Practice guideline on the prevention and treatment of central line-associated bloodstream infection: Part 1––Diagnosis and prevention

Abstract

Central line-associated bloodstream infection (CLABSI) is a severe complication of indwelling intravascular catheters and a leading cause of healthcare-associated infections in intensive care units (ICUs). CLABSI significantly increases mortality, prolongs hospital stays, and elevates healthcare costs. Recent large-scale clinical studies of CLABSI diagnosis, treatment, and prevention, alongside evolving antibiotic resistance patterns and new antimicrobial developments, have necessitated updates to clinical management strategies. Consequently, the Chinese Society of Critical Care Medicine (CSCCM) has updated the 2007 guideline on intravascular catheter-related infections. This guideline was developed using Grading of Recommendations Assessment, Development, and Evaluation (GRADE) methodology for evidence assessment. A multidisciplinary working group formulated clinical questions, conducted systematic literature reviews, performed meta-analyses, and synthesized evidence to draft recommendations. The recommendations underwent iterative revisions through expert panel reviews, remote and in-person meetings, and two rounds of voting by the Society’s Standing Committee before finalization. The guideline comprises 52 recommendations, focusing on adult patients in ICU with central venous catheters. This document, Part 1, addresses diagnosis and prevention, formulated using GRADE methodology. The aims of the guideline are to standardize best practices, reduce CLABSI incidence, and improve the outcomes of patients in ICUs.

Practice Guideline Registration: Practice Guideline Registration for Transparency (PREPARE-2024CN873).

Introduction

Recent rapid advances in critical care medicine have led to widespread use of intravascular catheters in intensive care units (ICUs). These catheters are essential for administering therapeutic drugs and blood products and also serve as critical tools for hemodynamic monitoring, blood purification, and extracorporeal membrane oxygenation (ECMO). The most severe complication of indwelling intravascular catheters is central line-associated bloodstream infection (CLABSI), which is a leading cause of ICU- or hospital-acquired infections.^[1]

Intravascular catheters are vascular access devices. Commonly used types of intravascular catheters include: (1) central venous catheters (CVCs) inserted via the internal jugular, subclavian, or femoral veins; (2) peripherally inserted central catheters (PICCs); and (3) implantable venous access ports (IVAPs), placed via the internal jugular or subclavian veins. Multiple randomized controlled trials (RCTs) comparing infection risks across insertion sites have demonstrated that femoral vein catheters carry the highest risk of CLABSI, followed by internal jugular and subclavian vein catheters.^[[2], [3], [4]^]

Over 50% of bloodstream infections in ICUs are catheter-related. In China, the reported incidence of CLABSI in ICUs is 1.5/1000 catheter-days, which is comparable to, or slightly lower than, rates in developed countries (1.8/1000 catheter-days to 5.2/1000 catheter-days);^[3,4] however, CLABSI contributes significantly to hospital-acquired infections in China, with marked regional and interhospital variability, and reducing CLABSI incidence remains a critical challenge.

The predominant pathogens causing CLABSI in China are gram-positive bacteria, particularly Staphylococcus species (27.07%), including coagulase-negative staphylococci, Staphylococcus aureus, and Enterococcus. S. aureus infections may correlate with contamination during emergency catheter placement.^[[3], [4], [5], [6], [7]^] Furthermore, infections caused by gram-negative bacteria are rising, most commonly Enterobacteriaceae (22.31%), followed by Klebsiella pneumoniae, Pseudomonas aeruginosa, and Acinetobacter baumannii.^[8] The widespread use of broad-spectrum antibiotics has led to increases in drug-resistant pathogens, such as vancomycin-resistant Enterococcus, methicillin-resistant coagulase-negative staphylococci, carbapenem-resistant Enterobacteriaceae, and carbapenem-resistant A. baumannii.^[9] Fungal infections, primarily Candida species, are also concerning, as they triple the mortality risk relative to non-Candida CLABSI.^[11] CLABSI increases sepsis risk by 4%–14% and mortality risk by 12%–25%, prolongs hospital stays, and escalates healthcare costs, imposing a substantial economic burden.^[8]

Given evolving medical technologies, antibiotic resistance trends, and novel antimicrobial agents, the Chinese Society of Critical Care Medicine (CSCCM) has updated the Guidelines for the Prevention and Treatment of Intravascular Catheter-Related Infections (2007). This revision integrates the latest evidence to provide recommendations for CLABSI prevention, diagnosis, and treatment, with the aim of standardizing best practices for central intravascular catheter management in adult ICU patients.

Guideline Development Process

Working group and development workflow

Initiated by the CSCCM, this guideline was developed by a multidisciplinary working group of 23 experts in critical care, biostatistics, evidence-based medicine, and medical informatics. The process was registered on the International Platform for Practice Guidelines Registration and Transparency (Registration ID: PREPARE-2024CN873). The steps taken to finalize the document were: initiation, proposing clinical questions, systematic literature searches and screening, meta-analysis and evidence synthesis based on evidence-based medicine, drafting preliminary recommendations, group review by correspondence, iterative revisions via remote/in-person meetings of the working group, and a final vote by the standing committee of the CSCCM.

Formulation of clinical questions

The guideline focuses on adult ICU patients with indwelling central intravascular catheters, addressing catheter insertion sites, techniques, selection, management, and CLABSI prevention, diagnosis, and treatment. Clinical questions were refined by reviewing prior guidelines and incorporating new high-priority issues aligned with recent advances. All clinical questions were formulated according to the PICO (population, intervention, control, outcome) principle.^[10]

Literature search and screening strategy

A systematic search of the PubMed, Embase, Ovid, Cochrane Library, and Web of Science databases, limited to Chinese and English publications, was conducted from January 1990 to February 2025. Search keywords included: catheter-related bloodstream infection, CLABSI, bacteremia, and their Chinese equivalents. Literature screening adhered to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines, was focused on adult ICU patients, and excluded studies related to children; therefore, the recommendations of this guideline are only applicable to adult ICU patients. Extension of these recommendations to children will require the retrieval of child-related literature and an additional evidence-based medicine evaluation.

Evidence and recommendation grading

According to the principles of the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) system, evidence quality was assessed on a scale from high to very low to determine the strength of recommendations (Table 1, Table 2).^[10]

Table 1.

Evidence quality assessment.

Methodological basis 1.High: RCTs or observational studies upgraded by two levels 2.Moderate: RCTs downgraded by one level or observational studies upgraded by one level 3.Low: RCTs downgraded by two levels or observational studies 4.Very Low: RCTs downgraded by three levels, observational studies downgraded by one level, case series, expert opinions, or other low-quality evidence

Factors that may decrease evidence strength 1.High risk of bias in available RCTs (methodological flaws suggesting significant bias) 2.Inconsistency of results, including discrepancies in subgroup analyses 3.Indirect evidence (e.g., different populations, interventions, controls, or outcomes) 4.Imprecision of results 5.High likelihood of reporting bias

Factors that may increase evidence strength 1.When the effect size is large, the quality of evidence is upgraded by one level (direct evidence, relative risk (RR) >2 and no suspected confounding factors) 2.When the effect size is very large, the quality of evidence is upgraded by two levels (direct evidence, RR >5 and no confounding factors affecting effectiveness) 3.Dose–response relationship

RCT: Randomized controlled trial.

Table 2.

Factors determining strong or weak recommendations.

Considerations	Recommendation process
Availability of high/moderate-quality evidence	Higher evidence quality increases the likelihood of a strong recommendation
Certainty of risk/benefit balance	Larger and more certain differences in outcomes between intervention and control groups favor a strong recommendation. Smaller net benefit or greater uncertainty leads to a weak recommendation
Certainty and consistency of effect estimates	More precise or consistent effect estimates (e.g., narrow confidence intervals) favor a strong recommendation
Cost of achieving desired benefits	Lower intervention costs compared to control (e.g., reduced resource use) favor a strong recommendation

Evidence levels: GRADE methodology evaluates evidence quality based on five aspects: (1) risk of bias, (2) inconsistency of results, (3) indirect evidence, (4) imprecision of results, and (5) reporting bias. The GRADE criteria for evaluating evidence quality are presented in (Table 1). The rating of high-quality evidence derived from RCTs can be downgraded due to factors listed in (Table 1); similarly, low-quality evidence from observational (non-randomized) studies can be upgraded based on these factors.

Recommendation strength: The GRADE methodology classifies recommendation levels into strong or weak recommendations; factors influencing recommendation strength are outlined in (Table 2). The collaboration group formulated recommendation levels based on evidence quality, intervention risk/benefit ratio, the certainty of outcomes, resource accessibility, and intervention feasibility and acceptability. When an intervention was clearly shown to have more benefits than harms, it was considered a strong recommendation. When an intervention may have more benefits than harms, but there were some uncertainties (such as low quality of evidence), it was considered a weak recommendation. When the evidence for benefits or risks was difficult to summarize or evaluate using the GRADE methodology, after discussion, the collaboration group formed a best practice statement (BPS), using the criteria outlined in (Table 3).

Table 3.

Best practice statement (BPS) criteria.

Criteria
1	Is the statement clear and actionable?
2	Is the recommendation necessary?
3	Is the net benefit (or harm) unequivocal?
4	Is the evidence difficult to collect or summarize?
5	Is the rationale for the recommendation explicit?
6	Is a BPS more appropriate than formal GRADE methods?

GRADE: Grading of Recommendations Assessment, Development, and Evaluation.

Expert voting

After discussing each recommendation in face-to-face or online meetings, the working group drafted the recommendations. All members then voted (agree/disagree/abstain) anonymously via a questionnaire system. If ≥75% of the guideline group members voted and the support rate was ≥80%, the recommendation was accepted. If no consensus was reached after the voting ended, voters could provide feedback for the working group to consider revising or discarding the recommendation.

Diagnosis of CLABSI

Question 3.1: How should specimens be collected from patients with suspected CLABSI?

When there is clinical suspicion of CLABSI, a standardized collection of two sets of blood samples (four in total, one set each of aerobic and anaerobic bottles) is necessary. One set should be collected from peripheral venous blood via peripheral venipuncture, and the other from venous blood via a catheter hub. If catheter removal is required, avoid contact with the patient’s skin during removal. Cut off approximately 5 cm of the catheter tip under sterile conditions, place it in a dry sterile container for inspection, and perform Maki’s semi-quantitative culture.^[12] For non-tunneled CVC, PICC, tunneled CVC, and IVAP, collection of at least two sets of blood cultures from patients suspected of having CLABSI is recommended, with at least one set being peripheral venous blood collected via skin puncture and another set of catheter blood collected via the catheter hub or IVAP.^[13] In addition to submitting the IVAP tip for inspection, reservoir submission may improve CLABSI diagnosis accuracy. If a multi-lumen catheter is used, it is recommended to take separate blood cultures from each lumen. Note that catheter blood and peripheral blood should be collected simultaneously and labeled to distinguish them.^[14]

Twenty-three percent of patients with catheter-related infections exhibit local infection symptoms and signs, manifesting as erythema, induration, fever, pain, tenderness, or purulent discharge ≤2 cm around the puncture site.^[15] Tunnel or pocket infections can present as complex local infections, with erythema and induration extending >2 cm from the puncture site, purulent discharge, and skin necrosis, among other features. Samples of local purulent discharge or exudate for Gram staining and culture can be collected by dry swab, while peripheral venous blood can be collected via percutaneous puncture for culture^[15]

Question 3.2: How should microbiological diagnosis of CLABSI be conducted?

Despite continuous advances in molecular diagnostic techniques, blood culture remains the gold standard for diagnosing CLABSI. Typically, each blood culture bottle should contain 8–10 mL of blood; <5 mL per bottle may result in false negatives or delayed microbial growth, while >10 mL may generate false positives due to the large amount of background CO₂ produced by white blood cells. It is recommended that blood culture specimens are sent to the laboratory within 2 h (not >4 h) after collection, transported at room temperature (20 °C–25 °C), and under conditions meeting biosafety requirements. If there is a delay in transportation, specimens should be stored at room temperature. Even under ideal disinfection conditions, 3%–5% of blood cultures may be contaminated with bacteria from the skin or environment; therefore, a comprehensive evaluation combining patient clinical symptoms and laboratory indicators is necessary when blood culture results are positive.^[16]

Detection of bloodstream infections in clinical laboratories currently involves both traditional culture and molecular biology (e.g., nucleic acid hybridization, nucleic acid amplification and DNA sequence analysis, gene chips, and matrix-assisted laser desorption ionization-time of flight mass spectrometry) methods. Clinicians select appropriate pathogen detection methods based on suspected pathogen types, patient epidemiology, and clinical presentation, and collect corresponding clinical specimens for testing. Laboratories should evaluate the methodological performance characteristics, turnaround time, influencing factors, etc., of different detection methods, and recommend the selection of optimal tests in communication with clinicians. For patients with suspected CLABSI, use of a combination of multiple detection techniques is recommended to improve pathogen detection and patient survival rates, and reduce the health economic burden of related treatments.^[11,17]

Question 3.3: What should the diagnostic criteria for CLABSI be?

CLABSI is currently more of an exclusionary diagnosis, referring to bacteremia or fungemia occurring ≤48 h after the insertion or removal of an intravascular catheter, accompanied by symptoms such as fever (>38 °C), chills, or hypotension, without other identifiable sources of infection, apart from the intravascular catheter. Peripheral venous blood cultures show positive for bacteria or fungi, or the same type of pathogen with the same antibiotic sensitivity results is cultured from both the catheter tip and peripheral blood.^[18,19]

Prevention of CLABSI

Question 4.1: How should the indwelling site for a CVC be chosen?

Recommendation: The subclavian vein is the preferred site for CVC insertion. (weak recommendation, low-quality evidence).

Rationale: Sites most commonly chosen for central venous catheterization include the subclavian, internal jugular, and femoral veins. The impact of puncture site on the risk of CLABSI is related to the density of skin flora at each site. Compared to the subclavian vein, bacterial colonization occurs earlier at the insertion points of femoral and internal jugular vein catheters, increasing the risk of CLABSI.

Although the subclavian vein is challenging to locate and carries a higher risk of pneumothorax during puncture, it has a long subcutaneous tunnel, making it easier to secure. The catheter insertion site has low mobility, is less prone to contamination from dressing displacement, and has the lowest skin flora density, making it the optimal site for central venous catheterization. The internal jugular vein is superficial and easy to locate, but neck movement can easily cause dressing displacement or detachment at the puncture site, increasing the likelihood of contact with microorganisms. The femoral vein catheter site is adjacent to the perineal area, which increases the risk of contamination by excreta. In addition, the secretions and skin folds in the perineum and groin areas contribute to high rates of bacterial colonization and infection. Axillary vein puncture has become more common with the widespread use of ultrasound guidance and is similar to the subclavian vein in terms of lowering CLABSI risk while reducing the risk of pneumothorax; however, research evidence from large-sample studies is lacking.

Evidence analysis: Twenty studies (eight RCTs and twelve observational studies) comparing the impact of different puncture sites on the risk of CVC-related infections in critically ill patients were identified through a systematic literature search. A multicenter RCT (3SITES) reported by Timsit et al.^[20] found that the risk of bloodstream infections was significantly higher in the femoral vein and internal jugular vein catheter groups than in the subclavian vein catheter group; however, there was a higher risk of pneumothorax with subclavian vein puncture. A 2017 meta-analysis by Arvaniti et al.^[21] revealed a similar risk of catheter-related bloodstream infections in the internal jugular and subclavian veins, while the risk was higher in the femoral vein than that in the subclavian vein (RR=2.44, 95% CI: 1.25 to 4.75, I²=61%) and lower in the internal jugular vein than in the femoral vein (RR=0.55, 95% CI: 0.34 to 0.89, I²=61%).

For this guideline, we conducted separate meta-analyses of eight RCTs^[20,[22], [23], [24], [25], [26], [27], [28]^] and 12 observational studies^[[29], [30], [31], [32], [33], [34], [35], [36], [37], [38], [39], [40]^], to compare the effects of different puncture sites on CLABSI occurrence in critically ill patients. Risk of catheter-related bloodstream infections was lower for the subclavian vein than for the internal jugular and femoral veins (RCT meta-analysis RR=0.51, 95% CI: 0.30 to 0.87, I²=46%; observational studies meta-analysis RR=0.65, 95% CI: 0.52 to 0.82, I²=55%).

Question 4.2: How should the vascular catheter placement site be chosen for blood purification treatment?

Recommendation: The preferred catheter insertion site for blood purification therapy is the right internal jugular vein, followed by the femoral vein (weak recommendation, low-quality evidence).

Rationale: Placement of indwelling blood purification catheters in the right internal jugular vein results in a low infection rate, poses no risk of thoracic duct injury compared to the left internal jugular vein, does not affect patient ability to get out of bed and perform rehabilitation activities, and is close to the right atrium, making it less likely to adhere to the vessel wall. Such placement is suitable for long-term indwelling, with a typical duration of ≤4 weeks; however, in ICU patients requiring blood purification, the internal jugular and subclavian veins are often needed for monitoring central venous pressure, rapid fluid replacement, intravenous nutrition, and vasoactive drug infusion, while also preserving sites for other life-support devices, such as ECMO. The femoral vein is an option for short-term blood purification catheterization because it is easy to puncture, has a low incidence of complications (such as hematoma) that are easily managed, and poses no risk of pneumothorax, with an indwelling time generally ≤2 weeks.

Evidence analysis: Systematic literature searches identified six studies comparing catheter insertion site selection and infection risk related to blood purification: one RCT and five observational studies. The RCT reported 87 infection-related events in 648 cases of femoral vein catheterization and 81 events in 626 non-femoral vein catheterization cases, which was not a significant difference (95% CI: 0.80 to 1.35, P=0.45, I²=0%).^[41] We conducted a meta-analysis of five observational studies^[[42], [43], [44], [45], [46]^] for this guideline, which indicated that the infection risk for femoral vein catheterization is significantly higher than that for non-femoral vein catheterization (95% CI: 1.16 to 2.60, P=0.008, I²=32%). Overall, the available data indicate that the incidence of infection-related events is significantly higher in patients catheterized via the femoral vein than in those where non-femoral vein sites are used (95% CI: 1.07 to 2.32, P=0.02, I²=38%).

Question 4.3: How should vascular catheter insertion site be chosen for ECMO treatment?

Recommendations: Vascular catheter puncture site should be chosen based on ECMO support mode, as follows: (1) For Veno-Arterial ECMO (VA-ECMO), peripheral cannulation is recommended, with the left/right femoral vein and left/right femoral artery as the first choices (BPS); (2) For Veno-Venous ECMO (VV-ECMO), right femoral vein-right internal jugular vein cannulation is recommended, with the right internal jugular vein as the first choice for dual-lumen single cannulation (BPS); (3) For Veno-Venous-Arterial ECMO (VV-A-ECMO) and Veno-Arterial-Venous ECMO (V-AV-ECMO), the right femoral vein and right internal jugular vein/left femoral artery are preferred cannulation sites (BPS).

Rationale: In VA-ECMO, peripheral vascular cannulation is simpler to perform and easier to manage than central cannulation via thoracotomy, causing less trauma, with lower risks of mediastinal infection and bleeding. Particularly in emergencies, such as cardiac arrest, the femoral vein is the preferred site for drainage cannulation. The puncture site for return cannulation can be the femoral artery or axillary artery. Axillary artery puncture is inconvenient and associated with a higher incidence of right hemisphere stroke; in the ICU, femoral artery puncture is more convenient and most commonly chosen. To reduce vascular complications and facilitate hemostasis by compression when removing the cannula, placement of the drainage and return cannulas in different limbs is optimal. Therefore, the left/right femoral vein and left/right femoral artery are preferred for VA-ECMO.

Traditional VV-ECMO cannulation generally involves placing two single-lumen cannulas, including femoral vein-jugular vein and femoral vein-femoral vein routes.^[47] For VV-ECMO, the femoral vein (drainage) and internal jugular vein (return) positions are usually chosen. Anatomically, the right internal jugular vein and right femoral vein provide more direct access to the superior and inferior vena cava and right atrium, and can be easily converted to VA mode if necessary. Thus, the right femoral and right internal jugular veins are preferred for VV-ECMO cannulation. If choosing dual-lumen single cannulation, the pathway from the right internal jugular vein through the right atrium to the inferior vena cava is straighter, with higher success rates for puncture and cannulation and fewer complications, as it provides the optimal path for oxygenated blood to enter the pulmonary artery from the right ventricle, making it the preferred cannulation site.

Evidence analysis: A retrospective cohort study conducted by Allou et al. in 2019 reported that the infection rate associated with femoral vein-femoral vein cannulation was significantly higher than that of the femoral vein-jugular vein.^[48] Therefore, clinically, the femoral vein-jugular vein is generally preferred as the cannulation site for VV-ECMO.

Meta-analyses by Raffa et al.^[49] and Mariscalco et al.^[50] showed that, relative to central cannulation, peripheral cannulation significantly reduced the risk of bleeding (P=0.02) and continuous blood purification (P=0.03); both meta-analyses had a moderate confidence level, according to AMSTAR-2. A meta-analysis conducted by Biancari et al.^[51] found that, among 556 patients with central cannulation and 713 with peripheral cannulation, central cannulation in VA-ECMO was associated with a higher in-hospital mortality rate than peripheral cannulation (70.8% vs. 64.5%, P=0.03); this meta-analysis also has a moderate confidence level (AMSTAR-2). These results may be related to complications such as massive bleeding/transfusion and bleeding/tamponade associated with central cannulation.^[52] Therefore, peripheral vessels are more often chosen as the preferred cannulation sites for VA-ECMO in the clinic.

Question 4.4: How should the insertion site be chosen for PICC placement?

Recommendation: The basilic vein is the preferred site for PICC placement (BPS).

Rationale: PICC insertion mainly involves the elbow veins, where the skin has a lower colony count, with relatively low oiliness and humidity, and the catheter is easy to secure, dressings are easy to change, and there is less chance of contamination. Commonly used elbow veins are the basilic, median cubital, and cephalic veins. The basilic vein is thick, straight, has fewer valves, is fixed in position, is closest to the superior vena cava, and has a high success rate for puncture. The median cubital vein is relatively thick and straight, but has more valves. The anterior cephalic vein is thicker than the posterior part of the vessel, with many branches, making it easy to inadvertently enter the axillary vein or internal jugular vein. The basilic vein has relatively low rates of CLABSI incidence. Therefore, the basilic vein, which has a low infection rate, is thick, straight, and has fewer valves, and should be the first choice for puncture and catheterization.

Evidence analysis: Literature search retrieved four studies comparing the impact of PICC puncture sites on CLABSI incidence: one retrospective cohort, one prospective cohort, and two observational studies. The PICC-related bloodstream infection rate was 2.31 per 100 catheter days (5.2%).^[53] Among the three different routes (basilic, median cubital, and cephalic veins), the CLABSI incidence rate for PICC placement in the basilic vein was lowest, but the difference was not significant.^[54,55]

Question 4.5: How should the material be chosen for indwelling CVCs?

Recommendations: Catheters made of polyurethane or silicone rubber are suggested, with silicone rubber preferred for long-term indwelling (weak recommendation, very low-quality evidence).

Rationale: The physical, chemical, mechanical, and surface properties of a catheter affect its performance, determine its cellular adhesion and microbial colonization, and depend on the material from which the catheter is manufactured. Silicone rubber and polyurethane are preferred materials for indwelling catheters due to their superior biocompatibility and durability relative to other materials. Polyurethane catheters are relatively harder than those made from silicone, and long-term indwelling catheters can cause friction and irritation at the puncture site, making it difficult for damaged tissues to heal and increasing local infection rates. Silicone rubber is relatively more flexible, which can reduce the risk of catheter-induced vascular injury.

Evidence analysis: Systematic literature search identified five studies comparing the impact of different catheter materials on CLABSI incidence: one observational study, three RCTs, and one meta-analysis. The observational study showed that PVC catheters are more prone to microbial colonization than PTFE catheters.^[56] An RCT involving 104 critically injured patients indicated that PVC catheters were more susceptible to bio-colonization than PTFE catheters; positive culture rate 24.6% vs. 6.9%, respectively (P <0.001).^[57] Another RCT found that polyurethane catheters had the same risk of catheter-related infections as those made of PTFE (5.4% [95% CI: 3.8% to 7.6%] vs. 6.9% [95% CI: 4.9% to 9.6%]).^[58] A systematic review including 13 studies compared the incidence of post-insertion complications between polyurethane and silicone rubber catheters, with an AMSTAR-2 assessment of very low credibility. The results showed no significant difference in the overall complication rates between these two materials, with both having relatively low risks of CLABSI.^[59] In summary, there is no significant difference in infection rates among catheters made of polyurethane, silicone rubber, and PTFE; however, an RCT involving 208 patients showed that PTFE catheters had a higher incidence and score of phlebitis than polyurethane catheters (P <0.001).^[60]

There is no significant difference in the probability of CLABSI occurrence among catheters made of polyurethane, silicone rubber, and PTFE; however, considering factors other than infection, silicone rubber and polyurethane are recommended as the preferred materials for indwelling catheters. Overall, the quality of relevant studies is variable, and large-scale clinical trials are needed for confirmation.

Question 4.6: Should an antimicrobial-coated central vascular catheter be chosen?

Recommendations: Use of antimicrobial-coated central vascular catheters is suggested (weak recommendation, high-quality evidence).

Rationale: Microbial adhesion to the catheter surface is a critical step in the pathogenesis of CLABSI. Use of catheters coated with antimicrobial materials or drugs applied to the catheter surface is a potentially effective measure to prevent CLABSI.

Evidence analysis: A systematic literature search identified three meta-analyses evaluating the efficacy of coated catheters in preventing CLABSI. A 2016 meta-analysis published by Lai et al. included 10,405 patients, had low risk of bias, low heterogeneity, and high credibility (AMSTAR-2), and demonstrated that antimicrobial-coated catheters significantly reduced CLABSI and microbial catheter colonization (aRR=2, 95%: CI: 1 to 3, RR=0.62, 95% CI: 0.52 to 0.74).^[61] Wang et al.^[62] conducted a meta-analysis involving 78,362 patients with high AMSTAR-2 credibility, which showed that antimicrobial-coated catheters were superior to non-coated catheters in preventing CLABSI and microbial colonization (RR=0.70, 95% CI: 0.53 to 0.91). Overall, antimicrobial-coated catheters were associated with reduced CLABSI relative to non-coated catheters. A network meta-analysis by Chong et al.,^[63] which evaluated 13 antimicrobial coatings (including chlorhexidine/silver sulfadiazine, minocycline/rifampin, miconazole/rifampin, vancomycin, teicoplanin, cefazolin, benzalkonium chloride, chlorhexidine, and 5-fluorouracil) across 17,255 catheters (AMSTAR-2 credibility rated low), found that minocycline/rifampin-coated catheters were the most effective in preventing CLABSI.

To reduce the incidence of CLABSI, the use of antimicrobial-coated catheters is recommended. Current evidence suggests that minocycline/rifampin-coated catheters are an effective approach for CLABSI prevention, but conclusions remain controversial, necessitating further research.

Question 4.7: How should the number of lumens for indwelling central vascular catheters be determined?

Recommendation: Select the catheter with the fewest lumens necessary to meet clinical diagnostic and therapeutic needs (weak recommendation, low-quality evidence).

Rationale: The use of multi-lumen catheters increases the risk of CLABSI, likely because such catheters provide more opportunities for bacteria and other microorganisms to contact the catheter and cause infection; however, the increased infection risk associated with multi-lumen catheters may be offset by their clinical convenience. Therefore, catheter necessity must be carefully evaluated, and the minimum number of lumens used.

Evidence Analysis: A systematic literature search identified three studies comparing the impact of lumen number on catheter-related infection rates: one observational study and two meta-analyses. The observational study included 1162 patients and showed that the use of multi-lumen catheters significantly increased the risk of CLABSI. For each additional lumen in a CVC, the risk increased (HR=4.4, 95% CI: 2.5 to 7.7), while permanently closing unused lumens appeared to have a protective effect (HR=0.3, 95% CI: 0.1 to 0.7).^[64] A meta-analysis conducted by Dezfulian et al.,^[65], including 6199 patients with low heterogeneity but low AMSTAR-2 credibility, demonstrated that CLABSI was more common with multi-lumen catheters (OR=2.15, 95% CI: 1.00 to 4.66). Another meta-analysis reported by Zürcher et al.^[66] involved 530 patients (low AMSTAR-2 credibility) and concluded that the number of lumens in CVCs influences the incidence of catheter-related bloodstream infections, with single-lumen catheters associated with a lower CLABSI risk than multi-lumen catheters.

The likelihood of CLABSI occurring varies according to the use of single-, double-, and triple-lumen CVCs. Comparatively, single-lumen catheters may have a lower risk of infection, but this conclusion requires further evidence-based medical support. Current recommendations emphasize selecting the catheter with the fewest lumens, based on clinical requirements and availability.

Question 4.8: Is maximum sterile barrier protection required during intravascular catheter insertion?

Recommendation: Maximum sterile barrier precautions must be followed during intravascular catheter insertion (BPS).

Rationale: Patient skin flora or inadequate operator hand hygiene can contribute to CLABSI. The use of extensive sterile barriers and strict aseptic precautions during catheter insertion minimizes microbial transmission, thereby reducing microbial catheter colonization, hospital-acquired infections, and healthcare costs. Operators must wear a surgical cap, surgical mask, perform hand hygiene, and don sterile gloves, sterile gowns, or sterile isolation gowns, with a large sterile drape covering the patient’s entire body. Contaminated or torn gloves must be replaced immediately during the procedure. Assistants should also wear a surgical cap, surgical mask, and perform hand hygiene.

Evidence analysis: A systematic literature search identified three studies comparing the impact of maximum sterile barriers vs. standard sterile barriers on CLABSI rates. A 2013 meta-analysis of eight studies (2887 patients, AMSTAR-2 rating very low) found that maximum sterile barriers significantly reduced CLABSI incidence (OR=0.35, 95% CI: 0.24 to 0.53, P <0.01);^[67] however, an RCT (996 patients) showed no statistically significant difference in CLABSI rates between maximum and standard sterile barriers.^[68] Another study reported that high compliance with maximum sterile barriers (≥95%), as part of a bundled prevention strategy, is critical for reducing CLABSI.^[69] A 2019 Japanese prospective multicenter observational study of 2383 patients indirectly supported that maximum sterile barriers reduce catheter-related infections (aHR=0.20, 95% CI: 0.05 to 0.84).^[70]

While evidence on the effectiveness of maximum sterile barriers remains inconclusive, the clinical and economic consequences of CLABSI justify their use during central catheter insertion, alongside comprehensive preventive measures, to minimize infection risk.

Question 4.9: What skin disinfectant should be selected for use at the insertion site before intravascular catheter placement?

Recommendation: Skin disinfection with alcohol-based chlorhexidine solution is strongly recommended before catheter insertion (strong recommendation, high-quality evidence).

Rationale: The most common route of catheter-related infection involves migration of skin-colonizing bacteria from the insertion site along the catheter surface to the tip.^[71] Using a potent antiseptic that inhibits microbial growth during pre-insertion skin disinfection helps prevent microbial contamination at the insertion site, thereby delaying or reducing catheter colonization and subsequent infectious complications. The three most commonly used disinfectants (chlorhexidine, povidone-iodine, and alcohol) all exhibit broad-spectrum antimicrobial activity.^[72]

Evidence analysis: A systematic search identified eight studies comparing the efficacy of different disinfectants in preventing catheter-related infections. A 2002 meta-analysis (8 studies, AMSTAR-2 credibility very low) found that chlorhexidine gluconate reduced CLABSI risk by 49% relative to 10% aqueous povidone-iodine (RR=0.51, 95% CI: 0.27 to 0.97).^[73] Another meta-analysis (9 studies, 6209 catheters, AMSTAR-2 credibility low) demonstrated that 2% chlorhexidine-alcohol significantly reduced CLABSI risk vs. povidone-iodine (RR=0.23, 95% CI: 0.09 to 0.57).^[74] Furthermore, a 2021 meta-analysis (2815 catheters, AMSTAR-2 credibility high) concluded that 1% chlorhexidine-alcohol was superior in reducing CLABSI risk compared with both 0.5 % chlorhexidine-alcohol (RR=0.40, 95% CI: 0.16 to 0.98) and 10% aqueous povidone-iodine (RR=0.31, 95% CI: 0.15 to 0.63).^[75] No direct comparisons between 2% chlorhexidine-alcohol and 1% chlorhexidine-alcohol have been reported.

CLABSI rates vary depending on the use of chlorhexidine, povidone-iodine, or alcohol for skin disinfection. Current evidence indicates that alcohol-based formulations are superior to aqueous solutions, with chlorhexidine-alcohol significantly reducing CLABSI incidence; however, the optimal concentration of chlorhexidine-alcohol requires further investigation.

Question 4.10: Should ultrasound guidance be used for central vascular catheterization?

Recommendation: We strongly recommend routine use of ultrasound guidance for central vascular localization and catheterization (strong recommendation, moderate-quality evidence).

Rationale: Traditional central venous catheterization relies on the identification of surface anatomical landmarks and palpating vessels. Due to individual variability, traditional localization methods are somewhat inaccurate. Ultrasound-guided central venous catheterization can precisely display the position of blood vessels and their adjacent structures, as well as promptly detect abnormalities in target vessels, such as intravascular thrombosis. Therefore, compared to traditional methods, ultrasound-guided catheterization leads to higher one-time puncture success rates, fewer puncture attempts, and shorter catheterization times. These advantages may reduce the incidence of catheter-related complications, including CLABSI, and increase safety. Hence, ultrasound-guided central venous catheterization is recommended; healthcare professionals performing the procedure should undergo standardized ultrasound training.

Evidence analysis: A systematic literature search identified 15 single-center RCTs comparing ultrasound-guided central venous catheterization with conventional methods.^[[76], [77], [78], [79], [80], [81], [82], [83], [84], [85], [86], [87], [88], [89], [90]^] Among these, 13 studies used anatomical landmark localization as the control, one used blind puncture, and one used conventional methods. Our meta-analysis of these 15 studies, involving a total of 3852 patients, showed that, relative to the control group, ultrasound-guided catheterization had a higher overall success rate (99.1% vs. 88.6%, P <0.001); higher one-time puncture success rate (77.3% vs. 43.9%, P <0.001); fewer puncture attempts (OR=–1.31, 95% CI: –1.71 to –0.92); shorter catheterization time (OR=–85.84 s, 95% CI: –113.87 to –57.81); and lower risk of complications (4.9% vs. 23.4%, P <0.001). Since all of the RCTs were single-center studies with high heterogeneity and bias risk, the quality of evidence is assessed as moderate.

Guideline working group of Chinese Society of Critical Care

Yuetian Yu (Department of Critical Care Medicine, Renji Hospital, School of Medicine, Shanghai Jiao Tong University).

Xiaoqing Li (Department of Critical Care Medicine, Zhongda Hospital of Southeast University).

Bin Ouyang (Department of Critical Care Medicine, The First Affiliated Hospital of SunYatSen University).

Wei Fang (Department of Critical Care Medicine, Shandong Provincial Hospital Affiliated to Shandong First Medical University).

Bo Hu (Department of Critical Care Medicine, Zhongnan Hospital of Wuhan University).

Fen Liu (Department of Critical Care Medicine, the First Affiliated Hospital, Jiangxi Medical College, Nanchang University).

Jiao Liu (Department of Critical Care Medicine, Ruijin Hospital, School of Medicine, Shanghai Jiao Tong University).

Jinglun Liu (Department of Critical Care Medicine, the First Affiliated Hospital of Chongqing Medical University).

Xiuling Shang (Department of Critical Care Medicine, Fuzhou University Affiliated Provincial Hospital, Fujian Provincial Hospital).

Aijun Pan (Department of Critical Care Medicine, The First Affiliated Hospital of USTC, Division of Life Science and Medicine, University of Science and Technology of China).

Changsong Wang (Department of Critical Care Medicine, the First Affiliated Hospital of Harbin Medical University).

Xianghong Yang (Department of Intensive Care Unit, Emergency & Intensive Care Unit Center, Zhejiang Provincial People’s Hospital, Affiliated People’s Hospital, Hangzhou Medical College).

Sheng Zhang (Department of Critical Care Medicine, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine).

Ling Liu (Department of Critical Care Medicine, Zhongda Hospital of Southeast University).

Managing Editor: Jingling Bao/Zhiyu Wang