VOL: 97, ISSUE: 15, PAGE NO: 36
Robert Pratt, MSc, RN, RNT, DN, is professor of nursing at the Richard Wells Research Centre, Thames Valley University, London
Carol Pellowe, MA Ed, RN, RNT, is deputy director and principal lecturer;Peter Harper, BSc, DipN, RN, RNT, CertEd, is senior lecturer (research);Heather Loveday, MA, RN, RNT, is principal lecturer (research); andNicola Robinson, PhD, BA, BSc, is reader; all at the Richard Wells Research Centre, Thames Valley University, LondonBloodstream infections associated with the use of central venous catheters (CVCs) are among the most severe complications that can occur (Box 1). They endanger health, prolong the duration of hospital stays and increase the cost of care (Emerson et al, 1996; Arnow et al, 1993). Every year, almost 6,000 patients in the UK acquire a catheter-related bloodstream infection (CR-BSI) (Fletcher and Bodenham, 1999a).
Bloodstream infections associated with the use of central venous catheters (CVCs) are among the most severe complications that can occur (Box 1). They endanger health, prolong the duration of hospital stays and increase the cost of care (Emerson et al, 1996; Arnow et al, 1993). Every year, almost 6,000 patients in the UK acquire a catheter-related bloodstream infection (CR-BSI) (Fletcher and Bodenham, 1999a).
Guidelines on preventing CR-BSI
Using the methods described by Harper et al (2001), recommendations for the prevention of CR-BSI have been developed and grouped in seven categories (Box 2). These recommendations are linked to different types of appraised evidence, but each is equally endorsed and none is regarded as optional.
For a wider discussion of the evidence and the evidence category for each recommendation, refer to the complete guideline publication (Pratt et al, 2001a) and the associated Technical Report (Pratt et al, 2001b). It is essential that the standard principles for the prevention of infection are always used (Loveday et al, 2001).
The right catheter for the right patient
Selecting the most appropriate catheter for each patient reduces the risk of subsequent infection.
CVCs with more than one lumen
Clinicians often use multilumen CVCs because they allow for the concurrent administration of fluids and medications and haemodynamic monitoring of critically ill patients. However, multilumen CVCs are thought to be more prone to infection than single-lumen catheters as a result of increased trauma at the insertion site and because multiple ports increase the frequency of CVC manipulation. There is conflicting evidence, but most shows that multilumen catheters carry a higher risk of infection (Pearson, 1996; Pratt et al, 2001b).
- Use a single-lumen catheter unless multiple ports are essential for patient management;
- If total parenteral nutrition is being given, use one CVC or lumen exclusively for that purpose.
Tunnelled CVCs and totally implantable devices
Surgically implanted (tunnelled) CVCs - Hickman catheters, for example - are commonly used to give stable anchorage and provide vascular access to patients who need long-term intravenous therapy. Alternatively, totally implantable intravascular devices - Port-A-Cath, for example - are also tunnelled under the skin but have a subcutaneous port or reservoir with a self-sealing septum that is accessible by needle puncture through intact skin.
Most studies report lower rates of infection in patients with tunnelled CVCs. Most also state that totally implantable devices have the lowest reported rates of CR-BSI (Pearson, 1996; Pratt et al, 2001b).
A recent meta-analysis focused on the value of tunnelling short-term CVCs to prevent catheter-related infections (CR-infections). Data synthesis showed that tunnelling reduced catheter colonisation (Box 3) by 39% and CR-BSI by 44% in comparison with non-tunnelled placement. Reviewers concluded that tunnelling decreased CR-infections, but their synthesis of the evidence does not support routine subcutaneous tunnelling of short-term subclavian venous catheters until its efficacy is evaluated at different placement sites and relative to other interventions (Randolph et al, 1998a).
- Use a tunnelled catheter or an implantable vascular access device for patients in whom long-term (>30 days) vascular access is anticipated.
Using an antimicrobial-coated CVC
Some CVCs are impregnated or coated with an antimicrobial substance or substances and some evidence suggests that these reduce the incidence of catheter colonisation and CR-BSI in some situations (Pratt et al, 2001b). Two types of antimicrobial-impregnated CVCs are available in the UK. One bonds the antimicrobial agent (silver ions) with the polyurethane; the other bonds two antimicrobial agents (chlorhexidine and silver sulphadiazine) onto the catheter surface.
A meta-analysis showed that chlorhexidine- and silver-sulphadiazine-bonded CVCs reduced catheter colonisation and CR-BSI by 40% in patients at high risk, such as those in intensive care units and those receiving total parenteral nutrition (Veenstra et al, 1999a). Another study found that the use of these CVCs in high-risk patients reduces the incidence of CR-BSI and death and provides a significant cost reduction (Veenstra et al, 1999b).
- Consider the use of an antimicrobial-impregnated CVC for adult patients who require short-term (<10 days)="" catheterisation="" and="" who="" are="" thought="" to="" be="" at="" high="" risk="" of="" cr-bsi.="">10>
The best insertion site to minimise risk
Several factors are assessed to determine the best catheter insertion site (Box 4).
The insertion site may influence the risk of infection. CVCs are generally inserted into the subclavian, jugular or femoral veins, or peripherally inserted into the superior vena cava by way of the cephalic and basilar veins of the antecubital space.
Subclavian, jugular and femoral placements
Multiple studies concluded that CVCs inserted into subclavian veins carry a lower risk of infection than those inserted into the jugular or femoral veins, but none of these tests was a randomised controlled trial (Pearson, 1996; Pratt et al, 2001b). Internal jugular insertion sites may pose a greater risk of infection because of the proximity of oropharyngeal secretions and because CVCs at this site are difficult to immobilise. However, mechanical complications associated with catheterisation may be less common with internal jugular insertion than with insertion into the subclavian vein.
We found no further evidence from randomised controlled trials that assessed the risks of infection associated with catheter insertion into the subclavian, internal jugular, or femoral veins (Mermel, 2000). All factors being equal, subclavian vein insertion has the lowest the risk of infection.
Peripherally inserted central catheters (PICCs) may be used as an alternative to subclavian or jugular vein catheterisation. PICCs are inserted into the superior vena cava via the cephalic and basilar veins of the antecubital space. They are cheaper, tend to have fewer mechanical complications, such as thrombosis or haemothorax, and are easier to maintain than short peripheral venous catheters because the need for frequent site rotation is reduced. There is also evidence to suggest that PICCs are associated with lower rates of infection than are other non-tunnelled CVCs (Pearson, 1996; Pratt et al, 2001b). This may be because the antecubital fossa is less colonised by micro-organisms and is less oily and moist than the chest or neck. Antecubital placement situates the catheter away from endotracheal and nasal secretions.
- In selecting an appropriate insertion site, assess the risks of infection against the risks of mechanical complications;
- Unless medically contraindicated, use the subclavian site in preference to the jugular or femoral sites for non-tunnelled catheter placement;
- Consider the use of peripherally inserted catheters as an alternative to subclavian or jugular vein catheterisation.
Optimum aseptic technique reduces risk
Strict adherence to hand decontamination and aseptic technique is widely accepted as the cornerstone of the prevention of infection. While this alone seems enough to prevent infections associated with the insertion of short peripheral venous catheters, it is recognised that central venous catheterisation creates a significantly greater risk of infection.
Evidence shows that if maximal barrier precautions are used during CVC insertion, catheter contamination and subsequent CR-BSI is significantly minimised (Raad et al, 1994; Pearson, 1996; Pratt et al, 2001b). Experts who have examined this evidence agree that maximal sterile barrier precautions are essential during CVC placement to reduce the risk of infection (Ward et al, 1997; Fletcher and Bodenham, 1999b).
It is generally agreed that CVCs inserted in the operating theatre pose a lower risk of infection than those inserted on inpatient wards or other patient-care areas. However, evidence suggests that the difference in the risks of infection depends more on the extent of the barrier protection used during catheter insertion than on the surrounding environment (Raad et al, 1994; Mermel et al, 1991).
- Use optimum aseptic technique, including a sterile gown, gloves and a large sterile drape, for the insertion of the CVC.
Preparation of insertion sites
Micro-organisms that colonise catheter hubs and the skin around the CVC insertion site are the cause of most CR-BSIs. Skin cleansing/antisepsis of the insertion site is one of the most important measures for the prevention of catheter colonisation and CR-BSI (Pearson, 1996). Evidence showed that 2% aqueous chlorhexidine was superior to both 10% povidone-iodine and 70% alcohol in the prevention of both central venous and arterial CR-infections (Maki et al, 1991; Mimoz et al, 1996).
There have been no direct comparisons of aqueous versus alcoholic solutions of chlorhexidine in relation to cutaneous antisepsis for the prevention of CR-infections, but an alcoholic solution of chlorhexidine combines the benefits of rapid action with excellent residual activity (Larsen, 1988).
There is no evidence to show that organic solvents such as acetone or ether, which are sometimes used to 'defat' the skin (remove skin lipids), have any impact on the incidence of infection. Their use could also greatly increase local inflammation and patient discomfort (Pearson, 1996). Nor is there evidence to show any benefit from using antimicrobial ointments on the insertion site at the time of catheter insertion (Pearson, 1996).
- Clean the skin site with an alcoholic chlorhexidine gluconate solution before CVC insertion. Use an alcoholic povidone-iodine solution for patients with a history of chlorhexidine sensitivity. Allow the antiseptic to dry before inserting the catheter;
- Do not apply organic solvents, such as acetone or ether, or antimicrobial ointment to the skin before catheter insertion.
Using good catheter and catheter-site care
Maintenance of a safe CVC and care of the catheter site are essential to prevent CR-infections. Good practice includes caring for the catheter hub and connection port, the use of an appropriate catheter-site dressing regimen and the use of flush solutions to maintain the patency of the catheter.
The catheter hub and connection port
Evidence suggests that contamination of the catheter hub is an important contributor to intraluminal microbial colonisation of catheters, particularly in the case of long-term catheters (Pearson, 1996; Pratt et al, 2001b).
Frequent catheter-hub manipulation increases the risk of microbial contamination and is the main cause of CR-BSI in long-term use (Mermel, 2000; Raad et al, 1993).
Catheter hubs and sampling ports should be disinfected before they are accessed; both povidone-iodine and chlorhexidine are effective for this purpose (Maki et al, 1997).
Some catheter and hub materials, such as polyurethane and silicone, may be incompatible with alcohol or iodine, and it is important to adhere to the manufacturer's recommendations.
- Before gaining access to the system, disinfect the external surfaces of the catheter hub and connection ports with an aqueous solution of chlorhexidine gluconate or povidone-iodine, unless this is contraindicated by the manufacturer's recommendations.
Choosing the right dressing for CVC sites
Occlusive dressings keep the skin moist and provide an ideal environment for the growth of local microflora, so CVC dressings must be permeable to water vapour (Fletcher and Bodenham, 1999a). The two most common types of dressing used for CVC sites are sterile, transparent, semipermeable polyurethane dressings (transparent dressings) and gauze-and-tape dressings. Transparent dressings, such as Opsite IV3000 and Tegaderm, are popular because they reliably secure the CVC, permit continuous inspection of the catheter site, allow patients to bathe and shower without saturating the dressing and need to be changed less often than standard gauze-and-tape dressings. They also save time. Both can be used safely (Pratt et al, 2001b).
Studies that focused on the use of antimicrobial ointment applied beneath the dressing to the catheter insertion site to prevent CVC-related infection did not clearly demonstrate efficacy (Maki and Band, 1981).
- Use either a sterile gauze or transparent dressing to cover the catheter site;
- If a gauze-and-tape catheter-site dressing is used, it must be replaced when the dressing becomes damp, loose or soiled, or when inspection of the insertion site is necessary;
- Do not apply antimicrobial ointment to CVC insertion sites as part of routine catheter-site care.
Prevention of catheter thrombosis
The relationship between vascular thrombosis, microbial adherence and CR-related infection is widely recognised (Pearson, 1996). See Box 5.
A recent meta-analysis concluded that heparin significantly reduces bacterial colonisation and shows a strong but non-significant trend towards reduction of CR-BSI (Randolph et al, 1998b).
Some types of tunnelled CVCs, such as Groshong catheters, may not need to be routinely flushed with an anticoagulant.
- Indwelling CVCs should be flushed routinely with an anticoagulant unless the manufacturer recommends otherwise.
Appropriate replacement of CVCs
CVC replacement comprises two elements: the frequency and the method of replacement.
Frequency of replacement
Some studies recommend replacement of CVCs at specified intervals to reduce infection; others state that the daily risk of infection remains constant and that routine replacement of CVCs, without a clinical indication of the need for it, reduces neither the rate of catheter colonisation nor the rate of CR-BSI.
Conclusions from a recent systematic review agree that exchange of catheters by any method every three days is not beneficial compared with an as-needed basis (Cook et al, 1997).
Two methods are used: a new catheter is either placed over a guide wire at the existing site (Box 6) or inserted percutaneously at another site. Studies of the risks of infection associated with guide-wire insertions showed conflicting evidence (Pearson, 1996; Pratt et al, 2001b).
A recent systematic review concluded that guide-wire exchange was associated with a trend towards higher rates of catheter colonisation, catheter exit-site infection and CR-BSI. However, guide-wire exchange was associated with fewer mechanical complications (Cook et al, 1997).
- Do not routinely replace non-tunnelled CVCs as a method of preventing CR-infection;
- Use guide-wire-assisted catheter exchange to replace a malfunctioning catheter or exchange an existing catheter if there is no evidence of infection at the catheter site or proven CR-BSI;
- If CR-infection is suspected but there is no evidence of infection at the catheter site, remove the existing catheter and insert a new one with a guide wire. If tests reveal CR-infection, the newly inserted catheter should be removed and, if still required, a new catheter inserted at a different site;
- Do not use guide-wire-assisted catheter exchange for patients with CR-infection. If continued vascular access is required, remove the implicated catheter and replace it with another one at a different insertion site.
Replacement of intravenous administration sets
The evidence examined showed that replacing administration sets 72 hours or more after initiation of use is safe and cost-effective. Certain intravenous fluids, such as blood, blood products and lipid emulsions, are more likely than other parenteral fluids to support microbial growth if contaminated, and more frequent replacement of intravenous tubing may be needed (Pearson, 1996; Pratt et al, 2001b). See Box 7.
- Replace all tubing when the vascular device is replaced;
- Replace intravenous tubing and stopcocks no more often than at 72-hour intervals, unless clinically indicated;
- Replace the intravenous tubing used to administer blood, blood products or lipid emulsions at the end of the infusion or within 24 hours of initiating the infusion.
Avoid unnecessary antibiotic prophylaxis
Prophylactic administration of systemic antimicrobials has been used to reduce the incidence of CR-BSI. The evidence is inconclusive and this practice may select for resistant micro-organisms (Pearson, 1996; Pratt et al, 2001b).
- Do not administer systemic antimicrobials routinely to prevent catheter colonisation.
CR-BSI is one of the most serious complications in an already seriously ill patient. Incorporation of these recommendations into local protocols and routine clinical practice will help to bring about a significant reduction in the incidence of CR-BSI in all NHS acute care trusts. The evidence base will be reviewed in 2002.
- This is the final article in our series on the recently published national evidence-based infection prevention guidelines. Parts one and two, both published in the March 29 issue of Nursing Times, describe the development of the guidelines and outline the standard principles. Part three, published in the April 5 issue, looks at guidelines concerned with preventing infections associated with the use of indwelling urethral catheters.