introduction of fluid down an obstructed lumen may brands of urokinase. cause catheter rupture. Catheter occlusion may An alternative to urokinase is Cathflo Activase include blockage resulting from kinking of the cath- (Alteplase), which is a recombinant human tissue eter, ‘pinch off syndrome’, occlusion of the catheter plasminogen activator (t-PA; Deicher et al., 2002). tip on the vessel wall, fibrin sheath or fibrin flap or Again, this is available on a named patient basis and luminal thrombus, or migration of the tip into a smal- is manufactured by Genentech. ler vessel. Plain X-ray or a catheter contrast study Other reasons for catheter malfunction can include may be helpful in confirming the diagnosis. Initially, damage to the catheter. For example, ‘pinch off’ as fibrin sheaths manifest with catheter dysfunction, pro- described earlier, or kinking of the catheter. Occasion- gressing to complete failure. They are usually discov- ally, the tip of the catheter can migrate, particularly if ered 1–2 weeks after placement (Crain, Horton & the catheter is short and the tip initially lies in the Mewissen, 1998). Infusion substances can penetrate upper superior vena cava or brachiocephalic vein. This between the catheter wall and the fibrin sheath in a may result in the catheter ceasing to function. Repeat retrograde manner, along the catheter to the site of chest X-rays may help in diagnosing these problems. venous insertion, and even out to perivascular and Internal repair of a damaged catheter is no longer subcutaneous layers. This can lead to cutaneous or recommended because of risk of infection and/or air subcutaneous necrosis. Untreated fibrin sheaths are embolus. External repairs of damaged catheters can be associated with increased risk of complications, but performed using kits provided by the manufacturers. interventional radiologists may be able to temporarily salvage catheter function by using percutaneous, Catheter-related thrombosis intravascular fibrin sheath stripping via a trans- femoral approach (Knutstad, Hager & Hauser, 2003). Catheter-associated thrombosis may be spontaneous, Where catheter occlusion is due to thrombus with- or may result from a prothrombotic state associated out symptomatic thrombosis, instillation of Hepsal either with underlying malignancy or treatment, par- (heparin sodium 10 U/ml) may be effective. If not, ur- ticularly with L-asparaginase, thalidomide or lenalida- okinase 10 000 U/ml reconstituted in 4 ml normal sal- mide. The catheter will normally require removal if ine may be tried, using 2 ml of solution into each thrombosis is confirmed. Intraluminal thrombosis may catheter lumen and ensuring that intraluminal vol- be prevented by adhering to appropriate flushing pro- umes only are instilled. Urokinase is manufactured by tocols and ensuring good placement of the catheter

2007 The Authors Journal compilation 2007 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2007, 29, 261–278 274 L. BISHOP ET AL. GUIDELINES ON CENTRAL VENOUS ACCESS DEVICES FOR USE IN ADULTS

tip (Table 2). The use of low dose warfarin is now catheter-related thrombus, (iv) damaged catheter, and contraindicated as it has been shown to be of no (v) the end of treatment. Removal of a skin-tunnelled apparent benefit for the prophylaxis of symptomatic catheter requires local anaesthetic and minor surgical catheter-related thrombosis in patients with cancer cut-down to remove the cuff if the catheter has been (Couban et al., 2005; Young et al., 2005). Dose-adjus- in situ for more than approximately 3 weeks. The ted warfarin may be superior but at the cost of an patient should lie down to avoid air embolus. Simple increased risk of bleeding. There are no published data traction can remove the catheter and cuff in catheters, concerning ideal levels of anticoagulation in thrombo- which have been in less than three weeks. Gentle cytopenic patients or on the recommended duration traction can be attempted, but if difficulty is encoun- of anticoagulant therapy in catheter-related thrombo- tered it is important to stop prior to the catheter sis. If the catheter is removed because of confirmed breaking. Removal should be undertaken by experi- thrombosis, therapeutic doses of low molecular weight enced personnel. The platelet count should be heparin and warfarin should be given in nonthrombo- >50 · 109/l and the INR less than 1.5. Ideally, 12 h cytopenic patients. In thrombocytopenic patients low should have elapsed after prophylactic low molecular molecular weight heparin may be used, adjusting the weight heparin, and 18 h after a therapeutic dose. dose in accordance with the level of thrombocytope- The skin-tunnelled catheter can be removed under nia. Full doses can be given if the platelet count local anaesthetic. Use of a local anaesthetic agent con- exceeds 80 · 109/l (BCSH, 2006), in the absence of taining epinephrine (w/v/1 : 200 000) may be helpful bleeding and where renal function is normal. With in the presence of thrombocytopenia and deranged platelet counts below this, the decision regarding hep- clotting, to minimize local bleeding. A small incision arin dose should be based on clinical need, the pres- (2 cm should be adequate) is made alongside the cuff ence or absence of bleeding, and whether or not the and blunt dissection used to free the cuff and avoid platelet count increments with platelet transfusion. catheter damage prior to removal. If the cuff is at the Renal function should be regularly monitored during exit site it can be removed by enlarging the exit site treatment. wound. The intravascular portion should be removed Anticoagulation should be continued for a period safely prior to cutting the catheter (Galloway & Bod- of approximately 3 months in uncomplicated cases, enham, 2004). If the catheter tip is sheared off during with a target INR of 2.5 (range: 2.0–3.0) when war- removal, it is likely to embolize into the right heart or farin is being used. If there is clinical or radiological pulmonary artery and will require urgent retrieval by evidence of persistent thrombus, anticoagulation vascular radiologists using a snare, under fluoroscopic should be continued for a longer period. Mechanical guidance (Bessoud et al., 2003). Internal repair of a clot lysis or local application of thrombolytic drugs to damaged catheter is no longer recommended because rapidly restore vein patency can be effective if the of risk of infection and/or air embolus. External vein is occluded with fresh thrombus. Collaboration repairs to damaged catheters can be performed using with vascular surgical or interventional radiology kits provided by the manufacturers. teams is advised. It is important to remove the catheter in the direc- If the patient has a PICC, any swelling of the arm tion of the tunnel. The catheter should be inspected should be monitored. Swelling alone does not confirm carefully after removal to ensure that it is complete, thrombosis, and if suspected it must be confirmed ra- and, if infection is suspected, the tip should be sent to diologically, by Doppler ultrasound, CT scanning or the microbiology department for culture. The cut- other imaging. If confirmed, the PICC should be down site should be sutured with a fine 3/0 or 4/0 removed and anticoagulants commenced as described monofilament suture. Large incisions and thick previously. sutures should be avoided as these scar badly. After removal, pressure should be applied to the exit point, tunnel and venotomy site and an occlusive dressing Technique of catheter removal placed over the exit site to avoid air embolism. Ports Indications for catheter removal include: (i) catheter- require surgical removal in theatre or equivalent. related infection, (ii) persistent catheter occlusion, (iii) Peripherally inserted central catheters and short term,

2007 The Authors Journal compilation 2007 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2007, 29, 261–278 L. BISHOP ET AL. GUIDELINES ON CENTRAL VENOUS ACCESS DEVICES FOR USE IN ADULTS 275

nontunnelled catheters can be removed at the patients ACKNOWLEDGEMENT bedside and pressure applied. The tips should be sent We are very grateful to Michaele Pocock for reading for microscopy and culture if clinically indicated. the manuscript and for providing helpful suggestions and further references. Recommendations for audit A locally based audit should include patient identiﬁca- DISCLAIMER tion data, diagnosis, date of catheter insertion, num- While the advice and information in these guidelines ber of previous catheters, operator and department is believed to be true and accurate at the time of where the catheter was inserted, complications associ- going to press, neither the authors, the British Society ated with the catheter, date of and reason for for Haematology nor the publishers accept any legal removal. Each unit should monitor their infection responsibility for the content of these guidelines. rates/1000 catheter days to observe any changes or trends in infection rates and be mindful of the emer- gence of resistant bacteria.

REFERENCES Roche A. (2003) Experience at a single catheter replacement strategies: A sys- institution with endovascular treatment tematic review of the literature. Critical Allon M. (2003) Prophylaxis against of mechanical complications caused by Care Medicine 25, 1417–1424. dialysis catheter-related bacteremia with implanted central venous access devices Cortelezzi A.N., Moia M., Falanga A., Po- a novel antimicrobial lock solution. in paediatric and adult patients. Amer- gliani E.M., Agnelli G., Bonizzoni E., Clinical and Infectious Diseases 36, ican Journal of Radiology 180, 527–532. Gussoni G., Barbui T. & Mannucci P.M. 1539–1544. Betjes M.G. & van Agteren M. (2004) (2005) CATHEM Study Group. Inci- Ansell J., Hirsh J., Poller L., Bussey H., Prevention of dialysis-related sepsis dence of thrombotic complications in Jacobson A. & Hylek E. (2004) The with a citrate-taurolidine-containing patients with haematological malig- pharmacology and management of the lock solution. Nephrology Dialysis and nancies with central venous catheters: a vitamin K antagonists: the seventh Transplantation 19, 1546–1551. prospective multicentre study. British ACCP conference on antithrombotic Camp-Sorrell D. (1992) Implantable ports: Journal of Haematology 129, 811–817. and thrombolytic therapy. Chest 126, everything you always wanted to know. Couban S., Goodyear M., Burnell M., Do- 204–233. Journal of Intravenous Nursing 15, lan S., Wasi P., Barnes D., Macleod D., Baglin T.P., Keeling D.M. & Watson H.G. 262–273. Burton E., Andreou P. & Anderson D.R. For the British Committee for Standards Carratala J. (2002) The antibiotic-lock (2005) Randomized placebo-controlled in Haematology (2005) BCSH Guide- technique for therapy of ‘highly nee- study of low-dose warfarin for the pre- lines on oral anticoagulation (warfarin): ded’ infected catheters. Clinical Micro- vention of central venous catheter-asso- third edition – 2005 update. British biological Infections 8, 282–289. ciated thrombosis in patients with Journal of Haematology 132, 277–285. Chernecky C., Macklin D., Nugent K. & cancer. Journal of Clinical Oncology 20, Barrera R., Mina B., Huang Y. & Groeger Waller J.L. (2002) The need for shared 4063–4069. J.S. (1996) Acute complications of cen- decision making in the selection of vas- Crain M.R., Horton M.G. & Mewissen tral line placement in profoundly cular access device: an assessment of M.W. (1998) Fibrin sheaths complicat- thrombocytopenic cancer patients. Can- patients and clinicians. Journal of Vas- ing central venous catheters. American cer 78, 2025–2030. cular Access Devices 7, 34–39. Journal of Radiology 171, 341–346. BCSH (2003) Guidelines for the use of Cicalini S., Palmieri F. & Petrosillo N. Crnich C.J. & Maki D.G. (2002) The pro- platelet transfusions. British Journal of (2004) Clinical review: new technolo- mise of novel technology for the pre- Haematology 122, 10–23. gies for prevention of intravascular vention of intravascular device-related BCSH (2006) Guidelines on the use and catheter-related infections. Critical Care blood stream infections II long-term monitoring of heparin. British Journal 8, 157–162. devices. Clinical and Infectious Disease of Haematology 133, 19–34. Conn C. (1993) The importance of syringe 1, 1232–1242. BCSH Guidelines (1997) Guidelines on size when using an implanted vascular Deicher S.R., Fesen M.R., Kiproff P.M., Hill the insertion and management of cen- access device. Journal of Vascular P.A., Li X., McCluskey E.R. & Semba C.P. tral venous lines. British Journal of Access Nursing 3, 11–18. (2002) Safety and efﬁcacy of alteplase for Haematology 98, 1041–1047. Cook D., Randolph A., Kernerman P., restoring function in occluded central Bessoud B., de Baere T., Kuoch V., Desru- Cupido C., King D., Soukup C. & Brun- venous catheters: results of the cardio- ennes E., Cosset M.F., Lassau N. & Buisson C. (1997) Central venous vascular thrombolytic to open occluded

2007 The Authors Journal compilation 2007 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2007, 29, 261–278 276 L. BISHOP ET AL. GUIDELINES ON CENTRAL VENOUS ACCESS DEVICES FOR USE IN ADULTS

lines trial. Journal of Clinical Oncology Fletcher S.J. & Bodenham A.R. (2000) T.E. & Armstrong D. (1993) Infectious 20, 317–324. Safe placement of central venous cathe- morbidity associated with use of long- Dezfulian C., Lavelle J., Nallamothu B.K., ters: where should the tip of the cathe- term venous access devices in patients Kaufman S.R. & Saint S. (2003) Rates ter lie? British Journal of Anaesthesia with cancer. Annals of Internal Medi- of infection for single lumen versus 85, 188–191. cine 119, 1168–1174. multilumen central venous catheters: a Frey A.M. & Schears G.J. (2006) Why are Hadaway L.C. (1995) Comparison of vas- meta-analysis. Critical Care Medicine we stuck on tape and suture? A review cular access devices. Seminars in Oncol- 31, 2385–2390. of catheter securement devices. Journal ogy Nursing 11, 154–166. Doerfler M.E., Kaufman B. & Goldenberg of Infusional Nursing 29, 34–38. Hamilton H. (2000) Selecting the correct A.S. (1996) Central venous in patients Gabriel J. (1999) Long-term central intravenous device: nursing assessment. with disorders of hemostasis.catheter venous access. In: Intravenous Therapy British Journal of Nursing 9, 968–978. placement. Chest 110, 185–188. in Nursing Practice (eds L. Dougherty & Hamilton H. (2004) Advantages of a Dougherty L. (2004) Vascular Access J. Lamb), pp. 301–333. Churchill Liv- nurse-led central venous access vascular Devices. In: Manual of Clinical Nursing ingstone Ltd, Edinburgh. service. The Journal of Vascular Access Procedures (eds L. Dougherty & S. Lis- Galloway S. & Bodenham A.R. (2004) 5, 109–112. ter), 6th edn, pp. 724–773. Blackwell Long-term central venous access. British Hamilton H. & Fermo K. (1998) Assess- Science, Oxford. Journal of Anaesthesia 92, 722–734. ment of patients requiring IV therapy Douketis J.D., Johnson J.A. & Turpie A.G. Garland J.S., Alex C.P., Henrickson K.J., via a central venous route. British Jour- (2004) Low molecular weight heparin McAuliffe T.L. & Maki D.G. (2005) A nal of Nursing 7, 451–460. as bridging anticoagulation during inter- vancomycin-heparin lock solution for Hatfield A. & Bodenham A.R. (1999) Por- ruption of warfarin. Archives of Internal prevention of nosocomial bloodstream table ultrasound for difficult venous Medicine 164, 1319–1326. infection in critically ill neonates with access. British Journal of Anaesthesia Evans G., Luddington R. & Baglin T. peripherally inserted central venous 82, 822–826. (2001) Beriplex P/N reverses severe catheters: a prospective, randomized Hind D., Calvert N., McWilliams R., warfarin-induced over anticoagulation trial. Pediatrics 116, 198–205. Davidson A., Paisley S., Beverley C. & immediately and completely in patients Giacometti A., Cirioni O., Ghiselli R., Thomas S. (2003) Ultrasonic locating presenting with major bleeding. Orlando F., Mocchegiani F., Silvestri C., devices for central venous cannulation: British Journal of Haematology 115, Licci A., De Fusco M., Provinciali M., a meta-analysis. British Medical Journal 998–1001. Saba V. & Scalise G. (2005) Compara- 327, 361–368. Farkas J.C., Liu N., Bleriot J.P., Chevret tive efficacies of quinupristin–dalfopris- Joynt G.M., Kew J., Gomersall C.D., Leu- S., Goldstein F.W. & Carlet J. (1992) tin, linezolid, vancomycin, and ng V.Y. & Liu E.K. (2000) Deep vein Single versus triple-lumen central ciprofloxacin in treatment, using the thrombosis caused by femoral venous catheter-related sepsis: a prospective antibiotic-lock technique, of experimen- catheters in critically ill adult patients. randomised study in a critically ill tal catheter-related infection due to Sta- Chest 117, 178–183. population. American Journal of Medi- phylococcus aureus. Antimicrobial Agents Kim S.H., Song K.I., Chang J.W., Kim cine 93, 277–282. in Chemotherapy 49, 4042–4045. S.B., Sung S.A., Jo S.K., Cho W.Y. & Fernandez-Hidalgo N., Almirante B., Call- Gillies D., O’Riordan E., Carr D., O’Brien Kim H.K. (2006) Prevention of uncuffed eja R., Ruiz I., Planes A.M., Rodriguez I., Frost J. & Gunning R. (2003) Central hemodialysis catheter-related bacter- D., Pigrau C. & Pahissa A. (2006) Anti- venous catheter dressings: a systematic emia using an antibiotic lock technique: biotic-lock therapy for long-term intra- review. Journal of Advanced Nursing a prospective, randomized clinical trial. vascular catheter-related bacteraemia: 44, 623–632. Kidney International 69, 161–164. results of an open, non-comparative Goetz A., Wagener M., Miller J.M. & Knutstad K., Hager B. & Hauser M. study. The Journal of Antimicrobial Muder R.R. (1998) Risk of infection due (2003) Radiological diagnosis and man- Chemotherapy 57, 1172–1180. to central venous catheters: effect of site agement of complications related to Fisher N.C. & Mutimer D.J. (1999) Cen- of placement and catheter type. Infec- central venous access. Acta Radiologica tral venous cannulation in patients with tion Control and Hospital Epidemiology 44, 508–516. liver disease and coagulopathy – a pros- 19, 842–845. Lameris J.S., Post P.J., Zonderland H.M., pective audit. Intensive Care Medicine Goodman M. (2000) Chemotherapy: prin- Gerritsen P.G., Kappers-Klunne M.C. & 25, 481–485. ciples of administration. In: Cancer Nur- Schutte H.E. (1990) Percutaneous pla- Fitzsimmons C., Gilleece M., Ranson M., sing (ed. C. Henke Yarbro et al.), 5th cement of Hickman catheters: compar- Wardley A., Morris C. & Scarffe J. edn, 385–443. Jones and Bartlett, Bos- ison of sonographically guided and (1997) Central venous catheter place- ton. blind techniques. American Journal of ment: extending the role of the nurse. Goodwin M.L. & Carlson I. (1993) The Radiology 155, 1097–1099. Journal of the Royal College of Physi- peripherally inserted central catheter: a Laxson C.J. & Titler M.G. (1994) Drawing cians of London 31, 533–535. retrospective look at 3 years insertions. coagulation studies from arterial lines: Fletcher S.J. (2005) Catheter related Journal of Intravenous Nursing 16, 92– an integrative literature review. Amer- blood stream infection. Continuing 103. ican Journal of Critical Care 3, 16–22. Education in Anaesthesia. Critical Care Groeger J.S., Lucas A.B., Thaler H.T., Ley B.E., Jalil N., McIntosh J., Smart A., & Pain 5, 49–451. Friedlander-Klar H., Brown A.E., Kiehn Wilson M., Foot A.B. & Millar M.R.

2007 The Authors Journal compilation 2007 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2007, 29, 261–278 L. BISHOP ET AL. GUIDELINES ON CENTRAL VENOUS ACCESS DEVICES FOR USE IN ADULTS 277

(1996) Bolus of infusion teicoplanin for Miller J., Goetz A., Squier C. & Muder R. Keidan R.D., Boraas M. & Weese J. intravascular-associated infections in (1996) Reduction in nosocomial intra- (1992) Comparison of infections with immunocompromised patients? Journal venous device-related bacteraemias Hickman and implanted port catheters of Antimicrobial Chemotherapy 38, after institution of an intravenous ther- in adult solid tumor patients. Journal of 1091–1095. apy team. Journal of Intravenous Nur- Surgical Oncology 49, 156–162. Logghe C., Van Ossel C., D’Hoore W., Ez- sing 19, 103–106. Pellowe C.M., Pratt R.J., Harper P., Love- zedine H., Wauters G. & Haxhe J.J. Moller T., Borregaard N., Tvede M. & day H.P., Robinson N., Jones S.R., Mac- (1997) Evaluation of chlorhexidine and Adamsen L. (2005) Patient education–a Rae E.D., Mulhall A., Smith G.W., Bray silver-sulfadiazine impregnated central strategy for prevention of infections J., Carroll A., Chieveley Williams S., venous catheters for the prevention of caused by permanent central venous Colpman D., Cooper L., McInnes E., bloodstream infection in leukaemic catheters in patients with haematologi- McQuarrie I., Newey J.A., Peters J., patients: a randomized controlled trial. cal malignancies: a randomized clinical Pratell,i N., Richardson G., Shah P.J., Journal of Hospital Infection 37, 145– trial. Journal of Hospital Infection 61, Silk D. & Wheatley C. (2003) Guideline 156. 330–341. Development Group. (2003) Evidence- Maki D.G., Ringer M. & Alvarado C.J. Morris P., Grace S., Glackin V., Akers N., based guidelines for preventing health- (1991) Prospective randomized trial of McNamarra G., Seale T. & Nicol A. care-associated infections in primary povidone iodine, alcohol and chlorhex- (1995) Audit of skin-tunnelled catheters and community care in England. Jour- idine for prevention of infection asso- in neutropenic patients. Bone Marrow nal of Hospital Infection 55 (Suppl. 2), ciated with central venous and arterial Transplantation 15 (Suppl. 2) Abstract S2–S127. catheters. Lancet 338, 339–343. 685, S168. Polderman K. & Girbes A. (2002) Central Maki D.G., Stolz S.M., Wheeler S. & Mer- Motonaga G.K., Lee K.K. & Kirsch J.R. venous catheter use: part 1: mechanical mel L.A. (1997) Prevention of central (2004) The efficacy of the arrow staple complications. Intensive Care Medicine venous catheter-related bloodstream device for securing central venous 28, 1–17. infection by use of antiseptic impreg- catheters to human skin. Anesthesia & Pratt R.J., Pellowe C., Loveday H.P., Rob- nated catheter. A randomised controlled Analgesia 99, 1436–1439. inson N., Smith G.W., Barrett S., Davey trial. Annals of Internal Medicine 127, Mumtaz H., Williams V., Hauer-Jenson P., Harper P., Loveday C., McDougall 257–266. M., Rowe M., Henry-Tillman R.S., He- C., Mulhall A., Privett S., Smales C., Makris M., Greaves M., Phillips W.S., Kit- aton K., Mancino A.T., Muldoon R.L., Taylor L., Weller B. & Wilcox M. (2001) chen S., Rosendaal F.R. & Preston E.F. Kilmberg V.S., Broadwater J.R., West- The epic project: developing national (1997) Emergency oral anticoagulant brook K.C. & Lang N.P. (2000) Central evidence-based guidelines for prevent- reversal: the relative efficacy of infu- venous catheter placement in patients ing healthcare associated infections. sions of fresh frozen plasma and clotting with disorders of Hemostasis. American Phase I: guidelines for preventing hospi- factor concentrate on correction of the Journal of Surgery 180, 503–505. tal-acquired infections. Department of coagulopathy. Thrombosis and Haemos- NICE (2002) Guidance on the Use of Health (England). Journal of Hospital tasis 77, 477–480. Ultrasound Locating Devices for Central Infection 47 (Suppl.), S1–S82. Mayer T. & Wong D.G. (2002) The use of Venous Catheters. NICE Technology Primhak R.H. (1998) Pressures used to polyurethane PICCs an alternative to Appraisal 49. National Institute for Clin- flush central venous catheters. Archives other materials. Journal of Vascular ical Excellence , London. of Disease in Childhood Fetal & Neona- Access Devices 3, 26–29. NICE (2003) Infection Control Prevention tal Edition 78, F234. Medicines and Healthcare Products Regu- of Healthcare Associated Infection in Raad I.I., Hohn D.C., Gilbreath B.J., latory Agency (MHRA) (2005) Medical Primary and Community Care. DOH, Suleiman N., Hill L.A., Bruso P.A., Device Alert on All Brands of Needle National Institute for Clinical Excel- Marts K., Mansfield P.F. & Bodey G.P. Free Intra Vascular Connectors. MDA/ lence, London, Clinical Guidelines 2. (1994) Prevention of central venous 2005/030, London. O’Grady N.P., Alexander M., Dellinger catheter-related infections by using Meier P., Fredrickson M., Catney M. & E.P., Gerberdin J.L., Heard S.O., Maki maximal sterile barrier precautions dur- Nettleman M. (1998) Impact of a dedi- D.G., Masur H., McCormick R.D., Mer- ing insertion. Infection Control and cated intravenous therapy team on mel L.A., Pearson M.L., Raad I.I., Rand- Hospital Epidemiology 15, 231–238. nosocomial bloodstream infection rates. olph A. & Weinstein R.A. (2002) Randolph A., Cook D.J., Gonzales C.A. & American Journal of Infection Control Guidelines for the prevention of intra- Brun-Buisson C. (1998) Tunnelling 26, 388–392. vascular catheter-related infections. short-term central venous catheters to Mermel L.A., McCormack R.D., Spring- American Journal of Infection Control prevent catheter-related infection: a man S.R. & Maki D.G. (1991) The 30, 476–489 (review). meta-analysis of randomized controlled pathogenesis and epidemiology of Passannante A. & Macik B.G. (1998) Case trials. Critical Care Medicine 26, 1452– catheter-related infection with pulmon- report: the heparin flush syndrome: a 1457. ary artery Swan-Ganz catheters: a pro- cause of iatrogenic haemorrhage. Amer- Ray C.E. Jr & Shenoy S.S. (1997) Patients spective study using molecular sub- ican Journal of Medical Science 296, with thrombocytopenia: Outcome of typing. American Journal of Medicine 71–73. radiological placement of central venous 91 (Suppl. 3B), 197S–205S. Pegues D., Axelrod P., McClarren C., Ei- access devices. Radiology 204, 97–99. senberg B.L., Hoffman J.P., Ottery F.D.,

2007 The Authors Journal compilation 2007 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2007, 29, 261–278 278 L. BISHOP ET AL. GUIDELINES ON CENTRAL VENOUS ACCESS DEVICES FOR USE IN ADULTS

RCN (2005) Royal College of Nursing IV venous catheter complications with an access devices. Current Problems in Sur- Therapy Forum. Standards for Infusion intravenous therapy team. Archives of gery 23, 309–388. Therapy. Royal College of Nursing, Lon- Internal Medicine 158, 473–477. Yamamoto A.J., Solomon J.A., Soulen don. Avaliable at: http://www.rcn.org. Solomon S. & Stoddard G. (2001) The M.C., Tang J., Parkinson K., Lin R. & uk/publications/pdf/002179.pdf (accessed infusion nurse and patient complication Schears G.J. (2002) Sutureless secure- on 22 June 2007). rates of peripheral-short catheters. Jour- ment devices reduces the complications Reynolds M.G., Tebbs S.E. & Elliott T.S.J. nal of Intravenous Nursing 24, 113– of peripherally inserted central venous (1997) Do dressings with increased per- 123. catheters. Journal of Vascular and Inter- meability reduce the incidence of cen- Trerotola S.O. (2000) Hemodialysis cath- ventional Radiology 13, 77–81. tral venous catheter related sepsis. eter placement and management. Radi- Yasaka M., Sakata T., Minematsu K. & Intensive and Critical Care Nursing 13, ology 215, 651–656. Naritomi H. (2002) Correction of INR 26–29. Treston-Aurand J., Olmsted R.N., Allen- by prothrombin complex concentrate Richet H., Hubert B., Nitemberg G., An- Bridson K. & Craig C.P. (1997) Impact and vitamin K in patients with warfarin dremont A., Buu-Hoi A., Ourbak P., of dressing materials on central venous related hemorrhagic complication. Galicier C., Veron M., Boisivon A. & catheter infection rates. Journal of Thrombosis Research 108, 25–30. Bouvier A.M. (1991) Prospective multi- Intravenous Nursing 20, 201–206. Yebenes J.C., Vidaur L., Serra-Prat M., centre study of vascular-catheter-related Trottier S.J., Veremakis S.C., O’Brien J. & Sirvent J.M., Batlle J., Motje M., Bonet complications and risk factors for posi- Auer A.I. (1995) Femoral deep vein A. & Palomar M. (2004) Prevention of tive central catheter cultures in inten- thrombosis associated with central catheter-related bloodstream infection sive care unit patients. Journal of venous catheterization: Results from a in critically ill patients using a disin- Clinical Microbiology 28, 2520–2525. prospective randomized trial. Critical fectable, needle-free connector: a ran- Rijnders B.J., Van Wijngaerden E., Care Medicine 23, 52–59. domized controlled trial. American Vandecasteele S.J., Stas M. & Peeter- Vesely T.M. (2003) Central venous cath- Journal of Infection Control 32, 291– mans W.E. (2005) Treatment of eter tip position: A continuing contro- 295. long-term intravascular catheter-related versy. Journal of Vascular and Young A.M., Begum G., Billingham L.J., bacteraemia with antibiotic lock: rando- Interventional Radiology 14, 527–534. Hughes D.J., Kerr D., Rea A., Stanley mized, placebo-controlled trial. Journal Watson H.G., Baglin T., Laidlaw S.L., A., Sweeney K. & Wheatley J.W. of Antimicrobial Chemotherapy 55, 90– Makris M. & Preston F.E. (2001) A (2005) WARP Collaborative Group, UK, 94. comparison of the efﬁcacy and rate of 2005, WARP – a multicentre prospect- Sharma A., Bodenham A.R. & Mallick A. response to oral and intravenous Vita- ive randomised controlled trial (RCT) of (2004) Ultrasound guided infra- min K in reversal of over-antic- thrombosis prophylaxis with warfarin in clavicular axillary vein cannulation for oagulation with warfarin. British cancer patients with central venous central venous access. British Journal of Journal of Haematology 115, 145–149. catheters (CVCs). ASCO Meeting Anaesthesia 9, 188–192. Whitman E.D. (1996) Complications asso- Abstracts June 23 (16S), LBA8004. Soifer N., Borzak S., Edlin B. & Weinstein ciated with the use of central venous R. (1998) Prevention of peripheral

2007 The Authors Journal compilation 2007 Blackwell Publishing Ltd, Int. Jnl. Lab. Hem. 2007, 29, 261–278