Low molecular weight heparin for prevention of central venous catheter‐related thrombosis in children

Abstract

Background

The prevalence of children diagnosed with thrombotic events has been increasing in the last decades. The most common thrombosis risk factor in neonates, infants and children is the placement of a central venous catheter (CVC). It is unknown if anticoagulation prophylaxis with low molecular weight heparin (LMWH) decreases CVC‐related thrombosis in children. This is an update of the Cochrane Review published in 2014.

Objectives

To determine the effect of LMWH prophylaxis on the incidence of CVC‐related thrombosis and major and minor bleeding complications in children. Further objectives were to determine the effect of LMWH on occlusion of CVCs, number of days of CVC patency, episodes of catheter‐related bloodstream infection (CRBSI), other side effects of LMWH (allergic reactions, abnormal coagulation profile, heparin‐induced thrombocytopaenia and osteoporosis) and mortality during therapy.

Search methods

The Cochrane Vascular Information Specialist searched the Cochrane Vascular Specialised Register, CENTRAL, MEDLINE, Embase and CINAHL databases and World Health Organization International Clinical Trials Registry Platform and ClinicalTrials.gov trials registers to 7 May 2019. We undertook reference checking of identified trials to identify additional studies.

Selection criteria

We included randomised controlled trials (RCTs) and quasi‐randomised trials comparing LMWH to no prophylaxis (placebo or no treatment), or low‐dose unfractionated heparin (UFH) either as continuous infusion or flushes (low‐dose UFH aims to ensure the patency of the central line but has no systemic anticoagulation activity), given to prevent CVC‐related thrombotic events in children. We selected studies conducted in children aged 0 to 18 years.

Data collection and analysis

Two review authors independently identified eligible studies, which were assessed for study methodology including bias, and extracted unadjusted data where available. In the data analysis step, all outcomes were analysed as binary or dichotomous outcomes. The effects of interventions were summarised with risk ratios (RR) and their respective 95% confidence intervals (CI). We assessed the certainty of evidence for each outcome using the GRADE approach.

Main results

One additional study was included for this update bringing the total to two included studies (with 1135 participants). Both studies were open‐label RCTs comparing LMWH with low‐dose UFH to prevent CVC‐related thrombosis in children. We identified no studies comparing LMWH with placebo or no treatment. Meta‐analysis found insufficient evidence of an effect of LMWH prophylaxis in reducing the incidence of CVC‐related thrombosis in children with CVC, compared to low‐dose UFH (RR 0.68, 95% CI 0.27 to 1.75; 2 studies; 787 participants; low‐certainty evidence). One study (158 participants) reported symptomatic and asymptomatic CVC‐related thrombosis separately and detected no evidence of a difference between LMWH and low‐dose UFH (RR 1.03, 95% CI 0.21 to 4.93; low‐certainty evidence; RR 1.17, 95% CI 0.45 to 3.08; low‐certainty evidence; for symptomatic and asymptomatic participants respectively). There was insufficient evidence to determine whether LMWH impacts the risk of major bleeding (RR 0.27, 95% CI 0.05 to 1.67; 2 studies; 813 participants; low‐certainty evidence); or minor bleeding. One study reported minor bleeding in 53.3% of participants in the LMWH arm and in 44.7% of participants in the low‐dose UFH arm (RR 1.20, 95% CI 0.91 to 1.58; 1 study; 158 participants; very low‐certainty evidence), and the other study reported no minor bleeding in either group (RR: not estimable). Mortality during the study period was reported in one study, where two deaths occurred during the study period. Both were unrelated to thrombotic events and occurred in the low‐dose UFH arm. The second study did not report mortality during therapy per arm but showed similar 5‐year overall survival (low‐certainty evidence). No additional adverse effects were reported. Other pre‐specified outcomes (including CVC occlusion, patency and CRBSI) were not reported.

Authors' conclusions

Pooling data from two RCTs did not provide evidence to support the use of prophylactic LWMH for preventing CVC‐related thrombosis in children (low‐certainty evidence). Evidence was also insufficient to confirm or exclude a difference in the incidence of major and minor bleeding complications in the LMWH prophylaxis group compared to low‐dose UFH (low and very low certainty respectively). No evidence of a clear difference in overall mortality was seen. Studies did not report on the outcomes catheter occlusion, days of catheter patency, episodes of CRBSI and other side effects of LMWH (allergic reactions, abnormal coagulation profile, heparin‐induced thrombocytopaenia and osteoporosis). The certainty of the evidence was downgraded due to risk of bias of the included studies, imprecision and inconsistency, preventing conclusions in regards to the efficacy of LMWH prophylaxis to prevent CVC‐related thrombosis in children.

Plain language summary

Do blood thinners prevent blood clots in children who are treated using central lines?

Background

Central lines are thin flexible tubes that are inserted into a person’s vein to provide medical care. They are used to deliver medicines, fluids or nutrients (to feed the patient) directly into the bloodstream. However, they can cause the blood to clot (form into a small lump). This prevents blood from flowing normally in the blood vessels. Blood clots can cause symptoms such as pain or swelling, but can also happen without any symptoms. They can lead to serious health problems and death, if they move and become stuck in major veins or in the lungs.

Since the 1990s, blood clots in children in general have become more common, notably due to the rise in blood clots in children treated with central lines. We reviewed the existing research to find out if giving children a blood thinning medicine called low molecular weight heparin (LMWH) protects against blood clots. We also wanted to know whether LMWH increases the risk of minor or severe bleeding, death or other unwanted (adverse) side effects.

What did we find?

We searched the scientific literature in May 2019. We found two studies (one more than when we last searched in 2014). The studies compared what happens in children (aged up to 18 years old) treated using central lines who received LMWH and those who did not. A total of 1135 children were followed for between 30 and 64 days. During that time, the number of blood clots, bleeding episodes and deaths were recorded. One study also looked at the number of deaths five years after treatment.

The two studies we found did not allow us to determine with certainty whether or not LMWH protects children with central lines from getting blood clots (with or without symptoms).

The studies also did not allow us to determine whether children on LMWH were more likely to experience bleeding (minor or severe), or whether LMWH increases or reduces the risk of death.

The two studies did not report any additional adverse effects caused by LMWH.

Certainty of the evidence

We judged the certainty of the evidence to be low for blood clots, low for major bleeding, very low for minor bleeding, and low for mortality. The certainty of the evidence was low or very low because:

⦁ there were differences in the way blood clots, bleeding and mortality were measured between studies;

⦁ there were few studies and events; and

⦁ the researchers and children in the studies knew who was receiving LMWH and who was not, which can influence results.

It is likely that future studies will have an important impact on our understanding of the role of LMWH to prevent blood clots and its side effects.

Conclusion

We need more studies on whether LMWH prevents blood clots in children who are treated using central venous lines.

Summary of findings

Summary of findings 1. Does low molecular weight heparin prevent central venous catheter‐related thrombosis compared to control in children?

Low molecular weight heparin (LMWH) for prevention of central venous catheter (CVC)‐related thrombosis in children
Patient or population: children with a CVC Setting: paediatric tertiary care hospital Intervention: LMWH Comparison: controla
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with control	Risk with LMWH
CVC‐related thrombosis (follow up ranged from 30 to 64 days)	Study population		RR 0.68 (0.27 to 1.75)	787 (2 RCTs)	⊕⊕⊝⊝ LOW b
	89 per 1000	58 per 1000 (24 to 156)
Symptomatic CVC‐related thrombosis (assessed with venography; follow up ranged from 30 to 44 days)	Study population		RR 1.03 (0.21 to 4.93)	158 (1 RCT)	⊕⊕⊝⊝ LOW c
	38 per 1000	39 per 1000 (8 to 185)
Asymptomatic CVC‐related thrombosis (assessed with venography; follow up ranged from 30 to 44 days)	Study population		RR 1.17 (0.45 to 3.08)	158 (1 RCT)	⊕⊕⊝⊝ LOW d
	88 per 1000	102 per 1000 (39 to 305)
Major bleeding (assessed with clinical findings; follow up ranged from 30 days to 64 days)	Study population		RR 0.27 (0.05 to 1.67)	813 (2 RCTs)	⊕⊕⊝⊝ LOW e
	12 per 1000	3 per 1000 (1 to 21)
Minor bleeding (assessed with clinical findings; follow up ranged from 30 to 64 days)	Study population		‐	813 (2 RCTs)	⊕⊝⊝⊝ VERY LOW f	Massicotte 2003 reported minor bleeding in 53.3% of patients in the reviparin arm and in 44.7% of patients in the low‐dose UFH arm (RR 1.20, 95% CI 0.91 to 1.58). Greiner 2018 reported no minor bleeding in either arm.
	see comment
Mortality during study period (follow up ranged from 30 to 64 days)	Study population		‐	813 (2 RCTs)	⊕⊕⊝⊝ LOW g	One study reported two deaths during the study period. Both were unrelated to VTE and occurred in the low‐dose UFH arm (Massicotte 2003). One study, including paediatric patients with acute lymphoblastic leukaemia, shows similar 5‐year mortality (Greiner 2018).
	see comment
*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; CVC: central venous catheter; LMWH: low molecular weight heparin; RR: risk ratio; UFH: unfractionated heparin
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.

^a Control: In Greiner 2018 the control group received low‐dose UFH and in Massicotte 2003 the control group received low‐dose UFH flushes
^b Downgraded two steps because of serious risk of bias and inconsistency. Both studies were open‐label trials with no blinding of intervention allocation. In Greiner 2018, substantial deviation from the intended intervention might affect results. Both studies reported differing effects of LMWH
^c Downgraded two steps because of serious risk of bias and imprecision. Underpowered study; Massicotte 2003 was closed prematurely due to slow recruitment rate
^d Downgraded two steps because of serious risk of bias and imprecision. Underpowered study; Massicotte 2003 was closed prematurely due to slow recruitment rate
^e Downgraded two steps because of serious risk of bias and imprecision (attributable to small number of events)
^f Downgraded three steps because of serious risk of bias and imprecision (attributable to small number of events) and inconsistency (attributable to important differences in outcome definition)
^g Downgraded two steps because of inconsistency and imprecision (attributable to small number of events)

Background

Description of the condition

Central venous catheters (CVC) are widely used for a variety of indications such as monitoring of haemodynamic status and the administration of parenteral nutrition, blood products and chemotherapeutic agents, or the infusion of other fluids (Klerk 2003). They are usually divided between non‐tunnelled CVC (centrally‐ and peripherally‐inserted), tunnelled CVC and totally‐implanted CVCs. Although advantageous, and at times life saving, the use of CVCs is associated with mechanical and infectious complications. The major mechanical complications observed are occlusion and thrombosis of the catheter (Bagnall‐Reeb 1990; Bona 2003; Finkelstein 2004; Klerk 2003; Male 2003; Steele 2001; Vidler 1999). The incidence of catheter‐related thrombosis is associated with different factors. Several retrospective and prospective clinical studies have reported incidences of catheter‐related thrombosis ranging between 4% and 66% (Andrew 1995; Fratino 2005; Krafte‐Jacobs 1995; Monagle 2012; Vidal 2014). More recent studies continue to show the rise of thrombotic events in children, particularly due to the higher incidence of CVC‐related thrombosis (Raffini 2009). Several factors have been implicated in causing CVC‐related thrombosis. These include underlying disease, for example, malignancies (Harlev 2010; Meister 2008; Mitchell 2010); the nature of the substances administered, such as chemotherapeutic agents or hyperosmolar solutions that can cause damage to the vascular endothelium; the site of the catheter, with a higher risk in femoral catheters than those inserted in jugular or subclavian veins (Klerk 2003); and the type of catheter (Carde 1989; Mclean 2005). The thrombosis related to CVCs can lead to pulmonary embolism (PE), sepsis, stroke or post‐thrombotic syndrome (Anton 2001), and may require long‐term anticoagulants.

Description of the intervention

In light of the potential morbidity associated with thrombosis‐related complications, thrombosis has become a very clinically relevant complication (Raffini 2011) and anticoagulation prophylaxis has increasingly been considered. Unfractionated heparin (UFH), administered intermittently or as a continuous infusion, has been evaluated for the prevention of thrombotic complications in a few paediatric studies (Randolph 1998; Shah 2008). However, the benefits of using UFH have to be weighted against the risks of allergic reactions, heparin‐induced thrombocytopaenia and bleeding complications secondary to errors in the dosage. On the other hand, low molecular weight heparin (LMWH) has the advantages of having a longer half‐life, increased bioavailability, ease of administration and capability for careful monitoring in children (Anton 2001; Monagle 2012). In the absence of convincing prospective paediatric studies, recent guidelines do not support primary prophylaxis for children with CVCs (Chalmers 2011; Monagle 2012).

How the intervention might work

LMWH, an anticoagulant drug from the family of heparinoids that is available in several formulations, has been successfully used as a prophylactic regimen for CVCs in adults with cancer (Kahale 2018). For example, at a dose of 2500 IU daily given subcutaneously, the LMWH dalteparin proved to be effective and safe in preventing the development of CVC‐related deep venous thrombosis (DVT) in adult cancer patients. The incidence of venography‐proven thrombosis was 6% in the prophylaxis group versus 62% in the placebo group (relative risk (RR) of thrombosis with prophylaxis: 0.15, 95% confidence interval (CI) 0.02 to 0.95; P = 0.002) and no bleeding complications developed (Monreal 1996).

Why it is important to do this review

The aim of this review was to evaluate the efficacy and safety of LMWH for the prevention of CVC‐related thrombosis in children. A previous systematic review did not show a beneficial effect of LMWH to reduce CVC‐related thrombosis (Vidal 2014). We aimed to reflect the addition of recently conducted studies to aid decision making for healthcare professionals and patients. This is an update of the Cochrane Review published in 2014 (Brandao 2014).

Objectives

To determine the effect of LMWH prophylaxis on the incidence of CVC‐related thrombosis and major and minor bleeding complications in children. Further objectives were to determine the effect of LMWH on occlusion of CVCs, number of days of CVC patency, episodes of catheter‐related bloodstream infection (CRBSI), other side effects of LMWH (allergic reactions, abnormal coagulation profile, heparin‐induced thrombocytopaenia, and osteoporosis) and mortality during therapy.

Methods

Criteria for considering studies for this review

Types of studies

We included randomised controlled trials (RCTs) of LMWH in children with CVCs. Studies that used alternative methods of randomisation such as alternate days of the week, odd or even date of birth, or hospital number (quasi‐randomised trials) were eligible for inclusion. We did not include studies that used historical controls. For any studies in which the unit of randomisation was the catheter, we planned to contact the primary authors to obtain data for the first catheter after randomisation, but this was not required.

Types of participants

We included children (age one month to 18 years) who required CVCs during their course of treatment either while in hospital or after discharge. The definition of CVC encompassed single‐ and multi‐lumen non‐tunnelled catheters (centrally‐ and peripherally‐inserted central catheters), tunnelled catheters and totally‐implanted catheters, with or without heparin‐bonding or antimicrobial impregnation.

Types of interventions

We included trials comparing LMWH versus low‐dose UFH. The use of low‐dose UFH continuous infusion or flushes (intermittent injection of a small amount of UFH) is a common clinical practice to ensure the patency of the central line, despite the lack of evidence of its beneficial effect for the prevention of CVC occlusion in both children and adults (Bradford 2020; López‐Briz 2018). We also planned to include trials comparing LMWH versus no prophylactic heparin (placebo or no treatment), but this practice was not used in either of the included trials. LMWH, placebo or UFH must have been administered during the entire time the catheter was in place.

Types of outcome measures

We included trials that reported on one or more of the following outcomes amongst the randomised participants.

Primary outcomes

• CVC‐related thrombosis (along the length of, or at the tip of, the catheter) as determined by either colour duplex Doppler ultrasonography or contrast venography, with or without clinical suspicion (symptomatic or asymptomatic). This was determined as dichotomous data (yes or no). We intended to obtain the incidence of CVC‐related thrombosis (number of thrombosis by 1000 catheter‐days), if data were available. A CVC‐related thrombosis was considered symptomatic if clinical signs or symptoms of thrombosis were present, including but not restricted to oedema, pain or erythema (Lee 2006).

• Major and minor bleeding complications, defined as follows (Mitchell 2011):

  • Major bleeding: fatal bleeding, clinically overt bleeding associated with a decrease in haemoglobin of at least 20 g/L in a 24 hours period, bleeding that is retroperitoneal, pulmonary, intracranial, or otherwise involves the central nervous system; and bleeding that requires surgical intervention in an operating suite

  • Minor bleeding: any overt or macroscopic evidence of bleeding that does not fulfil the above criteria for major bleeding

Secondary outcomes

• Occlusion of the catheter. This was determined as dichotomous data (yes or no). The catheter occlusion could be partial (blood cannot be aspirated through the catheter but infusion is possible) or complete (neither aspiration nor infusion possible) (Baskin 2009). We also intended to obtain the incidence of catheter occlusion using the number of occlusions/1000 catheter‐days if data were available.

• Days of catheter patency (duration of patency of first catheter, in days).

• Episodes of catheter‐related bloodstream infection (CRBSI), defined as a bloodstream infection in a patient with an intravascular catheter with at least one positive blood culture and with clinical manifestations of infections (i.e., fever, chills, and/or hypotension) and no apparent source of infection, in addition to adequate microbiological evidence of catheter‐related infection (Mermel 2009).

• Other side effects of LMWH (allergic reactions, abnormal coagulation profile, heparin‐induced thrombocytopaenia, osteoporosis). For the purpose of this review, heparin‐induced thrombocytopaenia was defined as an intermediate or high clinical probability of heparin‐induced thrombocytopaenia using the 4Ts score coupled with a positive immunoassay or functional assay (Cuker 2018).

• Mortality during the period of therapy.

• Any other reported adverse outcomes (not pre‐specified).

Search methods for identification of studies

Electronic searches

The Cochrane Vascular Information Specialist conducted systematic searches of the following databases for randomised controlled trials and controlled clinical trials without language, publication year or publication status restrictions:

• the Cochrane Vascular Specialised Register via the Cochrane Register of Studies (CRS‐Web searched on 7 May 2019);

• the Cochrane Central Register of Controlled Trials (CENTRAL) Cochrane Register of Studies Online (CRSO 2019, 4);

• MEDLINE (Ovid MEDLINE® Epub Ahead of Print, In‐Process & Other Non‐Indexed Citations, Ovid MEDLINE® Daily and Ovid MEDLINE®) (searched from 1 January 2017 to 7 May 2019);

• Embase Ovid (searched from 1 January 2017 to 7 May 2019);

• CINAHL Ebsco (searched from 1 January 2017 to 7 May 2019);

• AMED Ovid (searched from 1 January 2017 to 7 May 2019).

The Information Specialist modelled search strategies for other databases on the search strategy designed for CENTRAL. Where appropriate, they were combined with adaptations of the highly sensitive search strategy designed by the Cochrane Collaboration for identifying RCTs and controlled clinical trials (as described in the Cochrane Handbook for Systematic Reviews of Interventions Chapter 6, Lefebvre 2011). Search strategies for major databases are provided in Appendix 1.

The Information Specialist searched the following trials registries on 7 May 2019:

• the World Health Organization International Clinical Trials Registry Platform (who.int/trialsearch);

• ClinicalTrials.gov (clinicaltrials.gov).

Searching other resources

In addition, we searched the reference lists of identified trials.

Data collection and analysis

Selection of studies

Two review authors (MPM, NA) independently assessed all published articles identified as potentially relevant by the literature search for inclusion in the review. Discrepancies were resolved by consensus, with consultation of a third review author (LRB or MLA) if necessary. In order to be included, the trial had to meet the following criteria:

• the study population were children (age one month to 18 years);

• the intervention was LMWH compared with placebo, no treatment or low‐dose UFH (flushes or continuous infusion) for CVCs;

• the study was a randomised or quasi‐randomised controlled trial;

• one or more primary or secondary outcome measures were reported.

Data extraction and management

Two review authors (MPM, NA) independently extracted data from the retrieved articles. We contacted the primary authors of any articles for which there was inadequate information, or where relevant data could not be abstracted. Discrepancies were resolved by consensus, with consultation of a third review author (LRB or MLA) if necessary.

Assessment of risk of bias in included studies

We assessed the risk of bias of the included studies using Cochrane's 'Risk of bias' tool (Higgins 2017). The risk of bias was classified as low, high or unclear for each of the following domains: bias in selection of participants onto the study, performance bias, detection bias, bias due to missing data, and bias in selection of the reported results. Two review authors independently performed this evaluation (MPM, NA), and discussed disagreements to reach a consensus.

Measures of treatment effect

For dichotomous outcomes, we calculated RRs and 95% CIs as a measure of treatment effect. For continuous scales of measurements, we planned to analyse using mean difference (MD) with a 95% CI, where the same scales were used.

Unit of analysis issues

The individual participant was the unit of analysis. There were no studies that included a cross‐over or cluster‐randomised design. If there are such studies in future updates, we will include and analyse them as appropriate. We planned to perform analysis separately for studies using CVC as the unit of analysis, but no included study used this approach.

Dealing with missing data

We contacted the corresponding authors via e‐mail up to three times in order to obtain missing data. In the meta‐analysis, we used a complete‐case analysis approach, where participants considered to have missing data were excluded from the analysis. This type of analysis assumes that data is missing at random. In the event of concerns that data appeared to be not missing at random and that a difference was found between groups, we planned to perform sensitivity analysis as described below.

Assessment of heterogeneity

We tested heterogeneity and variability between trials using the Chi‐square test for heterogeneity (with the significance threshold set at P < 0.10), as well as the I² statistic. I² values above 50% indicate the possibility of substantial heterogeneity. We evaluated meta‐analyses exceeding this threshold and inspected their individual studies to identify possible sources of heterogeneity.

Assessment of reporting biases

We planned to investigate publication bias using funnel plots if 10 or more studies were identified.

Data synthesis

Data were presented descriptively. We utilised Review Manager software to perform data analysis (Review Manager 2014). When we were able to pool data we used a random‐effects method based on the inverse variance approach under the assumption that the effects being estimated in the different studies are not identical due to the clinical and methodological heterogeneity of the included studies.

Subgroup analysis and investigation of heterogeneity

We planned to perform subgroup analyses according to the control intervention:

• LMWH versus placebo or no treatment;

• LMWH versus low‐dose UFH (flushes or continuous infusion).

Given the limited number of studies retrieved, we did not carry out subgroup analyses. If sufficient numbers of studies are available in the future, we will perform these.

Sensitivity analysis

We planned to perform a sensitivity analysis to assess the impact of risk of bias on the effects of the intervention by excluding studies with one or more domains at high risk of bias from the analysis.

To test the robustness of findings to bias arising from missing data, in the event of concerns that data appeared to be not missing at random, we also planned to perform sensitivity analysis, with and without inclusion of studies with a substantial drop‐out rate. We defined 'substantial' as a drop‐out rate equal to or greater than the event rate of the control group,

Given the limited number of studies retrieved, we did not carry out sensitivity analyses. If in future updates sufficient studies are included, we will carry out sensitivity analyses.

Summarising findings and assessment of certainty of the evidence

We included a 'Summary of findings' table in this review to present the most important findings, including the certainty of evidence for the outcomes considered to be most clinically relevant. These outcomes were chosen in accordance with a suggested core set of outcomes for trials on the prevention of thromboembolic events (Mitchell 2011), identified in the Core Outcome Measures in Effectiveness Trials (COMET) database. Certainty of the evidence was determined using the GRADE approach, which considers the overall risk of bias of the included studies, the directness of the evidence, inconsistency within the results, precision of the effect estimate and the risk of publication bias (Balshem 2011; Guyatt 2008). The outcomes included in the 'Summary of findings' table were CVC‐related thrombosis; CVC‐related symptomatic thrombosis; CVC‐related asymptomatic thrombosis; major bleeding; minor bleeding and mortality. We created the 'Summary of findings' table using GRADEpro GDT 2015 software. See Table 1.

Results

Description of studies

Results of the search

See Figure 1.

1.

Study flow diagram.

In this update, we included and assessed one new study (Greiner 2018). We excluded 15 studies (Barekatain 2018; Cefali 2017; Del Principe 2012; Diamond 2018; Korkemaz 2017; Kurtkoti 2016; Lamontagne 2014; Lavau‐Denes 2013; Li 2016; Lutkin 2018; McDonald 1984; Pardun 2017; Rocha 2006; Thirunavukkarasu 2016; Wang 2019). We identified two ongoing studies (Klassen 2017; NCT03003390).

Included studies

We identified one new study for this update (Greiner 2018). Our review thus includes two studies with a total of 1135 participants enrolled (Greiner 2018; Massicotte 2003). Both studies compared LMWH (reviparin or enoxaparin) with low‐dose unfractionated heparin (UFH). We identified no studies comparing LMWH with placebo or no treatment. See Characteristics of included studies.

Massicotte 2003 included 186 children randomised to receive either LMWH (reviparin) prophylaxis (mean age 6.1 yr, standard deviation (SD) 5.2 yr) or standard care, that is, low‐dose UFH < 3 IU/kg body weight/hour (mean age 6.4 yr, SD 5.0 yr) for prevention of CVC‐related thrombosis (symptomatic or asymptomatic, or both) or thrombosis‐related death. Additionally, the most common and relevant complication related to LMWH use, bleeding, was analysed between study arms. The study's inclusion criteria for participants were neonates, infants and children admitted to the hospital with a newly placed CVC that were subsequently screened by examination or by imaging at CVC removal. During CVC placement, patients randomised to the intervention arm received an age‐appropriate LMWH dose to prevent thrombosis development. Patients randomised to the standard care arm received low‐dose heparin flushes or infusions (not equivalent to systemic anticoagulation). The outcomes reported were the rate of CVC‐related thrombosis, symptomatic and asymptomatic confirmed CVC‐related thrombosis, death due to VTE, major and minor bleeding events in both arms within 30 days of CVC placement or at the time of CVC removal (+ 14 days if < 30 days). Separate data for neonates and non‐neonates were not presented separately and are therefore presented together in this review.

Greiner 2018 recruited 949 children with newly‐diagnosed acute lymphoblastic leukaemia receiving induction chemotherapy as per the treatment protocols ALL‐BFM 2000 (NCT00430118) or AIEOP‐BFM‐ALL 2009 (NCT01117441) in a three‐arm open‐label RCT in Germany and Switzerland. Participants were randomised to either activity‐adapted antithrombin preparation, which will not be discussed further in this review, prophylactic LMWH (enoxaparin) at 80 to 100 IU/kg from day 8 to day 33 of induction (age: 1 to < 6 yrs: 49.5%; 6 to < 10 yrs: 22.9%; ≥ 10 yrs: 27.8%), or low‐dose UFH at a dose of 2 IU/kg body weight/hour (1 to < 6 yrs: 55.8%;, 6 < 10 yrs: 18.3%; ≥ 10 yrs: 26.0%), for prevention of CVC‐related thrombosis. The outcomes included: thrombosis (objectively proven by imaging following clinical suspicion), bleeding events, overall survival and event‐free survival, as measured over an observation period of 64 days.

Excluded studies

See Characteristics of excluded studies.

We excluded 15 studies in this update (Barekatain 2018; Cefali 2017; Del Principe 2012; Diamond 2018; Korkemaz 2017; Kurtkoti 2016; Lamontagne 2014; Lavau‐Denes 2013; Li 2016; Lutkin 2018; McDonald 1984; Pardun 2017; Rocha 2006; Thirunavukkarasu 2016; Wang 2019); bringing the total number of excluded studies to 31. Exclusion reasons were: trials run exclusively in the adult care setting (Abdelkefi 2004; De Cicco 2009; Del Principe 2012; Karthaus 2006; Kurtkoti 2016; Lamontagne 2014; Lavau‐Denes 2013; Li 2016; Mismetti 2003; Monreal 1996; Niers 2007; Rocha 2006; Thirunavukkarasu 2016; Verso 2005; Wang 2019) or in neonates aged less than 30 days old (Barekatain 2018; McDonald 1984); prospective cohort studies (Korkemaz 2017; Mitchell 2010; Raffini 2011; Trame 2010); retrospective cohort studies (Cavo 2010; Cefali 2017; Diamond 2018; Harlev 2010; Harney 2010; Lutkin 2018; Pardun 2017; Sandoval 2008; van Ommen 2010) and mixed design (Vegting 2012).

Risk of bias in included studies

The risk of bias of the included studies (Massicotte 2003, Greiner 2018) is shown in Figure 2 and is explained in detail in Characteristics of included studies.

2.

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

Allocation

In both studies, participants were randomised using a computer‐derived protocol and so are at low risk of bias in this domain. Allocation concealment was judged to be at low risk of bias in Greiner 2018 and at unclear risk of bias in Massicotte 2003 as it was not specifically addressed and the study author has not provided this information.

Blinding

No blinding of participants and study personnel occurred in either study, leading to a high risk of performance bias in both studies (Greiner 2018; Massicotte 2003). Greiner 2018 had an unclear risk for detection bias. Physicians and radiologists were aware of the randomisation arm but appropriate methods were used to diagnose the outcomes (i.e. appropriate radiological studies for thrombosis and transparent clinical criteria for bleeding, based on standardized definitions (Schulman 2005; Levine 2001; Committee for proprietary medicinal products 2016)). These methods might have reduced (without removing entirely) the risk of bias. The study by Massicotte was felt to be at low risk of detection bias because of the placement of an independent and blinded outcome adjudication process (Massicotte 2003).

Incomplete outcome data

The risk of attrition bias was considered high in one study (Massicotte 2003). For 28 participants (15%) no data were available regarding the efficacy outcome (venogram proven VTE), because the required venogram could not be performed (23 participants, 12 in the intervention arm and 11 in the control arm) or was inconclusive (two in the intervention arm and three in the control arm). In the second study, the rates of attrition were low (seven patients, two in the enoxaparin arm and five in the UFH arm) and was caused by deaths of participants (Greiner 2018).

Selective reporting

Both studies were felt to be at low risk of reporting bias. The study protocol of Massicotte 2003 was not available, but the published report included all expected outcomes. The study protocol of Greiner 2018 was provided as an online supplement and the report included all expected outcomes.

Other potential sources of bias

In Massicotte 2003, the study was stopped early due to the low enrolment rate (not due to data‐driven reasons) and was underpowered. However, this was not thought to introduce any potential bias.

Two additional sources of bias were identified in Greiner 2018. First, refusal of antithrombotic treatment, especially by children less than 6 years of age, led to substantial deviations from the intended interventions (10/312 (3%) in participants randomised to UFH, 105/317 (33%) in participants assigned to enoxaparin). These deviations might lead to under‐estimation of the efficacy of the enoxaparin and under‐estimation of the risk of bleeding. Additionally, the study protocol of Greiner 2018 was amended to allow participation of patients without a CVC, which has an unclear effect on the results. The study authors have not provided additional information about the allocation of included patients without a CVC, leading to a judgement of high risk of other potential bias.

Effects of interventions

See: Table 1

All CVC‐related thrombosis

Individually, the studies reported conflicting results. Massicotte 2003 found no clear evidence of a difference in the incidence of CVC‐related thrombosis between patients in the LMWH (reviparin) arm and the low‐dose UFH arm (RR 1.13, 95% CI 0.51 to 2.50). However, the study was not powered to detect the efficacy of LMWH prophylaxis to prevent CVC‐related thrombosis in children, given the early closure of the study. Conversely, in Greiner 2018, the risk of thromboembolism was moderately reduced by LMWH (enoxaparin) compared to the low‐dose UFH arm (RR 0.43, 95% CI 0.22 to 0.88).

Pooling the data in a meta‐analysis showed insufficient evidence of an effect of LMWH prophylaxis in reducing the incidence of CVC‐related thrombosis compared to low‐dose UFH (RR 0.68, 95% CI 0.27 to 1.75; 2 trials; 787 participants; low‐certainty evidence; Analysis 1.1). Of note, I² was 68%, suggesting substantial heterogeneity. The certainty of evidence was downgraded two steps because of serious risk of bias and inconsistency.

1.1. Analysis.

Comparison 1: LWMH versus control, Outcome 1: All CVC‐related thrombosis

One study reported symptomatic and asymptomatic CVC‐related thrombosis separately and showed no evidence of a difference in both instances (Massicotte 2003) (symptomatic thrombosis: RR 1.03, 95% CI 0.21 to 4.93; 1 trial; 158 participants; low‐certainty evidence; Analysis 1.2; asymptomatic: RR 1.17, 95% CI 0.45 to 3.08; 1 trial; 158 participants; low‐certainty evidence; Analysis 1.3). For both symptomatic and asymptomatic CVC‐related thrombosis, the certainty of evidence was downgraded two steps because of serious risk of bias and imprecision.

1.2. Analysis.

Comparison 1: LWMH versus control, Outcome 2: Symptomatic CVC‐related thrombosis

1.3. Analysis.

Comparison 1: LWMH versus control, Outcome 3: Asymptomatic CVC‐related thrombosis

No further information was available regarding the remaining pre‐specified secondary outcomes, including catheter occlusion, days of catheter patency, episodes of CRBSI, and other side effects of LMWH (allergic reactions, abnormal coagulation profile, heparin‐induced thrombocytopaenia and osteoporosis).

Major bleeding

Both studies reported on major bleeding, with six major bleeding events encountered among 813 participants (0.7%) (one in the LMWH group, five in the low‐dose UFH group). There is insufficient evidence to determine whether LMWH impacts the risk of bleeding due to imprecision and risk of bias in the included studies (RR 0.27, 95% CI 0.05 to 1.67; 2 trials; 813 participants, low‐certainty evidence; Analysis 1.4). No heterogeneity was detected for this analysis (I² = 0%).

1.4. Analysis.

Comparison 1: LWMH versus control, Outcome 4: Major bleeding

Minor bleeding

In Massicotte 2003, minor bleeding was found in 53.3% of participants in the reviparin arm and in 44.7% of participants in the low‐dose UFH arm (RR 1.20, 95% CI 0.91 to 1.58; 1 trial; 158 participants; very low‐certainty evidence; Analysis 1.5). No minor bleeding in either arms was reported in Greiner 2018 (RR: not estimable). The results were not pooled because of important differences in baseline rates of minor bleeding, likely attributable to differences in outcome definition. Massicotte 2003 defined minor bleeding as any bleeding event exceeding what was usually expected, where Greiner 2018 used more stringent criteria to define minor bleeding that were clinically relevant but did not met the criteria of a major bleeding.

1.5. Analysis.

Comparison 1: LWMH versus control, Outcome 5: Minor bleeding

Mortality

Both studies reported mortality but did not find evidence of clear differences between study arms. Massicotte 2003 reported two deaths during the study period. Both were unrelated to VTE and occurred in the low‐dose UFH arm. Greiner 2018, which included paediatric patients with acute lymphoblastic leukaemia, showed similar 5‐year all‐cause mortality (5‐year overall survival: enoxaparin: 90.9 ± 1.6%, low‐dose UFH: 92.4 ± 1.5%). The number of deaths during the intervention was not available per arm.

We were unable to pool the data for this outcome because of the differences in outcome reporting. We judged the certainty of the evidence for this outcome to be low due to inconsistency and imprecision (attributable to small number of events).

Discussion

Summary of main results

This review update found insufficient evidence of an effect of LMWH prophylaxis in reducing the incidence of CVC‐related thrombosis in children with a CVC, compared to low‐dose UFH (low‐certainty evidence). Only one study reported asymptomatic and symptomatic thrombosis, again showing no evidence of a difference between LWMH prophylaxis compared to heparin flushes (low‐certainty evidence). Similarly, there is insufficient evidence of any effect regarding the risk of major bleeding (low‐certainty evidence); minor bleeding (very low‐certainty evidence); or overall mortality following thromboprophylaxis with LMWH. Reported deaths were not thrombosis‐related. We were limited by the number of trials meeting the inclusion criteria. The two studies which were included did not report on other pre‐specified outcomes including CVC occlusion, days of CVC patency, CRBSI or other adverse effects of LMWH. See Table 1.

Overall completeness and applicability of evidence

The included studies compared LMWH with low‐dose UFH. We identified no studies comparing LMWH with placebo or no treatment. This reflects the current practice, as the use of low‐dose UFH continuous infusion or flushes is commonly used to ensure the patency of the CVC.

Children requiring a CVC are a highly complex and heterogenous population. They have different underlying conditions and may be exposed to different types of CVCs, with variable risk of CVC‐related complications and thrombosis. The included studies recruited participants with a limited number of conditions, thus leading to an under‐representation of several patient populations and CVCs. Massicotte 2003 included children with a variety of health conditions requiring CVC placement but was underpowered due to the early closure of the trial, while Greiner 2018 recruited exclusively children with acute lymphoblastic leukaemia receiving induction chemotherapy. Thus, the findings might not be applicable to all paediatric patients admitted to tertiary care hospitals with a CVC, such as children admitted to the paediatric intensive care unit. A recent cross‐sectional international survey has shown that among children admitted to the paediatric intensive care unit, 52.8% had at least one CVC. Only 17.0% received some form of pharmacological thromboprophylaxis (Faustino 2014). Unfortunately, we found no data to evaluate several secondary outcomes of this review, including the impact of the intervention on catheter patency.

Quality of the evidence

The included trials had an adequate methodological design (Massicotte 2003; Greiner 2018) and thorough report, including the study limitations and deviations from the protocol. Some limitations should however be highlighted. One study had an important deviation from the intended intervention based on refusal of subcutaneous injections in the LMWH arm (Greiner 2018). Additionally, both were open‐label studies, due to the ethical and practical challenges of using a placebo for subcutaneous injections, which in theory may increase the estimate of intervention effects. Nonetheless, in Massicotte 2003, a committee for outcome adjudication was in place, decreasing the risk of bias in this respect. No system was in place to blind the attending physician or radiologist to the treatment arm in Greiner 2018, and radiological assessment was solely based on clinical suspicion, thereby leading to some bias in outcome detection. However, investigators used appropriate standardized definitions of the outcomes.

We found substantial heterogeneity in our main analysis. The small number of studies did not allow further analysis to explore potential causes of heterogeneity. Possible explanations for heterogeneity include differences in underlying conditions of the study populations (varied versus acute lymphoblastic leukaemia), differences in LMWH used (reviparin versus enoxaparin) and differences in outcome assessment (systematic venogram at study exit versus radiologic evaluation based on clinical assessment). The International Society on Thrombosis and Haemostasis (ISTH) recommends the inclusion of any radiologically confirmed DVT, whether clinically apparent or asymptomatic, because acutely asymptomatic VTE not uncommonly presents with functional sequelae; and because of the insensitivity and non‐specificity of the clinical diagnosis of DVT in children (Faustino 2018, Mitchell 2011).

As a result, the risk of bias and the imprecision of the results prompted us to downgrade the certainty of the evidence from high to low for CVC‐related thrombosis and from high to low or very low for major and minor bleeding, respectively (Table 1).

Potential biases in the review process

We searched for published and unpublished abstracts and manuscripts in a comprehensive literature review designed and run by the Cochrane Vascular Information Specialist. This search was not limited to a particular language. We contacted study authors to obtain data that were missing in the original publications or abstracts. However, a limitation of this review is the fact that it was not possible to collect all relevant information from the included trials: 1) outcomes not included in the study such as efficacy data regarding CVC patency and safety data regarding osteoporosis, coagulopathy, heparin‐induced thrombocytopaenia and allergic reactions secondary to LMWH use, 2) data on relevant subset of patients not available, e.g. separate data for patients who did not have a CVC in place in Greiner 2018 and separate data for neonates and non‐neonates in Massicotte 2003.

Agreements and disagreements with other studies or reviews

This review found insufficient evidence of a difference in the incidence of CVC‐related thrombosis despite LMWH prophylaxis. Four other studies suggested differently and warrant a closer look (Diamond 2018; Harlev 2010; Mitchell 2010; Vegting 2012). They were reports from single institution studies (Harlev 2010;Vegting 2012; Diamond 2018), and from one multicentre collaborative study (Mitchell 2010). These studies were not randomised, increasing study limitations and the potential for bias.

Populations carrying different risks due to their different nature could vary in their thrombosis risks, providing a reasonable explanation for the differences seen between the studies. The studies by Harlev 2010 and Mitchell 2010 analysed the incidence of thrombosis in children with the most common form of paediatric cancer, acute lymphoblastic leukaemia (ALL). Harlev 2010 was a study including children treated over a nine‐year period (1999 to 2008). Their overall thrombosis incidence was 7.5%, and patients carrying an inherited thrombotic risk factor (factor V Leiden or prothrombin gene mutation) received upfront LMWH prophylaxis (enoxaparin 1 mg/kg body weight/dose, once daily) during the most thrombogenic phases of their chemotherapy protocol. The group with thrombophilia and LMWH had a thrombosis incidence of 16.6%, much lower than the previous historical report in the same subpopulation but without LMWH (~ 47%) (Nowak‐Göttl 1999). The incidence in the low‐risk patients was 4.5% (Harlev 2010). Half of those thrombotic events occurred in patients with CVCs. Mitchell 2010 also looked at children with ALL. Thrombosis risk stratification was done using a newly developed thrombosis risk stratification score. Four hundred fifty‐six participants were available for scoring and 339 for prospective validation of the score. Subsequently, 8/19 children scoring at high thrombotic risk received LMWH primary prophylaxis prior to CVC placement; LMWH was instituted based on physicians' preferences. This practice led to a significant increased thrombosis‐free survival in patients on LMWH (P = 0.02, log rank test). Therefore, both studies suggested a beneficial preventive antithrombotic effect in children with ALL and CVC placement. These results align with the included study in this review by Greiner 2018, which also reported a reduction of thrombosis in children with ALL. Conversely, the proportions of the two main underlying conditions in Massicotte 2003 were congenital heart disease (22.5%) and cancer (50%). It is possible that ALL is a more homogenous and potentially more thrombogenic condition, in which thromboprophylaxis might be more effective. In fact, even within the same underlying condition, as shown in Mitchell 2010, thrombosis risk stratification is likely to be required for the institution of LMWH‐based or any other type of anticoagulant‐based prophylaxis. This information is likely to be very relevant for all paediatric populations with a higher risk of thrombosis development, given that CVCs are potentially the most relevant thrombosis risk factor in place (O'Brien 2011).

Vegting 2012 was a single centre observational study from children requiring total parenteral nutrition (TPN) administered through a CVC. In this study, patients on TPN receiving either LMWH (for example, nadroparin in a single dose of 80 IU/kg targeting an anti‐Xa between 0.1 and 0.3) or an oral vitamin K antagonist (VKA) (acenocumarol, targeting an international normalised ratio (INR) between 2.0 and 3.0) were compared to patients from the same centre in whom anticoagulation prophylaxis was not used. Of note, approximately 75% of patients included in the prophylaxis arm were previously included in the non‐prophylaxis group as the clinic's anticoagulation policy was changed. The CVC‐related outcomes of this cohort were CVC‐related thrombosis, infection and obstruction. In summary, patients receiving thromboprophylaxis (n = 18; 16 LMWH, 2 VKA) had a lower rate of thrombosis (6%) in comparison to the children not receiving prophylaxis (n = 14; 33%; P = 0.034). Moreover, per 1000 TPN days, the prophylaxis and non‐prophylaxis groups had 0.1 and 2.6 CVC occlusions (P = 0.04) and 2.1 and 4.6 infections (P = 0.06), respectively, with a three‐year infection‐free survival of 46% and 19%, respectively (P = 0.03). No bleeding occurred in either group and the study authors concluded that anticoagulation prophylaxis significantly decreased CVC‐related complications in children receiving long‐term TPN. Diamond 2018 retrospectively compared the rate of CVC‐related thrombosis in children hospitalised for inflammatory bowel disease requiring a CVC with and without thromboprophylaxis with enoxaparin (< 40 kg: 0.5 mg/kg subcutaneously every 12 h; 40 kg and above: 40 mg subcutaneously daily). CVC‐related thrombosis occurred in 0/24 (0%) of cases with LMWH, compared to 5/23 (22%) of cases where no prophylaxis was given. This study did not find differences between groups in terms of red cell transfusion requirement or haemoglobin nadir, as a proxy of bleeding complications.

The differing conclusions between these studies (Diamond 2018; Vegting 2012) and this systematic review may be explained by differences in the baseline thrombosis risk of the study population and in the study methodology. More specifically, the population included in Vegting 2012 was a much more homogeneous one, recognizably amongst the highest risk groups for CVC‐related thrombosis development (Andrew 1995). In such patients, the nature of the infusate, leading to local vessel wall inflammation, in addition to protein precipitation within the catheter lumen contribute to a likely high rate of CVC‐related complications. Diamond 2018 also included a patients at high risk of thrombosis, combining multiple risk factors such as active inflammatory bowel disease, hospitalisation and a CVC. Their studies also included older children and adolescents (median age: 14 years), who are at higher risk of thrombosis (Andrew 1994).

Depending on the time of CVC insertion, thrombosis risk factors will have greater or lesser importance. Given the study design of Vegting 2012, the interpretation of the results is limited as 13/18 patients from the thromboprophylaxis arm had their CVC inserted and been exposed to total parenteral nutrition for a considerable time before they entered "the intervention study arm", likely leading to exposure to risk factors in an unbalanced and different manner. In Diamond 2018, patients received thromboprophylaxis at the primary care team's discretion, which may over‐estimate the effectiveness of LMWH, as patients with unmeasured variables that increased their thrombosis risk are more likely to be exposed to the intervention.

In comparison, several meta‐analyses have reported a beneficial effect of LMWH to prevent venous thromboembolism in adult populations, for example in hospitalised patients (Lederle 2011), in adult patients with cancer and a CVC (Kahale 2018), and in adults hospitalised in medical‐surgical intensive care units (Alhazzani 2013). However, any benefit on mortality remains unclear.

Authors' conclusions

Implications for practice.

Pooling data from two RCTs provided insufficient evidence of an effect of LMWH prophylaxis in reducing the incidence of CVC‐related thrombosis in children with a CVC, compared to low‐dose UFH (low‐certainty evidence). Evidence was insufficient to determine whether there were differences in the incidence of major and minor bleeding in the LMWH prophylaxis group compared to low‐dose UFH group (low‐ and very low‐certainty respectively). No evidence of a clear difference in overall mortality was seen. Studies did not report on the following outcomes: catheter occlusion, days of catheter patency, episodes of CRBSI, other side effects of LMWH (allergic reactions, abnormal coagulation profile, heparin‐induced thrombocytopaenia and osteoporosis) or adverse effects. The certainty of the evidence was downgraded due to risk of bias of the included studies and inconsistency, preventing conclusions in regards to the efficacy of LMWH prophylaxis to prevent CVC‐related thrombosis in children. Because of the uncertain efficacy of thromboprophylaxis in children, clinicians should practice with caution.

Implications for research.

Since the reported incidence of symptomatic CVC‐related thrombosis in children is approximately 10%, a trial evaluating the superiority of LMWH prophylaxis in comparison to standard care or low‐dose UFH will require at least 856 patients in total (that is 428 patients per arm in a one‐to‐one allocation) to achieve 80% power to detect a 50% relative risk reduction of symptomatic thrombosis, which can be considered clinically meaningful. The two‐sided significance level (1‐alpha) was set at 95. We used Open‐Epi software to make this sample size calculation (OpenEpi).

Many lessons can be learned from the included studies. Enrolment was problematic in both studies, either because of the relatively large number of exclusions or because of refusal to receive daily subcutaneous injections with LMWH, especially in young children. This highlights the difficulties of conducting RCTs in paediatric thrombosis.

Given the increasing placement of CVCs in modern paediatric care, and that paediatric hospital accreditation may be conditioned to the development of age‐ and scenario‐specific prophylaxis protocol development in the near future, there will be a continued need to conduct trials evaluating this problem. Trials should aim to evaluate the role of patient age, CVC properties, risk factors, and of additional prophylactic modalities such as new oral anticoagulant agents, warfarin and CVC flushes. Lastly, future studies should also address the potential costs of any prophylactic intervention.

Feedback

Anticoagulant feedback, February 2011

Summary

Feedback received on this protocol, and other reviews and protocols on anticoagulants, is available on the Cochrane Editorial Unit Website at http://www.editorial-unit.cochrane.org/anticoagulants-feedback.

What's new

Date	Event	Description
10 September 2019	New citation required but conclusions have not changed	New search run. One new study included, 15 new studies excluded and one new ongoing study identified. Text updated to reflect current Cochrane standards. New authors joined team. No change to conclusions.
10 September 2019	New search has been performed	New search run. One new study included, 15 new studies excluded and one new ongoing study identified.

History

Protocol first published: Issue 2, 2006
Review first published: Issue 3, 2014

Date	Event	Description
1 May 2013	Amended	Link to anticoagulant feedback added
30 October 2008	Amended	Converted to new review format.

Appendices

Appendix 1. Database search strategies

Source	Search strategy	Hits retrieved
CENTRAL via CRSO	#1 MESH DESCRIPTOR Catheterization, Central Venous EXPLODE ALL AND CENTRAL:TARGET #2 MESH DESCRIPTOR Catheters EXPLODE ALL AND CENTRAL:TARGET #3 broviac AND CENTRAL:TARGET #4 port adj3 cath AND CENTRAL:TARGET #5 hickman AND CENTRAL:TARGET #6 catheter* AND CENTRAL:TARGET #7 TCVC or PICC or CVC or CVAD AND CENTRAL:TARGET #8 venous adj3 (line or device) AND CENTRAL:TARGET #9 cannula* AND CENTRAL:TARGET #10 portacath AND CENTRAL:TARGET #11 #1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8 OR #9 OR #10 #12 MESH DESCRIPTOR Thrombosis EXPLODE ALL AND CENTRAL:TARGET #13 thrombo* or clot* AND CENTRAL:TARGET #14 #12 OR #13 #15 #11 AND #14 #16 MESH DESCRIPTOR Heparin, Low‐Molecular‐Weight EXPLODE ALL AND CENTRAL:TARGET #17 heparin* or LMWH or nadroparin* or fraxiparin* or enoxaparin or Clexane or klexane or lovenox or dalteparin or Fragmin or ardeparin or normiflo or tinzaparin or logiparin or Innohep or certoparin or sandoparin or reviparin or clivarin* or danaproid or danaparoid or bemiparin or bioparin or Alphaparin or Troparin AND CENTRAL:TARGET #18 antixarin or ardeparin* or bemiparin* or Zibor or cy 222 or embolex or monoembolex or Mono‐embolex or parnaparin* or "rd 11885" or tedelparin or Kabi‐2165 or Kabi 2165 196 #19 emt‐966 or emt‐967 or pk‐10169 or pk10169 #19 emt‐966 or emt‐967 or pk‐10169 or pk10169 AND CENTRAL:TARGET #20 cy‐216 or cy216 or seleparin* or tedegliparin or seleparin* or tedegliparin* or tedelparin or Boxol or Liquemine AND CENTRAL:TARGET #21 fr‐860 AND CENTRAL:TARGET #22 wy90493 or wy‐90493 AND CENTRAL:TARGET #23 kb‐101 or kb101 or lomoparan or orgaran AND CENTRAL:TARGET #24 parnaparin or fluxum or lohepa or lowhepa or "op 2123" or parvoparin AND CENTRAL:TARGET #25 AVE5026 or M118 or RO‐14 AND CENTRAL:TARGET #26 #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24 OR #25 #27 #15 AND #26 #28 01/01/2013_TO_07/05/2019:CRSINCENTRAL AND CENTRAL:TARGET #29 #27 AND #28	472
Clinicaltrials.gov	heparin | Thrombosis | Catheter OR catheterization OR catheterisation | Start date on or after 01/01/2013 | Last update posted on or before 04/07/2019	17
ICTRP Search Portal	heparin | Thrombosis | Catheter OR catheterization OR catheterisation	5
Medline (Ovid MEDLINE® Epub Ahead of Print, In‐Process & Other Non‐Indexed Citations, Ovid MEDLINE® Daily and Ovid MEDLINE®) 1946 to present 2017, 2018 and 2019 only	1 exp Catheterization, Central Venous/ 2 exp Catheters/ 3 broviac.ti,ab. 4 (port adj3 cath).ti,ab. 5 hickman.ti,ab. 6 catheter*.ti,ab. 7 (TCVC or PICC or CVC or CVAD).ti,ab. 8 (venous adj3 (line or device)).ti,ab. 9 cannula*.ti,ab. 10 portacath.ti,ab. 11 or/1‐10 12 exp Thrombosis/ 13 (thrombo* or clot*).ti,ab. 14 or/12‐13 15 11 and 14 16 exp Heparin, Low‐Molecular‐Weight/ 17 (heparin* or LMWH or nadroparin* or fraxiparin* or enoxaparin or Clexane or klexane or lovenox or dalteparin or Fragmin or ardeparin or normiflo or tinzaparin or logiparin or Innohep or certoparin or sandoparin or reviparin or clivarin* or danaproid or danaparoid or bemiparin or bioparin or Alphaparin or Troparin).ti,ab. 18 (antixarin or ardeparin* or bemiparin* or Zibor or cy 222 or embolex or monoembolex or Mono‐embolex or parnaparin* or "rd 11885" or tedelparin or Kabi‐2165 or Kabi 2165 196 #19 emt‐966 or emt‐967 or pk‐10169 or pk10169).ti,ab. 19 (emt‐966 or emt‐967 or pk‐10169 or pk10169).ti,ab. 20 (cy‐216 or cy216 or seleparin* or tedegliparin or seleparin* or tedegliparin* or tedelparin or Boxol or Liquemine).ti,ab. 21 fr‐860.ti,ab. 22 (wy90493 or wy‐90493).ti,ab. 23 (kb‐101 or kb101 or lomoparan or orgaran).ti,ab. 24 (parnaparin or fluxum or lohepa or lowhepa or "op 2123" or parvoparin).ti,ab. 25 (AVE5026 or M118 or RO‐14).ti,ab. 26 or/16‐25 27 15 and 26 28 randomized controlled trial.pt. 29 controlled clinical trial.pt. 30 randomized.ab. 31 placebo.ab. 32 drug therapy.fs. 33 randomly.ab. 34 trial.ab. 35 groups.ab. 36 or/28‐35 37 exp animals/ not humans.sh. 38 36 not 37 39 27 and 38 40 (2017* or 2018* or 2019*).ed. 41 39 and 40 42 from 41 keep 1‐87	87
EMBASE 2017, 2018 and 2019 only	1 exp central venous catheterization/ 2 exp catheter/ 3 broviac.ti,ab. 4 (port adj3 cath).ti,ab. 5 hickman.ti,ab. 6 catheter*.ti,ab. 7 (TCVC or PICC or CVC or CVAD).ti,ab. 8 (venous adj3 (line or device)).ti,ab. 9 cannula*.ti,ab. 10 portacath.ti,ab. 11 or/1‐10 12 exp thrombosis/ 13 (thrombo* or clot*).ti,ab. 14 or/12‐13 15 11 and 14 16 exp low molecular weight heparin/ 17 (heparin* or LMWH or nadroparin* or fraxiparin* or enoxaparin or Clexane or klexane or lovenox or dalteparin or Fragmin or ardeparin or normiflo or tinzaparin or logiparin or Innohep or certoparin or sandoparin or reviparin or clivarin* or danaproid or danaparoid or bemiparin or bioparin or Alphaparin or Troparin).ti,ab. 18 (antixarin or ardeparin* or bemiparin* or Zibor or cy 222 or embolex or monoembolex or Mono‐embolex or parnaparin* or "rd 11885" or tedelparin or Kabi‐2165 or Kabi 2165 196 #19 emt‐966 or emt‐967 or pk‐10169 or pk10169).ti,ab. 19 (emt‐966 or emt‐967 or pk‐10169 or pk10169).ti,ab. 20 (cy‐216 or cy216 or seleparin* or tedegliparin or seleparin* or tedegliparin* or tedelparin or Boxol or Liquemine).ti,ab. 21 fr‐860.ti,ab. 22 (wy90493 or wy‐90493).ti,ab. 23 (kb‐101 or kb101 or lomoparan or orgaran).ti,ab. 24 (parnaparin or fluxum or lohepa or lowhepa or "op 2123" or parvoparin).ti,ab. 25 (AVE5026 or M118 or RO‐14).ti,ab. 26 or/16‐25 27 15 and 26 28 randomized controlled trial/ 29 controlled clinical trial/ 30 random$.ti,ab. 31 randomization/ 32 intermethod comparison/ 33 placebo.ti,ab. 34 (compare or compared or comparison).ti. 35 ((evaluated or evaluate or evaluating or assessed or assess) and (compare or compared or comparing or comparison)).ab. 36 (open adj label).ti,ab. 37 ((double or single or doubly or singly) adj (blind or blinded or blindly)).ti,ab. 38 double blind procedure/ 39 parallel group$1.ti,ab. 40 (crossover or cross over).ti,ab. 41 ((assign$ or match or matched or allocation) adj5 (alternate or group$1 or intervention$1 or patient$1 or subject$1 or participant$1)).ti,ab. 42 (assigned or allocated).ti,ab. 43 (controlled adj7 (study or design or trial)).ti,ab. 44 (volunteer or volunteers).ti,ab. 45 trial.ti. 46 or/28‐45 47 27 and 46 48 (2017* or 2018* or 2019*).dc. 49 47 and 48 50 from 49 keep 1‐302	302
CINAHL 2017, 2018 and 2019 only	S45 S43 AND S44 S44 EM 2017 OR EM 2018 OR 2019 EM S43 S27 AND S42 S42 S28 OR S29 OR S30 OR S31 OR S32 OR S33 OR S34 OR S35 OR S36 OR S37 OR S38 OR S39 OR S40 OR S41 S41 MH "Random Assignment" S40 MH "Triple‐Blind Studies" S39 MH "Double‐Blind Studies" S38 MH "Single‐Blind Studies" S37 MH "Crossover Design" S36 MH "Factorial Design" S35 MH "Placebos" S34 MH "Clinical Trials" S33 TX "multi‐centre study" OR "multi‐center study" OR "multicentre study" OR "multicenter study" OR "multi‐site study" S32 TX crossover OR "cross‐over" S31 AB placebo* S30 TX random* S29 TX trial* S28 TX "latin square" S27 S15 AND S26 S26 S16 OR S17 OR S18 OR S19 OR S20 OR S21 OR S22 OR S23 OR S24 OR S25 S25 TX parnaparin or fluxum or lohepa or lowhepa or "op 2123" or parvoparin S24 TX AVE5026 or M118 or RO‐14 S23 TX kb‐101 or kb101 or lomoparan or orgaran S22 TX wy90493 or wy‐90493 S21 TX fr‐860 S20 TX cy‐216 or cy216 or seleparin* or tedegliparin or seleparin* or tedegliparin* or tedelparin or Boxol or Liquemine S19 TX emt‐966 or emt‐967 or pk‐10169 or pk10169 S18 TX antixarin or ardeparin* or bemiparin* or Zibor or cy 222 or embolex or monoembolex or Mono‐embolex or parnaparin* or "rd 11885" or tedelparin or Kabi‐2165 or Kabi 2165 196 #19 emt‐966 or emt‐967 or pk‐10169 or pk10169 S17 TX heparin* or LMWH or nadroparin* or fraxiparin* or enoxaparin or Clexane or klexane or lovenox or dalteparin or Fragmin or ardeparin or normiflo or tinzaparin or logiparin or Innohep or certoparin or sandoparin or reviparin or clivarin* or danaproid or danaparoid or bemiparin or bioparin or Alphaparin or Troparin S16 (MH "Heparin, Low‐Molecular‐Weight+") S15 S11 AND S14 S14 S12 OR S13 S13 TX thrombo* or clot* S12 (MH "Thrombosis+") S11 S1 OR S2 OR S3 OR S4 OR S5 OR S6 OR S7 OR S8 OR S9 OR S10 S10 TX portacath S9 TX cannula* S8 TX venous n3 (line or device) S7 TX TCVC or PICC or CVC or CVAD S6 TX catheter* S5 TX hickman S4 TX port n3 cath S3 TX broviac S2 (MH "Catheters+") S1 (MH "Catheterization, Central Venous+")	16
AMED 2017, 2018 and 2019 only	1 exp catheter/ 2 broviac.ti,ab. 3 (port adj3 cath).ti,ab. 4 hickman.ti,ab. 5 catheter*.ti,ab. 6 (TCVC or PICC or CVC or CVAD).ti,ab. 7 (venous adj3 (line or device)).ti,ab. 8 cannula*.ti,ab. 9 portacath.ti,ab. 10 or/1‐9 11 exp Thrombosis/ 12 (thrombo* or clot*).ti,ab. 13 or/11‐12 14 12 and 13 15 (heparin* or LMWH or nadroparin* or fraxiparin* or enoxaparin or Clexane or klexane or lovenox or dalteparin or Fragmin or ardeparin or normiflo or tinzaparin or logiparin or Innohep or certoparin or sandoparin or reviparin or clivarin* or danaproid or danaparoid or bemiparin or bioparin or Alphaparin or Troparin).ti,ab. 16 (antixarin or ardeparin* or bemiparin* or Zibor or cy 222 or embolex or monoembolex or Mono‐embolex or parnaparin* or "rd 11885" or tedelparin or Kabi‐2165 or Kabi 2165 196 #19 emt‐966 or emt‐967 or pk‐10169 or pk10169).ti,ab. 17 (emt‐966 or emt‐967 or pk‐10169 or pk10169).ti,ab. 18 (cy‐216 or cy216 or seleparin* or tedegliparin or seleparin* or tedegliparin* or tedelparin or Boxol or Liquemine).ti,ab. 19 fr‐860.ti,ab. 20 (wy90493 or wy‐90493).ti,ab. 21 (kb‐101 or kb101 or lomoparan or orgaran).ti,ab. 22 (parnaparin or fluxum or lohepa or lowhepa or "op 2123" or parvoparin).ti,ab. 23 (AVE5026 or M118 or RO‐14).ti,ab. 24 or/15‐23 25 14 and 24 26 exp CLINICAL TRIALS/ 27 RANDOM ALLOCATION/ 28 DOUBLE BLIND METHOD/ 29 Clinical trial.pt. 30 (clinic* adj trial*).tw. 31 ((singl* or doubl* or trebl* or tripl*) adj (blind* or mask*)).tw. 32 PLACEBOS/ 33 placebo*.tw. 34 random*.tw. 35 PROSPECTIVE STUDIES/ 36 or/26‐35 37 25 and 36 38 ("2017" or "2018" or "2019").yr. 39 37 and 38	0

Data and analyses

Comparison 1. LWMH versus control.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 All CVC‐related thrombosis	2	787	Risk Ratio (IV, Random, 95% CI)	0.68 [0.27, 1.75]
1.2 Symptomatic CVC‐related thrombosis	1		Risk Ratio (IV, Fixed, 95% CI)	Totals not selected
1.3 Asymptomatic CVC‐related thrombosis	1		Risk Ratio (IV, Fixed, 95% CI)	Totals not selected
1.4 Major bleeding	2	813	Risk Ratio (IV, Random, 95% CI)	0.27 [0.05, 1.67]
1.5 Minor bleeding	2		Risk Ratio (IV, Random, 95% CI)	Totals not selected

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Greiner 2018.

Study characteristics
Methods	Open‐label RCT of LMWH, activity‐adapted antithrombin or low‐dose UFH for prevention of CVC‐related thrombosis
Participants	Children with newly diagnosed ALL in induction, treated on the ALL‐BFM 2000 or AIEOP‐BFM ALL 2009 protocols and having a CVC inserted by day 8 of induction and remaining in place at least until day 33. The choice of CVC was at the discretion of the physician. No of participants: 949 (total); enoxaparin arm: 317; antithrombin arm: 320; control/comparison arm: 312
Interventions	Intervention: enoxaparin: 80 to 100 IU/kg daily sc, from day 8 to day 33 of induction chemotherapy Control/comparison: UFH 2 IU/kg/hour (as long as an infusion drip was running) or activity‐adapted plasma‐derived antithrombin preparation. The activity‐adapted plasma‐derived antithrombin preparation is not considered in this review
Outcomes	Primary: objectively confirmed thromboembolism Secondary: bleeding complications, including major and minor bleeding; overall survival; event‐free survival
Notes	The study was conducted in participating hospitals from Switzerland and Germany. In August 2004, the protocol was amended to allow participation of patients without CVC, leading to a small number of patients (5.6%) with no CVC in site.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated random number list, provided after patient's enrolment into the study
Allocation concealment (selection bias)	Low risk	Randomisation centrally performed, in a 1:1:1 ratio using permuted blocks of six patients, stratified by country and the glucocorticoid preparation administered during induction, not accessible to the participating centres
Blinding of participants and personnel (performance bias) All outcomes	High risk	No systematic provision was made for blinding the participants and families and attending physicians to the randomisation arm.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	No systematic provision was made for blinding the attending physicians or radiologists to the randomisation arm. However, appropriate methods were used to diagnose the outcomes (appropriate radiological studies for thrombosis and transparent clinical criteria for bleeding, based on standardised definitions) that might reduce (without removing entirely) the risk of bias.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Low levels of attrition which were explained by study authors
Selective reporting (reporting bias)	Low risk	The study protocol was provided as an online supplement and the report included all expected outcomes
Other bias	High risk	Effect of deviations from the intended interventions: the fact that the protocol was amended to include patients without CVCs might have altered the results. High rates of refusal of the antithrombotic treatment, especially in the enoxaparin arm, might under‐estimate the reduction in CVC‐related thrombosis in ITT analysis

Massicotte 2003.

Study characteristics
Methods	Open‐label RCT of LMWH for prevention of CVC‐related thrombosis in children
Participants	Newborns > 36 weeks gestational age to children < 18 years; if age < 36 weeks, then weight > 5 kg at randomisation, admitted to the hospital, with a CVC placed for the first time The neonates, infants and children with a newly placed CVC were subsequently screened by examination or by imaging at CVC removal (i.e. day + 30) No. of participants: 186 (total); intervention arm: 92; control/comparison arm: 94 The choice of CVC was at the discretion of the physician; CVCs were classified as subcutaneous (27% of patients), exteriorised (51%), peripherally‐inserted central catheters (20%) or unknown (2%).
Interventions	Intervention: reviparin: 30 IU/kg if < 3 months; 50 IU/kg if ≥ 3 months (dose as per previous age‐appropriate pharmacokinetic findings) Control/comparison: very low doses of UFH (flushes or infusion of UFH (< 3 IU/kg/hour)) (standard care)
Outcomes	Efficacy: CVC‐related VTE detected by an exit venogram at day + 30 (+ 14 d) or at the time of CVC removal (+ 14 days) if < 30 days, a confirmed symptomatic VTE within 30 d of CVC placement, a confirmed asymptomatic VTE within 30 d of CVC placement, or death due to thrombosis during the study period Safety: major bleeding and minor bleeding events
Notes	This study was conducted in 20 participating hospitals from five different countries (Australia, Canada, Germany, United Kingdom, United States of America). Underpowered, closed prematurely due to low enrolment rate.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "They were randomly assigned by a computer‐derived protocol to receive..."
Allocation concealment (selection bias)	Unclear risk	Not enough information available
Blinding of participants and personnel (performance bias) All outcomes	High risk	Quote: "The PROTEKT trial was an open‐label randomised controlled trial"
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "An independent and blinded Central Adjudication Committee assessed all efficacy and safety outcomes."
Incomplete outcome data (attrition bias) All outcomes	High risk	High rate of attrition for the primary outcome (quote: "For 23 patients (12 receiving reviparin‐sodium and 11 standard care), the mandatory venogram was not done, usually because of an inability of the patient to be transported to the radiology department. For the other five patients (two reviparin‐sodium, three standard care), the venograms were indeterminate.")
Selective reporting (reporting bias)	Low risk	The study protocol is not available, but the published report includes all expected outcomes
Other bias	Low risk	The study was stopped early due to low enrolment rate (not due to data‐driven reasons)

ALL: acute lymphoblastic leukaemia
CVC: central venous catheter
ITT: intention‐to‐treat
LMWH: low molecular weight heparin
RCT: randomised controlled trial
sc: subcutaneously
SD: standard deviation
UFH: unfractionated heparin
VTE: venous thromboembolism

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Abdelkefi 2004	Prospective randomised trial in adults with haemato‐oncological conditions receiving heparin infusion for thrombosis prevention
Barekatain 2018	Prospective RCT comparing the effect of low dose and high dose UFH on central catheter patency in very low birth neonates
Cavo 2010	Retrospective cohort study of patients receiving LMWH prophylaxis identified as per ICD‐9‐CM codes
Cefali 2017	Retrospective cohort study to assess the association between LMWH and thrombotic events in adult cancer patients with a CVC
De Cicco 2009	Prospective randomised trial comparing short‐term prophylaxis with either dalteparin or low dose acenocumarine in adults with cancer for prevention of venography‐proven catheter‐related thrombosis
Del Principe 2012	Prospective randomised trial to compare LMWH to no anticoagulation for prevention of catheter‐associated thrombosis in adults with acute myeloid leukaemia
Diamond 2018	Retrospective review comparing the incidence of CVC‐related thrombosis in children with inflammatory bowel disease requiring a CVC with and without thromboprophylaxis with enoxaparin
Harlev 2010	Retrospective cohort study of paediatric patients with acute lymphoblastic leukaemia, either receiving or not LMWH prophylaxis according to the practice of different periods of time
Harney 2010	Retrospective cohort study in paediatric patients with CVC admitted to ICU
Karthaus 2006	Prospective, double‐blinded, placebo‐controlled study in adult cancer patients receiving dalteparin to prevent CVC‐related thrombosis
Korkemaz 2017	Prospective, observational cohort study to compare alteplase use following tinzaparin or UFH use in chronic haemodialysis patients
Kurtkoti 2016	Prospective, open‐label, randomised cross‐over trial to compare venous and arterial enoxaparin in adults on haemodialysis or haemodiafiltration to prevent extra‐corporeal circuit thrombosis
Lamontagne 2014	Prospective study to compare to risk of catheter‐associated thrombosis associated with PICCs versus CVCs in adults
Lavau‐Denes 2013	Prospective, open‐label, RCT comparing thromboprophylaxis with LMWH, low‐dose warfarin or no anticoagulation in adult cancer patients with CVC
Li 2016	Prospective RCT comparing LMWH and UFH in adult haemodialysis patients
Lutkin 2018	Retrospective cohort study describing the effectiveness and safety of dalteparin as thromboprophylaxis for children undergoing haemodialysis
McDonald 1984	Prospective, open‐label, randomised study to evaluate thromboprophylaxis with UFH for neonates with a umbilical artery catheter
Mismetti 2003	Prospective, open‐label, randomised trial comparing the LMWH nadroparin to low dose warfarin in adult cancer patients with CVC
Mitchell 2010	Prospective paediatric cohort study for validation of a clinical predictive model for thrombosis development in children with ALL
Monreal 1996	Prospective open‐label, randomised trial enrolling only adults with cancer and central catheter devices receiving dalteparin as prophylaxis
Niers 2007	Prospective, randomised, double‐blinded, placebo‐controlled study in adult patients with hematologic cancer receiving nadroparin for prevention of catheter‐related thrombosis
Pardun 2017	Case‐control study on efficacy and safety of LMWH in infants and children with complex congenital heart disease
Raffini 2011	Prospective study to assess compliance to thromboprophylaxis guidelines in children
Rocha 2006	Protocol for a RCT to evaluate thromboprophylaxis with bemiparin in adult cancer patients with a CVC
Sandoval 2008	Retrospective cohort study on risk factors for thrombosis in hospitalised children
Thirunavukkarasu 2016	Prospective study to evaluate thromboprophylaxis with dalteparin versus UFH in adults undergoing haemodialysis
Trame 2010	Prospective study for the evaluation of pharmacokinetic of enoxaparin in children
van Ommen 2010	Retrospective cohort study for prophylaxis candidates among children on long‐term parenteral nutrition
Vegting 2012	Observational study for anticoagulation prophylaxis of children receiving total parenteral nutrition, including a retrospective and a partly prospective source of data
Verso 2005	Randomised, double‐blinded, placebo‐controlled study in adult cancer patients receiving enoxaparin for CVC‐related thrombosis prevention
Wang 2019	RCT in adult cancer patients that compared rivaroxaban versus LMWH

ALL: acute lymphoblastic leukaemia
CVC: central venous catheter
ICU: intensive care unit
ICD‐9‐CM: International Classification of Diseases, Ninth Revision, Clinical Modification
LMWH: low molecular weight heparin
PICCs: peripherally‐inserted central catheters
RCT: randomised controlled trial
UFH: unfractionated heparin

Characteristics of ongoing studies [ordered by study ID]

Klassen 2017.

Study name	Thromboprophylaxis in children treated for Acute Lymphoblastic Leukemia with low molecular weight heparin: a multicenter randomized controlled trial (NTR4707)
Methods	Multicenter open‐label RCT
Participants	Children between 1 and 19 years of age with primary ALL, who are treated within the treatment protocol DCOG ALL‐11 or subsequent treatment studies
Interventions	LMWH (nadroparin) sc once daily, in a standardised prophylactic weight‐ and anti‐Xa adjusted dose of 85 IU/kg with a maximum of 5700 IU/ day, with an aimed anti‐Xa level of 0.3 – 0.4 IU/mL (measured after 5 days)
Outcomes	Primary efficacy endpoint: incidence of symptomatic objectified VTE during ALL treatment Secondary efficacy endpoints: incidence of the composite of symptomatic and asymptomatic objectified VTE, value of plasma coagulation assays to predict the risk of VTE Primary safety endpoint: major bleeding Secondary safety endpoints: incidence of clinically relevant non‐major bleeding and minor bleeding, the burden of LMWH injections, adverse skin reactions
Starting date	October 2014
Contact information	Dr CH van Ommen Department of Pediatric Oncology/Hematology, Erasmus MC‐SophiaChildren’s Hospital, Rotterdam, The Netherlands
Notes

NCT03003390.

Study name	Prevention of Central Venous Catheter‐associated Thrombosis in Critically Ill Children: A Multicenter Phase 2b Trial (NCT03003390)
Methods	Multcentre open‐label RCT
Participants	Children up to 18 years old, anticipated to stay in a paediatric intensive care unit for at least 48 h, with an untunnelled CVC inserted within 24h, anticipated to be required at least 24 h
Interventions	LMWH (enoxaparin) sc twice daily, in a standardised prophylactic weight‐based dose, with an aimed anti‐Xa level of 0.2 – 0.5 IU/mL
Outcomes	Primary endpoints: catheter‐associated deep venous thrombosis, up to CVC removal, endogenous thrombin potential Secondary endpoints: other thromboembolic events, length of stay in paediatric intensive care unit, length of stay in hospital, presence of clinically relevant bleeding, presence of heparin‐induced thrombocytopaenia, mortality, etc.
Starting date	December 2016
Contact information	Dr E Vincent S Faustino Department of Pediatrics, Yale School of Medicine, New Haven, Connecticut, United States
Notes	Recruitment terminated; met study protocol defined stopping rule for futility

ALL: acute lymphoblastic leukaemia
DCOG: Dutch Childhood Oncology Group
LMWH: low molecular weight heparin
RCT: randomised controlled trial
sc: subcutaneously
VTE: venous thromboembolism

Differences between protocol and review

The planned assessments of the methodological quality of the included studies has been replaced by Cochrane's 'Risk of bias' tool (Higgins 2017).

We used a random‐effects, rather than a fixed‐effect, meta‐analysis, under the assumption that the effects being estimated in the different studies are not identical, to reflect the clinical and methodological heterogeneity of the included studies.

We reordered major and minor bleeding from a secondary outcome to a primary outcome to reflect clinical importance and amended the objective accordingly.

For completeness, we have added definitions and references for our primary and secondary outcomes.

We rephrased 'catheterisation‐related thrombosis' to 'catheter‐related thrombosis' throughout the review, including the title, to reflect the fact that the review deals with indwelling catheters and not the process of inserting the catheter.

Contributions of authors

MPM: selected trials, assessed trial quality, extracted data and wrote and edited the final review
NA: selected trials, assessed trial quality, extracted data and edited the final review
MLA: provided methodological expertise and edited the final review
LRB: provided methodological expertise and edited the final review

Sources of support

Internal sources

• Department of Hematology and Oncology, Haemostasis and Thrombosis Program, The Hospital For Sick Children, Toronto, Canada

• Canadian Institutes of Health, Canada

  Marie‐Claude Pelland‐Marcotte received support from a Canadian Institutes of Health Research Doctoral Research Award to honour Nelson Mandela

External sources

• Chief Scientist Office, Scottish Government Health Directorates, The Scottish Government, UK

  The Cochrane Vascular editorial base is supported by the Chief Scientist Office

Declarations of interest

MPM: none known
NA: none known
MLA: none known
LRB: has declared that his Institution received payment from Boehringer Ingelheim, for an ongoing study of dabigatran etexilate, a direct oral anticoagulant (DOAC) used to treat VTE. Literature on DOACs in children is preliminary and scarce. This is not relevant/cited in this review. DOACs are not used for prophylaxis of central lines due to concerns of inferiority identified in adults with mechanical heart valves. No known conflicts.

New search for studies and content updated (no change to conclusions)