Scolaris Content Display Scolaris Content Display

Antimicrobial lock solutions for preventing catheter‐related infections in haemodialysis

Contraer todo Desplegar todo

Abstract

Background

Patients undergoing haemodialysis (HD) through a central venous catheter (CVC) are exposed to several risks, being a catheter‐related infection (CRI) and a CVC lumen thrombosis among the most serious. Standard of care regarding CVCs includes their sealing with heparin lock solutions to prevent catheter lumen thrombosis. Other lock solutions to prevent CRI, such as antimicrobial lock solutions, have proven useful with antibiotics solutions, but not as yet for non–antibiotic antimicrobial solutions. Furthermore, it is uncertain if these solutions have a negative effect on thrombosis incidence.

Objectives

To assess the efficacy and safety of antimicrobial (antibiotic, non‐antibiotic, or both) catheter lock solutions for preventing CRI in participants undergoing HD with a CVC.

Search methods

We searched the Cochrane Kidney and Transplant Specialised Register up to 18 December 2017 through contact with the Information Specialist using search terms relevant to this review. Studies in the Register are identified through searches of CENTRAL, MEDLINE, and EMBASE, conference proceedings, the International Clinical Trials Register (ICTRP) Search Portal, and ClinicalTrials.gov.

Selection criteria

We included all randomised or quasi‐randomised control trials (RCTs) comparing antimicrobial (antibiotic and non‐antibiotic) lock solutions to standard lock solutions, in participants using a CVC for HD, without language restriction.

Data collection and analysis

Two authors independently assessed studies for eligibility, and two additional authors assessed for risk of bias and extracted data. We expressed results as rate ratios (RR) per 1000 catheter‐days or 1000 dialysis sessions with 95% confidence intervals (CI). Statistical analyses were performed using the random‐effects model.

Main results

Thirty‐nine studies, enrolling 4216 participants, were included in this review, however only 30 studies, involving 3392 participants, contained enough data to be meta‐analysed. Risk of bias was low or unclear for most domains in the majority of the included studies.

Studies compared antimicrobial lock solutions (antibiotic and non‐antibiotic) to standard sealing solutions (usually heparin) of the CVC for HD. Fifteen studies used antibiotic lock solutions, 21 used non‐antibiotic antimicrobial lock solutions, and 4 used both (antibiotic and non‐antibiotic) lock solutions. Studies reported the incidence of CRI, catheter thrombosis, or both.

Antimicrobial lock solutions probably reduces CRI per 1000 catheter‐days (27 studies: RR 0.38, 95% CI 0.27 to 0.53; I2 = 54%; low certainty evidence), however antimicrobial lock solutions probably makes little or no difference to the risk of thrombosis per 1000 catheter days (14 studies: RR 0.79, 95% CI 0.52 to 1.22; I2 = 83%; very low certainty evidence). Subgroup analysis of antibiotic and the combination of both lock solutions showed that both probably reduced CRI per 1000 catheter‐days (13 studies: RR 0.30, 95% CI: 0.22 to 0.42; I2 = 47%) and risk of thrombosis per 1000 catheter‐days (4 studies: RR 0.26, 95% CI: 0.14 to 0.49; I2 = 0%), respectively. Non‐antibiotic antimicrobial lock solutions probably reduced CRI per 1000 catheter‐days for tunnelled CVC (9 studies: RR 0.60, 95% CI 0.40 to 0.91) but probably made little or no difference with non‐tunnelled CVC (4 studies: RR 0.93, 95% CI 0.48 to 1.81). Subgroup analyses showed that antibiotic (5 studies: RR 0.76, 95% CI 0.42 to 1.38), non‐antibiotic (8 studies: RR 0.85, 95% CI 0.44 to 1.66), and the combination of both lock solutions (3 studies: RR 0.63, 95% CI 0.22 to 1.81) made little or no difference to thrombosis per 1000 catheter‐days compared to control lock solutions.

Authors' conclusions

Antibiotic antimicrobial and combined (antibiotic‐non antibiotic) lock solutions decreased the incidence of CRI compared to control lock solutions, whereas non‐antibiotic lock solutions reduce CRI only for tunnelled CVC. The effect on thrombosis incidence is uncertain for all antimicrobial lock solutions. Our confidence in the evidence is low and very low; therefore, better‐designed studies are needed to confirm the efficacy and safety of antimicrobial lock solutions.

PICO

Population
Intervention
Comparison
Outcome

El uso y la enseñanza del modelo PICO están muy extendidos en el ámbito de la atención sanitaria basada en la evidencia para formular preguntas y estrategias de búsqueda y para caracterizar estudios o metanálisis clínicos. PICO son las siglas en inglés de cuatro posibles componentes de una pregunta de investigación: paciente, población o problema; intervención; comparación; desenlace (outcome).

Para saber más sobre el uso del modelo PICO, puede consultar el Manual Cochrane.

Antimicrobial lock solutions for preventing infections in patients using a catheter for haemodialysis

What is the issue?

Most of the people presenting end‐stage kidney disease use haemodialysis (HD) to replace kidney function. Frequently, a central venous catheter (CVC) is needed to begin HD. In between HD sessions, the CVC needs a sealing solution to avoid catheter thrombosis (an obstruction due to clots), and this is frequently heparin.

In addition to catheter thrombosis, another frequent complication is catheter‐related infection (CRI). CRI originates in the catheter and then spreads to the blood or other organs.

Heparin prevents clot formation but does not prevent infections. Therefore, instead of heparin, the use of sealing solutions that can reduce CRIs has been proposed. These antimicrobial lock solutions could be divided into antibiotic (e.g. vancomycin) and non‐antibiotic (e.g. citrate) solutions. Antimicrobial lock solutions should fill the catheter lumen and then be locked in the catheter during in‐between HD sessions with or without heparin.

What did we do?

We did a systematic review to assess the question whether antimicrobial (antibiotic or non‐antibiotic) lock solutions were better than heparin to prevent CRIs in patients undergoing HD through a CVC and thrombosis compared to heparin. We searched the literature up until 18 December 2017 and identified 39 studies enrolling 4216 patients that met our inclusion criteria.

What did we find?

We included 39 studies, including 3,945 participants undergoing HD through a CVC. The studies compared CVC sealing solutions with heparin to antimicrobial lock solutions. Fifteen studies used only antibiotic lock solutions, 21 used non‐antibiotic lock solutions, and 4 used both (antibiotic and non‐antibiotic) lock solutions. Studies measured the incidence of CRIs and catheter thrombosis, or both. Overall quality of the studies was low for CRIs and very low for thrombosis. There was no information on funding sources for most of the studies.

In general antimicrobial lock solutions are likely superior to standard solutions in preventing CRIs among patients undergoing HD through a CVC, but non‐antibiotic solutions did not prove to reduce CRI. They are no worse than heparin at preventing thrombosis. Other adverse effects were not reported in most studies. Our confidence in these results is low due to the quality of the studies.

Conclusion

Some antimicrobial (antibiotic and the combination of antibiotic‐non antibiotic) lock solutions decrease the incidence of CRIs compared to heparin. Their effect on CVC permeability remains unclear. The quality of the studies is low and very low, respectively; therefore, more studies are needed to confirm the benefits and harms of antimicrobial lock solutions.

Authors' conclusions

Implications for practice

Our review shows that antibiotic antimicrobial and combined (antibiotic plus non‐antibiotic) lock solutions decreased the incidence of CRI compared to control lock solutions, usually heparin, but non‐antibiotic lock solutions did not significantly reduce CRI in participants undergoing HD through a CVC. This beneficial effect was also better established for tunnelled CVC. The effect on the incidence of thrombosis is uncertain for all antimicrobial lock solutions due to imprecision and heterogeneity of results. Our confidence on the evidence is low and very low ; better designed studies are needed to confirm the efficacy and safety of antimicrobial lock solutions. In addition, it is necessary to consider our uncertainty in other potential adverse effects such as ototoxicity from the use of gentamycin, risk of antibiotics resistance, among others, that were not assessed in the studies included in this review.

Implications for research

Although the available data show that some antimicrobial lock solutions are effective in preventing CRI, additional large, well‐designed RCTs are required to evaluate incidence of CRI, thrombosis, and adverse events. These studies should consider a follow‐up of patients sufficiently long to detect microbial resistance in the case of antimicrobial antibiotic solutions and other adverse effects in both the antibiotic and non‐antibiotic antimicrobial lock solutions.

Summary of findings

Open in table viewer
Summary of findings for the main comparison. Antimicrobial lock solutions vs control for preventing catheter‐related infections in patient undergoing haemodialysis

Antimicrobial lock solutions vs control for preventing catheter‐related infections in patient undergoing haemodialysis

Patient or population: CVC‐related infection
Setting: haemodialysis therapy
Intervention: antimicrobial lock solutions
Comparison: heparin and other lock solutions

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with heparin and other lock solutions

Risk with antimicrobial lock solutions

CVC ‐ related infections
assessed with: per 1000 days/catheter

Low

RR 0.38
(0.27 to 0.53)

2994
(27 RCTs)

⊕⊕⊝⊝
LOW 1 2 3 4 5

43 per 1.000

16 per 1.000
(12 to 23)

High

260 per 1.000

99 per 1.000
(70 to 138)

Thrombosis
assessed with: per 1000 days/catheter

Low

RR 0.79
(0.52 to 1.22)

2080
(14 RCTs)

⊕⊝⊝⊝
VERY LOW 6 7 8 9 10

6 per 1.000

5 per 1.000
(3 to 7)

High

330 per 1.000

261 per 1.000
(172 to 403)

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; RR: Risk ratio; OR: Odds ratio;

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

1 This judgment is based on the lack of information regarding the following criteria: random sequence generation, allocation concealment and blinding of outcome assesment

2 the statistics do not show serious heterogeneity and confidence intervals overlap.

3 the evidence is direct because the studies are in hemodialysis patients and the same sealing solutions the question of this review are used.

4 no imprecision is observed; because the decision regarding the use of antimicrobial lock solution is sealed better than heparin along the confidence interval

5 It is suspected a degree of publication bias

6 30% of the studies presented insufficient information to assess citerios of:random sequence generation, allocation concealment and blinding of outcome assesment

7 the statistical test showed a high heterogeneity and the confidence intervals do not overlap

8 the evidence is not indirect because the studies are in hemodialysis patients and the same sealing solutions the question of this review are used.

9 The 95% confidence interval of the pooled estimate ranges from 0.41 to 1.39, which is not narrow enough for a confident judgment of the effect size.

10 Publication bias is suspected as the funnel plot for this outcome shows asymmetry

Background

Description of the condition

Currently, arteriovenous fistulas (AVF) are the preferred access for haemodialysis (HD) patients; nevertheless, when AVFs are not available, a central venous catheter (CVC) has to be installed to start HD. The use of CVC for HD varies among different countries and changes over time in the same country (Ethier 2008). In the United States, the use of CVC has decreased from 27% in 1997 to 15% in 2013 (Pisoni 2015). Using a CVC for HD exposes participants to events such as infections and thrombosis. The use of CVC also increases the risk of mortality and treatment costs (Beathard 2008; Bradbury 2007; Foley 2009; Little 2001; Napalkov 2013). Catheters have a high likelihood of providing an adequate environment for bacterial growth, and leucocytes are unable to surround or phagocytise bacteria. Furthermore, proteins and glycocalyx biofilm coatings of catheters may protect bacteria from antibiotics and leucocytes (Ash 2000). A common complication among participants who undergo HD through CVCs is catheter‐related infection (CRI). The incidence of CRI varies according to different settings and different definitions of CRI, but they are reported at rates of 2.5 to 5.5 cases/1000 catheter‐days, or 0.9 to 2 episodes/patient/year (Katneni 2007; Napalkov 2013). CRI increase treatment costs and adversely affect participants' quality of life, since they often require hospitalisation to remove the CVC and to initiate intravenous antibiotic therapy according to blood cultures (Klevens 2008; NKF 2006). Several interventions to prevent CRI have been described, being the most important the decrease in the number of patients using a HD catheter and the use of strict aseptic protocols, if it is impossible to remove the catheter. Other interventions are the different types of medicated or impregnated dressings (Ullman 2015), catheter impregnation or coating (Lai 2016), application of mupirocin ointment (McCann 2010), and the use of an anti‐infective solution in each CVC lumen while not being used (lock solutions) in order to help prevent colonisation of the intraluminal surface by micro‐organisms that can form a biofilm on the inner wall of the CVC (Labriola 2008; Weijmer 2002).

Description of the intervention

Current standard of care for maintenance of HD CVC is to use lock solutions containing high concentrations of heparin to help prevent thrombosis (Ash 2000). However, the use of heparin may cause complications due to its systemic anticoagulant effect. Heparin can also antagonise bactericidal properties of some antibiotics and may promote biofilm formation (Moran 2008; Vanholder 2010). Locking catheter lumens using anti‐infective solutions – either antibiotic lock solutions (e.g. gentamicin, vancomycin, minocycline, cefazolin, cefotaxime) or non‐antibiotic lock solutions (e.g. sodium citrate, taurolidine) – can help prevent CRI (Betjes 2011; Labriola 2008; Yahav 2008). Antimicrobial (either antibiotic or on‐antibiotic) lock solutions may be administered in combination with an anticoagulant (usually heparin) and sometimes with another antibiotic. Citrate is often administered in combination with taurolidine and heparin (Grudzinski 2015; Labriola 2008; Yahav 2008).

How the intervention might work

It has been demonstrated that, compared to standard care, antibiotic lock solutions can reduce the risk of CRI (Yahav 2008). However, antibiotic lock solutions can increase the likelihood of adverse effects such as ototoxicity (associated with gentamicin). Antibiotic lock solutions can also cause potential antimicrobial resistance, which is the main reason this intervention has not been adopted widely (Venditto 2010). In contrast, non‐antibiotic antimicrobial lock solutions do not present these specific side effects and they are cheaper than heparin, however, their capacity to reduce CRI is uncertain (Grudzinski 2015; Yahav 2008). In vitro studies have indicated that high concentrations of trisodium citrate for locking catheters reduce antimicrobial activity (Weijmer 2005).

Why it is important to do this review

Although the effects of antimicrobial lock solutions for CVC used in HD have been assessed for the last decades, results and recommendations are controversial. A systematic review including eight randomised controlled trials (RCTs) (Labriola 2008) concluded that, compared to heparin, antimicrobial lock solutions reduced the risk of infection (RR 0.32, 95% CI 0.10 to 0.42), but the results did not differentiate between antibiotic lock solutions and non‐antibiotic antimicrobial lock solutions or evaluate the safety of the intervention.

Another systematic review (Yahav 2008) including 16 studies concluded that, compared to heparin, antibiotic lock solutions reduced the risk of CRI (RR 0.39, 95% CI 0.31 to 0.50), whereas the results for non‐antibiotic antimicrobial lock solutions were heterogeneous and effective only when associated with other measures to prevent CRI, such as nasal mupirocin or exit site topical iodine/chlorhexidine (RR 0.37, 95% CI 0.30 to 0.47). A more recent systematic review including five RCTs compared citrate to heparin and found no significant difference for bacteraemia, CVC permeability, and bleeding (Grudzinski 2015).

Although evidence for antibiotic lock solutions looks promising, this is not yet a standard of care due to uncertainty on safety issues. Furthermore, current evidence on non‐antibiotic antimicrobial lock solutions is conflicting and insufficient to recommend their use. Therefore, a new systematic review should assess the safety and efficacy of these interventions in order to help prevent CRIs.

Objectives

To assess the efficacy and safety of antimicrobial (antibiotic, non‐antibiotic, or both) catheter lock solutions for preventing CRI in participants undergoing HD with a CVC.

Methods

Criteria for considering studies for this review

Types of studies

We included RCTs and quasi‐RCTs in which allocation to treatment was obtained by alternation, use of alternate medical records, date of birth or other predictable methods, looking at antimicrobial (antibiotic, non‐antibiotic, or both) catheter lock solutions for preventing CRI and thrombosis in people undergoing HD through a CVC.

Types of participants

Inclusion criteria

Adults or children with acute kidney injury or end‐stage kidney disease undergoing HD using a CVC.

Exclusion criteria

We excluded studies of participants using a CVC for anything other than HD.

Types of interventions

  • All antimicrobial lock solutions: antibiotic (e.g. gentamicin, vancomycin, cefotaxime, and minocycline), non‐antibiotic (e.g. citrate, taurolidine, and ethanol) or both compared to heparin, tissue plasminogen activator, and other lock solutions with unknown antimicrobial properties

  • Studies investigating non‐lock solution interventions were excluded.

  • Studies investigating the treatment of CRI with antimicrobial‐lock solutions were excluded.

Types of outcome measures

Primary outcomes

  • CRI: defined as the presence of symptoms and signs suggesting systemic infection, such as fever (temperature ≥ 38°C) or hypotension, accompanied by positive blood cultures drawn from the catheter and a peripheral vein. Growth of the same micro‐organism in blood cultures drawn through both the catheter and a peripheral vein, without other bacteraemia sources of infection than the CVC.

Secondary outcomes

  • CVC‐related thrombosis, defined as a persistent inability to maintain a blood flow of > 250 mL/min or the need of thrombolytic therapy, or CVC removal due to occlusion.

  • CVC colonisation, defined as a positive culture by any methods in participants with or without signs of infection.

  • Bacteraemia from any sources.

  • Survival of CVC without thrombosis or infection, defined as the number of days the catheter is permeable and free of infection or thrombosis.

  • All‐cause mortality.

  • Adverse effects such as bacterial antibiotic resistance, bleeding episodes, or pulmonary embolism.

Search methods for identification of studies

Electronic searches

We searched the Cochrane Kidney and Transplant Specialised Register up to 18 December 2017 through contact with the Information Specialist using search terms relevant to this review. The Specialised Register contains studies identified from the following sources:

  1. Monthly searches of the Cochrane Central Register of Controlled Trials (CENTRAL)

  2. Weekly searches of MEDLINE OVID SP

  3. Handsearching of kidney‐related journals and the proceedings of major kidney conferences

  4. Searching of the current year of EMBASE OVID SP

  5. Weekly current awareness alerts for selected kidney and transplant journals

  6. Searches of the International Clinical Trials Register (ICTRP) Search Portal and ClinicalTrials.gov.

Studies contained in the Specialised Register are identified through search strategies for CENTRAL, MEDLINE, and EMBASE based on the scope of Cochrane Kidney and Transplant. Details of these strategies, as well as a list of handsearched journals, conference proceedings and current awareness alerts, are available in the Specialised Register section of information about Cochrane Kidney and Transplant.

See Appendix 1 for search terms used in strategies for this review.

Searching other resources

  1. Reference lists of included studies and previous relevant systematic reviews.

  2. Abstracts from major conferences and meetings for the past seven years between 2010 and 2017.

Data collection and analysis

Selection of studies

The search strategy was used to obtain titles and abstracts of studies that might be relevant to the review. Retrieved titles and abstracts were reviewed by two authors, and studies considered by any of the reviewers as potentially relevant were initially selected. Full text articles of these studies were further reviewed for eligibility by both authors, who had to both agree for including them in the review. Disagreements not resolved by discussion between authors were referred to a third author.

Data extraction and management

Studies reported in languages other than English or Spanish were translated before assessment. When more than one publication of the same study was found, reports were analysed to select the publication with the most complete data to be included in the analyses. Only when relevant outcomes were published in an earlier version, this version was used.

Two authors carried out data extraction, independently using a data extraction form, devised to record details of participant characteristics, interventions and outcome measures for each included study.

Assessment of risk of bias in included studies

Included studies were independently assessed for methodological quality by two authors. We assessed the following items using the risk of bias assessment tool (Higgins 2011) (Appendix 2).

  • Was there adequate sequence generation (selection bias)?

  • Was allocation adequately concealed (selection bias)?

  • Was knowledge of the allocated interventions adequately prevented during the study?

    • Participants and personnel (performance bias)

    • Outcome assessors (detection bias)

  • Were incomplete outcome data adequately addressed (attrition bias)?

  • Were reports of the study free of suggestion of selective outcome reporting (reporting bias)?

  • Was the study apparently free of other problems that could put it at a risk for bias?

Measures of treatment effect

Studies expressed the main outcomes either as rates of events per catheter‐days or per dialysis sessions, we used the generic inverse‐variance method to pool the results, reporting the rate ratio (RR) with 95% confidence intervals (95% CI) for each comparison (Higgins 2011).

Unit of analysis issues

We used catheter‐days or dialysis‐sessions for data reporting and analysis.

Dealing with missing data

Missing data was requested from the authors of included studies, by up to three e‐mails to the corresponding author; all relevant information obtained was included in the review.

Assessment of heterogeneity

We explored evidence of statistical heterogeneity across studies was explored using the I2 and the Chi2 test for heterogeneity, P‐value for statistical significance of chi‐square was set at 0.05. I2 of 0% to 25%, 26% to 50% and over 51%; corresponded to low, medium and high levels of heterogeneity respectively (Higgins 2011).

Assessment of reporting biases

Our search strategy aimed to minimize publication bias. We used funnel plots to assess publication bias for CVC‐related infection.

Data synthesis

We pooled data using random‐effects model.

Subgroup analysis and investigation of heterogeneity

Planned a priori subgroup analyses were used to explore possible sources of heterogeneity. Heterogeneity was explored according to the type of lock solutions and the type of catheters (tunnelled or non‐tunnelled).

Sensitivity analysis

Sensitivity analyses were performed on the main outcomes and comparisons excluding studies with high risk of bias and separating randomised from quasi‐randomised studies.

'Summary of findings' tables

We presented the main results of the review in 'Summary of findings' tables. These tables present key information concerning the quality of the evidence, the magnitude of the effects of the interventions examined, and the sum of the available data for the main outcomes (Schünemann 2011a). The 'Summary of findings' tables also include an overall grading of the evidence related to each of the main outcomes using the GRADE (Grades of Recommendation, Assessment, Development and Evaluation) approach (GRADE 2008). The GRADE approach defines the quality of a body of evidence as the extent to which one can be confident that an estimate of effect or association is close to the true quantity of specific interest. The quality of a body of evidence involves consideration of within‐trial risk of bias (methodological quality), directness of evidence, heterogeneity, precision of effect estimates and risk of publication bias (Schünemann 2011b)). We presented the outcomes 1) CVC‐related infections per 1000 catheter‐days, and 2) thrombosis per 1000 catheter‐days summary of findings Table for the main comparison.

Results

Description of studies

Results of the search

After searching the Register we identified 196 records. Titles and abstracts were screened and we retrieved 88 full‐text articles for further assessment. Of these, 39 studies (58 records) were included and 22 studies (29 records) were excluded. One study was completed prior to publication of this review (CLOCK 2017) and will be assessed in a future update (Figure 1).


Study selection flow diagram

Study selection flow diagram

Included studies

We included 39 studies (4216 participants) in this review (Al‐Hwiesh 2007; AZEPTIC 2011; Betjes 2004; Bleyer 2005; Buturovic 1998; Campos 2011; CHARTS 2008; CITRIM 2017; Corbett 2013; Cooper 1999; Davanipur 2011; D'Avella 2007; Dogra 2002; Geron 2008; Hendrickx 2001; Hermite 2012; Kanaa 2015; Kim 2006a; Kokenge 2010; Lange 2007; Lustig 2011; McIntyre 2004; Moghaddas 2015; Moran 2012; Mortazavi 2011; Nori 2006; Oguzhan 2012; Pervez 2002; Plamandon 2005a; Power 2009; Saxena 2006; Saxena 2012; Shirzad 2013; Sofroniadou 2012; Solomon 2010; Vercaigne 2016a; Weijmer 2005; Zhang 2009c; Zwiech 2016a).

Of these, nine studies either did not provide enough data, or the relevant data could not be extracted, and therefore were not meta‐analysed (Corbett 2013; Kanaa 2015; Kokenge 2010; Lange 2007; Lustig 2011; Plamandon 2005a; Power 2009; Shirzad 2013; Zwiech 2016a).

Study design

Thirty‐seven studies were parallel RCTs and two were quasi‐randomised (CHARTS 2008; Power 2009).

Setting

Twenty‐five studies were undertaken in an outpatient setting, eight in an inpatient setting (Al‐Hwiesh 2007; Betjes 2004; Bleyer 2005; CITRIM 2017; Hermite 2012; Kim 2006a; Moghaddas 2015; Sofroniadou 2012), and six did not describe a specific setting (D'Avella 2007; Geron 2008; Kokenge 2010; Lange 2007; Lustig 2011; Plamandon 2005a).

Participants

All studies included adults with ESKD undergoing HD through a CVC; none of the studies included children. Twenty‐four studies used tunnelled catheters (Al‐Hwiesh 2007; AZEPTIC 2011; CHARTS 2008; Cooper 1999; Corbett 2013; Dogra 2002; Geron 2008; Hendrickx 2001; Kanaa 2015; Kokenge 2010; McIntyre 2004; Moghaddas 2015; Mortazavi 2011; Moran 2012; Nori 2006; Oguzhan 2012; Pervez 2002; Plamandon 2005a; Power 2009; Saxena 2006; Saxena 2012; Solomon 2010; Vercaigne 2016a; Zhang 2009c), six studies used non‐tunnelled catheters (Buturovic 1998; CITRIM 2017; Davanipur 2011; Hermite 2012; Kim 2006a; Sofroniadou 2012), six studies used both types of catheters (Betjes 2004; Bleyer 2005; Campos 2011; Lange 2007; Weijmer 2005; Zwiech 2016a), and three studies did not describe the type of catheter (D'Avella 2007; Lustig 2011; Shirzad 2013).

Interventions
Antibiotic lock solutions versus heparin

Fifteen studies used antibiotic lock solutions.

Non‐antibiotic antimicrobial lock solutions versus heparin and other solutions

Twenty‐one studies used non‐antibiotic antimicrobial lock solutions. Twenty studies compared non‐antibiotic antimicrobial lock solutions versus heparin.

One study compared 46.7% citrate solution to 0.9% saline solution (Hermite 2012).

Antibiotic lock solutions plus non‐antibiotic antimicrobial lock solutions versus heparin

Four studies used antibiotic lock solutions plus non‐antibiotic antimicrobial lock solutions.

  • Gentamicin (320 µg/mL) in sodium citrate (4%) (Moran 2012)

  • Gentamicin (40 mg/mL) and 1 mL citrate (3.13%) (Dogra 2002)

  • Gentamicin (40 mg/mL) plus tricitrasol (46.7%) (Pervez 2002)

  • Gentamicin (4 mg/mL) in citrate (3.13%) (Nori 2006).

Outcomes

All included studies reported at least one of our outcomes. Seventeen studies reported only CRI (Al‐Hwiesh 2007; CHARTS 2008; Cooper 1999; Davanipur 2011; D'Avella 2007; Dogra 2002; Geron 2008; Hendrickx 2001; Kim 2006a; Lustig 2011; McIntyre 2004; Mortazavi 2011; Saxena 2012; Shirzad 2013; Sofroniadou 2012; Vercaigne 2016a; Zwiech 2016a); four studies reported only thrombosis (Buturovic 1998; Kokenge 2010; Lange 2007; Plamandon 2005a); 15 studies reported CRI and thrombosis (AZEPTIC 2011; Campos 2011; CITRIM 2017; Corbett 2013; Hermite 2012; Moghaddas 2015; Moran 2012; Nori 2006; Oguzhan 2012; Pervez 2002; Power 2009; Saxena 2006; Solomon 2010; Weijmer 2005; Zhang 2009c); two studies reported CRI and colonisation (Betjes 2004; Bleyer 2005); and one study reported CRI, thrombosis, and colonisation (Kanaa 2015).

The 39 studies included were further divided into three groups according to the type of antimicrobial lock solutions: antibiotic, non‐antibiotic, and the combination of both. One study with three arms (gentamicin/citrate, minocycline/EDTA and heparin), was included in the two comparisons (antibiotic lock solutions plus non‐antibiotic antimicrobial lock solutions and antibiotic lock solutions) (Nori 2006).

See Characteristics of included studies.

Excluded studies

We excluded 22 studies after reviewing the full‐text reports (see Characteristics of excluded studies). Four studies had different populations (Beigi 2010; Khosroshahi 2006b; Khosroshahi 2015; Onder 2008); 13 studies used different interventions (Chen 2014b; Chu 2016, Coli 2010; HEALTHY‐CATH 2009; Hryszko 2013; Hu 2011; Malo 2010; Mohammad 2016; Oran 2008; PreCLOT 2006; Ray 1999; Sishir 2014; Thomson 2011); one study used a different comparison (Meeus 2005); one study reported different outcomes (Bosma 2010); one study was terminated early (NCT01989091); one study has not been verified since 2009 and no results have been published (NCT00862966); and one study protocol could not be located (ISRCTN27307877).

Risk of bias in included studies

Risk of bias was low or unclear for most domains in the majority of studies. Most authors were contacted for additional or missing information regarding their included studies; however, we only obtained response from one author (Hermite 2012).

Allocation

Sequence generation

Two studies were considered at high risk of bias due to the method used for sequence generation (CHARTS 2008; Power 2009). Seventeen studies reported an appropriate method of sequence generation and were judged to be at low risk of bias (AZEPTIC 2011; Betjes 2004; CITRIM 2017; Dogra 2002; Kanaa 2015; Kim 2006a; Moran 2012; Mortazavi 2011; Oguzhan 2012; Pervez 2002; Saxena 2006; Saxena 2012; Sofroniadou 2012; Solomon 2010; Vercaigne 2016a; Weijmer 2005; Zwiech 2016a). Twenty studies reported insufficient information to estimate this risk of bias (Al‐Hwiesh 2007; Bleyer 2005; Buturovic 1998; Campos 2011; Cooper 1999; Corbett 2013; D'Avella 2007; Davanipur 2011; Geron 2008; Hendrickx 2001; Hermite 2012; Kokenge 2010; Lange 2007; Lustig 2011; McIntyre 2004; Moghaddas 2015; Nori 2006; Plamandon 2005a; Shirzad 2013; Zhang 2009c).

Allocation concealment

Seven studies reported adequate allocation concealment, representing a low risk of bias (Betjes 2004; CITRIM 2017; Kanaa 2015; Saxena 2006; Saxena 2012; Vercaigne 2016a; Weijmer 2005). One study was at high risk of bias as they did not use allocation concealment (CHARTS 2008). The remaining 31 studies had insufficient details of the method of allocation concealment and were judged to be at unclear risk of bias (Al‐Hwiesh 2007; AZEPTIC 2011; Betjes 2004; Buturovic 1998; Campos 2011; Cooper 1999; Corbett 2013; D'Avella 2007; Davanipur 2011; Dogra 2002; Geron 2008; Hendrickx 2001; Hermite 2012; Kim 2006a; Kokenge 2010; Lange 2007; Lustig 2011; McIntyre 2004; Moghaddas 2015; Moran 2012; Mortazavi 2011; Nori 2006; Oguzhan 2012; Pervez 2002; Plamandon 2005a; Power 2009; Shirzad 2013; Sofroniadou 2012; Solomon 2010; Zhang 2009c; Zwiech 2016a).

Blinding

Participants and personnel

Nine studies were at low risk of bias (Bleyer 2005; CITRIM 2017; Dogra 2002; Hermite 2012; Saxena 2006; Saxena 2012; Solomon 2010; Weijmer 2005; Zhang 2009c). Seven studies showed a high risk of bias because either study participants or personnel were unblinded (Al‐Hwiesh 2007; AZEPTIC 2011; CHARTS 2008; Kanaa 2015; McIntyre 2004; Moghaddas 2015; Nori 2006). In the remaining 23 studies insufficient information was provided to determine who were blind and therefore were judged unclear risk of bias (Betjes 2004; Buturovic 1998; Campos 2011; Cooper 1999; Corbett 2013; D'Avella 2007; Davanipur 2011; Geron 2008; Hendrickx 2001; Kim 2006a; Kokenge 2010; Lange 2007; Lustig 2011; Moran 2012; Mortazavi 2011; Oguzhan 2012; Pervez 2002; Plamandon 2005a; Power 2009; Shirzad 2013; Sofroniadou 2012; Vercaigne 2016a; Zwiech 2016a).

Outcome assessment

In seven studies the outcome assessment was blinded (AZEPTIC 2011; Dogra 2002; Hermite 2012; Kanaa 2015; Saxena 2006; Saxena 2012; Solomon 2010). In four studies the outcome assessment was not blinded and was judged to be at high risk of bias (CHARTS 2008; Kim 2006a; Moghaddas 2015; Zhang 2009c). Twenty‐eight studies did not provide enough information to be judged, therefore this risk of bias was uncertain (Al‐Hwiesh 2007; Betjes 2004; Bleyer 2005; Buturovic 1998; Campos 2011; CITRIM 2017; Cooper 1999; Corbett 2013; D'Avella 2007; Davanipur 2011; Geron 2008; Hendrickx 2001; Kokenge 2010; Lange 2007; Lustig 2011; McIntyre 2004; Moran 2012; Mortazavi 2011; Nori 2006; Oguzhan 2012; Pervez 2002; Plamandon 2005a; Power 2009; Shirzad 2013; Sofroniadou 2012; Vercaigne 2016a; Weijmer 2005; Zwiech 2016a).

Incomplete outcome data

Twenty‐two studies had a low risk of bias, reporting complete data charts (AZEPTIC 2011; Betjes 2004; Campos 2011; CHARTS 2008; Dogra 2002; Hendrickx 2001; Kim 2006a; McIntyre 2004; Moghaddas 2015; Moran 2012; Mortazavi 2011; Nori 2006; Oguzhan 2012; Pervez 2002; Saxena 2006; Saxena 2012; Sofroniadou 2012; Solomon 2010; Vercaigne 2016a; Weijmer 2005; Zhang 2009c; Zwiech 2016a). One study was rated high risk of bias due to incomplete outcome data (Bleyer 2005). Sixteen studies did not provide sufficient information to determine whether incomplete outcome data were adequately addressed (Al‐Hwiesh 2007; Buturovic 1998; CITRIM 2017; Cooper 1999; Corbett 2013; D'Avella 2007; Davanipur 2011; Geron 2008; Hermite 2012; Kanaa 2015; Kokenge 2010; Lange 2007; Lustig 2011; Plamandon 2005a; Power 2009; Shirzad 2013).

Selective reporting

Twenty‐eight studies were judged to be at low risk of reporting bias (Al‐Hwiesh 2007; AZEPTIC 2011; Betjes 2004; Bleyer 2005; Campos 2011; CHARTS 2008; CITRIM 2017; Davanipur 2011; Dogra 2002; Hendrickx 2001; Hermite 2012; Kanaa 2015; Kim 2006a; McIntyre 2004; Moghaddas 2015; Moran 2012; Mortazavi 2011; Nori 2006; Oguzhan 2012; Pervez 2002; Saxena 2006; Saxena 2012; Sofroniadou 2012; Solomon 2010; Vercaigne 2016a; Weijmer 2005; Zhang 2009c; Zwiech 2016a). One study was judged to be at high risk of bias (Power 2009) and 10 studies were judged to have unclear risk of bias (Buturovic 1998; Cooper 1999; Corbett 2013; D'Avella 2007; Geron 2008; Kokenge 2010; Lange 2007; Lustig 2011; Plamandon 2005a; Shirzad 2013).

Other potential sources of bias

One study was judged as high risk of bias because researchers declared conflict of interest related to research funding (Moran 2012). Three studies were judged as low risk of bias (Al‐Hwiesh 2007; AZEPTIC 2011; Hermite 2012), and the remaining 35 studies reported insufficient information and they were judged to have unclear risk of bias (Betjes 2004; Bleyer 2005; Buturovic 1998; Campos 2011; CHARTS 2008; CITRIM 2017; Cooper 1999; Corbett 2013; D'Avella 2007; Davanipur 2011; Dogra 2002; Geron 2008; Hendrickx 2001; Kanaa 2015; Kim 2006a; Kokenge 2010; Lange 2007; Lustig 2011; McIntyre 2004; Moghaddas 2015; Mortazavi 2011; Nori 2006; Oguzhan 2012; Pervez 2002; Plamandon 2005a; Power 2009; Saxena 2006; Saxena 2012; Shirzad 2013; Sofroniadou 2012; Solomon 2010; Vercaigne 2016a; Weijmer 2005; Zhang 2009c; Zwiech 2016a).

Effects of interventions

See: Summary of findings for the main comparison Antimicrobial lock solutions vs control for preventing catheter‐related infections in patient undergoing haemodialysis

See summary of findings Table for the main comparison.

Although 39 studies met the inclusion criteria, only 30 could be incorporated into the meta‐analyses, as the other nine studies did not present enough data to be included in the meta‐analysed (Corbett 2013; Kanaa 2015; Kokenge 2010; Lange 2007; Lustig 2011; Plamandon 2005a; Power 2009; Shirzad 2013; Zwiech 2016a) so all further descriptions refer to the 30 meta‐analysed studies.

Antimicrobial lock solutions (antibiotic, non‐antibiotic or both) versus control

Catheter‐related infection

Twenty seven studies reported the incidence of CRI per 1000 catheter‐days (AZEPTIC 2011; Betjes 2004; Bleyer 2005; Campos 2011; CHARTS 2008; CITRIM 2017; Cooper 1999; D'Avella 2007; Davanipur 2011; Dogra 2002; Geron 2008; Hermite 2012; Kim 2006a; McIntyre 2004; Moghaddas 2015; Moran 2012; Mortazavi 2011; Nori 2006; Oguzhan 2012; Pervez 2002; Saxena 2006; Saxena 2012; Sofroniadou 2012; Solomon 2010; Vercaigne 2016a; Weijmer 2005; Zhang 2009c). Antimicrobial lock solutions probably reduces the incidence of CRI per 1000 catheter days compared to control (Analysis 1.1: RR 0.38, 95% CI 0.27 to 0.53; I2 = 54%; moderate certainty evidence).

Two studies measured the incidence of CRI per 1000 dialysis sessions (Al‐Hwiesh 2007; Hendrickx 2001). Antimicrobial lock solutions may make little or no difference to the reduction of CRI compared with heparin lock solutions (Analysis 1.2: RR 0.47, 95% CI 0.07 to 3.13; I2 = 81%; low certainty evidence). Heterogeneity was high for this analysis.

Thrombosis

Fourteen studies reported the incidence of thrombosis, per 1000 catheter‐days (AZEPTIC 2011; Buturovic 1998; Campos 2011; CITRIM 2017; Hermite 2012; Moghaddas 2015; Moran 2012; Nori 2006; Oguzhan 2012; Pervez 2002; Saxena 2006; Solomon 2010; Weijmer 2005; Zhang 2009c). Antimicrobial lock solutions probably make little or no difference to the risk of thrombosis compared to control (Analysis 1.3: RR 0.79, 95% CI 0.52 to 1.22; I2 = 83%; moderate certainty evidence). Heterogeneity was high for this analysis.

Colonisation

Two studies reported the incidence of colonisation (Betjes 2004; Bleyer 2005). Antimicrobial lock solutions may make little or no difference to the reduction in colonisation compared to control (Analysis 1.4: RR 0.37 95% CI 0.04 to 3.36; I2 = 70%; low certainty evidence), with high heterogeneity.

Subgroup analyses

To explore sources of heterogeneity, subgroup analysis by CVC type was carried out for CRI and lumen thrombosis.

Eighteen studies used only tunnelled catheters (AZEPTIC 2011; CHARTS 2008; Cooper 1999; D'Avella 2007; Dogra 2002; Geron 2008; McIntyre 2004; Moghaddas 2015; Moran 2012; Mortazavi 2011; Nori 2006; Oguzhan 2012; Pervez 2002; Saxena 2006; Saxena 2012; Solomon 2010; Vercaigne 2016a; Zhang 2009c) and five used only non‐tunnelled catheters (CITRIM 2017; Davanipur 2011; Hermite 2012; Kim 2006a; Sofroniadou 2012). Four used both types of catheters (Betjes 2004; Bleyer 2005; Campos 2011; Weijmer 2005) however only three delivered data separating tunnelled and non‐tunnelled catheters (Betjes 2004; Campos 2011; Weijmer 2005). For the purpose of the analysis, the studies including both types were included separately in the tunnelled or non‐tunnelled group.

Due to insufficient data on included studies, it was not possible to explore other sources of heterogeneity, such as subgroups of diabetic participants or co‐interventions to prevent CRI.

Catheter‐related infection (tunnelled and non‐tunnelled catheters)

Twenty‐one studies reported data for tunnelled catheters (AZEPTIC 2011; CHARTS 2008; Cooper 1999; D'Avella 2007; Geron 2008; McIntyre 2004; Moghaddas 2015; Moran 2012; Mortazavi 2011; Nori 2006; Oguzhan 2012; Pervez 2002; Saxena 2006; Saxena 2012; Solomon 2010; Vercaigne 2016a; Zhang 2009c, Betjes 2004; Campos 2011; Dogra 2002; Weijmer 2005). Antimicrobial lock solutions probably reduce the incidence of CRI per 1000 catheter days compared to control in tunnelled catheters (Analysis 1.5: RR 0.40, 95% CI 0.30 to 0.53; I2 = 14%).

Eight studies reported data for non‐tunnelled catheters (Betjes 2004; Campos 2011; CITRIM 2017; Davanipur 2011; Hermite 2012; Kim 2006a; Sofroniadou 2012; Weijmer 2005). Antimicrobial lock solutions probably reduce the incidence of CRI per 1000 catheter days compared to control in non‐tunnelled catheters (Analysis 1.6: RR 0.37, 95% CI 0.16 to 0.86; I2 = 67%). Heterogeneity was high.

Heterogeneity disappears if only tunnelled catheters are analysed, therefore some heterogeneity in the main analysis might be explained on the basis of type of CVC.

Thrombosis (tunnelled and non‐tunnelled catheters)

Ten studies reported data for tunnelled catheters (AZEPTIC 2011; Moghaddas 2015; Moran 2012; Nori 2006; Oguzhan 2012; Pervez 2002; Saxena 2006; Solomon 2010; Weijmer 2005; Zhang 2009c). Antimicrobial lock solutions probably make little or no difference to the risk of thrombosis in tunnelled catheters (Analysis 1.7: RR 0.83, 95% CI 0.50 to 1.37; I2 = 76%). High heterogeneity was observed.

Four studies reported data for non‐tunnelled catheters (Buturovic 1998; CITRIM 2017; Hermite 2012; Weijmer 2005). Antimicrobial lock solutions probably make little or no difference to the risk of thrombosis in non‐tunnelled catheters (Analysis 1.8: RR 0.66, 95% CI 0.25 to 1.72; I2 = 92%). High heterogeneity was observed.

For this outcome, the type of catheter does not explain the observed heterogeneity.

Sensitivity analyses

We carried out a sensitivity analyses by excluding the studies deemed as to be at high risk of bias (for any domain), to determine whether the results of the meta‐analysis regarding the main outcomes were robust.

Catheter‐related infection

Eighteen studies with low or unclear risk of bias were analysed (Betjes 2004; Campos 2011; CITRIM 2017; Corbett 2013; D'Avella 2007; Davanipur 2011; Dogra 2002; Geron 2008; Hermite 2012; Mortazavi 2011; Oguzhan 2012; Pervez 2002; Saxena 2006; Saxena 2012; Sofroniadou 2012; Solomon 2010: Weijmer 2005). There was no change to the results; there was less precision and heterogeneity remained the same (Analysis 2.1: (RR 0.47, 95% CI 0.32 to 0.68; I2 = 58%).

Thrombosis

Nine studies with low or unclear risk of bias were analysed (Buturovic 1998; Campos 2011; CITRIM 2017; Hermite 2012; Oguzhan 2012; Pervez 2002; Saxena 2006; Solomon 2010; Weijmer 2005). There was no change to the results; either in direction or magnitude of effect (Analysis 2.2: (RR 0.85, 95% CI 0.48 to 1.52; I2 = 88%).

Antibiotic lock solutions versus control

Catheter‐related infection

Thirteen studies reported CRI per 1000 catheter‐days (Bleyer 2005; Campos 2011; Cooper 1999; Davanipur 2011; Kim 2006a; McIntyre 2004; Moghaddas 2015; Mortazavi 2011; Nori 2006; Saxena 2006; Saxena 2012; Sofroniadou 2012; Zhang 2009c). Antibiotic lock solutions probably reduce the incidence of CRI per 1000 catheter days compared to control (Analysis 3.1: RR 0.30, 95% CI 0.22 to 0.42; I2 = 18%; moderate certainty evidence). Heterogeneity was low.

One study (Al‐Hwiesh 2007) reported antibiotic lock solutions reduced the incidence in CRI per dialysis session compared to control (Analysis 3.2: RR 0.18, 95% CI 0.05 to 0.59).

Thrombosis

Five studies reported thrombosis per 1000 catheter‐days (Campos 2011; Moghaddas 2015; Nori 2006; Saxena 2006; Zhang 2009c). Antibiotic lock solutions probably make little or no difference to the risk of thrombosis per 1000 catheter‐days compared to control (Analysis 3.3; RR 0.76, 95% CI 0.42 to 1.38; I2 = 47%; moderate certainty evidence). Moderate heterogeneity was observed.

Colonisation

One study (Bleyer 2005) reported that antibiotic lock solutions reduced the incidence of catheter colonisation compared to heparin (Analysis 3.4: RR 0.10, 95% CI 0.01 to 0.79).

Subgroup analyses
Catheter‐related infection

Nine studies reported data for tunnelled catheters (Campos 2011; Cooper 1999; McIntyre 2004; Moghaddas 2015; Mortazavi 2011; Nori 2006; Saxena 2006; Saxena 2012; Zhang 2009c). Antibiotic lock solutions probably reduce the incidence of CRI per 1000 catheter‐days compared to control in tunnelled catheters (Analysis 3.5: RR 0.30, 95% CI 0.18 to 0.50; I2 = 35%). Moderate heterogeneity was observed.

Four studies reported data for non‐tunnelled catheters (Campos 2011; Davanipur 2011; Kim 2006a; Sofroniadou 2012). Antibiotic lock solutions probably reduce the incidence of CRI per 1000 catheter days compared to control in non‐tunnelled catheters (Analysis 3.6: RR 0.14, 95% CI 0.05 to 0.36; I2 = 0%).

Thrombosis

Three studies reported data for tunnelled catheters (Moghaddas 2015; Saxena 2006; Zhang 2009c). Antibiotic lock solutions probably make little or no difference to the risk of thrombosis per 1000 catheter‐days compared to control for tunnelled catheters (Analysis 3.7: RR 0.61, 95% CI 0.31 to 1.23; I2 = 45%).

No analysis was performed for non‐tunnelled catheters as no separate data was reported in the three studies that used both types of catheters.

Sensitivity analyses

Sensitivity analyses excluding studies deemed as to be at high risk of bias (for any domain) were undertaken.

Catheter‐related infection

Seven studies with low or unclear risk of bias were included in this analysis (Campos 2011; Cooper 1999; Davanipur 2011; Mortazavi 2011; Saxena 2006; Saxena 2012; Sofroniadou 2012). There was no change to the results with less heterogeneity and higher precision (Analysis 4.1: RR 0.37, 95% CI 0.28 to 0.48; I2 = 4%).

Thrombosis

Two studies with low or unclear risk of bias were included in this analysis (Campos 2011; Saxena 2006). There was no change to results, however heterogeneity was higher (Analysis 4.2: RR 0.79, 95% CI 0.24 to 2.63; I2 = 82%).

Non‐antibiotic antimicrobial lock solutions versus control

Catheter‐related infection

Eleven studies reported CRI per 1000 catheter‐days (AZEPTIC 2011; Betjes 2004; CHARTS 2008; CITRIM 2017; D'Avella 2007; Geron 2008; Hermite 2012; Oguzhan 2012; Solomon 2010; Vercaigne 2016a; Weijmer 2005). Non‐antibiotic antimicrobial lock solutions probably make little or no difference to the incidence of CRI per 1000 catheter days compared to control (Analysis 5.1: RR 0.65, 95% CI 0.41 to 1.05; I2 = 51%). Heterogeneity was moderate.

One study (Hendrickx 2001) reported no difference in the incidence of CRI per 1000 dialysis sessions between non‐antibiotic antimicrobial lock solutions and control (Analysis 5.2: RR 1.22, 95% CI 0.38 to 3.88).

Thrombosis

Eight studies reported thrombosis per 1000 catheter‐days (AZEPTIC 2011; Buturovic 1998; CHARTS 2008; CITRIM 2017; Hermite 2012; Oguzhan 2012; Solomon 2010; Weijmer 2005). Seven studies used heparin as control and one study (Hermite 2012) used saline solution 0.9% as control lock solution. Non‐antibiotic antimicrobial lock solutions probably make little or no difference to the risk of thrombosis per 1000 catheter‐days compared to control (Analysis 5.3: RR 0.85, 95% CI 0.44 to 1.66; I2 = 89%). Heterogeneity is high.

One study (Hendrickx 2001) reported a decrease in the incidence of thrombosis per 1000 dialysis sessions with non‐antibiotic antimicrobial lock solutions versus heparin (Analysis 5.4 RR 0.11, 95% CI 0.04 to 0.32).

Colonisation

One study (Betjes 2004) reported no difference in the reduction of catheter colonisation between non‐antibiotic antimicrobial lock solutions and control (Analysis 5.5: RR 0.99, 95%CI 0.27 to 3.68).

Subgroup analyses

Catheter‐related infection

Nine studies reported data for tunnelled catheters (AZEPTIC 2011; Betjes 2004; CHARTS 2008; D'Avella 2007; Geron 2008; Weijmer 2005; Oguzhan 2012; Solomon 2010; Vercaigne 2016a). Non‐antibiotic antimicrobial lock solutions probably reduce the incidence of CRI per 1000 catheter‐days compared to control in tunnelled catheters (Analysis 5.6: RR 0.60, 95% CI 0.40 to 0.91; I2 = 0%).

Four studies reported data for non‐tunnelled catheters (Betjes 2004; CITRIM 2017; Hermite 2012; Weijmer 2005). Non‐antibiotic antimicrobial lock solutions probably make little or no difference to the risk of CRI per 1000 catheter‐days compared to control for tunnelled catheters (Analysis 5.7: RR 0.93, 95% CI 0.48 to 1.81; I2 = 41%). Heterogeneity is moderate.

Thrombosis

Five studies reported data for tunnelled catheters (AZEPTIC 2011; CHARTS 2008; Oguzhan 2012; Solomon 2010; Weijmer 2005). Non‐antibiotic antimicrobial lock solutions probably make little or no difference to the risk of thrombosis per 1000 catheter‐days compared to control for tunnelled catheters (Analysis 5.8: RR 1.17, 95% CI 0.57 to 2.41; I2 = 72%). Heterogeneity was high.

Four studies reported data for non‐tunnelled catheters (Buturovic 1998; CITRIM 2017;Hermite 2012; Weijmer 2005) Non‐antibiotic antimicrobial lock solutions probably make little or no difference to the risk of thrombosis per 1000 catheter‐days compared to control for non‐tunnelled catheters (Analysis 5.9: RR 0.66, 95% CI 0.25 to 1.72; I2 = 92%). Heterogeneity was high.

Sensitivity analyses

Sensitivity analyses excluding studies deemed as to be at high risk of bias (for any domain) were undertaken.

Catheter‐related infection

Nine studies with low or unclear risk of bias were included in this analysis (Betjes 2004; CITRIM 2017; Geron 2008; Hermite 2012; Oguzhan 2012; Solomon 2010; Vercaigne 2016a; Weijmer 2005). The analysis showed a similar magnitude of effect but with less precision and higher heterogeneity (Analysis 6.1: RR 0.65, 95% CI 0.38 to 1.12; I2 = 61%).

Thrombosis

Six studies with low or unclear risk of bias were included in this analysis (Buturovic 1998; CITRIM 2017; Hermite 2012; Oguzhan 2012; Solomon 2010; Weijmer 2005). The analysis showed a similar magnitude of effect but with less precision and same heterogeneity (Analysis 6.2: RR 0.90, 95% CI 0.43 to 1.91; I2 = 92%).

Combined antibiotic plus non‐antibiotic antimicrobial lock solutions versus control

Catheter‐related infection

Four studies reported CRI per 1000 catheter‐days (Dogra 2002; Moran 2012; Nori 2006; Pervez 2002). Antibiotic plus non‐antibiotic antimicrobial lock solution probably reduce the incidence of CRI per 1000 catheter‐days compared to control (Analysis 7.1: RR 0.26, 95% CI 0.14 to 0.49; I2 = 0%; moderate certainty evidence).

Thrombosis

Three studies reported thrombosis per 1000 catheter‐days (Moran 2012; Nori 2006; Pervez 2002). Antibiotic plus non‐antibiotic antimicrobial lock solution probably make little or no difference to the incidence of thrombosis per 1000 catheter‐days compared to control (Analysis 7.2: RR 0.63, 95% CI 0.22 to 1.81; I2 = 0%; moderate certainty evidence).

Colonisation

No study reported catheter colonisation.

Subgroup analyses

All four studies used tunnelled catheters.

Sensitivity analyses

Sensitivity analyses excluding studies deemed as to be at high risk of bias (for any domain) were undertaken.

Catheter‐related infection

Two studies with low or unclear risk of bias were included in this analysis (Dogra 2002; Pervez 2002). The result no longer showed a reduction in CRI (Analysis 8.1: RR 0.16, 95% CI 0.02 to 1.03; I2 = 0%).

Thrombosis

One study (Pervez 2002) showed a similar magnitude of effect but with less precision (Analysis 8.2: RR 0.66, 95% CI 0.18 to 2.44).

Other planned outcomes

Survival of the catheter, infection and thrombosis‐free days, were reported in a small number of included studies, however the data was not presented in enough detail to be meta‐analysed.

Mortality and adverse effects were not reported.

Discussion

Summary of main results

Antimicrobial lock solutions reduced CRI in patients using CVC for HD, compared to standard lock solutions, usually heparin (moderate certainty evidence). This beneficial effect was best for antibiotic lock solutions and for combined (antibiotic plus non‐antibiotic antimicrobial) lock solutions. Heterogeneity was low for antibiotic lock solutions and for combined lock solutions (I2 = 18% and 0% respectively) but high for all antimicrobial lock solutions and non‐antibiotic antimicrobial lock solutions (I2 = 54% and 51% respectively). Therefore the type of antimicrobial lock solution might explain some of the heterogeneity in the main analysis. Type of catheter (tunnelled vs non‐ tunnelled) might also influence the effect of antimicrobial lock solutions. For both tunnelled (RR 0.40; 95% CI 0.30 to 0.53; I2 = 14%) and non‐tunnelled catheters, antimicrobial lock solutions (RR 0.37, 95% CI 0.16 to 0.86; I2 = 67%) reduced CRI.

The difference between both types of catheters could be related to the pathogenesis of infection. In the case of non‐tunnelled catheters the infection could be developed through the either external or the internal surface of the catheter. By contrast, the route for the development of infection of tunnelled catheters corresponds to the intraluminal surface, which is in contact with the antimicrobial lock solutions. However the type of catheter alone, does not explain heterogeneity, probably due to differences in characteristics of patients using one or the other type of catheter. Due to the quality of the evidence, our confidence in the effect of antimicrobial lock solutions on CRI is moderate.

The efficacy of antimicrobial lock solutions in reducing the incidence of CRI is still present in the sensitivity analysis, using the studies classified as having low or unclear risk of bias, however, even though these results are robust, our confidence in the effect on CRI is moderate due to the quality of the evidence.

Regarding the safety of antimicrobial lock solutions, they are no worse than heparin or other lock solutions on the incidence of thrombosis of CVC for HD. These results remain unaltered when analysed according to type of antimicrobial lock solutions, type of CVC, and quality of studies.

It is important to note that there is less evidence for thrombosis, due to a small number of studies reporting this outcome; therefore, these estimates are imprecise. These results are also heterogeneous, which might be due to a study that used saline solution instead of heparin as control lock solution or due to different definitions of thrombosis throughout the studies. Due to the quality of the evidence, our confidence in the effect of antimicrobial lock solutions on lumen thrombosis is low.

It is not possible to draw conclusions regarding CVC colonisation because this outcome was reported in only two studies that used different antimicrobial lock solutions. Survival of the catheter, infection‐ and thrombosis‐free periods, mortality, and adverse effects were not reported.

Overall completeness and applicability of evidence

We included thirty studies in the analyses with a total of 3392 participants. Studies were conducted between 1998 and 2017 in adult participants undergoing HD through a CVC, mainly in an ambulatory setting. The studies compared antimicrobial with other control lock solutions, mainly heparin, in order to prevent CRI without increasing thrombosis.

Our review considered all types of antimicrobial lock solutions together and subsets of antibiotic, non‐antibiotic and combination of both solutions. Other included outcomes were thrombosis and colonisation. Most studies rarely reported other adverse effects, follow‐up were short or not reported so we do not have information regarding long term effects, such as antibiotic resistance.

Our review presents some limitations to its applicability, mainly because adverse effects other than thrombosis were not assessed in studies for all antimicrobial lock solutions.

It is important to remember that some literature shows that sodium citrate is associated with some adverse effects such as hypocalcaemia, ventricular arrhythmias, and few cases of sudden death (Aguinaga 2011), none of which were reported in our included studies. Additionally, antibiotic solutions could cause microbial resistance, especially when used for long periods(Korkor 2009). Antibiotic antimicrobial lock solutions were quite different among studies, so we were unable to determine if any of them was better than the others.

Quality of the evidence

Thirty nine studies were selected for this review including 4216 participants. Thirty‐seven were RCTs and two quasi‐randomised, 17 studies (44%) presented an appropriate sequence of randomizations whereas seven studies (18%) had low risk of bias for allocation concealment. Blinding method was reported in nine studies for participants and health personnel (23%), and seven studies for outcome assessment (18%). Complete data was reported in 22 studies (56%). Reporting bias was low in 28 studies (72%). It was not possible to determine others type of bias in 35 studies (90%), mainly because conflict of interest or funding was not declared (Figure 2; Figure 3)


Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.


Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Publication bias is likely to occur in all comparisons based on the funnel plots (Figure 4; Figure 5; Figure 6).


Funnel plot of comparison: 1 Antimicrobial (Antibiotic plus non‐antibiotic) solutions versus control, outcome: 1.1 CVC‐related infection (per 1000 catheter‐days).

Funnel plot of comparison: 1 Antimicrobial (Antibiotic plus non‐antibiotic) solutions versus control, outcome: 1.1 CVC‐related infection (per 1000 catheter‐days).


Funnel plot of comparison: 1 Antimicrobial (antibiotic plus non‐antibiotic) lock solutions versus control, outcome: 1.3 Thrombosis (per 1000 catheter‐days).

Funnel plot of comparison: 1 Antimicrobial (antibiotic plus non‐antibiotic) lock solutions versus control, outcome: 1.3 Thrombosis (per 1000 catheter‐days).


Funnel plot of comparison: 3 Antibiotic lock solutions versus control, outcome: 3.1 Catheter‐related infection (per 1000 catheter‐days).

Funnel plot of comparison: 3 Antibiotic lock solutions versus control, outcome: 3.1 Catheter‐related infection (per 1000 catheter‐days).

Potential biases in the review process

A comprehensive search of the Cochrane Kidney and Transplant's Specialised Register was undertaken, thereby decreasing the likelihood of overlooking published studies. However, some studies may have been missed in the process, or may have been presented in some conferences not included in our search. Data was not complete for many studies; we contacted the authors via e‐mail regarding any missing information, however only one responded.

This review was performed independently by two authors in the stages of selection of articles and two others in the extraction of data. A third author participated in both processes when no agreement was reached through consensus. Sensitivity analysis was performed on the main results. As several studies contained insufficient information to make a judgment of quality, we contacted the authors; unfortunately, with no response from them.

Agreements and disagreements with other studies or reviews

Yahav 2008 published a systematic review evaluating antimicrobial lock solutions for the prevention of HD CRI. This review included 16 studies using antibiotic lock solutions or non‐antibiotic lock solutions. We included in our review 9 of the 11 comparing antibiotic lock solutions and 4 of the 5 comparing non‐antibiotic lock solutions. The studies not included in our review did not meet of our inclusion criteria (i.e. inappropriate design or insufficient data reported). Snaterse 2010 published a review analysing the use of catheter locking with antimicrobial, antibiotic and non‐antibiotic solutions for preventing the infection of CVC used for an extended period of time, including multiple indications of prolonged use. All studies that included participants who used catheter for HD were included in the review.

Zhao 2014 evaluated whether citrate is superior to heparin in the permeability of HD CVC and the prevention of infections. Eleven of the 12 studies included in the review are considered in the our systematic review. One study was not included because it was not randomised.

Grudzinski 2015 published a review whose objective was to evaluate the benefits and risks of catheter locks for HD catheters with sodium citrate. The five included studies are included in our review; however, it is not possible to compare these with our review because their aims are the benefits and harms of using citrate versus heparin as a CVC locking solution.

The results of our review are similar regarding the effect of antibiotic lock solutions to those reported by Yahav 2008 and Snaterse 2010, showing that antibiotic lock solutions for HD CVC reduce the incidence of CRI per 1000 catheter‐days. In addition, Zhao 2014 found that citrate in combination with other antimicrobials (non‐antibiotic antimicrobial lock solutions and antibiotic lock solutions) compared with heparin were associated with a lower incidence of CRI; however citrate alone was not shown to be better than heparin. Our review confirmed these findings.

Study selection flow diagram
Figuras y tablas -
Figure 1

Study selection flow diagram

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Figuras y tablas -
Figure 2

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Figuras y tablas -
Figure 3

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Funnel plot of comparison: 1 Antimicrobial (Antibiotic plus non‐antibiotic) solutions versus control, outcome: 1.1 CVC‐related infection (per 1000 catheter‐days).
Figuras y tablas -
Figure 4

Funnel plot of comparison: 1 Antimicrobial (Antibiotic plus non‐antibiotic) solutions versus control, outcome: 1.1 CVC‐related infection (per 1000 catheter‐days).

Funnel plot of comparison: 1 Antimicrobial (antibiotic plus non‐antibiotic) lock solutions versus control, outcome: 1.3 Thrombosis (per 1000 catheter‐days).
Figuras y tablas -
Figure 5

Funnel plot of comparison: 1 Antimicrobial (antibiotic plus non‐antibiotic) lock solutions versus control, outcome: 1.3 Thrombosis (per 1000 catheter‐days).

Funnel plot of comparison: 3 Antibiotic lock solutions versus control, outcome: 3.1 Catheter‐related infection (per 1000 catheter‐days).
Figuras y tablas -
Figure 6

Funnel plot of comparison: 3 Antibiotic lock solutions versus control, outcome: 3.1 Catheter‐related infection (per 1000 catheter‐days).

Comparison 1 All antimicrobial (antibiotic plus non‐antibiotic plus the combination) solutions versus control, Outcome 1 Catheter‐related infection (per 1000 catheter‐days).
Figuras y tablas -
Analysis 1.1

Comparison 1 All antimicrobial (antibiotic plus non‐antibiotic plus the combination) solutions versus control, Outcome 1 Catheter‐related infection (per 1000 catheter‐days).

Comparison 1 All antimicrobial (antibiotic plus non‐antibiotic plus the combination) solutions versus control, Outcome 2 Catheter‐related infection (per 1000 dialysis sessions).
Figuras y tablas -
Analysis 1.2

Comparison 1 All antimicrobial (antibiotic plus non‐antibiotic plus the combination) solutions versus control, Outcome 2 Catheter‐related infection (per 1000 dialysis sessions).

Comparison 1 All antimicrobial (antibiotic plus non‐antibiotic plus the combination) solutions versus control, Outcome 3 Thrombosis (per 1000 catheter‐days).
Figuras y tablas -
Analysis 1.3

Comparison 1 All antimicrobial (antibiotic plus non‐antibiotic plus the combination) solutions versus control, Outcome 3 Thrombosis (per 1000 catheter‐days).

Comparison 1 All antimicrobial (antibiotic plus non‐antibiotic plus the combination) solutions versus control, Outcome 4 Colonisation.
Figuras y tablas -
Analysis 1.4

Comparison 1 All antimicrobial (antibiotic plus non‐antibiotic plus the combination) solutions versus control, Outcome 4 Colonisation.

Comparison 1 All antimicrobial (antibiotic plus non‐antibiotic plus the combination) solutions versus control, Outcome 5 Subgroup analysis: catheter‐related infection (per 1000 catheter‐days) in tunnelled catheters.
Figuras y tablas -
Analysis 1.5

Comparison 1 All antimicrobial (antibiotic plus non‐antibiotic plus the combination) solutions versus control, Outcome 5 Subgroup analysis: catheter‐related infection (per 1000 catheter‐days) in tunnelled catheters.

Comparison 1 All antimicrobial (antibiotic plus non‐antibiotic plus the combination) solutions versus control, Outcome 6 Subgroup analysis: catheter‐related infection (per 1000 catheter‐days) in non‐tunnelled catheters.
Figuras y tablas -
Analysis 1.6

Comparison 1 All antimicrobial (antibiotic plus non‐antibiotic plus the combination) solutions versus control, Outcome 6 Subgroup analysis: catheter‐related infection (per 1000 catheter‐days) in non‐tunnelled catheters.

Comparison 1 All antimicrobial (antibiotic plus non‐antibiotic plus the combination) solutions versus control, Outcome 7 Subgroup analysis: thrombosis (per 1000 catheter‐days) in tunnelled catheters.
Figuras y tablas -
Analysis 1.7

Comparison 1 All antimicrobial (antibiotic plus non‐antibiotic plus the combination) solutions versus control, Outcome 7 Subgroup analysis: thrombosis (per 1000 catheter‐days) in tunnelled catheters.

Comparison 1 All antimicrobial (antibiotic plus non‐antibiotic plus the combination) solutions versus control, Outcome 8 Subgroup analysis: thrombosis (per 1000 catheter‐days) in non‐tunnelled catheters.
Figuras y tablas -
Analysis 1.8

Comparison 1 All antimicrobial (antibiotic plus non‐antibiotic plus the combination) solutions versus control, Outcome 8 Subgroup analysis: thrombosis (per 1000 catheter‐days) in non‐tunnelled catheters.

Comparison 2 Sensitivity analysis: all antimicrobial lock solutions versus control excluding studies deemed with high‐risk bias, Outcome 1 Catheter‐related infection (per 1000 catheter‐days).
Figuras y tablas -
Analysis 2.1

Comparison 2 Sensitivity analysis: all antimicrobial lock solutions versus control excluding studies deemed with high‐risk bias, Outcome 1 Catheter‐related infection (per 1000 catheter‐days).

Comparison 2 Sensitivity analysis: all antimicrobial lock solutions versus control excluding studies deemed with high‐risk bias, Outcome 2 Thrombosis (per 1000 catheter‐days).
Figuras y tablas -
Analysis 2.2

Comparison 2 Sensitivity analysis: all antimicrobial lock solutions versus control excluding studies deemed with high‐risk bias, Outcome 2 Thrombosis (per 1000 catheter‐days).

Comparison 3 Antibiotic lock solutions versus control, Outcome 1 Catheter‐related infection (per 1000 catheter‐days).
Figuras y tablas -
Analysis 3.1

Comparison 3 Antibiotic lock solutions versus control, Outcome 1 Catheter‐related infection (per 1000 catheter‐days).

Comparison 3 Antibiotic lock solutions versus control, Outcome 2 Catheter‐related infection (per 1000 dialysis sessions).
Figuras y tablas -
Analysis 3.2

Comparison 3 Antibiotic lock solutions versus control, Outcome 2 Catheter‐related infection (per 1000 dialysis sessions).

Comparison 3 Antibiotic lock solutions versus control, Outcome 3 Thrombosis (per 1000 catheter‐days).
Figuras y tablas -
Analysis 3.3

Comparison 3 Antibiotic lock solutions versus control, Outcome 3 Thrombosis (per 1000 catheter‐days).

Comparison 3 Antibiotic lock solutions versus control, Outcome 4 Colonisation.
Figuras y tablas -
Analysis 3.4

Comparison 3 Antibiotic lock solutions versus control, Outcome 4 Colonisation.

Comparison 3 Antibiotic lock solutions versus control, Outcome 5 Subgroup analysis: catheter‐related infection (per 1000 catheter‐days) in tunnelled catheters.
Figuras y tablas -
Analysis 3.5

Comparison 3 Antibiotic lock solutions versus control, Outcome 5 Subgroup analysis: catheter‐related infection (per 1000 catheter‐days) in tunnelled catheters.

Comparison 3 Antibiotic lock solutions versus control, Outcome 6 Subgroup analysis: catheter‐related infection (per 1000 catheter‐days) in non‐tunnelled catheters.
Figuras y tablas -
Analysis 3.6

Comparison 3 Antibiotic lock solutions versus control, Outcome 6 Subgroup analysis: catheter‐related infection (per 1000 catheter‐days) in non‐tunnelled catheters.

Comparison 3 Antibiotic lock solutions versus control, Outcome 7 Subgroup analysis: thrombosis per 1000 catheter‐days in tunnelled catheters.
Figuras y tablas -
Analysis 3.7

Comparison 3 Antibiotic lock solutions versus control, Outcome 7 Subgroup analysis: thrombosis per 1000 catheter‐days in tunnelled catheters.

Comparison 4 Sensitivity analysis: antibiotic lock solutions versus control excluding studies judged to be at high‐risk bias, Outcome 1 Catheter‐related infection (per 1000 catheter‐days).
Figuras y tablas -
Analysis 4.1

Comparison 4 Sensitivity analysis: antibiotic lock solutions versus control excluding studies judged to be at high‐risk bias, Outcome 1 Catheter‐related infection (per 1000 catheter‐days).

Comparison 4 Sensitivity analysis: antibiotic lock solutions versus control excluding studies judged to be at high‐risk bias, Outcome 2 Thrombosis (per 1000 catheter‐days).
Figuras y tablas -
Analysis 4.2

Comparison 4 Sensitivity analysis: antibiotic lock solutions versus control excluding studies judged to be at high‐risk bias, Outcome 2 Thrombosis (per 1000 catheter‐days).

Comparison 5 Non‐antibiotic antimicrobial lock solutions versus control, Outcome 1 Catheter‐related infection (per 1000 catheter‐days).
Figuras y tablas -
Analysis 5.1

Comparison 5 Non‐antibiotic antimicrobial lock solutions versus control, Outcome 1 Catheter‐related infection (per 1000 catheter‐days).

Comparison 5 Non‐antibiotic antimicrobial lock solutions versus control, Outcome 2 Catheter‐related infection (per 1000 dialysis sessions).
Figuras y tablas -
Analysis 5.2

Comparison 5 Non‐antibiotic antimicrobial lock solutions versus control, Outcome 2 Catheter‐related infection (per 1000 dialysis sessions).

Comparison 5 Non‐antibiotic antimicrobial lock solutions versus control, Outcome 3 Thrombosis (per 1000 catheter‐days).
Figuras y tablas -
Analysis 5.3

Comparison 5 Non‐antibiotic antimicrobial lock solutions versus control, Outcome 3 Thrombosis (per 1000 catheter‐days).

Comparison 5 Non‐antibiotic antimicrobial lock solutions versus control, Outcome 4 Thrombosis (per 1000 dialysis sessions).
Figuras y tablas -
Analysis 5.4

Comparison 5 Non‐antibiotic antimicrobial lock solutions versus control, Outcome 4 Thrombosis (per 1000 dialysis sessions).

Comparison 5 Non‐antibiotic antimicrobial lock solutions versus control, Outcome 5 Colonisation.
Figuras y tablas -
Analysis 5.5

Comparison 5 Non‐antibiotic antimicrobial lock solutions versus control, Outcome 5 Colonisation.

Comparison 5 Non‐antibiotic antimicrobial lock solutions versus control, Outcome 6 Subgroup analysis: catheter‐related infection (per 1000 catheter‐days) in tunnelled catheters.
Figuras y tablas -
Analysis 5.6

Comparison 5 Non‐antibiotic antimicrobial lock solutions versus control, Outcome 6 Subgroup analysis: catheter‐related infection (per 1000 catheter‐days) in tunnelled catheters.

Comparison 5 Non‐antibiotic antimicrobial lock solutions versus control, Outcome 7 Subgroup analysis: catheter‐related infection (per 1000 catheter‐days) in non‐tunnelled catheters.
Figuras y tablas -
Analysis 5.7

Comparison 5 Non‐antibiotic antimicrobial lock solutions versus control, Outcome 7 Subgroup analysis: catheter‐related infection (per 1000 catheter‐days) in non‐tunnelled catheters.

Comparison 5 Non‐antibiotic antimicrobial lock solutions versus control, Outcome 8 Subgroup analysis: thrombosis (per 1000 catheter‐days) in tunnelled catheters.
Figuras y tablas -
Analysis 5.8

Comparison 5 Non‐antibiotic antimicrobial lock solutions versus control, Outcome 8 Subgroup analysis: thrombosis (per 1000 catheter‐days) in tunnelled catheters.

Comparison 5 Non‐antibiotic antimicrobial lock solutions versus control, Outcome 9 Subgroup analysis: thrombosis (per 1000 catheter‐days) in non‐tunnelled catheters.
Figuras y tablas -
Analysis 5.9

Comparison 5 Non‐antibiotic antimicrobial lock solutions versus control, Outcome 9 Subgroup analysis: thrombosis (per 1000 catheter‐days) in non‐tunnelled catheters.

Comparison 6 Sensitivity analysis: non‐antibiotic lock solutions versus control excluding studies judged to be at high‐risk bias, Outcome 1 Catheter‐related infection (per 1000 catheter‐days).
Figuras y tablas -
Analysis 6.1

Comparison 6 Sensitivity analysis: non‐antibiotic lock solutions versus control excluding studies judged to be at high‐risk bias, Outcome 1 Catheter‐related infection (per 1000 catheter‐days).

Comparison 6 Sensitivity analysis: non‐antibiotic lock solutions versus control excluding studies judged to be at high‐risk bias, Outcome 2 Thrombosis (per 1000 catheter‐days).
Figuras y tablas -
Analysis 6.2

Comparison 6 Sensitivity analysis: non‐antibiotic lock solutions versus control excluding studies judged to be at high‐risk bias, Outcome 2 Thrombosis (per 1000 catheter‐days).

Comparison 7 Combined antimicrobial lock solutions versus control, Outcome 1 Catheter‐related infection (per 1000 catheter‐days).
Figuras y tablas -
Analysis 7.1

Comparison 7 Combined antimicrobial lock solutions versus control, Outcome 1 Catheter‐related infection (per 1000 catheter‐days).

Comparison 7 Combined antimicrobial lock solutions versus control, Outcome 2 Thrombosis (per 1000 catheter‐days).
Figuras y tablas -
Analysis 7.2

Comparison 7 Combined antimicrobial lock solutions versus control, Outcome 2 Thrombosis (per 1000 catheter‐days).

Comparison 8 Sensitivity analysis: combined antimicrobial lock solutions versus control excluding studies deemed with high‐risk bias, Outcome 1 Catheter‐related infection (per 1000 catheter‐days).
Figuras y tablas -
Analysis 8.1

Comparison 8 Sensitivity analysis: combined antimicrobial lock solutions versus control excluding studies deemed with high‐risk bias, Outcome 1 Catheter‐related infection (per 1000 catheter‐days).

Comparison 8 Sensitivity analysis: combined antimicrobial lock solutions versus control excluding studies deemed with high‐risk bias, Outcome 2 Thrombosis (per 1000 catheter‐days).
Figuras y tablas -
Analysis 8.2

Comparison 8 Sensitivity analysis: combined antimicrobial lock solutions versus control excluding studies deemed with high‐risk bias, Outcome 2 Thrombosis (per 1000 catheter‐days).

Summary of findings for the main comparison. Antimicrobial lock solutions vs control for preventing catheter‐related infections in patient undergoing haemodialysis

Antimicrobial lock solutions vs control for preventing catheter‐related infections in patient undergoing haemodialysis

Patient or population: CVC‐related infection
Setting: haemodialysis therapy
Intervention: antimicrobial lock solutions
Comparison: heparin and other lock solutions

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with heparin and other lock solutions

Risk with antimicrobial lock solutions

CVC ‐ related infections
assessed with: per 1000 days/catheter

Low

RR 0.38
(0.27 to 0.53)

2994
(27 RCTs)

⊕⊕⊝⊝
LOW 1 2 3 4 5

43 per 1.000

16 per 1.000
(12 to 23)

High

260 per 1.000

99 per 1.000
(70 to 138)

Thrombosis
assessed with: per 1000 days/catheter

Low

RR 0.79
(0.52 to 1.22)

2080
(14 RCTs)

⊕⊝⊝⊝
VERY LOW 6 7 8 9 10

6 per 1.000

5 per 1.000
(3 to 7)

High

330 per 1.000

261 per 1.000
(172 to 403)

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

CI: Confidence interval; RR: Risk ratio; OR: Odds ratio;

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

1 This judgment is based on the lack of information regarding the following criteria: random sequence generation, allocation concealment and blinding of outcome assesment

2 the statistics do not show serious heterogeneity and confidence intervals overlap.

3 the evidence is direct because the studies are in hemodialysis patients and the same sealing solutions the question of this review are used.

4 no imprecision is observed; because the decision regarding the use of antimicrobial lock solution is sealed better than heparin along the confidence interval

5 It is suspected a degree of publication bias

6 30% of the studies presented insufficient information to assess citerios of:random sequence generation, allocation concealment and blinding of outcome assesment

7 the statistical test showed a high heterogeneity and the confidence intervals do not overlap

8 the evidence is not indirect because the studies are in hemodialysis patients and the same sealing solutions the question of this review are used.

9 The 95% confidence interval of the pooled estimate ranges from 0.41 to 1.39, which is not narrow enough for a confident judgment of the effect size.

10 Publication bias is suspected as the funnel plot for this outcome shows asymmetry

Figuras y tablas -
Summary of findings for the main comparison. Antimicrobial lock solutions vs control for preventing catheter‐related infections in patient undergoing haemodialysis
Comparison 1. All antimicrobial (antibiotic plus non‐antibiotic plus the combination) solutions versus control

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Catheter‐related infection (per 1000 catheter‐days) Show forest plot

27

Rate Ratio (Fixed, 95% CI)

0.45 [0.38, 0.54]

2 Catheter‐related infection (per 1000 dialysis sessions) Show forest plot

2

Rate Ratio (Fixed, 95% CI)

0.49 [0.21, 1.12]

3 Thrombosis (per 1000 catheter‐days) Show forest plot

14

Rate Ratio (Random, 95% CI)

0.79 [0.52, 1.22]

4 Colonisation Show forest plot

2

Rate Ratio (Random, 95% CI)

0.37 [0.04, 3.36]

5 Subgroup analysis: catheter‐related infection (per 1000 catheter‐days) in tunnelled catheters Show forest plot

21

Rate Ratio (Random, 95% CI)

0.40 [0.30, 0.53]

6 Subgroup analysis: catheter‐related infection (per 1000 catheter‐days) in non‐tunnelled catheters Show forest plot

8

Rate Ratio (Random, 95% CI)

0.37 [0.16, 0.86]

7 Subgroup analysis: thrombosis (per 1000 catheter‐days) in tunnelled catheters Show forest plot

10

Rate Ratio (Random, 95% CI)

0.83 [0.50, 1.37]

8 Subgroup analysis: thrombosis (per 1000 catheter‐days) in non‐tunnelled catheters Show forest plot

4

Rate Ratio (Random, 95% CI)

0.66 [0.25, 1.72]

Figuras y tablas -
Comparison 1. All antimicrobial (antibiotic plus non‐antibiotic plus the combination) solutions versus control
Comparison 2. Sensitivity analysis: all antimicrobial lock solutions versus control excluding studies deemed with high‐risk bias

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Catheter‐related infection (per 1000 catheter‐days) Show forest plot

18

Rate Ratio (Random, 95% CI)

0.47 [0.32, 0.68]

2 Thrombosis (per 1000 catheter‐days) Show forest plot

9

Rate Ratio (Random, 95% CI)

0.85 [0.48, 1.52]

Figuras y tablas -
Comparison 2. Sensitivity analysis: all antimicrobial lock solutions versus control excluding studies deemed with high‐risk bias
Comparison 3. Antibiotic lock solutions versus control

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Catheter‐related infection (per 1000 catheter‐days) Show forest plot

13

Rate Ratio (Random, 95% CI)

0.30 [0.22, 0.42]

2 Catheter‐related infection (per 1000 dialysis sessions) Show forest plot

1

Rate Ratio (Random, 95% CI)

Totals not selected

3 Thrombosis (per 1000 catheter‐days) Show forest plot

5

Rate Ratio (Random, 95% CI)

0.76 [0.42, 1.38]

4 Colonisation Show forest plot

1

Rate Ratio (Random, 95% CI)

Totals not selected

5 Subgroup analysis: catheter‐related infection (per 1000 catheter‐days) in tunnelled catheters Show forest plot

9

Rate Ratio (Random, 95% CI)

0.30 [0.18, 0.50]

6 Subgroup analysis: catheter‐related infection (per 1000 catheter‐days) in non‐tunnelled catheters Show forest plot

4

Rate Ratio (Random, 95% CI)

0.14 [0.05, 0.36]

7 Subgroup analysis: thrombosis per 1000 catheter‐days in tunnelled catheters Show forest plot

3

Rate Ratio (Random, 95% CI)

0.61 [0.31, 1.23]

Figuras y tablas -
Comparison 3. Antibiotic lock solutions versus control
Comparison 4. Sensitivity analysis: antibiotic lock solutions versus control excluding studies judged to be at high‐risk bias

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Catheter‐related infection (per 1000 catheter‐days) Show forest plot

7

Rate Ratio (Random, 95% CI)

0.37 [0.28, 0.48]

2 Thrombosis (per 1000 catheter‐days) Show forest plot

2

Rate Ratio (Random, 95% CI)

0.79 [0.24, 2.63]

Figuras y tablas -
Comparison 4. Sensitivity analysis: antibiotic lock solutions versus control excluding studies judged to be at high‐risk bias
Comparison 5. Non‐antibiotic antimicrobial lock solutions versus control

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Catheter‐related infection (per 1000 catheter‐days) Show forest plot

11

Rate Ratio (Random, 95% CI)

0.65 [0.41, 1.05]

2 Catheter‐related infection (per 1000 dialysis sessions) Show forest plot

1

Rate Ratio (Random, 95% CI)

Totals not selected

3 Thrombosis (per 1000 catheter‐days) Show forest plot

8

Rate Ratio (Random, 95% CI)

0.85 [0.44, 1.66]

4 Thrombosis (per 1000 dialysis sessions) Show forest plot

1

Rate Ratio (Random, 95% CI)

Totals not selected

5 Colonisation Show forest plot

1

Rate Ratio (Random, 95% CI)

0.99 [0.27, 3.68]

6 Subgroup analysis: catheter‐related infection (per 1000 catheter‐days) in tunnelled catheters Show forest plot

9

Rate Ratio (Random, 95% CI)

0.60 [0.40, 0.91]

7 Subgroup analysis: catheter‐related infection (per 1000 catheter‐days) in non‐tunnelled catheters Show forest plot

4

Rate Ratio (Random, 95% CI)

0.93 [0.48, 1.81]

8 Subgroup analysis: thrombosis (per 1000 catheter‐days) in tunnelled catheters Show forest plot

5

Rate Ratio (Random, 95% CI)

1.17 [0.57, 2.41]

9 Subgroup analysis: thrombosis (per 1000 catheter‐days) in non‐tunnelled catheters Show forest plot

4

Rate Ratio (Random, 95% CI)

0.66 [0.25, 1.72]

Figuras y tablas -
Comparison 5. Non‐antibiotic antimicrobial lock solutions versus control
Comparison 6. Sensitivity analysis: non‐antibiotic lock solutions versus control excluding studies judged to be at high‐risk bias

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Catheter‐related infection (per 1000 catheter‐days) Show forest plot

9

Rate Ratio (Random, 95% CI)

0.65 [0.38, 1.12]

2 Thrombosis (per 1000 catheter‐days) Show forest plot

6

Rate Ratio (Random, 95% CI)

0.90 [0.43, 1.91]

Figuras y tablas -
Comparison 6. Sensitivity analysis: non‐antibiotic lock solutions versus control excluding studies judged to be at high‐risk bias
Comparison 7. Combined antimicrobial lock solutions versus control

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Catheter‐related infection (per 1000 catheter‐days) Show forest plot

4

Rate Ratio (Random, 95% CI)

0.26 [0.14, 0.49]

2 Thrombosis (per 1000 catheter‐days) Show forest plot

3

Rate Ratio (Random, 95% CI)

0.63 [0.22, 1.81]

Figuras y tablas -
Comparison 7. Combined antimicrobial lock solutions versus control
Comparison 8. Sensitivity analysis: combined antimicrobial lock solutions versus control excluding studies deemed with high‐risk bias

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1 Catheter‐related infection (per 1000 catheter‐days) Show forest plot

2

Rate Ratio (Random, 95% CI)

0.16 [0.02, 1.03]

2 Thrombosis (per 1000 catheter‐days) Show forest plot

1

Rate Ratio (Random, 95% CI)

0.66 [0.18, 2.44]

Figuras y tablas -
Comparison 8. Sensitivity analysis: combined antimicrobial lock solutions versus control excluding studies deemed with high‐risk bias