1.6 Intravascular catheter sepsis.pdf - ARTICLES

Document Sample
1.6 Intravascular catheter sepsis.pdf - ARTICLES Powered By Docstoc

        INTRAVASCULAR CATHETER                                                 annually by clinics and hospitals in the USA. 1 This includes
                                                                               more than 5 million central venous and pulmonary artery
        SEPSIS                                                                 catheters.
                                                                                  Catheter-related infections (CRI) remain among the top three
        Mervyn Mer                                                             causes of hospital-acquired infections, with a mortality of up to
                                                                               25%, and result in prolonged hospitalisation (mean of 7 days)
                                                                               and increased medical costs.HO The estimated cost of treating
        lntravascular devices are an integral component of modem-day
                                                                               one episode 01 catheter-related bloodstream infection (CRBS1)
        medicaJ practice. They are used to administer intravenous
                                                                               in the USA ranged from 58 000 in 1988 to more than 528 000 for
        fluids, medications, blood products and parenteral nutrition. In
                                                                               intensive care patients in 1994. On the basis of these figures, the
        addition, they serve as a valuable monitor of the haemo-
                                                                               economic burden from CRBSl is substantial.
        dynamic status of critically ill patients.
                                                                                  Central venous catheters (CVCS) account for an estimated
                                                                               90% of all CRBSl.lI Rates of bloodstream infection range from 4
        BACKGROUND AND HISTORY                                                 to 13 per 1 000 central catheter days,n with lower rates in
       Only a century ago, no means of vascular access existed for the         respiratory intensive care units and higher rates in bums units.
       lile-sustaining support 01 critically ill patients. In the late 1800s     Given the magnitude and seriousness of the problem of CRI,
       steel needles became available, and with the advancing                  it is essential for health care workers to have a clear
       knowledge of electrolyte physiology the therapeutic use of              understanding of the diagnosis, pathogenesis, prevention and
       intravenous fluid became established. In 1945, following the            treatment of this problem and the new developments in the
       advent of penicillin and the need for multiple intravenous              field. Most of these infections can be reversed with appropriate
       injections, plastic catheters for continuous vascular access were       diagnosis and treatment, and many can be prevented.
       describedY A further technological advance took place in 1%7
       when the placement of long nylon catheters into central veins,
                                                                               FORMS OF CATHETER SEPSIS
       to limit medication-associated phlebitis, was described in
       oncology patients.) These catheters were initially inserted by          Definitions
       peripheral cutdown techniques and later via percutaneous
       approaches into the subclavian and jugular veins. Over the past         Definitions relating to intravascular catheter sepsis have been
       15 years the focus of research and development has been on the          put forward by various workers, but many have complicated
       physicochemical properties 01 catheters, looking at such aspects        matters and been confusing. This has in part related to the fact
       as improved catheter materials, tensile strength, rupture               that definitions used for surveillance and research purposes
       resistance, biocompatibility and the creation of catheter micro-        have diffened from those used lor clinical diagnosis. The
       environments hostile to invading organisms.                             Centers for Disease Control and Prevention in Atlanta, Georgia,
                                                                               have suggested sensible definitions1J which allow for the use of
         Intravascular devices have therefore been a major advance in
                                                                               both clinical and laboratory evidence of catheter sepsis. These
       terms of patient comfort and care, but with them has come the
                                                                               should be universally used in the definition of intravascular
       burden of complications, including a variety of local and
                                                                               catheter sepsis and are documented in modified form in
       systemic infectious complications.
                                                                               Table I.
          In general, intravascular devices can be divided into those
       used lor short-term (temporary) vascular access and those used
       for long-term (indwelling) vascular access. Long-term
       intravascular devices usually require surgical insertion. while
                                                                                T~le   L Definitions for infections
       short-term devices can be inserted percutaneously. The main
       focus of this review relates to short-term catheters.                    Catheter colonisation.: growth of:;e 15 colony-forming units
                                                                                (semi-quantitative culture) &om a proximal or distaI catheter
                                                                                segment in the absence of local or systemic infection
       ~GNTnJDEOFTHEPROBLEM                                                     Local infection: erythema, tenderness,. induration or purulence
                                                                                within 2 cm of the skin insertion site of the catheter
El Although no Specific local statistics are available, more than               Catheter-re1ated bloodstre.a.m. infection: isolation of the same
       150 million intravascular devices are currently purchased                organism (i.e. the identical species as per antibiogram) &om
                                                                                culture (semi-quantitative or quantitative) of a catheter segment
                                                                                and from the blood (preferably drawn from a peripheral vein)
                                                                                of a patient with accompanying clinical symptoms and signs of
       lntmsitv Can' Unit, johanntsburg Hospital, and Departmmt of Medicine,    bloodstream.infection and no other apparent source of sepsis
       Unrverslty of the Wittl'atersrand, Johannesburg
       Mervyn Mer, MB BCh, Dip PEC (SA), FCP (SA)

       October 1999, VolIS, No. 2 SAICC
                                                                       Establishing a diagnosis of CRI involves both clinical and
The skin around the insertion site is the most common portal of        laboratory components. The clinical features are generally
entry. I .. ,.. The current Wlderstanding is that a fibrin sheath      nonspecific and indude fever, rigors, hypotension and
develops around the catheter which promotes the adherence of           confusion. If there is no apparent source of sepsis in a patient
pathogens. This is referred to as the biofiJrn layer. Skin             with an intravascular line (especially a central venous catheter)
organisms then migrate from the skin insertion site along the          and if the sepsis appears to be refractory to antimicrobial
external surface of the catheter to colonise the distaI                therapy or is of abrupt onset and associated with shock. the
intravascular tip and ultimately cause bloodstream infection.          possibility of line-related sepsis needs to be considered.
Contamination of the catheter hub during its manipulation by           Fundoscopy should always form part of the clinical
medical and nursing personnel is the second most common                examination as focal retinal lesions are common in patients
portal of entry of micro-organisms. These organisms migrate            \vith CVC-derived Candida infection, even when blood cultures
along the internal surface of the catheter leading to luminal
                                             l                         are negative. Inflammation or purulence at the catheter
colonisation and thence to bloodstream infection. u,.l7-1!1 Although   insertion site is seen in less than hall the cases.
much less common than either of the above two mechanisms,                The laboratory components include culture of blood and the
haematogenous dissemination from a distal infectious focus or          catheter. Blood cultures ;,re central to the diagnosis of CRBSI.
administration of contaminated infusate may also cause CRI.1lUl        Two to three 10 m.l samples, ideally from separate peripheral
Other sources such as contaminated transducer kits,                    venipuncture sites, should be sent to the laboratory. Paired
disinfectants and infusion lines are also rare causes.                 quantitative cultures, which involve taking blood from both the
                                                                       catheter and a peripheral site, may be particularly useful where
MICROBIOLOGICAL PROFILE OF CATHETER-                                   luminal colonisation is predominant. The diagnosis is
RELATED INFECTIONS (TABLE                        m                     suggested when fivefold or more colonies are isolated from the
                                                                       blood drawn from the vascular catheter as compared with the
The microbiology of CRI reflects a predominance of skin
                                                                       concurrent peripheral sample.1i1:UJ
organisms such as Staphylococcus epidermidis, S. aureus, Bacillus
species and Corynebacterium species (especially JK strains). jK           The most widely used laboratory teduUque for culturing the
bacteraemia occurs almost exdusively in severely                       catheter is the semiquantitative roll-plate metho(pt In this
immunosuppressed patients who are or have been receiving               method, cultures are obtained from a segment of the catheter
broad-spectrum antibiotics and who have indwelling                     after it has been removed from the patient by rolling the
intravascular devices.                                                 catheter segment across the surface of a blood-agar plate at
                                                                       least four times and then determining the number of colonies
                                                                       present after a period of incubation. Growth of ;;;: 15 colony-
 Table n. Common organisms ~ed with atheter--re1at:ed                  forming units from a proximal or distal catheter segment is
 infections                          .                                 regarded as significant. Quantitative teduUques for culturing
 Staphyloa>ccus epidmnidis        Enterobactu species                  the catheter include the sonication and vortexing methods,
 5. aulTUS                        Serratia rrrtlTasan5                 which involve extracting micro-organisrns from the catheter
 Cmuiida species                  Citro/><ldrr Jmutdii                 surface into a medium for culturing. This entails either flushing
 Arinetobocter species            Enterococcus species
                                                                       out the catheter segment and immersion in culture medium or
 Pseudomoruls tletUginosa         Bacitlus species
                                                                       placement of the segment in culture medium with
 Stenatrophomonas maltoplu1ia     Caryntbaderium (especially JK
 Klebsiella species               strains)                             sonication.m73 Quantitative culture is the most sensitive
                                                                       technique for diagnosis of catheter-related infection. Other
                                                                       techniques such as Gram staining of the catheter surface and
                                                                       culture of the tip in broth are associated with high false-
                                                                       positive rates. A newer diagnostic culture technique is that of
   Contamination from the hands of medical and nursing
                                                                       the endoluminal brush. This allows samples to be taken via the
personnel is frequently responsible for infection with such
                                                                       lumen of the catheter with a brush while the catheter remains
organisms as Pseudomonas aeruginosa, Acinetobacter species,
                                                                       in situ. A sensitivity of 95% and a specificity of 84% in the
Stenotrophomonas mllitaphilia and Candida species.c.:~
                                                                       diagnosis of CRI has been reported with this teduUque. 3 -" This
   Emerging pathogens, including species of Enterococcus,              technique does not require sacrifice of the cathet~ but there is
Micrococcus and Achromobacter, rapidly growing mycobacteria            still a delay before culture results are known. There are also
such as Mycobacterium Jorhlitum and M. chelonei, and fungal            concerns that the process of brushing may lead to embolisation
organisms such as Malassezia furfur, Rhodotorula species,              of infected. biofilm. The place of the endoluminal brush in
Fusarium species, Trichosporin species and Hansenula anomala           clinical practice is still to be determined.
have also caused catheter    infections.l~
                                                                   ARTICLES                                                                         J

        PREVENTIVE STRATEGIES FOR                   CRI                      Silver-chelated subcutaneous collagen cuffs
        Strict adherence to halldwashing and aseptic technique remains the   These cuffs may be attached to percutaneously inserted CVCs
        comeTstone of prevention of CRI. However, other measures may         and are designed to act as both a mechanical barrier to the
        confer additional protection and need to be considered in the        migration of micro-organisms and an antimicrobial deterrent
        preventive strategy. These include infusion therapy teams, use       (through the effect of silver ions). They have been shown to
        of barrier precautions during catheter insertion, cutaneous          lower the risk of catheter colonisation and CRBSI in critically ill
        antimicrobials and antiseptics, site of catheter insertion,          patients.X'.JI The anti-infective effect is short-lived, however, as
        tunnelling of eves, silver-chelated subcutaneous collagen            the collagen to which the silver ions are chelated is
        cuffs, antiseptic hubs, catheter-site dressings and the use of       biodegradable. Other drawbacks include cost and the need for
        antimicrobial impregnated catheters.                                 specialised training.

        Infusion therapy team                                                Antiseptic hubs
       The presence of an experienced infusion therapy team whose            These have been designed to protect against hub colonisation.
       task is to insert and maintain catheters has been shown to            A fourfold decrease in catheter-related sepsis has been
       decrease the rate of eRBSI up to eightfold and limit overall          demonstrated with their use.:J9 A major limitation, however, is
       costs.J1.J::! Similarly, strict adherence to peotocals for catheter   that protection is only conferred against organism migration
       insertion in the intensive care unit (lCU) and theatre are also       along the internal surface of the catheter. They do not protect
       beneficial in decreasing the rates of CRI.                            against the migration of skin organisms along the external
        Maximum sterile barriers
       Careful hand washing together with the use of sterile gloves, a
       mask. gown and cap and a large drape have been associated             There has been ongoing debate concerning the best method of
       with a greater than sixfold decrease in eve-related sepsi?            catheter dressing. This has essentially revolved around the
       and a fourfold decrease in the rate of bacteraemia related to         relative merit of gauze and tape dressings versus transparent
       pulmonary artery catheters." The use of this practice cannot be       films. In a meta-analysis of catheter dressing regimens, eves
       over-emphasised.                                                      on which a transparent dressing was used were assoriated with
                                                                             a significantly higher incidence of catheter tip colonisation but
       Cutaneous antimicrobials and antiseptics                              a non-significant increase in CRBSI..jD

       Given the important role of cutaneous microflora in the                 The preference in our unit is an adhesive gauze dressing
       pathogenesis of catheter-related infections, measures to reduce       with a central non-adherent pad.
       cutaneous colonisation of the insertion site are of vital
       importance.                                                           Antimicrobial coating of catheters
         For skin decontamination before catheter insertion in a three-      In recent years, antimicrobial substances have been effectively
       group comparing the efficacy of treatment, 2%                bonded to catheters, especially those designed for short-term
       ch10rhexidine gluconate was associated with a fourfold                use. Two coated cves are currently available, a
       decrease in CRBSI as compared with 10% povidone-iodine and            chlorhexidine/ silver sulphadiazine catheter and a
       70% alcohoL The use of PNB ointment (polymixin-neomycin-              minocycline/rifampicin catheter. Several studies have shown
       badtradn) at the skin entry site has been associated with a           potential benefits of such catheters in terms of reduction of
       lower rate of CRBS!; however, the overall protective effect is        catheter colonisation as well as CRBSLuo,.u-c
       offset by a higher risk of fungal colonisation and infecti.on.l3        A potential drawback of the chIorhexidine/silver
         It is the practice in our unit to use a chlorhexidine gluconate-    sulphadiazine catheter, however, is that the coating is applied
       containing solution for skin preparation.                             only to the external surfaces and does not protect against
                                                                             endoluminal colonisation as a result of hub contamination. The
       Tunnelling of evCs                                                    minocycline / rifampicin catheter is coated on both the external

lIB This involves placing the proximal segment of the catheter               and internal surfaces and may therefore be more effective."
                                                                             One of the concerns about the use of antimicrobial-
       under the skin at a distance from the point of entry to the vein.
                                                                             impregnated catheters relates to the possible development of
       A lower rate of CRBSI has been reported in one study in
                                                                             antimicrobial resistance, and where they are used continued
       critically ill patients.J6 More data are required to support this
       observation.                                                          surveillance for resistance is required.

       October 1999. Vol. 15. No. 2 SAJCC

TREATMENT PRINCIPLES OF CATHETER-                                   Central venous catheters
                                                                    cves account for an estimated 90% of all CRBSI.         on-
Treatment depends on the stage of infection and the pathogen.       tunnelled (percutaneously) inserted CVCS are the most
As a general rule if CRBSI is suspected the catheter must be
                                                                    commonly used central catheters. A host of risk factors for
removed and replaced only if necessary. Most of the infectious      evC-related infections have been reported, induding duration
complications are sell-limited and resolve after removal of the     of catheterisation,. location of the catheter (the internal jugular
catheter.                                                           having a higher rate of CRI than the subclavian vein), the
  Indications for antibiotic therapy include persistent sepsis      presence of sepsis, type of dressing, multi-lumen catheters
despite catheter removal, evidence of septic thrombosis of the      (increased frequency of manipulation), less stringent barrier
great veins, clinical or echocardiograpruc evidence of              precautions during placement, experience of personnel
endocarditis, metastatic foci of infection.. underlying valvular    inserting the device, and the administration of parenteral
heart disease (especially prosthetic valves), and an underlying     nutrition.
immunosuppressed state.                                               The duration of evc use remains controversial. Despite this,
  In terms of specific pathogens and CRBSI, S.   aUTellS   and      however, no catheter should be left in place longer than
Candida species require special mention. In the setting of          absolutely necessary. ThE:' duration of catheterisation has been
                                                                    shown to be a risk factor for infection in several studies!C·S!; and
uncomplicated S. Qureus eRBSI, the cathe~er should be
removed and 2 weeks of parenteral antibiotics given. There is a     scheduled replacement remains widely practised in most ICUs.
high relapse rate if these are given for a shorter time.e.~         In a recently performed study in mainland Britain, where 165
                                                                    ICUs were surveyed,>6 catheters were routinely replaced, the
  Systemic antifungal therapy (together with removal of the
                                                                    mean time being 6.5 days.
catheter) should be given in all cases of catheter-related
candidaemia in view of the potentially significant sequelae. 11       We have recently completed and analysed a evc study in
Amphotericin B or fluconazole (except for fluconazole-resistant     the multidisciplinary lCV at Johannesburg Hospital. The study
organisms such as Candida glabrata and C. krllsei) should be        was a prospective randomised double-blind study which
commenced. Fluconazole 400 mg daily for at least 14 days has        entailed comparison of a 14-day placement of standard triple-
                                                                    lumen versus antimicrobial-impregnated (chlorhexidinelsilver
been shown to be as effective as amphotericin B 0.5 mg/kg/d
                                                                    sulphadiazine) cves on the rates of CRI. Our aim was to
for 14 daysl with fluconazole being less toxic.-li
                                                                    determine whether we could safely increase the duration of
                                                                    insertion time from 7 days to 14 days and the influence of the
SPECIFIC CATHETER TYPES AND INFECT10                                antimicrobial-impregnated catheter on the incidence of CRI.
                                                                      One hundred and eighteen critically ill patients were
Short peripheral intravenous catheters
                                                                    included in the study, which spanned 3-1951.5 catheter hours (1
These remain the most commonly used intravascuJar devices.          456 catheter days). Sixty-two patients received a standard
There is a significant risk of contamination 72 hours after         catheter and 56 an antimicrobial·impregnated catheter. The insertion site should be an upper extremity or     Eighteen of the patients developed a CRBSI, 1 of whom died,
the external jugular vein. There is a greater risk of infection     and 5 patients demonstrated catheter colonisation. This rate of
with lower extremity sites and "vith cutdowns.                      CRBSI compares favourably with those previously reported, in
                                                                    which many of the catheters were in place for shorter periods
Peripheral arterial catheters                                       of time than in our study.lU9
Peripheral arterial catheters are associated with less infection      The most frequent source of infection was the skin.. followed
than pulmonary artery catheters (PACsl, CVCs and short              by hub and infusate contamination and lastly haematogenous
peripheral catheters.~ This may be explained by high arterial       seeding. The source; and organisms were identified with the
flow around the catheter, which probably decreases the              aid of restriction-fragment length DNA subtyping.
adherence of The Centers for Disease               We were unable to show any difference in CRJ rates behveen
Control and Prevention guideline u suggests that replacement        the two types of catheters in the study. Most importantly we
of catheters and relocation of insertion sites need take place no   were able to condude that standard     cves  can safely be left in
more frequently than every 4 days.                                  place for 14 days (with stringent infection control measures).
  It is our unit policy to keep peripheral arterial catheters in    Parenteral nutrition was not noted to be a risk factor for
place for up to 30 days prior to replacement and relocation,        catheter sepsis, and neither was the site of insertion (internal
unless otherwise indicated.                                         jugular vein versus subclavian vein).
                                                                      On the basis of the results of this study, it is now our practice
                                                                    to keep standard cves in place for 14 days unless there is an

                                                                               Peripherally inserted central venous catheters
         Table ID. Protocol for inse.rt:ion of central venous catheters        (PICes)
          • Oean the skin around the insertion site over a wide area by
             rubbing for 2 minutes with sterile gauze or rottonwool            PICCs provide an alternative to subclavian or jugular vein
             soaked in a ch.Iorhexidine gIuconate-containing solution.         catheterisation and are inserted into the superior vena cava or
             Sterile gloves must be worn..                                     right atrium \ria the cephalic and basilar veins of the
          • The doctor,. wearing a mask and Gip, scrubs up (using a            antecubital fossa. Compared with other cves they are
             chImhexidine gluconate-rontaining scrub solution) and then        associated with few mechanical complications, an apparent
             dons a sterile gown and gloves.
                                                                               lower rate of infectionl',oll and decreased cost. The length of
          • 1he doctor then deans the area again and drapes widely to
                                                                               time that these catheters can be left in place safely has not yet
            include the patienrs head, neck. chest,. limbs and torso down
            to the pelvis. Only the portion necessary for catheter insertion   been determined, although they have been used successfully
            should be left exposed.                                            for extended periods.
         • The 'flush' (heparin 1 (XX) ill in 19 ml sterile saline) is drawn
            up avoiding any contamination by the doctor after cleansing        G uidewire exchanges
            of the stopper on the heparin container. The doctor draws up
            the 'flush' with a sterile syringe and needle, while the           A recent meta·analysis of evc replacement strategies revealed
            assistant holds the vials.                                         that guidewire exchanges were associated with greater risk of
         • Once the line has been inserted, a sterile piece of gauze           CRI but fewer mechanical complications than new-site
            soaked in a chlorhexidirle gluconate-containing solution is        replacement.51 If guidewire exchange is used. metirulous
            applied over the insertion site and adjacent area for              aseptic technique is necessary. This procedure should not be
            approximately 30 seconds.                                          perfonned in the setting of confirmed or clinically suspected
         • The area is then dried with sterile gauze and an adhesive
                                                                               sepsis. In our unit we do not practice guidewire exchanges.
            gauze dressing with a central non-adherent pad applied.
         • The dressings are changed daily and the insertion site
            inspected and cleaned in a sterile manner. Cleaning includes       AOOITIONAL RECOMMENDATIONS TO LIMIT
            removal of old blood, dots, exudates and crusts and the            INFEcrION
            application of a chIomexidine gluconate-soa.ked piece of
            sterile gauze to the insertion site for approximately 30           On the basis of the results of our study, previous guidelines13
            seconds, before drying and dressing the area                       and the cumulative anecdotal experience in our unit, both
         • Any signs uf local infection (red. hot, swollen. painful,           nursing and medical, we now have a dedicated policy
            purulence) must be reported.                                       regarding the insertion, maintenance and use of intravascular
                                                                               devices. The basic principle revolves around strict adherence to
                                                                               aseptic technique at all times. It is our present policy to change
       indication for earlier removal. This practice goes hand-m-hand          central venous and haemodialysis catheters after 14 days,
       with a stringent protocol relating to aseptic insertion technique       peripheral venous catheters after 3 days and arterial lines after
       and care of the catheter. A modified fonn of this protocol              30 days, unless removal is indicated beforehand. Lines used for
       appeiUS in Table ffi.                                                   the administration of blood products must be replaced. within
                                                                               24 hours. [jpid-containing parenteral nutrition solutions
       Pulmonary artery catheters                                              should be completed within a 24-hour period. Parenteral
       Varying rates of infection have been reported with PACs                 nutrition must be administered via a single dedicated port with
       (Swan.(;anz catheters), but most are similar to CVCs. Where             the administration line being replaced at 24-hour intervals
       higher percentages have been reported, this has been attributed         (performed as a sterile procedure). Administration sets such as
       to the number of manipulations performed. The 'Hahds-Off                those used. for the delivery of inotropes and antibiotics are
       Catheter', which is endosed in a contamination·proof shield             replaced at 72·hour intervals, or before if clinically indicated.
       enabling the doctor to prepare, test and insert it without              The day on which lines are changed should be dearly noted on
       exposure to external contamination. has been associated with a          the rcv chart
       decrease in systemic infection.~ Most PACs are heparin-                    It is our policy to replace bridges and their attached lines,
       bonded, which reduces catheter thrombosis and microbial                 transducers and continuous flush devices every 7 days. This is
m:I adherence." The current Centers for Disease Control and                    longer than the recommended %-hour interval,u but it is our
       Prevention guideline recommends catheter replacement at least           experience that provided there is strict adherence to aseptic
       every 5 days. U It is our current practice to keep in PACs for up       technique, the infection risk is not increased. Aseptic technique
       to 7 days if necessary; by which time the patient &equently no          also extends to care of ports and caps attached to intravasrular
       longer requires this form of catheter.                                  devices and includes the spraying of a chlomexidine gluconate-
                                                                               containing solution following manipulations.

       October 1999, Vol1S,     0.2 SAJCC
                                                                                                          34..Mm DC. Ringer M. Alvando q. Prospective randomised trial of p m ~ 31ooho1
CONCLUSION                                                                                                    and chknhexidine for p......."tion of infa1ions ~ with Q5ltral venous and IlI12rial
                                                                                                              ca.theters. Lmat 1991; J38: 339-343.
Intravascular catheter-related ~psis remains a major problem.                                             35. Maki OC. Band JO. A <Dm.parative study of polyantibioticand iodopher oirttmaIts in
                                                                                                              preo..entioo. of vascular catheter-relared infection. NIl I Mtd 1961; 7Q: 739-744.
Stringent adherence to aseptic technique and infection control                                            36. TunsitJ-F. 5ebille V, Fadas J-C. rt lI1.. Effect of subcutani'OuS tunnelling on intftna1 jugular
measures remain the cornerstone of prevention.                                                                catheter-related sepsis in aiticaUy ill patients: a prospective randomized multicenter wdy.
                                                                                                                 lAMA      1996; 'I/6: 1416-142il.
                                                                                                          :rl.   Mm DC,. Cobb 1.. Gimnan JK. Shapiro JM.          Ringet M. HeJ~ Ra. An attachable silver-
                                                                                                                 impregnah.'d cuff for p.....·ention of infection with CVes: a prospective randomised
                                                                                                                 multicelter trial. Ant 1 Mc! 1981l; IlS: 307-31-t.
1. Zirnmennan B.lntravenous tubing for p<m'f'~ therapy. 5cirna 1945; 101;567-568.                         38.    FkJWer5  RH. SdIwenzer Kf, KopeI RF. d lI1.. Efficacy of an attachable subrutane0u5 cuff for
2. Mye.s L InttaVl'l1OU5 catheterization. Am J Nuts 1945; 45: 930-931.                                           the preo.·ention of intrava5Cu1ar ca.theter-rel.ated infection..JAMA 1989; 2liil: 878-883.
3. RDy RB. Willcimon RH, Bayliss CE. The utilization of long ny10n catheters for                          39. Segurn M. Alva=:-Lenna F. TeUado JM.. rt lI1.. Advannos in swgical technique: .. diniotl trial
   prolonged intr;1Y8\OUS infusions. Gm MaJ As5Dc J 1967; 96: 94-97.                                          on the preYention of ca.theter-rel.ated ~is using a new hub model J\Jut S"''i 1996; 213; 363-
4. Maki DG. 11Ifrdi<m CIl...w Iry ,..hvr1tIscu!aT Ckui«5: p,.~, Stnrkgios for Prn>crn-.                          369.
   London: RoyaI5ociety of Medicine Servi<.Es. 1991.                                                      40. Hoffman KK. Weber DJ, Samsa GP. Rutala WA Transparent po!yunothane film as an
5. Smith R4 ~ 5M.. 5imbedoff MS. Excess mortality in critically ill patients with                             intrave<lDlJ"> ~ dnos6ing: a meta-analysis cl the infection risks.. lAMA 1992; Ut7: 1I'!l2-
   nosocomial. bloodstream infections. C1IlSt 1991; 100: 16+-167.                                                2ll7.
6. Martin MA. Pf<iller MA. WenZ2I. RP. ~tivt,> staphylocDcc:al                                            41.                                         s.
                                                                                                                 Civetta JM. H~J. Ball De<::rea$ing catbet2r-rel.ated and hospital <XI5l$ by
      ~               Mortality and hospital stay. A1UIlrrtDn Md 1989; UO: 9-16-                                 continuous quality imPrD\'el\ent. Crit GIrl' Mtd 1996;!'l: 166().I665.

 7. Haley RW, Schabefg OR, Van Allmm. SO, McGowan JE jun. Estimating the extra d1a:Be5 and                42.    Raad L Oarouiche R,. Ha<:hem R. Mansouri M. Bodey Gp. The broad spectrum adivity and
    prolongation of hospitalization d~ to /IOI:IOCOlD.ial infections: a comparison of mi'thods. /                effiGlC}' of calheter5 mated with minocydine and
                                                                                                                                                               rifampicin. I Infrd Vis 1996; 17.1: 4]8424..
    mfrct Vis 1980; 141: 248-2:>7.                                                                        43. Raad L OarouIche R. Dupui:s J, d fll. Central VenolG ca.theters cooted with IIIinocvdine and
 8. Pillet D. Torrara 0, WenzeI RP. Nosocomial bloodstream infection in critiGilly ill patients:              rifampKin for the pre-vmtion of catheter relared cokxtizatiDn and bIood-weam We.:oon. Ann
    ex~ Imgth of stay. extra C05t5 and attributable mortality. JAMA l~ Z71; 1596-1601.
                                                                                                                 (n.!ern   Mtd 19'97; 121: 267-Z74_
 9. Arrow PM.<Sng EM, Brech M. Consequ~ of intravascular catheter sq>sis. Oin                       44.. Darouiche RO. Raad Il,. Heard SO. rt al. A comparison of two antimicrobial-impregnah.'d
    fnfrrt Dis 1993; 16: 778-784-.                                                                             e=traI verIOUSalherers. N Engt J....!t'd 1999; 34Q: 1-8.
10. HeiseInml D. Nosoromial bloodstream um..Don5 in the aitica11}' ill.1AMA 1994;?n; 1819-                45. Raad It Sabbagh MF. Optimal duration of thatpY for catheter-related ShIpIry!oa:lm<.s llllml$
                                                                                                               bacteremia: a study of 53 cases and review. Rn> Irrf«/. Vis 199'2; It: 75-82-
11. Mak:i DG. Irtfe<:tioo5; dUI' 10 infusion ther3py. In: Ilertn<!I:t IV. erachman PS, eds- H">pitw       46. Malano!W GJ. Samure 1.1H, Pefanis A. I<arduner AW. Shlplryloaxcu5 llUrtuS ~
    blfrctim5. 3nl ~ Boston, Mass:: little. Brown. 1992.                                                       associated b3cteremia: minimal effective therapy and unusual infuctious axnplicatiDns
                                                                                                               associated with artviaI shmth catheter$. ArdIl"fun MLJi 1995; 155: 1161-1166-
12. National No5ocDmiallnfections Surveillanc:e (NNIS) Systo= report. data ~ from
    0dDber 1956 _April 1998. i.ssua!d June 1996. Am Jlrrf«t Contl'lllI9'J8; 26: 522-533.                  47.    Rose HD. Venous catheter-associated candidemia. Am I MftI       sa 1978; 215; 265-269.
                                                                                                          48.    Rex JH.Benr.otI lE. Sugar AM. d fll. A randomized trial oxnparing £Iuc:nnazule with
l~ Pear!ion ML Hospital Infection Contn>l. Practices Advi50ry Committa'. Guideline fur
    prevention of irrtra~devio:- related infuctions./nfrd Ctmtrol Hosp Epi1kmial1996; 17:                        amphotericin B for !he treatment of candidemia in patim15 without neutropenia. N EJlgI I
                                                                                                                 Md 1994; 331: 1325-1330.
                                                                                                          49. CoIlin J, Collin C. Constable FK. Jolmston lD.lnfusHm thrombophlebitis and inkction with
14.. ~ LA" McConnick RD. Springrnan SR. Maki DC. The pathogmesis and
     epidemiology of catheter-related infection with puImonary artery Swan-Ganz catheterso a                  various cannuI.Js.1..rUlat 1975; 2: 150-153.
     prosp«tive study ut:ili:zing molecular subtyping. Am I MDll99l; 9: 1",-205.                          50. Card.- RM. 5d.wartz It Wong He. Brub JP. Pen.-utaneDus indwelling radial-artery
15.   Raad L Intrava:scular-eatheter-related ~ /.smal 1998; 351.: 9'J3.898.                                   ca.theten; for monitoring cardioYascular function. Prospective !itUdy of the ri5k of Ihrombosis
                                                                                                              and infedion. N Engf J Mnl197-l; 290: 1227-1231.
16. DC. SInlz SM. W!weIer 5. MennellA. Prevention ofceltral venous catheter-related
    bloodstrnm infection by use of an antiseptk--impregna!ed catheter. .. randomi5ed. mntmIIed            51. Samsoondar W. Freeman jB, CouItish L O:dey C CoIoniutiDn of intrava5CU1ar cathet2r$ in
                                                                                                               an intensive care unit. Am J 1985; 149: 7JG.7J2.
    trial. AIIn Intm< MDlI997; 1Z1: 'b7-2b6.
                                                                                                          S2.. Ril:het H. Hubm B. Nitemberg C. d al. Pl"ll5J"!dive multicalter study of ~ ~ ­
11. Sa1=Ian MB. Isenberg HO. Shilpiro JF, lipsitz PJ, Rubin LG. A p~ study of lho:-
                                                                                                                 related cultures in intensive care unit patients. I Oi" MiaobioII99l); 2&: 252l).2S25..
    catbeb!r hub as the portal of entry fur microorga.ni5lns causing ~·relatedsepsis in
    neonate. Jlnfrd Dis 1993; 1.67: 487-490.                                                              53. Gll RT. Kruse JA. Thill-8ah.arnzian. Carlson RW. Triple Y5 single-Iurnm...-rral venous
                                                                                                                 catheters. A prospectivestudy in a criliGilly ill population. ArdIlllfun Md 1989; 149: 1139-
18. lina.res J. Sitge>-s.-a A. GarralI. J. perez JL Martin R. Pathogenesis of athrter 5eJ!5i:s: ..
    prospedive study with quantitative and semiquantit;>ti'o... cultures of catheter hub and
    segments. I am Miaobio/l985; ll357..J60..                                                             Sol              n.
                                                                                                                 Milkr Vm.ll5 B. Mathru M. Comparison of the sterility of long-term. celtnll VmDU5
                                                                                                                 catheterization using single-lumen. tripJe.lumen and pulmonary artery ca.1heter1 cm Co:rr
19. Sitges-Sern A. L.inare1 Pere:l JL Jaurrieta E. I...orfttte L A randomi:zed rri.a.l on the effect of
                                                                                                                 Mnl19S1; 12:. 63oI-fi37.
    tubing cbangt5 an hub CllI\laminatiDn and catheter ~ during parentenl nutrition.. J
      PlUmttr EntD-W Nuu 1985; 9: 322-325.                                                                55. Ullman RF. Gwevich L SdJodl. PE. Cwlh.a BA Colonization and ~ related to
                                                                                                              duration of triple-lumen intra~1arcatheter placement. Am I lJlfrrl Controll99lt; 18: 201·
20. Maki DC. Martin WT. Nationwide epidemicaf ~ caU5ed. by contaminated infu5ion
    prud\lCl1.J Infrd Dis 1975; U1:. '11il-Z72..
                                                                                                          56. Cyna AM. HovendenJL LehmannA,. ~K.1WiaP. Routinr ~ofcentra.l
ZL DC,. Anda5Dn RI.. Shulman JA In--u5e contamination of intravenous infusion fluid..
                                                                                                              Vt>nOU'!i ca.t.herers: survey of intensive care unib in mainland Britain. BMI 1998; 316:
    Appl Miaobiol1974; 28: 77B-7St.
22. Raad It Darouiche RO. Cathetet" related septicemia: risk m:luction.. bffrrt Med 1996; U: t!I1l-
                                                                                                          Sl. Conen Y; I'os5e Jp. Ka.roubi Po rl fll. The'Hands-Off' catheter and the prevmtion of.syMemic
    812; 815-816; 823.
                                                                                                              infections ~ted with pu1mon;uy ilrtIery catheter. 1Ek5pir Cri, Gm Md 1998; 157: 284-
23.. ~ RI. Raad It BaIan:i A. Koo 1.. Rand K. ~yeu- ex~ with sonicated vascular
     ca.tbetercultures in .. c::IinkaI. microbiology laboratory. I Gin MiaobioII99O; 2S: 7&82.                   "".
                                                                                                          58. Mnmel LA, StoIz Sf\.t t.iaki DC. 5urfaa 3ntimiaobia.l activity of heparin-bonded 3nd
Z-t. Mala DC. Weise cr. S;mdin HW. A semiquanlitative culllm' method fur identifying                          antiseptic impregnated ~ ca~ I lnfrd Dt5' 1993; 167: ~4.
     inl:ra~ catbel2r infection. N Engt J Mnl1'H7; 2%:: 1305-1309.
                                                                                                           59.   Ryder MA.       P~ 3CO.'S5 options. Sllrg     Oin North Am 1995; 4: J95..4Z1.
25. Kiehn lE. Armsttong 0_ Cl\anges in the spednIm of organisms ca.using b.1dlon'mia and
                                                                                                           6ll   Raad L Davis S. M. d fll. Low infection rate and long dwability of nontunnried si1astic
     fwlgemia in immunocompromised patients due to ~ 3COI.'S!> devkei. Ewr I Oin MiaubiDI
                                                                                                                 athetes. A safe mst.-effedi..-e allenilm.·e for long-term venDU5 ~ Ndr Intml Mm 1993;
     loifect. cm 1990; 9: 869-812-
                                                                                                                 153: 1791-1796-
26. Raad L CVC·-mted fungao>m..ia in cancel" patients. CIinitJll GuiIk of Fwrpllrrfrd:ion 1995; 6:
                                                                                                           61. Cook O. Randolph A. Kememw1 P. d t>L Central venous catheter repIacmlent stra~ A
      "'-                                                                                                      $JSlem3t1c revlO!W of the literature..  ern
                                                                                                                                                      ~ MnlI997; 25: 1411-1<&2-1_
'I/. Cleri DJ. Conado ML. SeUgman 51. Quantitative culture of intravenous catheteD and otha"
     inl:raViilSCU1ar imerts. I lrrfrd Vis 1981; IU: 781-786.
28. Raad It Sabbagh MF. Rand KH. Sberertz RJ. Quantitative tip cuIlun' methods and the
    diagn0si5 of <:I!ItlRl venous catbetet related infections. DWgn Miaobiol trrfrrl  cm
                                                                                     1992; lS: 13-
29. Kite P, Dobbins BM. \\'"Ilmx MH. d fll. Evaluations of a novel. endoluminai brush lIW'thod for
    in sih4 d~of ca.tbeter-related sepsis. 1 Oin PIltfrolI91J7; SO: 218-282.
JO. light> MJ. Ki~ P. Fawley WN,    0, Mahon MJ. An endoluminaJ brmh to detect the
    inf£.ch.d. Ci.'Dtral VftIOU5 catho.t2r in sihl: .. pilot study. BM/l996; 3U: 1528-l529.
31- Faubioo. Wc. We5ley JR, KhaIdi N. SilvaJ. Total pan-nteral nutribDn catheter sqSs: impact
    of the team approach. I PIVmt6 Enknl NllU 1986; 1.D: 642-645.
32- Maki DC. Yes, V"upnia. ~ ~ is very important maximal barrier precautiom
    during inset't:UI reduoe      me
                                 risk of O!nb:al Vl!DDl!S catbeter-ma~~ Inf«t Qmtrol
    HDsp EpitlnrrioII994; 15: 227-23ll.
33. Raad It Hohn DC. GiIbreub Bf. d ill. Prevention cl Q5ltral wnou:s cathaer-related inNctions
    U5ing maxim3l sterile ~ precautiC1n5 dJuiog in5ertim- /nftd Control ffu;p EpidDnialI994;
      15.: 231-238.