Interventions other than anticoagulants and systemic antibiotics for prevention of central venous catheter‐related infections in children with cancer

Abstract

Background

Use of central venous catheters (CVC) in treatment of children with cancer is associated with infective complications. Current evidence‐based guidelines to prevent catheter‐related infections are mainly relevant to the adult population. They are not cancer (especially not childhood cancer) specific. Two existing Cochrane reviews have looked at prophylactic antibiotics and anticoagulants to prevent CVC‐related infections.

Objectives

The primary objective was to find which interventions, if any, were effective in preventing CVC‐related infections in children with cancer. Further objectives were to examine the effectiveness of each intervention in the following subgroups: implanted versus external catheters, haematological versus non‐haematological malignancies, and in those receiving haematopoietic stem cell transplants (HSCT) versus no HSCT.

Search methods

We searched the Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library 2008, Issue 4), MEDLINE (January 1950 to January 2009), EMBASE (January 1980 to January 2009) and CINAHL(R) (January 1982 to March 2009). We also searched reference lists of relevant articles and proceedings of relevant international conferences (2004 to 2008).

Selection criteria

Randomised and quasi‐randomised studies comparing any intervention (other than anticoagulants, systemic antibiotics and antibiotic lock techniques) versus no intervention, placebo or any other intervention to prevent CVC‐related infections in children with cancer.

Data collection and analysis

Two authors independently selected studies, assessed trial quality and extracted data. Where necessary, we contacted study authors for further data and clarification of methods.

Main results

Three trials involving two different interventions were included. Two trials involving 680 children compared flushing CVC with urokinase (with or without heparin) versus heparin alone. Neither of these trials reported on the primary outcome of catheter‐related blood stream infection (CRBSI). There was a non‐significantly decreased rate of catheter‐associated infection (CAI) (Rate Ratio 0.72, 95% confidence interval 0.12 to 4.41) in the urokinase (with or without heparin) arm compared with the heparin arm.

One trial involving 113 children compared frequency of catheter dressing change every 15 days versus every 4 days. It did not report on CRBSI or CAI. There were no premature catheter removals for infection in either of the trial arms.

Authors' conclusions

Three RCTs for only two types of interventions to prevent CVC‐related infections in children with cancer have been identified. Flushing CVC with urokinase (with or without heparin) compared to heparin alone possibly leads to decrease in CAI rates. Changing catheter dressings every 15 days versus every 4 days does not lead to more premature catheter removals due to infection although data were insufficient to assess if catheter‐related infection rates were changed.

Plain language summary

Interventions to prevent central venous catheter‐related infections in children with cancer

Treatment of children with cancer often involves giving drugs, fluids and blood products through veins. In addition, small amounts of blood from the child are frequently needed for testing in the laboratory. All this can be achieved by inserting a central venous catheter (CVC) which is a small tube inserted via skin into the blood vessel in the neck or the armpit. This allows repeated testing and treatment of the child with cancer over a period of months while minimising the discomfort. The presence of CVC in the veins also leads to an increased risk of infections which can be life‐threatening. Our review systematically assessed the research done on strategies to prevent these infections in children with cancer.

A total of three research studies were identified. Two studies showed that there may be a decrease in CVC‐related blood infections if the space in the CVC was washed and filled at regular intervals with urokinase (a drug which dissolves blood clots) with/without heparin (a drug which prevents the formation of blood clots) compared to heparin alone. One study showed that changing the dressing which covered the skin at the insertion of CVC every 15 days rather than every 4 days did not lead to an increased removal of the CVC because they had become infected. No research studies were identified for several other potential strategies which could reduce CVC‐related infections in children with cancer.

Background

Description of the condition

In the management of children with cancer, good and reliable venous access is essential. Currently this is achieved in the majority of patients by catheterisation of central veins which enables effective and safe delivery of chemotherapy, parenteral nutrition and blood products, where required, and also allows frequent blood sampling while minimising child discomfort by decreasing frequent peripheral venepuncture. Along with these advantages, central venous catheters (CVCs) are also associated with some problems: mechanical, infectious and thrombotic. The spectrum of infection problems ranges from localised infection at the exit site of the CVC to confirmed catheter‐related bloodstream infections (CRBSI).

The incidence of infections (conventionally expressed as infections per 1000 CVC days) depends on the kind of patient population, hospital setting, underlying diagnosis and type of catheter used. In one series of paediatric oncology patients the overall incidence of central‐line associated bloodstream infections (BSI) was 1.7 per 1000 CVC days, but it was much higher for inpatients (7.4 per 1000 CVC days) than for outpatients (0.3 per 1000 CVC days) (Simon 2000). This difference was a reflection of much higher frequencies of CVC use during periods of inpatient hospital treatment and the fact that some of the inpatients, in particular during high‐dose chemotherapy, were severely myelosuppressed (e.g. prolonged neutropenia) and immunosuppressed. Similarly, differences in the incidence of central‐line associated BSI have been shown among implanted and external catheters in children with cancer (Ingram 1991).  

Gram positive bacteria, particularly coagulase negative Staphylococcus are the commonest isolates from the blood in children with malignancies with indwelling CVC with febrile or septic episodes (Raymond 2000; Simon 2008). Other common pathogens causing infection in such a setting are Staphylococcus aureus, different species of aerobic Gram‐negative bacilli and Candida albicans (Laws 2006).

Description of the intervention

In 2002 the Centers for Disease Control and Prevention (CDC), USA recommended general guidelines to prevent catheter‐related infections (peripheral and central venous, arterial, peritoneal, umbilical and pulmonary arterial catheters) across all ages and patient groups (CDC 2002). The major areas of emphasis were: educating and training healthcare providers, using maximal sterile barrier precautions during insertion, using a 2% chlorhexidine preparation for skin antisepsis whenever the line is to be handled, avoiding routine replacement of CVCs as a strategy to prevent infection and using antiseptic/antibiotic impregnated short‐term central venous catheters if the rate of infection is high despite adherence to other strategies (i.e. education and training, maximal sterile barrier precautions and 2% chlorhexidine for skin antisepsis).

How the intervention might work

The most frequent route for catheter‐related infection is colonisation of the catheter by skin commensals. These may enter either through the hub of the catheter and subsequently colonise the inside lumen of the catheter, or they may enter from the exit site on the skin and colonise the outer surface of the catheter (Raad 1993; Salzman 1993). Other rarer routes of infection are through haematogenous seeding of infection from other sites in the body and from contaminated infusates. Once the microbe has gained entry onto the surface of the catheter, further progress depends on its ability to bind to the material and link with extracellular glycopolysaccharides, fibrin and fibronectin to form a biofilm (Darouiche 2001). Staphylococcus epidermidis produces specific adhesins to attach to the luminal surface of the catheter, and each other, forming a biofilm. Unlike S. epidermidis, adherence of Staphylococcus aureus is dependent on the presence of host‐tissue ligands, including fibronectin, fibrinogen and collagen. The biofilm enhances the survival of the microbe by attenuating the effect of both the body's immune response as well as that of the antibiotic. So, strategies to prevent the infections would need primarily to prevent colonisation of the catheter by skin commensals and prevent/disrupt the formation of a biofilm (Darouiche 2001; Simon 2006).

Why it is important to do this review

Cancer patients are at a higher risk of infections from disruption of mucocutaneous barriers (from vascular catheters, chemotherapy and radiotherapy), quantitative and qualitative impairment of phagocytic cells and cell‐mediated immunity. Children with cancer are a unique group because of their frequent episodes of neutropenia. Frequent intended or inadvertent catheter handling by healthcare professionals, parents, siblings and children themselves puts them at additional risk of catheter‐related infections. The CDC guidelines for catheter‐related infections (CDC 2002) did not consider cancer patients (adult or children) as a specific subgroup. Existing systematic reviews report on a few single interventions which have shown benefit in flushing the CVC with a vancomycin/heparin solution but no benefit from administering an antibiotic prior to the insertion of the catheter or using anticoagulant. None of these reviews have been able to include a child‐specific analysis (Akl 2007; van de Wetering 2007).  

In this review we aim to look systematically at all interventions, other than anticoagulants and systemic antibiotics, which are used in an attempt to prevent the CVC‐related infections in children with haematological and non‐haematological cancers. We also aim to subgroup the interventions into previously recognised categories, pool results and carry out meta‐analysis wherever appropriate and feasible (CDC 2002; Pratt 2007). This will help to provide the most up to date and clear evidence for oncology units, helping them to implement interventions and direct future research.

Objectives

The primary aim was to find which interventions (other than anticoagulants, systemic antibiotics and antibiotic lock techniques) were effective in preventing CVC‐related infections in children with cancer.

Further objectives were to examine the effectiveness of each intervention (for details see the section 'Types of interventions') across the following subgroups:

1. those with implanted versus external long‐term tunnelled catheters;

2. those with haematological versus non‐haematological malignancies; and

3. those children with cancer who have undergone haematopoietic stem cell transplant (HSCT) versus those who have not.

Methods

Criteria for considering studies for this review

Types of studies

All randomised controlled trials (RCTs) and quasi‐randomised controlled trials.

Types of participants

All children (less than 18 years of age) with cancer who had a long‐term tunnelled CVC. It is possible that some of the studies would have a proportion of patients with non‐malignant conditions (haematological and non‐haematological) who met other selection criteria. In that case, as long as the majority had malignant conditions, we would have included the studies and done a sensitivity analysis to study the effect of including such studies. Similarly, we would have included studies on children with cancer where some patients were older than 18 years if the majority were less than 18 years of age and performed a sensitivity analysis to study the effect of including such studies.

Types of interventions

Any intervention (other than anticoagulants, systemic antibiotics and antibiotic lock techniques) versus no intervention, placebo or any other intervention to prevent CVC‐related infections. We did not consider interventions to treat other catheter‐related complications. We made an effort to classify interventions on the basis of previously recognised categories wherever feasible and meaningful (CDC 2002; Pratt 2007).

• Education of healthcare workers and patients/carers.

• Type of catheter:

  • type of material, e.g. Teflon^®, polyurethane, polyvinyl chloride, polyethylene, silicone, etc.;

  • whether antimicrobial/antiseptic impregnated, e.g. chlorhexidine/silver sulfadiazine, minocycline/rifampin, platinum/silver and ionic silver cuffs;

  • single or multiple lumens;

  • implanted or non‐implanted.

• Site of catheter insertion: subclavian, jugular and femoral placements.

• Hand hygiene and general asepsis: decontamination of hands either by washing with an antimicrobial liquid soap and water, or by using an alcohol hand rub. Using aseptic non‐touch technique for catheter site care.

• Skin antisepsis: with chlorhexidine or povidone‐iodine.

• Catheter and catheter site care: this included good practice in caring for the patient’s catheter hub and connection port, the use of an appropriate CVC site dressing regimen (transparent polyurethane or gauze and tape dressings) and using flush solutions (heparin or saline) to maintain the patency of the CVC.

• In‐line filters.

• Catheter replacement strategies.

• Any other interventions other than anticoagulants and systemic antibiotics.

Types of outcome measures

We looked at the outcomes defined below. The definitions had been adopted from standard definitions used elsewhere (CDC 2002; Eggimann 2004; Simon 2006).

Primary outcomes

Catheter‐related blood stream infection

Bacteraemia/fungaemia in a patient with an intravascular catheter with at least one positive blood culture obtained from a peripheral vein, clinical manifestations of infection (i.e. fever, chills and/or hypotension), and no apparent source for the BSI except the catheter. One of the following should be present: a positive semiquantitative (> 15 Colony Forming Units (CFU)/catheter segment) or quantitative (> 10³ CFU/catheter segment) culture whereby the same organism (species and antibiogram) is isolated from the catheter segment and peripheral blood; simultaneous quantitative blood cultures with a > 5:1 ratio CVC versus peripheral; differential period of CVC culture versus peripheral blood culture positivity of > two hours.

Secondary outcomes

Catheter‐associated infection

Systemic infection, characterised by clinical signs such as fever, chills, lowering of arterial blood pressure, reduced blood flow through the circulatory periphery (capillary filling time > 3 seconds) and oliguria, with or without positive blood cultures, in a patient with a CVC and no other detectable focus.  Some authors of studies will also include one or more of the following: CVC used in the last 48 hours, defervescence of symptoms and signs after removal of catheter, CVC present for more than three days.

Exit site infection

Erythema, induration and/or tenderness within 2 cm of the catheter exit site (may be associated with other signs and symptoms of infection, such as fever or pus emerging from the exit site with or without positive culture from skin/pus at local site).

Tunnel infection

Tenderness, erythema and/or induration > 2 cm from the catheter exit site, along the subcutaneous tract of a tunnelled catheter (e.g. Hickman or Broviac catheter).

Pocket infection

Infected fluid in the subcutaneous pocket of a totally implanted intravascular device, often associated with tenderness, erythema and/or induration over the pocket; spontaneous rupture and drainage, or necrosis of the overlying skin may also occur.

Premature catheter removals for infection

Search methods for identification of studies

See: Cochrane Childhood Cancer Group methods used in reviews

Electronic searches

We searched the following databases: Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library 2008, Issue 4), MEDLINE (January 1950 to January 2009), EMBASE (January 1980 to January 2009) and CINAHL(R) (January 1982 to March 2009). There were no language restrictions. A detailed description of the search strategies is provided in Appendix 1, Appendix 2, Appendix 3 and Appendix 4.

Searching other resources

Proceedings of scientific meetings

We handsearched abstracts of conference proceedings of the following:

1. American Society of Clinical Oncology 2004 to 2008;

2. American Society of Pediatric Hematology/Oncology 2004 to 2008;

3. International Society of Paediatric Oncology 2004 to 2008; and

4. Multinational Association of Supportive Care in Cancer 2004 to 2008.

Reference lists

We searched reference lists of included studies and reviews for additional relevant studies.

Identifying ongoing unpublished trials

We identified ongoing research trials from the metaRegister of Controlled Trials (www.controlled‐trials.com/mrct/) and the National Research Register (portal.nihr.ac.uk/Pages/NRRArchive.aspx). If any relevant studies were found, we contacted the investigators to obtain more information.

Data collection and analysis

Selection of studies

The Cochrane Childhood Cancer Group ran the search strategy in CENTRAL, MEDLINE and EMBASE. The results from different databases were merged and duplicate reports of the same study removed. RSA and RR independently selected studies; they were not blinded to the journal, the authors or the institution.  The first stage of screening of the search results involved assessing titles and abstracts to determine whether each article met inclusion criteria. If the title and abstract did not provide enough information to make a decision then we obtained the full text of the article and applied the inclusion criteria. We resolved disagreement between RSA and RR by discussion and if no agreement was reached then we obtained the opinion of BP and/or TE. RSA corresponded with study investigators, where appropriate, to clarify study eligibility. At the end of the process RSA divided the included studies based on the previously decided categories of interventions.

Data extraction and management

We used specifically designed forms for data extraction and there was no blinding to the journal, the authors or the institution. We extracted data with the aim of producing an intention‐to‐treat (ITT) analysis. RSA and RR extracted data independently and then compared. We resolved disagreement by discussion and if no agreement was reached then we obtained the opinion of BP and/or TE.  RSA contacted the original investigators if there was a need to obtain additional data.

Assessment of risk of bias in included studies

RSA and RR independently assessed the methodological quality of selected trials using the specific tool for assessing risk of bias recommended by The Cochrane Collaboration (Higgins 2008). Other biases assessed were those due to early stoppage of trial and funding by pharmaceutical industry. We resolved disagreement by discussion and if no agreement was reached then we obtained the opinion of BP and/or TE. The findings of the risk of bias assessment were described in a narrative form for the individual domains in the results section.

Measures of treatment effect

For dichotomous outcomes, we expressed the estimate of effect of an intervention as risk ratio together with 95% confidence interval (CI). For continuous outcomes, we used weighted mean differences and used standard deviations to summarise the data for each group using mean differences and 95% CIs. For counts of rare events e.g. catheter infections which were conventionally expressed as per 1000 CVC days, we used rate ratio (rate of event in experimental arm/rate of event in control arm) as a summary statistic. Meta‐analysis of rate ratios was done by using the generic inverse‐variance approach in RevMan. For this purpose, the data were entered as natural logarithms (log rate ratio and the standard error of the log rate ratio).

Unit of analysis issues

In the case of any intervention trials other than those with a simple parallel design in which each individual was randomised to one of the intervention groups, such as cluster‐randomised trials or cross‐over trials, we took appropriate steps to avoid unit of analysis error.

Dealing with missing data

Whenever possible, RSA contacted the original investigators to request missing data. Where there were missing data, we 'imputed' and carried out a sensitivity analysis between studies in which data were 'imputed' for ITT analysis assuming that all missing participants experienced the event, or that all missing participants did not experience the event.

Assessment of heterogeneity

We assessed the significance of any discrepancies in the estimates of the treatment effects from the different trials with a random‐effects model and by using the I² statistic, which describes the percentage total variation across studies that is due to heterogeneity rather than to chance (heterogeneity was defined as I² > 50%).

Assessment of reporting biases

We assessed reporting bias by producing a funnel plot, where there were a sufficient number of included studies, as well as by carrying out a sensitivity analysis with a fixed‐effect and a random‐effects model. We took the following measures to reduce reporting bias:

• searching multiple electronic databases, proceedings of scientific meetings and trial registries to deal with location and time lag bias;

• keeping no language restriction in the search strategy; and

• excluding duplicate reports of the same study to avoid duplicate publication bias.

Data synthesis

The primary aim was to produce a meta‐analysis only if studies were of similar comparisons reporting the same outcome measures for each intervention category. If this was not feasible then we reported a narrative analysis for that intervention.

Subgroup analysis and investigation of heterogeneity

We assessed clinical heterogeneity by examining the types of participants and interventions for all outcomes in each study and by carrying out a subgroup analysis. We analysed the following subgroups:

1. those with implanted versus external catheter;

2. those with haematological versus non‐haematological malignancies; and

3. those children with cancer who have undergone HSCT versus those who have not.

Sensitivity analysis

We planned to undertake sensitivity analyses of the following to examine the effect of statistical heterogeneity:

1. analysis with and without quasi‐randomised studies;

2. analysis with and without studies where some patients were > 18 years of age;

3. analysis with and without studies in which a proportion of patients have non‐malignant conditions;

4. analysis with and without abstracts of studies obtained from scientific meetings but no details available from the original authors;

5. analysis with and without studies excluded after post‐hoc analysis;

6. analysis with a fixed‐effect and a random‐effects model; and

7. analysis between studies in which data were 'imputed' for ITT analysis, assuming that all missing participants experienced the event, or that all missing participants did not experience the event.

Results

Description of studies

Results of the search

The search strategy identified 876 citations, including 216 duplicates. We screened the titles and abstracts of the 660 unique citations and identified 28 as potentially eligible based on predetermined criteria. After full text screening we excluded 16 of the 28 because all or most of the participants in these trials were over 18 years of age and we excluded a further two because the intervention was not randomised. Of the remaining 10 studies for possible inclusion we excluded seven (Excluded studies) for various reasons as detailed in Characteristics of excluded studies. No additional information was obtained from searching reference lists of included studies and conference proceedings. We identified three trials which are currently recruiting patients and could be potentially eligible in the future on searching the research registers (Handrup 2008; Martin 2008; Wetering 2008) and their details can be found in Characteristics of ongoing studies.

Included studies

Three studies were finally included in the review (Included studies). In two of them the prophylactic intervention was flushing the CVC with urokinase (Aquino 2002; Dillon 2004) and in the other the prophylactic intervention was a longer interval (15 days) between changing the dressing of the CVC (Benhamou 2002). Details of these studies are described in Characteristics of included studies.

Two studies assessed urokinase as the prophylactic intervention

• Aquino 2002 enrolled 103 patients between ages 1 and 21 years with haematological and non‐haematological malignancies. Only those with implanted CVC were included. The patients were randomised to monthly catheter flushes with 3 ml of Urokinase‐Heparin (total doses 5000 IU of urokinase) versus heparin (total doses 300 units of heparin). Only those participants who received catheter flushes on at least six occasions at one month intervals were included in analysis. Outcomes were bacteraemia and premature catheter removal for infection.

• Dillon 2004 enrolled 577 patients between ages 3 months and 21 years with haematological and non‐haematological malignancies. Patients with implanted and external CVC were included. The patients were randomised to two weekly catheter flushes with urokinase (5000 IU/ml) versus heparin (100 units/ml) in volume sufficient to fill the entire catheter. Outcomes were complete catheter occlusion, partial catheter occlusion, catheter‐related infection, time to first catheter occlusion and time to first catheter‐related infection.

One study assessed frequency of dressing change as the prophylactic intervention

Benhamou 2002 enrolled 113 patients between ages 1 and 22 years with malignancy (mainly non‐haematological) who were candidates for high‐dose chemo/radiotherapy followed by bone marrow transplantation. Only those with external catheters were included. The patients were randomised to catheter dressing change every 15 days versus every 4 days. Outcomes were skin toxicity at dressing site, local pain, bacteraemia and premature catheter removal for infection.

Excluded studies

Three eligible RCTs did not report outcomes of interest (Freiberger 1992; Hinds 1991; Wiernikowski 1991). Freiberger 1992 randomised 60 children with CVC to four different dressing change techniques (betadine and tegaderm; betadine and gauze; hibiclens and tegaderm; hibiclens and gauze). Outcomes were irritation of skin (erythema, discharge and swelling) at dressing site and any bacterial growth at exit site of CVC. The authors did not differentiate whether signs of inflammation at the exit site were from irritation to the dressing or from infection. Systemic infections (CRBSI/CAI) were not measured as an outcome. Hinds 1991 randomised 42 children with single lumen Hickman catheters to a clean or unclean procedure during blood sampling from the catheters and subsequent reinfusion. Outcomes were any positive bacterial or fungal cultures in the blood collected during sampling. Children with any evidence of infection were specifically excluded form the intervention. Exit site infection, tunnel infection or systemic infections (CRBSI/CAI) were not measured. Wiernikowski 1991 randomised 33 children with Broviac catheters to weekly flushes with sterile saline or bacteriostatic saline which contained 1% benzyl alcohol. The primary outcome was catheter colonisation which was defined as positive blood culture from CVC as part of weekly surveillance or any clinically septic episode in the absence of a positive culture. This definition was too broad and we contacted authors to obtain further information on outcomes which may permit inclusion of the study. However, we obtained no further data.

The other four studies were excluded for the following reasons. One study had participants and outcomes of interest but the intervention was not prophylactic (van Eys 1982), one study was excluded because non‐tunnelled catheters were used (Henneberg 1996), and two studies (Freiberger 1991; Jones 2001) were earlier published abstracts of RCTs which we eventually included (Dillon 2004) or excluded (Freiberger 1992) in this review.

Risk of bias in included studies

The risk of bias was assessed using the specific tool recommended by The Cochrane Collaboration (Higgins 2008). In the three RCTs included in the review, one did not describe the method of sequence generation (Dillon 2004) and none described allocation concealment (Aquino 2002; Benhamou 2002; Dillon 2004). In only one study were the participants and personnel blinded (Aquino 2002) while in one other study blinding participants and personnel was not feasible due to the nature of the intervention (Benhamou 2002). None of the studies described blinding of outcome assessors (Aquino 2002; Benhamou 2002; Dillon 2004). None of the studies did intention to treat analysis (Aquino 2002; Benhamou 2002; Dillon 2004) and the percentage of participants included in analysis as compared to those enrolled was 72% (Aquino 2002) and 99% (Benhamou 2002; Dillon 2004). No study reported all the expected standard outcomes (Aquino 2002; Benhamou 2002; Dillon 2004). In Aquino 2002 and Dillon 2004, both of which were partly funded by the pharmaceutical industry, the projected number of participants were not enrolled as the study drug was removed from the market and enrolment of participants had to be discontinued. A more detailed description of risk of bias can be seen in Figure 1 and Figure 2 as well as in the risk of bias table in the Characteristics of included studies table.

1.

Methodological quality graph: review authors' judgements about each methodological quality item presented as percentages across all included studies.

2.

Methodological quality summary: review authors' judgements about each methodological quality item for each included study.

Effects of interventions

We found three RCTs for two different interventions ‐ urokinase (Aquino 2002; Dillon 2004) and frequency of dressing change (Benhamou 2002). For all other types of interventions which are categorised in the methods section (Types of interventions) we did not find any RCTs which could be included in this review.

Urokinase as the prophylactic intervention

Catheter‐related blood stream infection as an outcome

None of the two studies reported this as an outcome.

Catheter‐associated infection as an outcome

Aquino 2002 reported an overall rate of CAI for implanted catheters of 0.4 per 1000 CVC days in the urokinase‐heparin arm and 0.6 per 1000 CVC days in the heparin arm. They did not report on the statistical significance of this result and no P values or confidence intervals were given. Dillon 2004 reported an overall rate of CAI for all types of CVC of 1.6 per 1000 CVC days in the urokinase arm and 2.2 per 1000 CVC days in the heparin arm with a P value of 0.05 (Dillon 2004). On meta‐analysis of these two trials there was a non‐significantly decreased rate of CAI (rate ratio 0.72, 95% confidence interval 0.12 to 4.41) in the urokinase (with or without heparin) arm compared with the heparin arm (Figure 3). No statistical heterogeneity was identified in the analysis. Intention‐to‐treat analysis was not feasible due to insufficient information available from individual studies. Sensitivity analysis by fixed‐effect or random‐effects model made no difference to the rate ratio or the confidence intervals. Other sensitivity analysis were not feasible. A funnel plot to assess publication bias was not done as there was an insufficient number of included studies.

3.

Forest plot of comparison: 1 Urokinase (with or without Heparin) versus heparin, outcome: 1.1 Catheter‐associated infection.

Meta‐analysis was not possible for any of the subgroups. In one trial CAI rates were stratified by type of CVC (Dillon 2004). The rate of CAI for external catheters was 2.6 per 1000 CVC days in the urokinase arm and 3.9 per 1000 CVC days in the heparin arm with a P value of 0.04. The authors commented that there was no significant difference in the rate of CAI for implanted catheters in both arms but the actual CAI rates were not given.

Exit site infection as an outcome

Dillon 2004 did not report this as an outcome.

Tunnel infection as an outcome

Dillon 2004 did not report this as an outcome.

Pocket infection as an outcome

None of the two studies reported this as an outcome.

Premature catheter removals for infection as an outcome

Although both studies reported this as an outcome, only Aquino 2002 reported it in a complete and meaningful way. There were two premature catheter removals for infection in the urokinase‐heparin arm and one in the heparin arm. They did not report on the statistical significance of this result and no P values or confidence intervals were given.

Frequency of dressing change as the prophylactic intervention

Catheter‐related blood stream infection as an outcome

The study did not reported this as an outcome.

Catheter‐associated infection as an outcome

The study did not reported this as an outcome.

Exit site infection as an outcome

The study did not reported this as an outcome.

Tunnel Infection as an outcome

The study did not reported this as an outcome.

Pocket infection as an outcome

This outcome was not applicable to this trial of external catheters.

Premature catheter removals for infection as an outcome

This was reported as an outcome. There were no premature catheter removals for infection in the intervention arm (catheter dressing change every 15 days) or in the control arm (catheter dressing change every 4 days) (Benhamou 2002).

Discussion

The aim of the review was to systematically assess the effectiveness of interventions for prevention of CVC‐related infections in the specific high‐risk population of children with cancer. The range of interventions assessed is based on previously published American and English evidence‐based guidelines on prevention of catheter‐related infections (CDC 2002; Pratt 2007). The primary finding of this review is that no RCTs which fit the inclusion criteria could be identified for most of the intervention groups (Types of interventions). The three RCTs looking at two different interventions (use of urokinase and frequency of dressing change) which we finally included in the review did not report the primary outcome (CRBSI) and several of the secondary outcomes (exit site infection, tunnel infection, pocket infection).

Accurate diagnosis of catheter‐related infections remains a challenge. The criteria for CAI are broad and non‐specific; this presents a challenge in a clinical/trial setting, but is acceptable for epidemiological surveillance. CRBSI is specific and accepted a gold standard by the scientific community (CDC 2002; Pratt 2007). However, diagnosis of CRBSI requires either catheter removal or at least two positive blood cultures (at least one of which should be from the peripheral vein). Both of these are challenging in a paediatric haematology/oncology setting and often not feasible. Alternative techniques, which are equally sensitive and specific in diagnosing CRBSI but are more practical, have been suggested (Raad 2007) and their feasibility in the paediatric haematology/oncology setting has been demonstrated (Franklin 2004; Gaur 2005).

Urokinase as the prophylactic intervention

Urokinase disrupts biofilms by virtue of its fibrinolytic activity. Using this as a biological basis, several cases have been reported in which use of urokinase as an adjunct along with antibiotics was found to be useful in treating persistent/refractory CVC infections ( Ascher 1993; Fishbein 1990; Jones 1993) but this has not been proven in controlled clinical trials (Atkinson 1998; La Quaglia 1994). An extension of this role of urokinase is the prevention of formation of biofilms and thus catheter‐related infections. Kalmanti 2002 demonstrated a reduction in incidence of CRBSI in a cohort of 16 children with external CVC which were flushed with 10,000 IU of urokinase every week (0.46 per 1000 CVC days) compared with a historical cohort of 15 children with external CVC which received heparin flushes (1.08 per 1000 CVC days).

In this review, none of the included RCT reported CRBSI rates. Based on the meta‐analysis, the rate of CAI is 0.72 times with use of urokinase compared to use of heparin although the confidence intervals are wide and the two included RCT are clinically heterogenous with regard to the type of study drug, its concentration, frequency of flushing and types of catheters. Only one RCT in this review looked at children with cancer with external catheters (Dillon 2004). It did not measure incidence of CRBSI but of CAI (which is less specific) and found the incidence to be 2.6 per 1000 CVC days in the urokinase arm and 3.9 per 1000 CVC days in the heparin arm (P value of 0.04). There is a suggestion that urokinase may not be as effective for prevention of infections in implanted catheters as in external catheters but insufficient data and methodological limitations in the included studies (Aquino 2002; Dillon 2004) limit our ability to draw more definite inferences. Both the included studies that were partly funded by the pharmaceutical industry were closed prematurely due to withdrawal of the study drug (Abbokinase) from the market due to manufacturing violations (FDA 1999).

Frequency of dressing change as the prophylactic intervention

Current guidelines based on expert opinion recommend that transparent dressings over CVC should be changed every seven days, or sooner if they are damp, loose or soiled (CDC 2002; Pratt 2007). Benhamou 2002 randomised patients to catheter dressing change every 15 days versus every 4 days which resulted in less skin toxicity at dressing site in the 15‐day group and no premature catheter removals for infection in either group. Importantly, only 17% of those in the 15‐day group had dressing changed at pre‐specified frequency. Whereas the planned frequency was maintained in the 4‐day group (mean = 4 ± 1 days), it was usually shortened in the 15‐day group (mean = 8 ± 4 days), mainly because dressings had loosened. As adherence to intervention was so poor, the effectiveness of a 15‐day dressing frequency change regimen on prevention of CVC‐related infections cannot be judged.

Authors' conclusions

Implications for practice.

In the absence of good quality RCTs for most interventions to prevent CVC‐related infections in children with cancer, our practice presently is guided by other sources. One such source is the evidence extrapolated from RCTs (where available) from adult or general paediatric population. The other source is guidelines published by expert panels of paediatric oncologists based on available evidence from non‐randomised studies and cumulative clinical experience. Specifically with regard to urokinase, its use in regular flushing of CVCs (particularly external catheters) may be of some benefit although this is not conclusively proven yet.

Implications for research.

More RCTs looking at various interventions to prevent CVC‐related infections in children with cancer are needed. When designing such trials, care should be taken to include all relevant outcomes (including CRBSI) a priori based on standard definitions. There is also a need to standardise the definitions of these outcomes in a paediatric setting where catheter removal or peripheral vein culture may not be an option. Specifically with regard to regular flushing of CVCs with urokinase, the optimum dose, frequency of flushing and cost‐effectiveness needs to be established in future studies.

Appendices

Appendix 1. Search strategy for Medline (OVID)

1. For Children the following MeSH headings and text words were used:

1. Infant, Newborn/ or Infant/ 
 2. Child, Preschool/ or Child/ 
 3. Adolescent/ 
 4. Minors/ 
 5. Puberty/ 
 6. Pediatrics/ 
 7. Schools/ or Schools, Nursery/ 
 8. 1 or 2 or 3 or 4 or 5 or 6 or 7 
 9. infant.ti,ab. or infant$.mp. or infancy$.mp. 
 10. (newborn or newborn$ or (new adj born$)).mp. 
 11. (baby or baby$ or babies).mp. 
 12. (neonat$ or perinat$ or postnat$).mp. 
 13. (child or child$).mp. 
 14. (schoolchild or schoolchild$ or (school adj child) or (school adj child$)).mp. 
 15. (kid or kids).mp. 
 16. (toddler or toddler$).mp. 
 17. adolesc$.mp. 
 18. (teen or teens or teenager or teenager$).mp. 
 19. (boy or boy$).mp. 
 20. (girl or girl$).mp 
 21. (minors or minors$).mp. 
 22. (underag$ or (under adj ag$)).mp. 
 23. juvenil$.mp. 
 24. youth$.mp. 
 25. kindergar$.mp. 
 26. (puberty or puber$ or pubescen$).mp. 
 27. (prepuberscen$ or prepuberty$).mp. 
 28. (pediatrics or pediatric$ or paediatric$ or peadiatric$).mp. 
 29. (schools or preschool$ or (pre adj school$)).mp. 
 30. ((primary adj school$) or (secondary adj school$)).mp. 
 31. ((elementary adj school) or (elementary adj school$)).mp. 
 32. ((high adj school$) or highschool$).mp. 
 33. (nursery adj school$).mp. 
 34. (schoolage or (school adj age$) or schoolage$).mp. 
 35. or/9‐34 
 36. 8 or 35

2. For Cancer the following MeSH headings and text words were used:

1 (cancer or cancers or cancer$).mp. 
 2 (oncology or oncolog$).mp. 
 3 (neoplasm or neoplasms or neoplasm$).mp. or exp neoplasms/ 
 4 (carcinoma or carcinom$).mp. or exp carcinoma/ 
 5 (tumor or tumour or tumor$ or tumour$ or tumors or tumours).mp. 
 6 (malignan$ or malignant).mp. 
 7 (hematooncological or hemato oncological or hemato‐oncological or hematologic neoplasms or hematolo$).mp. or exp hematologic neoplasms/ 
 8 or/1‐7

3. For Childhood cancer the following MeSH headings and text words were used:

1 (leukemia or leukemi$ or leukaemi$ or acute lymphocytic leukemia).mp. 
 2 (AML or lymphoma or lymphom$ or hodgkin or hodgkin$ or T‐cell or B‐cell or non‐hodgkin).mp. 
 3 (sarcoma or sarcom$ or Ewing$ or osteosarcoma or osteosarcom$ or wilms tumor or wilms$).mp. 
 4 (nephroblastom$ or neuroblastoma or neuroblastom$ or rhabdomyosarcoma or rhabdomyosarcom$ or teratoma or teratom$ or hepatoma or hepatom$ or hepatoblastoma or hepatoblastom$).mp. 
 5 (PNET or medulloblastoma or medulloblastom$ or PNET$ or neuroectodermal tumors or primitive neuroectodermal tumor$ or retinoblastoma or retinoblastom$ or meningioma or meningiom$ or glioma or gliom$).mp. 
 6 (pediatric oncology or paediatric oncology).mp. 
 7 ((childhood adj cancer) or (childhood adj tumor) or (childhood adj tumors) or childhood malignancy or (childhood adj malignancies) or childhood neoplasm$).mp. 
 8 ((pediatric adj malignancy) or (pediatric adj malignancies) or (paediatric adj malignancy) or (paediatric adj malignancies)).mp. 
 9 ((brain adj tumor$) or (brain adj tumour$) or (brain adj neoplasms) or (brain adj cancer$) or brain neoplasm$).mp. 
 10 (central nervous system tumor$ or central nervous system neoplasm or central nervous system neoplasms or central nervous system tumour$).mp. 
 11 intracranial neoplasm$.mp. 
 12 LEUKEMIA/ or LYMPHOMA/ or brain tumor/ or central nervous system tumor/ or teratoma/ or sarcoma/ or osteosarcoma/ 
 13 nephroblastoma/ or neuroblastoma/ or rhabdomyosarcoma/ or hepatoblastoma/ or medulloblastoma/ or neuroectodermal tumor/ or retinoblastoma/ or meningioma/ or glioma/ or childhood cancer/ 
 14 or/1‐13

4. For Catheter the following MeSH headings and text words were used:

1 (broviac or port‐a‐cath or port acath or port a cath or hickman).mp. 
 2 (catheterisation central venous or tunnelled central venous catheter or TCVC or catheterization central venous or peripherally inserted central catheter or picc or central venous line or central venous device).mp. or catheterization, central venous/ 
 3 (central venous access device or CVAD or CVC or central venous catheter).mp. 
 4 (catheter or catheter$ or cathethers or catheterisation or catheterizations or catheterization$).mp. 
 5 exp catheter/ 
 6 exp catheterization/ or catheterization.mp. 
 7 (cannulation or cannulations or cannulation$ or cannula or cannulas or cannula$).mp. 
 8 or/1‐7

5. For Infection the following MeSH headings and text words were used:

1 (infection or infections or infection$).mp. or exp Infection/ or communicable diseases/ or (communicable disease or communicable diseases).mp. 
 2 (bacteremia or bacteremias or bacteremia$).mp. 
 3 exp Bacteremia/ 
 4 (sepsis or septic or septicemia or septicemi$ or specific bacterial infection).mp. 
 5 septic shock.mp. or exp septic shock/ 
 6 (shock, toxic or toxic shock or toxic shock syndrome or shock syndrome, toxic or shock syndromes, toxic or syndrome, toxic shock or syndromes, toxic shock or toxic shock syndromes or shock, endotoxic or endotoxic shock).mp. 
 7 exp sepsis/ 
 8 (severe sepsis or sepsis, severe or blood poisoning or poisoning, blood).mp. 
 9 (catheter‐related infection or catheter related infections or catheter related infection$).mp. 
 10 (endotoxemia or endotoxemias or endotoxemia$).mp. or exp endotoxemia/ 
 11 (toxemia or toxemias or toxemia$).mp. or exp toxemia/ 
 12 (endotoxin or endotoxins or endotoxin$).mp. or exp endotoxins/ 
 13 (hemorrhagic septicemia or Haemorrhagic Septicaemia or Septicaemia, Haemorrhagic or Hemorrhagic Septicaemia or Septicaemia, Hemorrhagic or Septicemia, Hemorrhagic or Haemorrhagic Septicemia or Septicemia, Haemorrhagic or Hemorrhagic Bacteremia or Haemorrhagic Bacteremia or Bacteremia, Haemorrhagic or Bacteremia, Hemorrhagic).mp. 
 14 exp hemorrhagic septicemia/ 
 15 or/1‐14

6. For Cochrane RCT/CCT the following MeSH headings and text words were used:

1. (randomized controlled trial or controlled clinical trial).pt. or randomized.ti,ab. or placebo.ti,ab. or drug therapy.fs. or randomly.ti,ab. or trial.ti,ab. or groups.ti,ab.) and humans.sh.

Combined search: 1 and (2 or 3) and 4 and 5 and 6

[mp = title, original title, abstract, name of substance word, subject heading word; / = MeSH heading; $ = 1 or more characters; ti,ab = title or abstract; fs = floating subheading; pt = publication type; sh = MeSH Subject Heading; RCT = randomized controlled trial; CCT = controlled clinical trial]

Appendix 2. Search strategy for Embase (OVID)

1. For Children the following Emtree terms and text words were used:

1 infant/ or infancy/ or newborn/ or baby/ or child/ or preschool child/ or school child/ 
 2 adolescent/ or juvenile/ or boy/ or girl/ or puberty/ or prepuberty/ or pediatrics/ 
 3 primary school/ or high school/ or kindergarten/ or nursery school/ or school/ 
 4 or/34‐36 
 5 (infant$ or newborn$ or (new adj born$) or baby or baby$ or babies or neonate$ or perinat$ or postnat$).mp. 
 6 (child$ or (school adj child$) or schoolchild$ or (school adj age$) or schoolage$ or (pre adj school$) or preschool$).mp. 
 7 (kid or kids or toddler$ or adoles$ or teen$ or boy$ or girl$).mp. 
 8 (minors$ or (under adj ag$) or underage$ or juvenil$ or youth$).mp. 
 9 (puber$ or pubescen$ or prepubescen$ or prepubert$).mp. 
 10 (pediatric$ or paediatric$ or peadiatric$).mp. 
 11 (school or schools or (high adj school$) or highschool$ or (primary adj school$) or (nursery adj school$) or (elementary adj school) or (secondary adj school$) or kindergar$).mp. 
 12 or/5‐11 
 13 4 or 12

2. For Cancer the following Emtree terms and text words were used:

1 (cancer or cancers or cancer$).mp. 
 2 (oncology or oncolog$).mp. or exp oncology/ 
 3 (neoplasm or neoplasms or neoplasm$).mp. or exp neoplasm/ 
 4 (carcinoma or carcinom$).mp. or exp carcinoma/ 
 5 (tumor or tumour or tumor$ or tumour$ or tumors or tumours).mp. or exp tumor/ 
 6 (malignan$ or malignant).mp. 
 7 (hematooncological or hemato oncological or hemato‐oncological or hematologic neoplasms or hematolo$).mp. or exp hematologic malignancy/ 
 8 or/1‐7

3. For Childhood cancer the following Emtree terms and text words were used:

1 (leukemia or leukemi$ or leukaemi$ or acute lymphocytic leukemia).mp. 
 2 (AML or lymphoma or lymphom$ or hodgkin or hodgkin$ or T‐cell or B‐cell or non‐hodgkin).mp. 
 3 (sarcoma or sarcom$ or Ewing$ or osteosarcoma or osteosarcom$ or wilms tumor or wilms$).mp. 
 4 (nephroblastom$ or neuroblastoma or neuroblastom$ or rhabdomyosarcoma or rhabdomyosarcom$ or teratoma or teratom$ or hepatoma or hepatom$ or hepatoblastoma or hepatoblastom$).mp. 
 5 (PNET or medulloblastoma or medulloblastom$ or PNET$ or neuroectodermal tumors or primitive neuroectodermal tumor$ or retinoblastoma or retinoblastom$ or meningioma or meningiom$ or glioma or gliom$).mp. 
 6 (pediatric oncology or paediatric oncology).mp. 
 7 ((childhood adj cancer) or (childhood adj tumor) or (childhood adj tumors) or childhood malignancy or (childhood adj malignancies) or childhood neoplasm$).mp. 
 8 ((pediatric adj malignancy) or (pediatric adj malignancies) or (paediatric adj malignancy) or (paediatric adj malignancies)).mp. 
 9 ((brain adj tumor$) or (brain adj tumour$) or (brain adj neoplasms) or (brain adj cancer$) or brain neoplasm$).mp. 
 10 (central nervous system tumor$ or central nervous system neoplasm or central nervous system neoplasms or central nervous system tumour$).mp. 
 11 intracranial neoplasm$.mp. 
 12 LEUKEMIA/ or LYMPHOMA/ or brain tumor/ or central nervous system tumor/ or teratoma/ or sarcoma/ or osteosarcoma/ 
 13 nephroblastoma/ or neuroblastoma/ or rhabdomyosarcoma/ or hepatoblastoma/ or medulloblastoma/ or neuroectodermal tumor/ or retinoblastoma/ or meningioma/ or glioma/ or childhood cancer/ 
 14 or/1‐13

4. For Catheter the following Emtree terms and text words were used:

1 (broviac or port‐a‐cath or port acath or port a cath or hickman).mp. 
 2 (catheterisation central venous or tunnelled central venous catheter or TCVC or catheterization central venous or peripherally inserted central catheter or picc or central venous line or central venous device).mp. or exp central venous catheterization/ or exp central venous catheter/ 
 3 (central venous access device or CVAD or CVC or central venous catheter).mp. 
 4 (catheter or catheter$ or cathethers or catheterisation or catheterizations or catheterization$).mp. 
 5 exp intravenous catheter/ or exp catheter/ 
 6 exp catheterization/ or catheterization.mp. 
 7 (cannulation or cannulations or cannulation$ or cannula or cannulas or cannula$).mp. or exp cannulation/ or exp cannula/ 
 8 or/1‐7

5. For Infection the following Emtree terms and text words were used:

1 (infection or infections or infection$).mp. or exp Infection/ or communicable disease/ or (communicable disease or communicable diseases).mp. 
 2 (bacteremia or bacteremias or bacteremia$).mp. 
 3 exp Bacteremia/ 
 4 (sepsis or septic or septicemia or septicemi$ or specific bacterial infection).mp. 
 5 septic shock.mp. or exp septic shock/ 
 6 (toxic shock or toxic shock syndrome or toxic shock syndromes or endotoxic shock).mp. 
 7 exp sepsis/ or exp septicemia/ 
 8 (severe sepsis or blood poisoning).mp. 
 9 (catheter‐related infection or catheter related infections or catheter related infection$).mp. or exp catheter infection/ 
 10 (endotoxemia or endotoxemias or endotoxemia$).mp. or exp endotoxemia/ 
 11 (toxemia or toxemias or toxemia$).mp. or exp toxemia/ 
 12 (endotoxin or endotoxins or endotoxin$).mp. or exp endotoxins/ 
 13 (hemorrhagic septicemia or haemorrhagic septicaemia or hemorrhagic septicaemia or haemorrhagic septicemia or hemorrhagic bacteremia or haemorrhagic bacteremia).mp. 
 14 exp hemorrhagic septicemia/ 
 15 or/1‐14

6. For RCT/CCT the following Emtree terms and text words were used:

1 Randomized Controlled Trial/ 
 2 Controlled Clinical Trial/ 
 3 randomized.ti,ab. 
 4 placebo.ti,ab. 
 5 randomly.ti,ab. 
 6 trial.ti,ab. 
 7 groups.ti,ab. 
 8 drug therapy.sh. 
 9 or/1‐8 
 10 Human/ 
 11 9 and 10

Combined search: 1 and (2 or 3) and 4 and 5 and 6

[mp = title, abstract, subject headings, heading word, drug trade name, original title, device manufacturer, drug manufacturer name; / = Emtree term; $ = 1 or more characters; ti,ab = title or abstract; sh = subheading: RCT = randomized controlled trial; CCT = controlled clinical trial]

Appendix 3. Search strategy for Cochrane Central Register of Controlled Trials (CENTRAL)

1. For Children the following text words were used:

(infant OR infan* OR newborn OR newborn* OR new‐born* OR baby OR baby* OR babies OR neonat* OR perinat* OR postnat* OR child OR child* OR schoolchild* OR schoolchild OR school child OR school child* OR kid OR kids OR toddler* OR adolescent OR adoles* OR teen* OR boy* OR girl* OR minors OR minors* OR underag* OR under ag* OR juvenil* OR youth* OR kindergar* OR puberty OR puber* OR pubescen* OR prepubescen* OR prepuberty* OR pediatrics OR pediatric* OR paediatric* OR peadiatric* OR schools OR nursery school* OR preschool* OR pre school* OR primary school* OR secondary school* OR elementary school* OR elementary school OR high school* OR highschool* OR school age OR schoolage OR school age* OR schoolage* OR infancy)

2. For Cancer the following text words were used:

(cancer OR oncology OR oncolog* OR neoplasms OR neoplas* OR carcinoma OR carcinom* OR tumor OR tumour OR tumor* OR tumour* OR cancer* OR malignan* OR hematooncological OR hemato oncological OR hemato‐oncological OR hematologic neoplasms OR hematolo*)

3. For Childhood cancer the following text words were used:

(leukemia OR leukemi* OR leukaemi* OR childhood ALL OR AML OR lymphoma OR lymphom* OR hodgkin OR hodgkin* OR T‐cell OR B‐cell OR non‐hodgkin OR sarcoma OR sarcom* OR Ewing* OR osteosarcoma OR osteosarcom* OR wilms tumor OR wilms* OR nephroblastom* OR neuroblastoma OR neuroblastom* OR rhabdomyosarcoma OR rhabdomyosarcom* OR teratoma OR teratom* OR hepatoma OR hepatom* OR hepatoblastoma OR hepatoblastom* OR PNET OR medulloblastoma OR medulloblastom* OR PNET* OR neuroectodermal tumors, primitive OR retinoblastoma OR retinoblastom* OR meningioma OR meningiom* OR glioma OR gliom* OR pediatric oncology OR paediatric oncology OR childhood cancer OR childhood tumor OR childhood tumors OR brain tumor* OR brain tumour* OR brain neoplasms OR central nervous system neoplasm OR central nervous system neoplasms OR central nervous system tumor* OR central nervous system tumour* OR brain cancer* OR brain neoplasm* OR intracranial neoplasm*)

4. For Catheter the following text words were used:

(broviac OR port‐a‐catch OR port acatch OR port a catch OR portacatch OR port OR port‐a‐cat OR portacat OR hickman OR catheterisation central venous OR tunnelled central venous catheter OR TCVC OR catheterization central venous OR peripherally inserted central catheter OR picc OR central venous line or central venous device OR central venous access device OR CVAD OR CVC or central venous catheter OR catheter OR catheterisation OR catheterization OR cannula OR cannulation)

5. For Infection the following text words were used:

(infection OR infections OR communicable disease OR communicable diseases OR bacteremia OR bacteremi* OR sepsis OR septic OR septic* OR septicemia OR septic shock OR specific bacterial infection OR toxic shock OR toxic shock syndrome OR endotoxic shock OR severe sepsis OR blood poisoning OR catheter‐related infection OR endotoxemia OR toxemia OR endotoxin OR hemorrhagic septicemia OR haemorrhagic septicaemia OR hemorrhagic septicaemia OR haemorrhagic septicemia OR hemorrhagic bacteremia OR haemorrhagic bacteremia)

Combined search: 1 and (2 or 3) and 4 and 5

[* = 1 or more characters]

The search will be performed in title, abstract or keywords

Appendix 4. Search strategy for CINAHL Plus (EBSCO)

1. For Children the following CINAHL headings and text words were used:

1 infant* or infancy* 
 2 newborn* or new W1 born 
 3 baby* or babies 
 4 neonat* or perinat* or postnat* 
 5 child* 
 6 schoolchild* or school W1 child*

7 kid or kids 
 8 toddler* 
 9 adolesc* 
 10 teen or teens or teenager* 
 11 boy* 
 12 girl* 
 13 minors* 
 14 underag* or under W1 age* 
 15 juvenile* 
 16 youth* 
 17 kindergar* 
 18 puber* or pubescen* 
 19 prepuberscen* or prepuberty* 
 20 pediatric* or paediatric* 
 21 (schools or preschool*) or pre W1 school* 
 22 primary W1 school* or secondary W1 school*

23 elementary W1 school* 
 24 high W1 school* or highschool* 
 25 nursery W1 school* 
 26 school W1 age* or schoolage* 
 27 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10 or 11 or 12 or 13 or 14 or 15 or 16 or 17 or 18 or 19 or 20 or 21 or 22 or 23 or 24 or 25 or 26

2. For Cancer the following CINAHL headings and text words were used:

1 cancer* 
 2 oncolog* 
 3 neoplasm* or (MH "Neoplasms+") 
 4 (MH "Carcinoma+") or carcinom* 
 5 tumour* or tumor* 
 6 malignan* 
 7 (hemato oncological OR hematooncological OR hemato‐oncological OR hematolo*) OR hematologic neoplasms OR (MH "Hematologic Neoplasms+")  
 8 1 or 2 or 3 or 4 or 5 or 6 or 7

3. For Childhood cancer the following CINAHL headings and text words were used:

1 (leukemi* or leukaemi*) or acute lymphocytic leukemia 
 2 AML or lymphom* or hodgkin* or T‐cell or B‐cell or non‐hodgkin 
 3 sarcom* or Ewing* or osteosarcom* or wilms* 
 4 nephroblastom* or neuroblastom* or rhabdomyosarcom* or teratom* or hepatom* or hepatoblastom* 
 5 medulloblastom* or PNET* or neuroectodermal tumors or primitive neuroectodermal tumor* or retinoblastom* or meningiom* or gliom* 
 6 pediatric oncology or paediatric oncology 
 7 childhood W1 cancer or childhood W1 tumor* or childhood malignancy or childhood W1 malignancies or childhood neoplasm* 
 8 pediatric W1 malignancy or pediatric W1 malignancies or paediatric W1 malignancy or paediatric W1 malignancies 
 9 brain W1 tumor* or brain W1 tumour* or brain W1 neoplasms or brain W1 cancer* 
 10 central nervous system tumor* or central nervous system neoplasm* or central nervous system tumour* 
 11 intracranial neoplasm*   

12 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10 or 11

4. For Catheter the following CINAHL headings and text words were used:

1 broviac OR port‐a‐cath OR port acath OR port a cath OR hickman

2 catheterisation central venous OR tunnelled central venous catheter OR TCVC OR catheterization central venous OR central venous line OR central venous device

3 central venous access device or CVAD or CVC or central venous catheter 
 4 catheter* or catheterisation or catheterization* 
 5 (MH "Catheters+")   
 6 (MH "Catheterization+")  

7 cannula* 
 8 1 or 2 or 3 or 4 or 5 or 6 or 7

5. For Infection the following CINAHL headings and text words were used:

1 infection* or (MH "Infection+") or communicable disease* 
 2 bacteremia* 
 3 (MH "Bacteremia") 
 4 (sepsis or septic or septicemia*) or specific bacterial infection 
 5 septic shock or (MH "Shock, Septic+") 
 6 toxic shock or toxic shock syndrome* or endotoxic shock 
 7 (MH "Sepsis+") 
 8 severe sepsis or blood poisoning 
 9 catheter‐related infection or catheter related infection* 
 10 endotoxemia* or (MH "Endotoxemia")  
 11 toxemia* or (MH "Toxemia+") 
 12 endotoxin* or (MH "Endotoxins") 
 13 hemorrhagic septicemia or Haemorrhagic Septicaemia or Hemorrhagic Septicaemia or Haemorrhagic Septicemia or Hemorrhagic Bacteremia or Haemorrhagic Bacteremia

14 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10 or 11 or 12 or 13

6. For RCT/CCT the following CINAHL headings and text words were used:

1 "randomized controlled trial"

2 (MH "Clinical Trials")

3 (MH "Cochrane Library")

4 "controlled clinical trial"

5 (MH "Clinical Trial Registry")

6 TI randomized or AB randomized

7 TI placebo or AB placebo

8 "drug therapy" or (MH "Drug Therapy")

9 TI randomly or AB randomly

10 TI trial or AB trial

11 TI groups or AB groups

12 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10 or 11

Combined search: 1 and (2 or 3) and 4 and 5 and 6

[RCT = randomized controlled trial; CCT = controlled clinical trial]

Data and analyses

Comparison 1. Urokinase (with or without heparin) versus heparin.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Catheter‐Associated Infection	2		Rate Ratio (Fixed, 95% CI)	0.72 [0.12, 4.41]

1.1. Analysis.

Comparison 1 Urokinase (with or without heparin) versus heparin, Outcome 1 Catheter‐Associated Infection.

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Aquino 2002.

Methods	Double blind randomisation
Participants	103 newly diagnosed patients with haematological and non‐haematological malignancies enrolled between August 1994 and July 1998; ages 1 to 21 years (mean age 9.2 ± 0.8 years in study group and 8.6 ± 0.8 years in control group); implanted catheters only; only patients whose catheter placement was anticipated to be > 6 months were enrolled.
Interventions	Intervention: Monthly catheter flushes with 3ml of Urokinase‐Heparin which had 5000 IU of urokinase (Abbokinase) Control: Monthly catheter flushes with 3ml of Heparin which had 300 units of heparin Only those participants who received catheter flushes on at least six occasions at one month intervals were included in the analysis.
Outcomes	Bacteraemia (CAI); premature catheter removal for infection Outcomes assessed on 74/103 patients as 7 did not receive first flush within 1 month of study entry and 22 received fewer than the mandated 6 flushes
Notes	Funding: Abbott Laboratories, National Institute of Health, Children's Cancer Fund for Dallas and Wipe Out Kids' Cancer Premature closure of trial due to removal of study drug from market. Planned target of enrolling 120 participants not met.
Risk of bias
Bias	Authors' judgement	Support for judgement
Adequate sequence generation?	Low risk	Quote: "Patients were randomised using a single 1:1 permuted block of length 120"
Allocation concealment?	Unclear risk	Allocation concealment not described
Blinding? All outcomes	Low risk	Quote: "The pharmacist was the only unblinded person involved in this study" Comment ‐ Probably done
Incomplete outcome data addressed? All outcomes	Unclear risk	29/103 of participants enrolled in the study excluded from analysis. Insufficient reporting of exclusions within control and study group to permit judgement. ‘As‐treated’ analysis done
Free of selective reporting?	Unclear risk	The study did not report on CRBSI or pocket infection
Free of other bias?	Unclear risk	Trial stopped early when the study drug urokinase was removed from the market by US Food and Drug Administration. However, bias for study drug is unlikely as the study reported no difference in outcomes between control and study groups

Benhamou 2002.

Methods	Randomisation. Blinding of participants and personnel not feasible. Blinding of outcome assessors not described
Participants	113 children with malignancy (mainly non‐haematological) who were candidates for high‐dose chemo/radiotherapy followed by bone marrow transplantation enrolled between July 1990 and April 1993; median age 5 years (range 1 to 22 years) in study group and 7 years (range 2 to 19 years) in control group; external catheters only
Interventions	Intervention: Catheter dressing changed every 15 days Control: Catheter dressing changed every 4 days Three types of dressings used depending on the condition of the underlying skin (grade 0/1 cutaneous toxicity ‐ tegaderm; grade 2/3 ‐ mefix type; grade 4 ‐ sterile gauze and tape)
Outcomes	Skin toxicity at dressing site; local pain; bacteraemia; premature catheter removal for infection
Notes	Only 17% of those in the intervention group compared to 76% of those in the control group had dressing changed at pre‐specified frequency. This was mainly due to loose dressing in the intervention group necessitating earlier dressing changes.
Risk of bias
Bias	Authors' judgement	Support for judgement
Adequate sequence generation?	Low risk	Quote: "A computer generated list was used" "Randomisation was stratified by the type of HDC (with or without busulfan)"
Allocation concealment?	Unclear risk	Allocation concealment not described
Blinding? All outcomes	Unclear risk	Blinding of participants and personnel not feasible. Blinding of outcome assessors not described
Incomplete outcome data addressed? All outcomes	Low risk	Quote: "One patient relapsed after randomisation and did not receive HDC (15‐day Group)" Comment ‐ Outcome reported for 112/113 randomised patients. Missing outcome data are unlikely to have a clinically meaningful impact on the intervention effect estimate.
Free of selective reporting?	Unclear risk	The study did not report on CRBSI or CAI
Free of other bias?	High risk	The frequency of dressing affects cutaneous toxicity which in turn affects material used for dressing. So any effect attributed to dressing frequency might actually be confounding for dressing material

Dillon 2004.

Methods	Randomisation stratified by type of CVC
Participants	577 patients with haematological and non‐haematological malignancies enrolled between July 1997 and December 1998; ages 3 months to 21 years (> 95% were < 18 years age); median age 4.5 years; implanted and external catheters
Interventions	Intervention: CVC flushed every two weeks with urokinase (Abbokinase 5000 IU/ml) in volume sufficient to fill the entire catheter Control: CVC flushed every two weeks with heparin (100 units/ml) in volume sufficient to fill the entire catheter The drug was required to remain in the catheter lumen for a minimum of 1 hour and a maximum of 14 days
Outcomes	Complete catheter occlusion; partial catheter occlusion; catheter‐related infection; time to first catheter occlusion; time to first catheter‐related infection
Notes	Funding: In part by Abbott Laboratories Premature closure of trial due to removal of study drug from market. Planned target of enrolling 680 participants not met.
Risk of bias
Bias	Authors' judgement	Support for judgement
Adequate sequence generation?	Unclear risk	Sequence generation not described
Allocation concealment?	Unclear risk	Allocation concealment not described
Blinding? All outcomes	Unclear risk	Blinding not described
Incomplete outcome data addressed? All outcomes	Unclear risk	8/577 of participants enrolled in the study excluded from analysis. Insufficient reporting of exclusions within control and study group to permit judgement. ‘As‐treated’ analysis done.
Free of selective reporting?	High risk	The study did not report on CRBSI. Also outcomes of interest (premature catheter removal for infection) are reported incompletely.
Free of other bias?	Unclear risk	Trial stopped early when the study drug urokinase was removed from the market by US Food and Drug Administration.

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Freiberger 1991	Earlier published abstract of Freiberger 1992 which was excluded for reasons mentioned below.
Freiberger 1992	The study compared different antiseptics used to clean the skin as well as different dressings used to cover the exit site. No information available on number of participants randomised to each arm. The study also did not report any outcomes of interest.
Henneberg 1996	This study compared non‐tunnelled catheters with tunnelled catheters.
Hinds 1991	The study randomised blood sampling from CVC and reinfusion of blood following the sampling in the usual clean way versus an exaggerated unclean way. It did not report any outcomes of interest.
Jones 2001	Earlier published abstract of Dillon 2004 which has been included in this review.
van Eys 1982	In this study children with malignancies who were not malnourished were randomised to receive intravenous hyperalimentation or regular diet. Although the study looked at catheter‐related infections as an outcome, the intervention was not prophylactic.
Wiernikowski 1991	The study compared flushing CVC with bacteriostatic saline (sterile saline mixed with 1% benzyl alcohol) versus sterile saline alone. The outcomes were bacterial colonisation and clinical sepsis. It did not report outcomes of interest.

Characteristics of ongoing studies [ordered by study ID]

Handrup 2008.

Trial name or title	A randomised study of Taurolock for the locking of tunnelled central venous catheters in children with malignant diseases
Methods	Randomised
Participants	Children aged 0 to 17 years with malignant disease requiring a tunnelled CVC
Interventions	When not in use the children's tunnelled CVC are locked with the liquid Taurolock or heparin
Outcomes	Primary Outcome Measures: Number of CRBSI/1000 CVC days in the Taurolock group versus the heparin group. Number of CVCs removed in the Taurolock group versus the heparin group. Secondary Outcome Measures: Biofilm formation in the CVCs treated with Taurolock compared with the biofilm formation in the CVCs treated with heparin.
Starting date	April 2008
Contact information	Mette M Handrup, MD. +45 8949 6749. Handrup@ki.au.dk
Notes	This study is currently recruiting participants. Outcomes are expected in November 2010.

Martin 2008.

Trial name or title	Ethanol lock therapy for the prevention of catheter‐related blood stream infections
Methods	Double‐blind randomised
Participants	Children aged 6 months to 21 years, central venous access and a history of three or more catheter‐related blood stream infections in the prior 6 months
Interventions	Experimental arm ‐ 25% ethanol lock, Control arm ‐ heparin lock
Outcomes	Number of episodes of catheter‐related blood stream infections
Starting date	August 2008
Contact information	Judith M Martin, MD. University of Pittsburgh
Notes	This study is currently recruiting participants. Outcomes are expected in December 2010.

Wetering 2008.

Trial name or title	Ethanol lock solution for the prevention of tunnelled central venous catheter infections in paediatric oncology patients, a randomised controlled trial
Methods	Double‐blind randomised
Participants	Paediatric oncology patients between 1 and 18 years of age with a newly inserted tunnelled central venous catheter (both internal and external devices)
Interventions	After insertion of the catheter an ethanol (70%)‐lock solution will be administered (3 ml) for a duration of 2 hours, once weekly or longer if the catheter is not locked in between. The control group will be locked with the standard heparin (100U/ml) solution (3 ml).
Outcomes	Primary outcome measures ‐ First catheter‐related bacteraemia, death of the patient, or removal of the catheter, whatever comes first. Secondary outcome measures ‐ fever, antibiotic use, days of hospital admission and thrombosis.
Starting date	October 2007
Contact information	Dr MD Wetering. +31 (0)20 5663050. m.d.vandewetering@amc.uva.nl
Notes	This study is currently recruiting participants. Outcomes are expected in October 2010.

Differences between protocol and review

The method for measurement of dichotomous and continuous outcomes had been specified in the protocol but not for count of rare events e.g. catheter infections per 1000 CVC days. This has been added to the methods section of the review.

Contributions of authors

RS Arora ‐ Conceiving, designing and co‐ordinating the review; designing search strategies and undertaking searches; searching grey literature; screening initial search and then filtering based on inclusion criteria; writing to authors of papers for additional information; quality assessment and data extraction; data analysis and interpretation; writing protocol and review. He will act as the guarantor of the review.

R Roberts ‐ Screening initial search and then filtering based on inclusion criteria; quality assessment and data extraction

TOB Eden ‐ Data analysis and interpretation; writing protocol and review; providing a clinical perspective; providing general advice on the review

B Pizer ‐ Data analysis and interpretation; writing protocol and review; providing a clinical perspective; providing general advice on the review

Sources of support

Internal sources

• No sources of support supplied

External sources

• Christie Hospital NHS Foundation Trust and Teenage Cancer Trust, UK.

  RS Arora is funded by a grant from the Paediatric Endowment Fund Christie Hospital NHS Foundation Trust and from the Teenage Cancer Trust

• Teenage Cancer Trust, UK.

  T Eden holds a programme grant from the Teenage Cancer Trust

Declarations of interest

None known

New