Reduction of catheter-related bloodstream infections
K B C C C i C I I S
c o r C
Intensive and Critical Care Nursing 43 (2017) 12–22
Contents lists available at ScienceDirect
Intensive and Critical Care Nursing
j ourna l ho m epage: www.elsev ier .com/ iccn
revention of central venous line associated bloodstream infections in dult intensive care units: A systematic review
iana Carolina Velasquez Reyesa,∗, Melissa Bloomerb, Julia Morpheta
Monash University, School of Nursing and Midwifery Peninsula campus, McMahons Road, Frankston VIC, 3199, Australia Deakin University, School of Nursing and Midwifery, PO Box 20000, Geelong, VIC, AUS 3217, Australia
r t i c l e i n f o
rticle history: eceived 27 February 2017 eceived in revised form 3 May 2017 ccepted 23 May 2017
eywords: lood stream infection prevention atheter atheterisation entral line associated blood stream
nfection entral venous line
nfection prevention-control ntensive care ystematic review
a b s t r a c t
Background: In adult Intensive Care Units, the complexity of patient treatment requirements make the use of central venous lines essential. Despite the potential benefits central venous lines can have for patients, there is a high risk of bloodstream infection associated with these catheters. Aim: Identify and critique the best available evidence regarding interventions to prevent central venous line associated bloodstream infections in adult intensive care unit patients other than anti-microbial catheters. Methods: A systematic review of studies published from January 2007 to February 2016 was undertaken. A systematic search of seven databases was carried out: MEDLINE; CINAHL Plus; EMBASE; PubMed; Cochrane Library; Scopus and Google Scholar. Studies were critically appraised by three independent reviewers prior to inclusion. Results: Nineteen studies were included. A range of interventions were found to be used for the preven- tion or reduction of central venous line associated bloodstream infections. These interventions included dressings, closed infusion systems, aseptic skin preparation, central venous line bundles, quality improve-
ment initiatives, education, an extra staff in the Intensive Care Unit and the participation in the ‘On the CUSP: Stop Blood Stream Infections’ national programme. Conclusions: Central venous line associated bloodstream infections can be reduced by a range of inter- ventions including closed infusion systems, aseptic technique during insertion and management of the central venous line, early removal of central venous lines and appropriate site selection.
© 2017 Elsevier Ltd. All rights reserved.
Implications for clinical practice
• Interventions other than high cost devices such as antimicrobial-coated catheters offer an alternative or complementary solution to central venous line associated bloodstream infections in adult Intensive Care Units.
suc s hav
• The findings in this study show that low cost interventions venous line bundles and aseptic management of these device bloodstream infections rates.
Patients admitted to Intensive Care Units (ICUs) require spe- ialised management of life threatening conditions. The complexity
f the treatment and the procedures that patients in ICU may equire, make central venous lines essential (College of Intensive are Medicine of Australia and New Zealand [CICM], 2011). High
∗ Corresponding author. E-mail address: email@example.com (D.C. Velasquez Reyes).
ttp://dx.doi.org/10.1016/j.iccn.2017.05.006 964-3397/© 2017 Elsevier Ltd. All rights reserved.
h as education, surveillance, checklists, reporting and central e positive outcomes in reducing central venous line associated
volume intravenous fluids, parenteral nutrition, cardiovascular measurements, medication administration and blood infusions all require the use of central venous lines (Walder et al., 2002; World Health Organization, 2014). Despite their potential benefits, the risk of central venous line associated bloodstream infections (CLABSI) is high (Siempos et al., 2009). A CLABSI is a laboratory-confirmed bloodstream infection (BSI) in a patient who had a central venous
line within the 48 hours prior to development of the BSI, not related to an infection at another site (Centre of Control and Disease Prevention, 2014; Fagan et al., 2013; Kallen et al., 2010; O’Grady et al., 2011, 2002).
i f H a 2 c e
a 2 s o f f ( i d v o
i o f r
a t 2 w C f o W o
D.C. Velasquez Reyes et al. / Intensive
A 2010 study conducted in the United States of America (USA) dentified that about 41,000 patients developed CLABSI (Centre or Disease Control and Prevention, 2011; Virginia Department of ealth, 2013). Around 18,000 of those affected were ICU patients, nd one in four may die (Centre for Disease Control and Prevention, 011; Virginia Department of Health, 2013). CLABSI is also asso- iated with increased cost in patient care (Walder et al., 2002), stimated at USD$33,000 (Stevens et al., 2014).
Consequently, there is increased interest in ways to reduce nd prevent CLABSI (Dumont and Nesselrodt, 2012; O’Grady et al., 011). In 2008, a systematic review was undertaken evaluating trategies other than antimicrobial-coated catheters to reduce risk f CLABSI in the ICU (Ramritu et al., 2008b). In 2011, the Centre or Disease Control and prevention published updated guidelines or the prevention of intravascular catheter associated infections O’Grady et al., 2011, 2002). Despite these guidelines, and advances n understanding related to infection patterns, pathogen agents, ifferent pathogenesis, epidemiology and new diagnosis and pre- ention techniques in the last decade (Kim et al., 2011), no update f this systematic review has been undertaken.
The aim of this systematic review was to identify all existing nterventions to prevent and/or reduce CLABSI in adults in ICU, ther than antimicrobial-coated catheters. Literature published rom 2007 was included, as that is when the previous systematic eview was undertaken (Ramritu et al., 2008b).
This systematic review followed the Cochrane Effective Practice nd Organisation of Care Review Group (EPOC) recommenda- ions to assess quality in systematic reviews (Chandler et al., 013). Randomised controlled trials and observational studies hich investigated interventions for the prevention or reduction of LABSI in adult ICU patients were included. The quality of evidence
or each included study was determined based on the Grades of Rec- mmendations, Assessment, Development and Evaluating (GRADE) orking Group (Schünemann et al., 2011). Only studies with a high
r moderate quality rating were included.
Studies conducted in ICUs with adult patient populations were included. All interventions which sought to prevent and/or reduce CLABSI including the CDC recommended interventions (Centre for Disease Control and Prevention, 2011; Centre of Control and Disease Prevention, 2014; O’Grady et al., 2011, 2002) and the Institute of Healthcare Improvement (IHI) compilation of bundle of interventions designated to work together to reduce CLABSI were included.
The following outcome measures were examined following the uggestions given by the Cochrane EPOC Review Group (Chandler t al., 2013).
Central venous line associated bloodstream infection rates (per 1000 central venous line days)
Identification and incidence rate of pathogen agents identified in the colonised central venous lines detected (laboratory test report data) Mortality and comorbidity rates related to CLABSI.
ritical Care Nursing 43 (2017) 12–22 13
• Increased length of hospitalisation (measured in days) caused by the presence of CLABSI.
• Measurement of the time (measured in days) from central venous line insertion to removal.
• Length of stay in ICU, measured from the day of admission to ICU to the day of discharge from ICU.
• Studies published in languages other than English. • Non-academic studies, conference abstracts, oral presentation or
not original research. • Characteristics of participants not reported, no baseline data,
studies with no clear description of the intervention applied. • Studies with unclear aim, methodology, or data collection, or
those with missing data were excluded. • Studies conducted wholly or in part with paediatric populations
where the results were not reported separately. • Studies where ICUs were included together with another ward
(e.g. emergency department, coronary care units) where the results were not reported separately.
• Studies where antimicrobial-coated catheters were used were excluded from this review because several systematic reviews have recently been published on this topic (Antonelli et al., 2012; Liu et al., 2014; Raad, 2012; Ramritu et al., 2008a).
MEDLINE; CINAHL Plus; EMBASE; PubMed; Cochrane Library, Google Scholar and Scopus databases were searched using the following keywords (or abbreviations) and MeSH search terms; ‘catheteri*ation-central venous’, infection*, prevention*, blood- stream*.
The Cochrane Collaboration RevMan software (The Cochrane Collaboration, 2014) was used for data entry and management. The data were extracted following The Cochrane’s manual check- list (Higgins and Green, 2011). One researcher screened the titles and abstracts of each study. The three researchers then analysed the full-text of 87 studies meeting the inclusion criteria for full text assessment.
Assessment of methodological quality
Three researchers assessed the quality of each study, to deter- mine inclusion, using the Meta-Analysis of Statistics Assessment and Review Instrument (MAStARI) tool (Joanna Briggs Institute, 2014).
Assessment of risk of bias in included studies
The Cochrane Risk and Bias Assessment (RoBANS) tool and guideline were applied to each of the included studies (Higgins et al., 2011).
Outcome measures and statistical analysis, such as relative risk (RR), probability (p), Pearson Correlation Coefficient, Confidence Interval (CI) and the statistical power of the mean were analysed.
14 D.C. Velasquez Reyes et al. / Intensive and Critical Care Nursing 43 (2017) 12–22
1,500 stud ies iden�fied 825 through database search & 675 through Google Scholar search
87 full text studies ass ess ed for eligibili ty
68 stud ies exclud ed a�er full text cri�cal app raisal:
3 Full text wri �en in a language other tha n E nglish
16 not in ICU se�ng, not central line catheters
14 Oral presenta�ons/ conference/posters stud ies
13 Par�al or full use of an �microbial coated catheters
3 Cha racteris�cs of par�cipants not reported, no baseline data
1 No clear descrip�on or correla�on of the interven�on
5 Poor methodology measurement of the interven�on related with the aim of stud y. No confound er discuss ed
5 Miss ing data, lack of central line ass ociated bloodstream infec�ons rate reported
1 Different bun dles applied during interven�on bu t not reported or considered in the outcomes measurements, no confoun ders considered
1 Different ty pes and brand s of cap s were used duri ng interven�on. Cha racteri s�cs of pa r�cipants not reported
5 Weak methodology, surveys an d self-reported da ta, correla�on betwee n interven�on and outcome not clear, miss ing data characteris�cs of pa r�cipants not reported.
1 Vari ous interven�ons were ap pli ed duri ng stud y, no clear indica�on which interven�on redu ced the CLABSI rate
1,41 3 stud ies exclud ed:
371 non central line catheters
221 non adu lt ICU se�ng
150 non-acad emic research stud ies
17 not Engli sh langua ge
3 not pub lished a�er 200 6
1,50 0 stud ies scree ned by �tle/ab stract
W t h s
19 stud ies includ ed in the systema�c review
Fig. 1. PRISM
hen the data was homogenous, a Forest Plot was created to illus- rate the strength of the effect of the intervention. Due to the eterogeneity of the interventions and their reported results, meta- ynthesis was undertaken of the qualitative research studies.
From the database searches, a total of 1500 studies were iden- ified (Fig. 1).
isk of bias in included studies
All the included randomised controlled studies had a low risk f bias (Kwakman et al., 2012; Marsteller et al., 2012; Mimoz et al.,
2007; Pedrolo et al., 2014; Speroff et al., 2011; Timsit et al., 2012; Timsit et al., 2009; Yousefshahi et al., 2013) due to their random sequence generation, blinding of outcome assessment and inter- vention exposure measurement. Eleven studies had unclear risk of bias, because the interventions could not be blinded (e.g., the nature of the interventions, or methodology of reporting. However, the studies were include as the confounders/bias were considered in the reporting of results (Bonello et al., 2008; Cherifi et al., 2013; Jeong et al., 2013; Maki et al., 2011; Marsteller et al., 2014; McLaws and Burrell, 2012; Parikh et al., 2012; Rangel-Frausto et al., 2010; Scheithauer et al., 2014; Tang et al., 2014; Vilins et al., 2009). None of the studies included in this review reported a high risk of bias.
After the full text quality appraisal was independently under-
taken by the three researchers, and following the risk of bias assessment, 19 studies were included in this systematic review (Fig. 1).
and Critical Care Nursing 43 (2017) 12–22 15
c 2 2 e e t 2 2 i M R 2 v c 2 2 b n i i S s 2 e t e
i P T n d e ( e 2 3
t d d ( d c C T 1 p s m 0 s d d d e s m c 2
D.C. Velasquez Reyes et al. / Intensive
nterventions of included studies
The studies examined a variety of interventions, including edu- ation based interventions (Bonello, Fletcher, 2008; Cherifi, Gerard, 013; Scheithauer et al., 2014; Speroff et al., 2011; Tang et al., 014), skill-mix based interventions (Bonello et al., 2008; Cherifi t al., 2013; Marsteller et al., 2014; Marsteller et al. 2012; Parikh t al., 2012, Scheithauer et al., 2014), equipment based interven- ions (Maki et al., 2011; Rangel-Frausto et al., 2010; Vilins et al., 009), various dressings (Pedrolo et al., 2014; Scheithauer et al., 014, Timsit et al., 2012; Timsit et al., 2009) and aseptic based
nterventions (Bonello et al., 2008; Jeong et al., 2013; Kwakman, uller, 2012; McLaws and Burrell, 2012; Mimoz, Villeminey, 2007;
angel-Frausto et al., 2010; Speroff et al., 2011; Yousefshahi et al., 013). Several studies evaluated the use of the different central enous line insertion bundles including the Institute for Health- are Improvement (IHI) bundle (Bonello et al., 2008; Cherifi et al., 013; Jeong et al., 2013; McLaws and Burrell, 2012; Parikh et al., 012; Pedrolo et al., 2012). Many of these studies combined the IHI undle with other interventions, including education (web semi- ars, monthly educational sessions, auditing and feedback based
nfection programs) (Yousefshahi et al., 2013), face to face meet- ngs, teleconferences and/or online in-services (Bonello et al., 2008; peroff et al., 2011; Tang et al., 2014), teamwork strategies and tandardised data collection tools (Bonello et al., 2008; Jeong et al., 013), surveillance interventions (Jeong et al., 2013; Scheithauer t al., 2014; Tang et al., 2014) and other care bundles (e.g. the ven- ilator associated pneumonia bundle) (Bonello et al., 2008; Speroff t al., 2011).
ressings Five studies examined the effect of various central venous line
nsertion site dressings on CLABSI rates (Kwakman et al.„ 2012; edrolo et al., 2014; Scheithauer et al., 2014; Timsit et al., 2012; imsit et al., 2009). Four studies compared chlorhexidine impreg- ated dressings with other dressings including medical honey ressing (Kwakman et al., 2012), standard dressing (Scheithauer t al., 2014; Timsit et al., 2009), sterile gauze and micropore tape Pedrolo et al., 2014), transparent Tegaderm 3 M dressing (Timsit t al., 2012), Tegaderm Transparent Film Dressing 3 M (Timsit et al., 012) and highly adhesive Tegaderm HP Transparent Film Dressing
M (Timsit et al., 2012). There was no significant difference in CLABSI and skin colonisa-
ion rates when chlorhexidine dressings were compared with other ressings (Table 1). However, one study found that Chlorhexidine ressings lowered the number of Gram-positive bacterial infections Scheithauer et al., 2014). In one study comparing chlorhexidine ressings with standard dressings, dressings in both cohorts were hanged at either three or seven days. The authors reported that LABSI was less common with chlorhexidine dressings (Table 1). he authors also reported a catheter colonisation of 142 out of 657catheters (7.8%) in the three day dressing change group (10.4 er 1000 catheter-days) and 168 of 1828 catheters (8.6%) in the even day dressing change group (11.0 per 1000 catheter-days), a ean absolute difference of 0.8% (95% CI, −1.78% to 2.15%) (HR,
.99; 95% CI, 0.77–1.28) (Timsit et al., 2009). The remaining four tudies reported changing the dressings every seven days unless ressing edges detached, dressing integrity was compromised or ischarge accumulated in the catheter opening (leaking or soiled ressings) (Kwakman et al., 2012; Pedrolo et al., 2014; Scheithauer t al., 2014; Timsit et al., 2012). In addition to the dressings, these
tudies also used other interventions to reduce CLABSI including aximal sterile barriers (Timsit et al., 2012; Timsit et al., 2009),
are bundles (Pedrolo et al.„ 2014; Timsit et al., 2012; Timsit et al., 009), and antiseptic skin preparation prior the insertion of the cen-
Fig. 2. Comparing open vs. closed infusion containers studies.
tral venous line (Kwakman et al., 2012; Pedrolo et al., 2014; Timsit et al., 2012; Timsit et al., 2009) (Table 1).
Open vs. closed infusion containers Three studies compared open and closed infusion containers
(Maki et al.„ 2011; Rangel-Frausto et al.„ 2010; Vilins et al., 2009). (Table 2, Fig. 2) Open infusion containers were defined as com- mercially available glass bottles, burettes and semi-rigid plastic containers that must admit air (air filter or needle) to empty, requir- ing external venting (Maki et al., 2011; Rangel-Frausto et al., 2010; Vilins et al., 2009). Closed infusion containers were defined as fully collapsible plastic containers that did not require any exter- nal venting to empty, with self-sealing injection ports (Viaflex or Viaflo) (Maki et al., 2011; Rangel-Frausto et al., 2010). Meta- analysis indicated that closed infusion container systems reduced CLABSI (Fig. 2). In each of these three studies, a standard central venous line management protocol was in place, including hand hygiene with alcohol based hand rub, and a care protocol docu- menting and evaluating the gauze dressing condition.
Aseptic skin preparation Chlorhexidine skin antiseptic was used as skin preparation prior
to central venous line insertion in ten studies (Bonello et al.„ 2008; Cherifi et al.„ 2013; Jeong et al., 2013; Kwakman et al., 2012; McLaws and Burrell, 2012; Mimoz et al.„ 2007; Rangel-Frausto et al., 2010; Speroff et al., 2011; Timsit et al., 2012; Yousefshahi et al., 2013). However, only two studies examined the effective- ness of the antiseptic applied at the central venous line insertion site in the reduction of central venous line pathogen colonization (Mimoz et al., 2007; Yousefshahi et al., 2013). One study com- pared topical Chlorhexidine with an antiseptic agent composed of hydrogen peroxide (H2O2) and silver (Sanosil 2%) (Yousefshahi et al., 2013). The antiseptic agent composed of hydrogen per- oxide (H2O2) and silver produced a lower number of positive central venous line tip cultures than the Chlorhexidine, however, there was no significant difference between the two (Yousefshahi et al., 2013) (Tables 3 and 4). Skin antisepsis with 0.1% octeni- dine dihydrochloride and 2% 2-phenoxyethanol (octenisept) was applied in one study in combination with Chlorhexidine dressings (Scheithauer et al., 2014). Alcohol povidone-iodine was also com- pared against Chlorhexidine in four studies (Mimoz et al., 2007; Timsit et al., 2012; Timsit et al.„ 2009; Yousefshahi et al., 2013) and only one study reported a higher prevention of CLABSI when using Chlorhexidine as skin antiseptic compared to alcohol povidone- iodine (Mimoz et al., 2007). One study compared a Chlorhexidine bath one day prior to central venous line insertion, with Povidone-
Iodine scrub immediately preceding central venous line insertion, with no reduction in CLABSI rates (Yousefshahi et al., 2013) (Table 3).
16 D.C. Velasquez Reyes et al. / Intensive and Critical Care Nursing 43 (2017) 12–22
Table 1 Examination of studies comparing dressings.
Author & year Group ICUs in sample
Patients in sample
CLABSI n (%) Positive Culture Swab n (%)
CLABSI/1000 catheter days
RR (95% CI) p
Kwakman et al. (2012) Control 1 106 – 36 (34) – NR 0.98 Revami (honey) and gauze 1 129 – 44 (34) –
Pedrolo et al. (2014) Chlorhexidine dressing 2 43 6 – – NR 0.52 (13.95)
Gauze and Micropore 2 42 5 – – (11.9)
Author & year Group ICUs in sample
Central line days
CLABSI n (%) Positive Culture Swab n (%)
CLABSI/1000 catheter days
RR (95% CI) p
Scheithauer et al. (2015) Chlorhexidine dressing 2 7282 11 – 1.5/1000 (0.75–2.70) <0.001 Standard dressing 2 4938 29 – 5.87/1000 (0.93–8.43)
Author & year Group ICUs in sample
Catheters in sample
CRBSI (n) Positive Culture Swab n (%)
CRBSI/1000 catheter days
RR (95% CI) p
Timsit et al. (2012) Chlorhexidine dressing 12 2108 9 75 0.5 1.284 0.45 (0.67–2.45)
Adhesive dressing 12 998 10 97 1.3 Standard dressing 12 1067 12 89 1.3
Timsit et al. (2009) Chlorhexidine dressing 7 1825 – – 1.3 0.24 0.05 (0.09–0.65)
Standard dressing 7 1953 – – 0.4
Note: ICU = Intensive Care Unit; NR = not reported; CLABSI = Central line associated blood stream infection; CRBSI = Catheter related blood stream infection; RR = Risk Ratio.
Table 2 Comparing open vs. closed infusion containers studies.
Author & Year Group No. ICUs in sample
No. of patients in sample
Rate of CLABSI per 1000 central line-days (infections/days)
Rate of CLABSI per 1000 central line-days (%)
Maki et al. (2011) Open 15 2237 153/15,189 10.2 0.33 <0.001 Closed 15 2136 45/13,456 3.3 (0.24–0.46)
Rangel-Frausto et al. (2010) Open 4 548 59/3661 16.1 0.20 <0.001 Closed 4 548 13/4055 3.2 (0.11–0.36)
Vilins et al. (2009) Open 3 483 28/4297 6.5 0.49 0.03 Closed 3 642 13/4041 3.2 (0.26–0.95)
Table 3 Examination of studies comparing antiseptics.
Author & Year Group ICUs in sample
Patients in sample
CRBSI Catheter Tip Positive Culture n (%)
CRBSI/1000 catheter days (%)
RR (95%CI) P
Yousefshai et al. (2013)
Chlorhexidine 3 113 – 29 – 1.05 0.75
(21.3) (0.76–1.45) Sanosil 2% 3 136 – 26 –
(23.0) Mimoz et al. (2007) Chlorhexidine 1 242 4 28 1.7 2.01 0.002–0.009
(11.6) (1.24–3.24) Alcohol based-Povidone-iodine 1 239 10 53 4.2 1.87
N = Risk
t M S e r r w l 2 e r
ote: CRBSI = Catheter Related Bloodstream Infection, ICU = Intensive Care Unit, RR
entral venous line bundles Seven studies used central venous line bundles as an interven-
ion (Bonello et al., 2008; Cherifi et al., 2013; Jeong et al., 2013; cLaws and Burrell, 2012; Parikh et al., 2012; Pedrolo et al., 2014;
cheithauer et al., 2014; Speroff et al., 2011; Tang et al., 2014; Timsit t al., 2012). A key focus of central venous line bundles is early emoval of central venous lines (Table 5). The heterogeneity of the eported data restricted meta-analysis, therefore meta-synthesis as conducted. Three studies implemented the central venous
ine bundle outlined by the IHI (Bonello et al., 2008; Jeong et al., 013; Tang et al., 2014). This bundle has been examined in differ- nt studies, and results indicated a positive reduction in CLABSI ates (Al-Tawfiq et al., 2012; Blot et al., 2014; Sacks et al., 2014)
(Table 6). Three studies used alternative bundles together with multiple interventions to reduce CLABSI, with all reporting a reduc- tion in infection rates (Cherifi et al., 2013; McLaws and Burrell, 2012; Yousefshahi et al., 2013) (Table 6). As a key component of the central venous line bundle, the early removal of central venous line was examined as an effective practice to reduce CLABSI (McLaws and Burrell, 2012). This study estimated probabilities for CLABSI at different dwell times to identify the dwell time that was closest to being infection free, less than1 in 100 chance of infection, they
reported the safest dwell time was the lowest cumulative probabil- ity of CLABSI, 1 in 100 chance, for a cumulative catheter dwell time of seven days giving an adjusted CLABSI rate of 1.8/1000 line days (McLaws and Burrell, 2012). This is consistent with other studies
D.C. Velasquez Reyes et al. / Intensive and Critical Care Nursing 43 (2017) 12–22 17
Table 4 Aseptic skin preparation used in each study.
Bonello et al. (2008) Not specified Cheriffi et al. (2013) Skin antisepsis was performed with 0.5% chlorhexidine in 70% alcohol (Cedium® , QUALIPHAR) or with 5% alcoholic
povidone-iodine (Iso-Betadine® solution hydroalcoolique, MEDA Pharma). Jeon et al. (2013) Skin antisepsis was performed with Chlorhexidine (2% chlorhexidine in 70% of one of the following: isopropyl alcohol,
alcohol, povidone-iodine, or a mixture of alcohol and povidoneiodine). Kwakman et al. (2012) Skin antisepsis was performed with 0.5% chlorhexidine in 70% alcohol. Mcklaws et al. (2012) Skin antisepsis was performed with 2% alcoholic chlorhexidine. Mimoz et al. (2007) Skin antisepsis was 5% povidone-iodine in 70% ethanol (Betadine Alcoolique; Viatris Pharmaceuticals, Meı́rignac,
France) or a combination of 0.25% chlorhexidine gluconate, 0.025% benzalkonium chloride, and 4% benzylic alcohol (Biseptine; Bayer HealthCare, Gaillard,France).
Rangel et al. (2010) Not specified. Speroff et al. (2011) Not specified use of chlorhexidine skin antisepsis. Timsit et al. (2012) Skin preparation was with alcoholic povidone-iodine (PVI) or alcoholic chlorhexidine solution in accordance to
standard procedure in each ICU. First, the insertion site was scrubbed with a detergent (4% aqueous PVI solution, Betadine Scrub; Viatris Pharmaceuticals, Merignac,France) or 4% chlorhexidine solution (Hibiscrub; Molnlycke Health Care, Wasquehal, France); rinsed with sterile water; and dried with sterile gauze. An alcohol-based antiseptic solution (5% PVI in 70% ethanol [Betadine Alcoholic Solution; Viatris Pharmaceuticals] or 0.5% chlorhexidine, 67% ethanol [Molnlycke Health Care]; or 0.25% Healthcare, Gaillard, France]) was
Yosefshahi et al. (2013) Skin preparation was bath with C
Table 5 Institution of Healthcare Improvement IHI central venous line bundles.
Central venous line bundles strategies compiled by the IHI Hand hygiene Maximal sterile barrier precautions during insertion Chlorhexidine skin antisepsis Daily assessment of central vascular catheter necessity
v t T f M
2 2 b e l a 2 S m o l t r 2
p S i t m
Prompt removal of central venous lines Avoidance of femoral site
xamining the effect of a central venous line bundle on CLABSI with imilar positive results (Furuya et al., 2011; , Loveday et al., 2014).
ombining multiple interventions with central venous line undles
A variety of initiatives that could be combined with central enous line bundles (Table 5) primarily focussed on staff educa- ion, surveillance, development of tools and teamwork practices. he combination of any of these interventions with bundles were ound to reduce CLABSI (Bonello et al., 2008; Cherifi et al., 2013;
cLaws and Burrell, 2012; Tang et al., 2014).
uality improvement initiatives These initiatives included compliance checklists (Bonello et al.,
008; Jeong et al., 2013; McLaws and Burrell, 2012; Tang et al., 014), surveillance (Cherifi et al., 2013; Tang et al., 2014), feed- ack (Jeong et al., 2013), hand hygiene education programs (Jeong t al., 2013; Speroff et al., 2011), visual promotion of central venous ine care campaigns (Jeong et al., 2013; Speroff et al., 2011), cre- tion of collaborative teams (Jeong et al., 2013; Speroff et al., 011), tools, guidelines or protocols (McLaws and Burrell, 2012; peroff et al., 2011). The virtual collaborative intervention included onthly educational conference calls, web seminars, individual
nline coaching and email report access related to central venous ine management (Speroff et al., 2011). These different combina- ions of interventions resulted in a significant reduction in CLABSI ates (Bonello et al., 2008; Cherifi et al., 2013; McLaws and Burrell, 012; Tang et al., 2014) (Table 6).
ducation In four studies the IHI bundle was combined with educational
rograms delivered in the ICU (Cherifi et al., 2013; Jeong et al., 2013;
peroff et al., 2011; Tang et al., 2014). It included interdisciplinary mprovement team implementing organisational changes related o the central venous line bundle, they introduced sharing goals and
ethods by collaborative charter, monthly conferences calls face
chlorhexidine, 0.025% benzalkonium chloride, 4% benzyl alcohol [Biseptine Bayer then applied for at least 1 min. hlorhexidine 2%. and 10% Povidone-Iodine.
to face, learning sessions at least (three per month), education in ‘Plan to Do, Study Act (PDSA) methodology, bed side checklists edu- cation and interdisciplinary interactive team rounds. Only two of these studies demonstrated a significant reduction in CLABSI rates (Cherifi et al., 2013, Tang et al., 2014) (Table 6).
Increasing ICU medical staffing One study examined the effect of staffing skill mix on infection
rates, by comparing infection rates before and after the introduction of an extra intensivist in the ICU (Parikh et al., 2012), demonstrating an intensivist as a cost effective and beneficial strategy in reducing CLABSI rates (Parikh et al., 2012) (Table 7). Not only were more lives saved when there was an extra intensivist in the ICU, but the cost of the intensivist was cheaper than the expenses associated with a higher CLABSI rate (Barnett et al., 2010; Parikh et al., 2012; Stevens et al., 2014).
Participating in on the CUSP Two studies examined the outcomes arising from participation
in On the CUSP, a national program in the USA, and its impact in reducing CLABSI (Marsteller et al., 2014; Marsteller et al., 2012). Both studies reported significant reductions in CLABSI after apply- ing the intervention. The first study additional of the On the CUSP national program examined the positive outcome of mandatory reporting and CLABSI reduction. After comparing CLABSI rates from hospitals where a mandatory reporting public reports of central venous line associated infections rates was in place (Table 8).
From the nineteen studies included in this review, fifteen reported a positive impact in the reduction of CLABSI (Bonello et al., 2008; Cherifi et al., 2013; Maki et al., 2011; Marstelleret al., 2014, Marsteller et al., 2012; McLaws and Burrell, 2012; Mimoz et al., 2007; Parikh et al., 2012; Rangel-Frausto et al., 2010; Tang et al., 2014; Timsit et al., 2012; Timsit et al., 2009; Vilins et al., 2009; Yousefshahi et al., 2013). Open infusion containers, chlorhexidine dressings, aseptic interventions for central venous line maintenance including central venous line bundles, manda- tory reporting, communication and continuing education based interventions were effective in reducing CLABSI rates. In addition,
checklists, facilitating feedback, regular rounds and supervision and extra intensivist in the ICU were also successful interventions. Chlorhexidine skin preparation prior to central venous line inser- tion was not shown to significantly reduce the rate of CLABSI. (Jeong
18 D.C. Velasquez Reyes et al. / Intensive and Critical Care Nursing 43 (2017) 12–22
Table 6 Examination of studies comparing central line bundles.
Author & year Group ICUs in sample
Patients in sample
CLABSI n (%) CLABSI/1000 catheter days
RR (95% CI) p
Jeong et al. (2013) Baseline 4 79 4.7 6/1290 0.39 0.76 Intervention 4 309 1.8 7/3899 (0.11–1.39)
McLaws et al. (2012) First 12 months 37 4166 3.8 27/7176 (2.5–5.5) 0.002 Last 6 months 37 NR 1.6 26/16,100 (1.0–2.4)
Tang et al. (2014) Baseline 5 NR 1.6 17/10,325 NR 0.03 Intervention 5 481 0.6 6/9388
Author & year Group ICUs in sample
Patients in sample
Mean CLABSI n (%)
Mean of CLABSI/1000 days
RR (95% CI) p
Bonello et al. (2008) First 3 months 12 NR 0.52 5.2/1000 NR NR Last 3 months 12 NR 0.27 2.7/1000
Cherifi et al. (2013) Before intervention 5 1354 4.00 24/1000 0.49 0.212 (0.24–0.98)
During intervention 5 1571 1.81 12/1000 After intervention 5 1439 2.73 16/1000 1.37 0.413
Author & year Group Hospitals in sample
Hospitals in Tool kit group
Hospitals in Virtual collaborative group
Median CLABSI/1000 catheter days in Tool kit group
Median CLABSI/1000 catheter days in Virtual collaborative group
RR (95% CI) p
Speroff et al. (2011) Baseline 59 29 30 2.42 1.84 NR NR (0.65–6.80) (0.00–3.83)
3 months 59 29 30 2.47 2.24 (1.48–5.35) (0.54–4.69)
6 months 59 29 30 2.54 2.28 (0.00–4.98) (0.00–3.73)
9 months 59 29 30 1.23 1.75 (0.00–3.93) (0.00–3.74)
12 months 59 29 30 1.17 1.18 (0.00–3.61) (0.00–2.71)
15 months 59 29 30 1.77 2.04 (0.00–3.30) (0.00–4.91)
18 months 59 29 30 1.16 2.76 (0.00–5.46) (0.00–4.67)
Note: ICU = intensive care unit; NR = not reported; CRBSI = Catheter related blood stream infection.
Table 7 Examination of a study increasing ICU medical staffing.
Author & Year Group No. of ICUs in sample
No. patients in sample
No. CLABSI per 1000 central line days (infection/days)
Rate of CLABSI per 1000 central line days (%)
RR (95%CI) p
Parikh et al. (2012) First year before intervention 1 1113 13/1531 8.5 8.32 0.0006 (1.91–36.28)
l C i s
e C W e h u ( b t
Last year after intervention 1 10
ote:CI = confidence interval, CLABSI = Central line bloodstream infection, ICU = Inte
t al., 2013; Kwakman et al., 2012; Pedrolo et al., 2014; Speroff et al., 011).
Important findings from this study should be considered in ICUs ooking to reduce CLABSI rates. Choice of dressing is important. hlorhexidine dressings were used in each study comparing dress-
ng’s types, and are shown to have positive outcomes in decreasing kin flora which decreases CLABSI incidence.
This study found that Chlorhexidine skin preparation was qually effective as alcohol-based povidone iodine for preventing LABSI in most cases as other studies have shown (Adams and ilson, 2012; Bashir et al., 2012), but that there was no differ-
nce between Chlorhexidine and the antiseptic agent composed of ydrogen peroxide (H2O2) and silver (Yousefshahi et al., 2013). Reg-
lar/daily bathing with Chlorhexidine had no effect on CLABSI rates Noto et al., 2015; Seyman et al., 2014), although there was reduced lood culture contamination (Popovich et al., 2010). One reason for his finding may be the conclusion that dressing changes should
Care Unit, RR = Risk Ratio.
be minimised to reduce CLABSI rates (O’Grady et al., 2011; Rupp et al., 2013). Only two studies reported side effect dermatitis when applying chlorhexidine dressings (Timsit et al., 2012; Timsit et al., 2009), no other data was collected regarding this important con- sideration. Further analysis of skin side effects are recommended in future studies.
Early removal of central venous lines is recommended, ideally before day eight (Exline et al., 2013; Mangum et al., 2013; Weeks et al., 2014). Early removal of central venous lines is one of the key features of central venous line bundles, and several studies included in this systematic review reported reduced CLABSI rates associated with bundle use (Bonello et al., 2008, McLaws and Burrell, 2012; Tang et al., 2014). Another important feature of the central venous
line bundles found to be reported was the aseptic skin preparation methods prior the insertion of central venous line. Hence central venous line bundles should be promoted across ICUs.
D.C. Velasquez Reyes et al. / Intensive and Critical Care Nursing 43 (2017) 12–22 19
Table 8 Examination studies analysing participation in On the CUSP: Stop BSI National Program and mandatory reporting.
Author, year & period No. of ICU in sample Mean rate of CLABSI per 1000 central line days according to study quarters
RR according to study quarters
Control group Inter group Control group Inter group Control group Inter.group n n M M RR RR
Marsteller et al. (2012) Baseline 22 23 2.71 4.48 1.00 1.00 1st Q NR 1.12 NR 0.25 2nd Q NR 1.83 NR 0.41 3rd Q 2.16 1.33 0.79 0.30 4th Q 0.56 0.96 0.21 0.21 5th Q 0.52 0.88 0.19 0.20 6th Q 0.83 0.85 0.31 0.19
No. of ICUs in sample Mean Rate of CLABSI per 1000 central line days
n Group PR policy<1 year Group PR policy<1 year Group VR policy Group NR Policy M M M M
Marsteller et al. (2014) 1046 Baseline 2.49 1.85 2.20 1.90 1stQ 2.33 1.52 1.89 1.65 2ndQ 2.00 1.22 1.46 1.32 3rdQ 1.61 1.29 1.83 0.96 4thQ 1.49 1.24 1.38 0.96 5thQ 0.90 1.21 1.91 1.11 6thQ NR 1.16 1.11 1.15
N sociat r 1 year V
q C 2 s p
( o c e e c t
o r e p i c e
w l o r 2
E v c s
ote: BSI = Bloodstream infection, CI = Confidence interval, CLABSI = Central line as eported status, PR < 1 year = Public reporting had begun for one year or less, PR > R = Voluntary reporting.
In addition, central venous line bundles in combination with uality improvement or educational interventions further reduces LABSI rates (Cherifi et al., 2013; Ramritu et al., 2008b; Tang et al., 014). Quality improvement interventions have previously been hown to reduce CLABSI rates (Blot et al., 2014), and are an inex- ensive way to improve patient outcomes in ICU.
Despite favourable outcomes with closed infusion systems Maki et al., 2011, Rangel-Frausto et al., 2010, Vilins et al., 2009), pen infusion systems continue to be used in some developing ountries such as Argentina, Brazil, Colombia and Mexico (Garrett t al., 2002; Macías et al., 1999; Maki et al., 2011; Rangel-Frausto t al., 2010; Rosenthal and Maki, 2004). Further research should be onducted to analyse and address the factors impeding the change owards closed infusion container systems in developing countries.
Only one study included in this review examined the impact f an additional intensivist, demonstrating a reduction on CLABSI ates (Parikh et al., 2012). There are however several other studies xamining a variety of healthcare associated infections in different opulations, which have similarly reported that additional staffing
s effective at reducing healthcare associated infections and health- are costs (Chordas, 2004; Peters and Locke Nagele, 2010; Spaeth t al., 2003; Terry, 2002).
Finally, the mandatory reporting of CLABSI rates was associated ith reduced CLABSI rates. This finding is consistent with other
iterature (Marsteller et al., 2014), which reported the incidence f methicillin resistant staphylococcus aureus (MRSA) is similarly educed by mandatory reporting (Biswal et al., 2015; Edge et al., 007; Pearson et al., 2009; Sheps and Birnbaum, 2012).
This systematic review included only studies published in
nglish. There was also a predominant heterogeneity in the inter- entions in the included studies. Consideration of the multiple onfounders which could influence the outcomes of the studies hould be taken into account.
ed bloodstream infection, ICU = Intensive Care Unit, Inter = Intervention, NR = not = Public reporting had begun for more than one year, Q = Quarter RR = Risk Ratio,
This systematic review identified several interventions able to reduce or prevent CLABSI. Aseptic technique, which includes the application of skin antiseptic pre-central venous line insertion, and aseptic central venous line maintenance is essential. Closed infusion systems should be used at all times. Central venous line bundles, which promote appropriate insertion site selection, asep- tic central venous line management and early removal; as well as increased intensivist staffing were both found to contribute to reduction and prevention of CLABSI and Quality improvement initiatives aimed at education and safety practices should be under- taken. Finally mandatory reporting CLABSI is imperative to the prevention or reduction of CLABSI in adult ICUs.
The authors have no sources of funding to declare.
Ethical statement not applicable the authors undertook a sys- tematic review, no ethical statements to declare.
Conflict of interest
The authors have no conflict of interest to declare.
The researchers have no acknowledgements to make.
• APPENDIX A KEYWORDS AND MeSH SEACRH.
“catheterization”[MeSH Terms] OR catheterization[Text Word] catheterisation”[MeSH Terms] OR catheterisation[Text Word].
2 and C
t v a t W
• • • • • • • • • • •
0 D.C. Velasquez Reyes et al. / Intensive
“catheterization, central venous”[MeSH Terms] OR cen- ral catheterization[Text Word] “catheterisation, central enous”[MeSH Terms] OR central catheterisation[Text Word] ssociated[All Fields] AND (“infection”[MeSH Terms] OR infec- ion[Text Word]) “infection”[MeSH Terms] OR infections[Text
ord]. “prevention and control”[Subheading] OR prevention[Text
Catheter, Central Venous Catheters, Central Venous Venous Catheter, Central Venous Catheters, Central Central Venous Catheter Catheter Related Infections Infection, Catheter-Related
Catheter-Associated Infections Infections, Catheter-Associated preventive therapy preventive measures
able B1 xcluded studies
Author/year Reason for exclusion
Al-Tawfiq et al. (2013) Characteristics of participants not reported, no base l Amarasingham et al. (2007) No clear description and correlation of the interventi
CLABSI rate prevention or reduction. Barrera et al. (2011) Poor methodology during intervention using alcohol
replacement of the bottles was done either the bottle Barsuk et al. (2014) Poster, no full text access. Berenholtz et al. (2014) Significant percentage of missing data. No measurem Cherry et al. (2011) Not conducted in ICU setting. 80 Not conducted in ICU setting. DePalo et al. (2010) Use of antibacterial-impregnated catheters. Dilek et al. (2012) Poor measurement of outcomes according to aim and Doherty and Axelrod (2011) No access to full text. Duane et al. (2009) Use of antibacterial-impregnated catheters. DuBose et al. (2008) Use of antibacterial-impregnated catheters. (ARROW
Chlorhexidine acetate and silver sulfadiazine cathete Flinchum et al. (2010) No access to full text. Ghonim et al. (2012) No access to full text. Hansen et al. (2014) Use of antibacterial-impregnated catheters. Hopfner et al. (2012) No access to full text. Jaggi et al. (2013) Intervention only partially applied. Use of antibacteri Khalid et al. (2013) Use of antibacterial-impregnated catheters. Leblebicioglu et al. (2013) Intervention only partially applied. Use of antibacteri Lin et al. (2013) Poor quality in outcomes measurements, not clear wh
characteristics of participants or confounders not me Matocha and Montero (2012) No access to full text. Matocha and Montero (2012) No access to full text. McMullan et al. (2013) Poor quality in the methodology, baseline was extend
No clear which intervention reduce CLABSI rates. Miller et al. (2010) Poor quality in reporting study outcomes, no confiden Ong et al. (2011) Use of antibacterial-impregnated catheters. (MultiMe
and benzalkonium chloride on both surfaces (Vantex Osorio et al. (2013) Study not written in English Parada et al. (2013) No access to full text. Popovich et al. (2010) Poor quality in reporting of outcomes. Characteristics
Different bundles applied during intervention but not Ramirez et al. (2012) Poor quality in methodology, different types and bran
reported. Render et al. (2011) Poor methodology quality, self-reported data, charact Rosenthal et al. (2010) Use of antibacterial-impregnated catheters. Sacks et al. (2014) Use of antibacterial-impregnated catheters. Triple lum
coated catheters were used. Saldanha et al. (2014) No access to full text. Seyman et al. (2014) Poor quality on the methodology, subjective interven Stone et al. (2007) Weak methodology, surveys and self-reported data, c
participants not reported. Thom et al. (2014) Poor methodology quality various interventions were
rate. Vigorito et al. (2011) Weak methodology self-reported data, demographics Weeks et al. (2014) Weak methodology self-reported data, change of tool
ritical Care Nursing 43 (2017) 12–22
• prevention • control • Catheters, Indwelling • in-dwelling catheters • Catheter-Associated Infection • Catheter-Related Infection • Intensive Care, Surgical • Critical care
See Table B1.
Appendix C. Supplementary data
Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.iccn.2017.05.006.
ine data only comparison with National Healthcare Safety Network data. on (development of the clinical information technology assessment tool) with
based hand rub bottles. The amount of alcohol based rub was not quantified. The s were half empty.
ent of CLABSI intervention compliance.
interventions of the study. No discussion of confounders.
gardBlue PLUS Multilumen CVC, antimicrobial surface coated using Chlorhexidine, rs)
al-impregnated catheters. ich intervention resulted in the reduction of CLABSI. Self-reported data,
asured or discussed, no confidence intervals provided.
ed during study not clear, outcomes measurements not correlated with study aim.
ce interval provided. Use of antibacterial-impregnated catheters. d CVC and Intro-Flex) or CVCs impregnated with silver platinum carbon amalgam
of participants not reported. Confidence interval only reported for one outcome. reported or considered in the outcomes measurements, no confounders considered. ds of caps were used during intervention. Characteristics of participants not
eristics of participants not reported.
en second generation antimicrobial catheter Arrow g + ard Blue Plus antimicrobial
tion, no characteristics of participants and confounders addressed. orrelation between intervention and outcome not clear, characteristics of
applied during study, no clear indication which intervention reduced the CLABSI
measured not relevant to possible confounders of the study. during study period. Missing data.
D.C. Velasquez Reyes et al. / Intensive
dams, A., Wilson, S., 2012. The impact of using chlorhexadine gluconate products in the adult critical care setting. Am. J. Infect. Control 40, e175–e176.
l-Tawfiq, J.A., Abed, M.S., Memish, Z.A., 2012. Peripherally inserted central catheter bloodstream infection surveillance rates in an acute care setting in Saudi Arabia. Ann. Saudi Med. 32, 169–173.
ntonelli, M., De Pascale, G., Ranieri, V.M., Pelaia, P., Tufano, R., Piazza, O., et al., 2012. Comparison of triple-lumen central venous catheters impregnated with silver nanoparticles (AgTive® ) vs conventional catheters in intensive care unit patients. J. Hosp. Infect. 82, 101–107.
arnett, A.G., Graves, N., Rosenthal, V.D., Salomao, R., Rangel-Frausto, M.S., 2010. Excess length of stay due to central line-associated bloodstream infection in intensive care units in Argentina, Brazil, and Mexico. Infect. Control Hosp. Epi- demiol. 31, 1106–1114.
ashir, M.H., Olson, L.K., Walters, S.A., 2012. Suppression of regrowth of normal skin flora under chlorhexidine gluconate dressings applied to chlorhexidine gluconate-prepped skin. Am. J. Infect. Control, 344–348.
iswal, M., Mewara, A., Appannanavar, S.B., Taneja, N., 2015. Mandatory public reporting of healthcare-associated infections in developed countries: how can developing countries follow. J. Hosp. Infect. 90, 12–14.
lot, K., Bergs, J., Vogelaers, D., Blot, S., Vandijck, D., 2014. Prevention of central line- associated bloodstream infections through quality improvement interventions: a systematic review and meta-analysis. Clin. Infect. Dis. 59, 96–105.
onello, R.S., Fletcher, C.E., Becker, W.K., Clutter, K.L., Arjes, S.L., Cook, J.J., et al., 2008. An intensive care unit quality improvement collaborative in nine Department of Veterans Affairs hospitals: reducing ventilator-associated pneumonia and catheter-related bloodstream infection rates. Joint Commission J. Qual. Patient Saf. 34, 639–645.
ollege of Intensive Care Medicine of Australia and New Zealand [CICM], 2011. College of Intensive Care Medicine of Australia and New Zealand Minimum Stan- dards for Intensive Care Units College of Intensive Care Medicine of Australia and New Zealand., pp. 1–18.
entre for Disease Control and Prevention, 2011. CDC Vitalsigns Report March 2011. CDC Centre for Disease Control and Prevention.
entre of Control and Disease Prevention, 2014. Central Line Associated Blood- stream Infection CLABSI. Device Associated Module CLABSI. Centre of Control and Disease Prevention.
handler, J., Churchill, R., Higgins, J., Lasserson, T., Tovey, D., 2013. Methodological Standards for the Conduct of Cochrane Intervention Reviews (Version 2.3, 02 December, 2013). Cochrane Effective Practice and Organization of Care Review Group (EPOC), The Cochrane Collaboration.
herifi, S., Gerard, M., Arias, S., Byl, B., 2013. A multicenter quasi-experimental study: impact of a central line infection control program using auditing and perfor- mance feedback in five Belgian intensive care units. Antimicrob. Resist. Infect. Control 2, 1–7.
hordas, L., 2004. Making the grade: insurers are increasingly turning to The Leapfrog Group and other organizations for hospital care and safety data to help prevent medical errors and encourage better-informed health-care decisions. (Patient Safety). Best’s Rev. 105, 63.
oherty, M., Axelrod, P., 2011. Can a new connector reduce the number of central line-associated bloodstream infections (CLABSIs)? Am. J. Infect. Control 39, E55.
umont, C., Nesselrodt, D., 2012. Preventing central line-associated bloodstream infections CLABSI. Nursing (Lond.) 42, 41–46.
dge, V.L., Odoi, A., Fyfe, M., Macdougall, L., Majowicz, S.E., Doré, K., et al., 2007. Physi- cian diagnostic and reporting practices for gastrointestinal illnesses in three health regions of British Columbia. Can. J. Public Health = Revue canadienne de santé publique. 98, 306.
xline, M.C., Ali, N.A., Zikri, N., Mangino, J.E., Torrence, K., Vermillion, B., et al., 2013. Beyond the bundle – journey of a tertiary care medical intensive care unit to zero central line-associated bloodstream infections. Crit. Care 17, 1–13.
agan, R.P., Edwards, J.R., Park, B.J., Fridkin, S.K., Magill, S.S., 2013. Incidence trends in pathogen-specific central line–associated bloodstream Infections in US inten- sive care units, 1990–2010. Infect. Control Hosp. Epidemiol. 34, 893–899.
uruya, E.Y., Dick, A., Perencevich, E.N., Pogorzelska, M., Goldmann, D., Stone, P.W., 2011. Central line bundle implementation in US intensive care units and impact on bloodstream infections. PLoS One 6, 15452–15458.
arrett, D.O., Clifford McDonald, A., Wanderley, C., Wanderley, P., Miller, J., Carr, M., et al., 2002. An outbreak of neonatal deaths in Brazil associated with contami- nated intravenous fluids. J. Infect. Dis. 186, 81–86.
iggins, J., Green, S., 2011. Cochrane handbook for systematic reviews of interven- tions. In: Collaboration TC editor. The Cochrane Collaboration.
iggins, J., Altman, D.G., Gøtzsche, P.C., Jüni, P., Moher, D., Oxman, A.D., et al., 2011. The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials. BMJ (Online) 343.
eong, I.S., Park, S.M., Lee, J.M., Song, J.Y., Lee, S.J., 2013. Effect of central line bundle on central line-associated bloodstream infections in intensive care units. Am. J. Infect. Control 41, 710–716.
oanna Briggs Institute, 2014. The Joanna Briggs Institute Joanna Briggs Institute Reviewers’ Manual, 2014 edition. The Joanna Briggs Institute.
allen, A.J., Patel, P.R., O’Grady, N.P., 2010. Preventing catheter-related bloodstream
infections outside the intensive care unit: expanding prevention to new settings. Clin. Infect. Dis. 51, 335–341.
im, J.S., Holtom, P., Vigen, C., 2011. Reduction of catheter-related bloodstream infections through the use of a central venous line bundle: epidemiologic and economic consequences. Am. J. Infect. Control 39, 640–646.
ritical Care Nursing 43 (2017) 12–22 21
Kwakman, P.H., Muller, M.C., Binnekade, J.M., Akker, J.P., Borgie, C.A., Schultz, M.J., et al., 2012. Medical-grade honey does not reduce skin colonization at central venous catheter-insertion sites of critically ill patients: a randomized controlled trial. Crit. Care 16 (5), 1–8, R214.
Liu, H., Liu, H., Deng, J., Chen, L., Yuan, L., Wu, Y., 2014. Preventing catheter-related bacteremia with taurolidine-citrate catheter locks: a systematic review and meta-analysis. Blood Purif. 37, 179–187.
Loveday, H.P., Wilson, J.A., Pratt, R.J., Golsorkhi, M., Tingle, A., Bak, A., et al., 2014. epic3:National evidence-based guidelines for preventing healthcare-associated infections in NHS hospitals in England. J. Hosp. Infect. 86, S1–S70.
Macías, A.E., Munoz, J.M., Bruckner, D.A., Galván, A., Rodríguez, A.B., Guerrero, F.J., et al., 1999. Parenteral infusions bacterial contamination in a multi-institutional survey in Mexico: considerations for nosocomial mortality. Am. J. Infect. Control 27, 285–290.
Maki, D.G., Rosenthal, V.D., Salomao, R., Franzetti, F., Sigfrido Rangel-Frausto, M., 2011. Impact of switching from an open to a closed infusion system on rates of central line-associated bloodstream infection: a meta-analysis of time-sequence cohort studies in 4 countries. Infect. Control Hosp. Epidemiol. 32, 50–58.
Mangum, D.S., Verma, A., Weng, C., Sheng, X., Kirchhoff, A.C., Fluchel, M., et al., 2013. A comparison of central lines in pediatric oncology patients: early removal and patient centered outcomes. Pediatr. Blood Cancer 60, 1890–1895.
Marsteller, J., Sexton, J.B., Hsu, Y.J., Hsiao, C.J., Holzmueller, C.G., Pronovost, P.J., et al., 2012. A multicenter, phased, cluster-randomized controlled trial to reduce cen- tral line-associated bloodstream infections in intensive care units. Crit. Care Med., 2933–2939.
Marsteller, J., Hsu, Y.-J., Weeks, K., 2014. Evaluating the impact of mandatory pub- lic reporting on participation and performance in a program to reduce central line–associated bloodstream infections: evidence from a national patient safety collaborative. Am. J. Infect. Control 42, 209–215.
Matocha, D., Montero, M., 2012. Reaching zero central line associated infections by improving compliance to aseptic technique. Am. J. Infect. Control 40, e132.
McLaws, M.L., Burrell, A.R., 2012. Zero risk for central line-associated bloodstream infection: are we there yet. Crit. Care Med. 40, 388–393.
Mimoz, O., Villeminey, S., Ragot, S., Dahyot-Fizelier, C., Laksiri, L., Petitpas, F., et al., 2007. Chlorhexidine-based antiseptic solution vs alcohol-based povidone- iodine for central venous catheter care. Arch. Intern. Med., 2066–2072.