Central Venous Access Device Complications in Children With Leukemia: A Systematic Review and Meta‐Analysis

ABSTRACT

Background

Children diagnosed with leukemia require prolonged central venous access for chemotherapy, transfusions, and supportive care. Due to their immunocompromised status, coagulopathies, and intensive treatment protocols, they may be at increased risk for central venous access device (CVAD)‐related complications, including infection, thrombosis, and mechanical failure. This study aimed to quantify the incidence and characterize the nature of CVAD‐associated complications in this high‐risk pediatric population.

Methods

A systematic review (1970–2025) was conducted using PRISMA guidelines, including studies describing CVAD‐associated complications in children (≤ 18 years old) undergoing treatment for leukemia. The Risk of Bias in Non‐Randomized Studies of Interventions (ROBINS‐I) tool was used to assess the studies' quality. Meta‐analysis and proportional meta‐analysis for non‐comparative studies (Freeman–Tukey transformation) using the random‐effects model were conducted.

Results

Six databases were searched identifying 1126 articles: 1102 were excluded and 24 studies with 13 879 CVADs were included. CVAD‐associated complications identified included: device failure 18.5% (95% CI 10.8–27.8), central line‐associated bloodstream infection (CLABSI) 31.3% (95% CI 23.6–39.6), local infection 5.4% (95% CI 2.5–9.3), CVAD‐associated venous thromboembolism (VTE) 5.2% (95% CI 2.9–8.3), dislodgement/migration 2.4% (95% CI 1.5–3.5), breakage/rupture 3% (95% CI 1.9–4.3), occlusion 5.3% (95% CI 1.5–11.2), wound dehiscence 4.4% (95% CI 2.2–7.3), and accidental removal 4.9% (95% CI 2.0–9.0). Tunneled central venous access devices (TCVADs) were associated with a higher proportion of complications when compared to totally implantable venous access devices (TIVADs).

Conclusion

Device failure and CLABSI are the CVAD‐associated complications with the highest incidence rates. Comparing CVAD subtypes, TCVADs may have a higher complication risk than TIVADs. Further large prospective studies with long‐term follow‐up are needed to investigate these findings and determine their impact on patient morbidity and mortality.

Abbreviations

CLABSI

central line‐associated bloodstream infection

CVAD

central venous access device

IJV

internal jugular vein

nt‐CVAD

non‐tunneled central venous access device

PRIMSA

Preferred Reporting Items for Systematic Reviews and Meta‐Analyses

ROBINS‐I

Risk of Bias in Non‐Randomized Studies of Interventions

TCVAD

tunneled central venous access devices

TIVAD

totally implantable venous access devices

VTE

venous thromboembolism

1. Introduction

Central venous access devices (CVADs) are indwelling catheters with distal tips positioned in large central veins, typically at the cavoatrial junction, to enable safe and sustained vascular access [1]. They are frequently used in pediatric patients with leukemia, providing intravenous access for the delivery of medications, fluids, and chemotherapy.

Several types of CVADs exist, including non‐tunneled (nt‐CVAD), tunneled (TCVAD), and totally implantable (TIVAD) [2]. nt‐CVADs, such as peripherally inserted central catheters (PICCs), are inserted percutaneously, with the catheter exiting the skin close to the cannulation site and providing relatively short‐term venous access, although they can stay in place for months. PICC ports represent an evolution of the traditional PICC, incorporating a subcutaneous port that enables intermittent access while maintaining the advantages of peripheral insertion [3].

TCVADs provide long‐term venous access; the catheter is tunneled subcutaneously to anchor it in place and exits the skin away from the venous cannulation site. Examples of such devices include Hickman and Broviac (Becton Dickinson, USA). TIVADs are TCVADs that are connected to a subcutaneous port, such as the Port‐a‐Cath (Smiths Medical, USA). As they are protected by overlying skin, they are easily concealed cosmetically, and they generally allow for a longer duration of use and reduced infection risk [2]. Device selection is guided by multiple factors, including the duration and type of treatment required, patient characteristics, and clinician preferences [2].

The use of CVADs in children can be associated with a variety of complications, with reviews indicating one in four children will experience a significant complication or device failure prior to completion of therapy [4, 5]. CVAD‐associated complications following insertion include, among others, device failure, bloodstream infections, venous thromboembolism, and various mechanical complications (e.g., rupture, occlusion). These complications often necessitate diverse interventions, which can delay therapy and, as a result, increase patient morbidity and mortality [6, 7].

Children with leukemia are a unique cohort who potentially have greater susceptibility to CVAD‐associated complications secondary to cancer pathogenesis, immunosuppression, and type of administered therapy [8]. Understanding these complications in this cohort is necessary to help guide treatment decision‐making and future research aimed at reducing these adverse outcomes. Hence, the aim of this systematic review and meta‐analysis was to determine the current evidence regarding the incidence of CVAD‐associated complications in children with leukemia.

2. Methods

We performed a systematic review and meta‐analysis to report the incidence of CVAD‐associated complications in children with leukemia.

2.1. Protocol and Registration

The Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA) guidelines were used in the formulation of this systematic review [9]. The study was registered with the PROSPERO (International Prospective Register of Systematic Reviews) database (ID: CRD420251103341).

2.2. Literature Search

A search of the literature published between January 1970 and July 2025 in the English language was performed by two independent investigators (M.S. and P.K.). They searched the following scientific databases: Ovid MEDLINE, Ovid Embase, Cochrane Central Register of Controlled Trials. Additionally, the following clinical trial registries were also searched: World Health Organization International Clinical Trials Registry Platform, ClinicalTrials.gov, and Australian New Zealand Clinical Trials Registry. Manual citation searching was also performed by screening the reference lists of articles returned from the search to identify additional relevant articles.

For scientific databases, the following search terms were used: Ovid MEDLINE—“Child* AND (central venous catheter* OR catheterization, central venous) AND (complication* OR adverse effect* OR CLABSI OR catheter‐related infections OR thromboembolism OR device removal OR surgical wound dehiscence) AND (leukemia OR hematologic neoplasms).” Ovid EMBASE—“Child* AND (central venous catheter* OR catheter) AND (complication* OR catheter related bloodstream infection OR catheter infection OR thromboembolism OR device removal OR surgical wound dehiscence) AND (leukemia OR hematologic malignancy).” Cochrane Central Register of Controlled Trials—“Child* AND (central venous catheter* OR catheterization, central venous) AND (complication* OR CLABSI OR catheter‐related infections OR thromboembolism OR device removal OR surgical wound dehiscence) AND (leukemia OR hematologic neoplasms).” For clinical trial registries, the following search terms were used: “Child* AND central venous catheter* AND leukemia.”

2.3. Screening and Quality Assessment

All English language studies published after January 1970 that described the incidence of CVAD‐associated complications in children with leukemia were considered for inclusion.

Studies were eligible for inclusion if they met the following criteria: (1) pediatric patients < 18 years of age, (2) patients with a diagnosis of leukemia, (3) CVAD inserted for any duration of treatment, (4) complications during the maintenance phase of treatment, (5) human retrospective or prospective studies. Publications of abstract‐only articles, case reports, letters to the editor, review articles, and study protocols were excluded to ensure we only included data from complete, original research.

Two authors (M.S. and P.K.) independently reviewed titles and abstracts identified by the search. Full‐text reports of potentially relevant studies were retrieved and independently assessed for eligibility by two authors (M.S. and P.K.). Two authors (M.S. and P.K.) independently extracted relevant data and assessed the quality of included studies using the ROBINS‐I (Risk of Bias in Non‐Randomized Studies of Interventions) tool [10]. We resolved any disagreements over the eligibility of studies, quality assessment, and data extraction by consensus and discussion with the senior author (M.P.).

For included studies, we extracted data describing study characteristics, participant characteristics, intervention characteristics, and the incidence of CVAD‐associated complications. Table 1 outlines the complications analyzed in this review and their respective definitions.

TABLE 1.

CVAD‐associated complications.

Complication	Definition
Device failure	Need for removal of the device prior to completion of planned therapy for any reason [11]
2CLABSI	A laboratory confirmed bloodstream infection that is not secondary to an infection at another body site [12]
3Local infection	Presence of erythema, swelling, tenderness at the insertion site ± fever [12]
4CVAD‐associated VTE	Clinical symptoms (pain/erythema/swelling/line dysfunction) and ultrasound evidence of CVAD‐associated VTE [13]
5Dislodgement/migration	Any movement resulting in the CVAD migrating out of a central vein [13]
6Breakage/rupture	Split or break in the CVAD material resulting in external leakage and/or radiographic evidence of internal leak [14]
7Occlusion	Unable to aspirate and/or inject into CVAD [13]
8Wound dehiscence	Separation of the margins of surgical incision with/without protrusion of device [15]
9Accidental removal	Removal of the device by the patient prior to completion of planned therapy [16]

Abbreviations: CLABSI, central line‐associated blood stream infection; CVAD, central venous access device; VTE, venous thromboembolism.

2.4. Statistical Analysis

Data are reported as the number of cases and proportion (%) with corresponding 95% confidence intervals (CIs). We conducted a meta‐analysis using a random‐effects model; for non‐comparative studies, we conducted a proportional meta‐analysis using the Freeman–Tukey double arcsine transformation and a random‐effects model. Heterogeneity was assessed using the I ² statistic, with values > 50% considered indicative of substantial heterogeneity. p < 0.05 were considered significant. These analyses were undertaken using MedCalc (version 23.2.6) [17]. Our primary outcome measure was the incidence of specific CVAD‐associated complications as described in Table 1. Subgroup analysis on CVAD type was performed where data were available.

3. Results

3.1. Search Results

Figure 1 shows the PRISMA flow diagram outlining our search results. A total of 1126 titles were identified from the literature search. Following the removal of duplicates, 915 titles and abstracts were screened, and 724 records were excluded as they did not meet the inclusion criteria. One hundred eighty‐five full‐text articles were then screened, resulting in 161 articles not meeting the eligibility criteria. Therefore, 24 studies were ultimately included.

FIGURE 1.

Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA) flow diagram.

3.2. Quality of Included Studies

All of the included studies were at moderate risk of bias. Table 2 reports the results of the risk of bias assessment.

TABLE 2.

Quality of included studies assessed using the Risk of Bias in Non‐Randomized Studies of Interventions (ROBINS‐I) tool.

Authors	Confounding	Selection of participants into the study	Classification of interventions	Deviations from intended interventions	Missing data	Measurement of outcomes	Selection of the reported result	Overall bias
Abbas et al. [18]	Moderate	Low	Low	Low	Low	Moderate	Low	Moderate
Ataseven et al. [19]	Moderate	Moderate	Low	Low	Low	Moderate	Low	Moderate
Bergmann et al. [20]	Moderate	Low	Low	Moderate	Low	Moderate	Low	Moderate
Berrueco et al. [21]	Moderate	Low	Low	Low	Low	Low	Low	Moderate
Carr et al. [22]	Moderate	Low	Low	Low	Low	Moderate	Low	Moderate
Cesca et al. [23]	Moderate	Low	Low	Low	Moderate	Moderate	Low	Moderate
Demircioglu et al. [24]	Moderate	Low	Low	Low	Low	Low	Low	Moderate
Gonzalez et al. [25]	Moderate	Low	Low	Low	Low	Low	Low	Moderate
Handrup et al. [26]	Moderate	Low	Low	Low	Low	Low	Low	Moderate
Jarvis et al. [27]	Moderate	Moderate	Low	Low	Moderate	Low	Low	Moderate
Junqueira et al. [28]	Moderate	Low	Low	Low	Low	Low	Low	Moderate
Katsibardi et al. [29]	Moderate	Moderate	Low	Low	Low	Low	Low	Moderate
Ligia et al. [16]	Moderate	Low	Low	Low	Low	Low	Low	Moderate
Loh et al. [30]	Moderate	Low	Low	Low	Low	Low	Low	Moderate
Madsen et al. [31]	Moderate	Low	Low	Low	Low	Low	Low	Moderate
Male et al. [32]	Low	Low	Low	Low	Moderate	Low	Low	Moderate
Martynov et al. [33]	Moderate	Low	Low	Low	Low	Low	Low	Moderate
McLean et al. [34]	Moderate	Low	Low	Low	Moderate	Low	Low	Moderate
NoaillyCharny et al. [35]	Moderate	Low	Low	Low	Low	Low	Low	Moderate
Ren et al. [36]	Moderate	Low	Low	Low	Low	Low	Low	Moderate
Rogers et al. [37]	Moderate	Low	Low	Low	Low	Low	Low	Moderate
Ruiz‐Llobet et al. [38]	Moderate	Low	Low	Low	Moderate	Low	Low	Moderate
SezginEvim et al. [39]	Moderate	Low	Low	Low	Low	Moderate	Low	Moderate
Zachariah et al. [40]	Moderate	Low	Low	Low	Low	Low	Low	Moderate

3.3. Characteristics of Included Studies

Characteristics of the included studies are outlined in Table 3. All included studies were cohort studies: 16 retrospective studies and eight prospective studies.

TABLE 3.

Characteristics of included studies.

Authors	Patients (n)	CVAD total (n)	CVAD type (n)	CVAD complications studied (as classified in Table 1)
Abbas et al. [18]	148	199	TCVAD (101) TIVAD (47)	1, 2, 3, 9
Ataseven et al. [19]	52	52	TCVAD (4) TIVAD (48)	1, 2, 3, 4, 6, 7, 8, 9
Bergmann et al. [20]	136	136	nt‐CVC (77) TCVAD (59)	1, 2, 3
Berrueco et al. [21]	73	73	TIVAD (73)	2
Carr et al. [22]	115	115	TCVAD (15) TIVAD (100)	3, 5, 6, 7, 9
Cesca et al. [23]	117	117	CVADs/All	2, 3
Demircioglu et al. [24]	30	37	TCVAD (31) TIVAD (6)	1, 3, 4, 5, 6, 9
Gonzalez et al. [25]	172	172	TCVAD (139) TIVAD (33)	1, 2, 4, 5, 7
Handrup et al. [26]	98	98	TCVAD (63) TIVAD (35)	1, 2
Jarvis et al. [27]	47	47	CVADs/All	4
Junqueira et al. [28]	179	192	TIVAD (192)	1, 2, 3, 8
Katsibardi et al. [29]	86	86	TCVAD (86)	1, 2
Ligia et al. [16]	129	196	PICC (196)	1, 2, 3, 4, 5, 6, 7, 9
Loh et al. [30]	80	80	TIVAD (80)	1, 2, 3, 7
Madsen et al. [31]	16	16	TCVAD (16)	1, 5, 6, 9
Male et al. [32]	85	85	CVADs/All	4
Martynov et al. [33]	350	498	TCVAD (469) TIVAD (29)	2, 3
McLean et al. [34]	362	362	CVADs/All	1, 4
NoaillyCharny et al. [35]	192	295	PICC (157) TCVAD (138)	4
Ren et al. [36]	2423	2423	PICC (2423)	4
Rogers et al. [37]	40	40	TCVAD (16) TIVAD (24)	2
Ruiz‐Llobet et al. [38]	652	652	CVADs/All	4
SezginEvim et al. [39]	262	310	TIVAD (250) TCVAD (60)	1, 2, 3, 4, 5, 6
Zachariah et al. [40]	29	42	TCVAD (14) TIVAD (28)	1, 2

Abbreviations: nt‐CVC, non‐tunneled central venous catheter; PICC, peripherally inserted central catheter; TCVAD, tunneled central venous catheter; TIVAD, totally implanted venous access device.

3.4. Outcomes

Table 4 outlines the incidence rates for all CVAD‐associated complications.

TABLE 4.

Incidence rates of CVAD‐associated complications.

Type of device	Studies	CVADs	Proportion	95% CI	I 2	p
Device failure
All	14	1979	18.522	10.800–27.764	95.73%	P < 0.0001
TCVAD	5	238	32.525	3.595–72.876	97.48%	P < 0.0001
TIVAD	5	368	12.067	4.829–21.982	81.61%	p = 0.0002
CLABSI
All	15	2291	31.316	23.628–39.558	94.00%	P < 0.0001
TCVAD	2	145	37.782	4.470–80.299	96.65%	P < 0.0001
TIVAD	3	345	24.570	7.097–48.219	94.98%	P < 0.0001
Local infection
All	11	1932	5.397	2.499–9.317	90.53%	P < 0.0001
TIVAD	4	420	6.032	2.078–11.861	76.34%	p = 0.0054
CVAD‐associated VTE
All	11	4631	5.229	2.861–8.257	91.72%	P < 0.0001
PICC	3	2276	4.257	0.643–10.830	93.78%	P < 0.0001
Dislodgement/migration
All	6	846	2.399	1.479–3.533	0.00%	P = 0.8754
Breakage/rupture
All	6	726	2.987	1.878–4.342	0.00%	p = 0.5145
TCVAD	2	31	11.722	3.165–24.693	0.00%	p = 0.5387
Occlusion
All	5	615	5.281	1.486–11.219	86.19%	P < 0.0001
TIVAD	2	180	4.379	1.897–7.822	0.00%	P = 0.9627
Wound dehiscence
All	2	244	4.402	2.200–7.314	0.00%	P = 0.4623
Accidental removal
All	6	615	4.858	1.958–8.964	71.38%	P = 0.0037
TCVAD	2	31	8.854	1.720 to 20.732	0.00%	p = 0.9566

3.4.1. Device Failure

There were 14 studies (1979 total CVADs) that reported the incidence of CVAD failure. Overall, 18.5% (95% CI 10.8–27.8; I ² = 95.7%, p < 0.0001) of CVADs failed prior to the completion of planned therapy (Figure 2). TCVADs had a higher proportion of device failure when compared to TIVADs (TCVADs 32.5% [95% CI 3.6–72.9], 238 devices; TIVADs 12.1% [95% CI 4.8–22.0], 368 devices).

FIGURE 2.

Proportion meta‐analysis for device failure.

3.4.2. CLABSI

There were 15 studies (2291 total CVADs) that reported the incidence of CLABSI. Overall, 31.3% (95% CI 23.6–39.6; I ² = 94.0%, p < 0.0001) of devices developed a CLABSI (Figure 3). TCVADs had a higher proportion of CLABSI when compared to TIVADs (TCVADs 37.8% [95% CI 4.5–80.3], 145 devices; TIVADs 24.6% [95% CI 7.1–48.2], 345 devices).

FIGURE 3.

Proportion meta‐analysis for CLABSI.

3.4.3. Local Infection

There were 11 studies (1932 total CVADs) that reported the incidence of local infection. Overall, 5.4% (95% CI 2.5–9.3; I ² = 90.5%, p < 0.0001) of CVADs developed a local infection (Figure 4). Four studies also reported that the proportion of TIVADs developing a local infection was 6.0% (95% CI 2.1–11.9; I ² = 76.3%, p < 0.0054).

FIGURE 4.

Proportion meta‐analysis for local infection.

3.4.4. CVAD‐Associated VTE

There were 11 studies (4631 total CVADs) that reported the incidence of CVAD‐associated VTE. Overall, 5.2% (95% CI 2.9–8.3; I ² = 91.7%, p < 0.0001) of CVADs were associated with VTE (Figure 5). Three studies also reported that the proportion of PICCs that were associated with VTE was 4.3% (95% CI 0.6–10.8; I ² = 93.8%, p < 0.0001).

FIGURE 5.

Proportion meta‐analysis for CVAD‐associated VTE.

3.4.5. Dislodgement/Migration

There were six studies (846 total CVADs) that reported the incidence of CVAD dislodgement/migration. Overall, 2.4% (95% CI 1.5–3.5; I ² = 0.0%, p = 0.8754) of CVADs experienced dislodgement or migration (Figure 6).

FIGURE 6.

Proportion meta‐analysis for dislodgement/migration.

3.4.6. Breakage/Rupture

There were six studies (726 total CVADs) that reported the incidence of CVAD breakage/rupture. Overall, 3.0% (95% CI 1.9–4.3; I ² = 0.0%, p = 0.5145) of CVADs experienced breakage and/or rupture (Figure 7).

FIGURE 7.

Proportion meta‐analysis for CVAD breakage/rupture.

3.4.7. Occlusion

There were five studies (615 total CVADs) that reported the incidence of CVAD occlusion. Overall, 5.3% (95% CI 1.5–11.2; I ² = 86.2%, p < 0.0001) of CVADs experienced occlusion (Figure 8). Two studies also reported that the proportion of TIVADs occluding was 4.4% (95% CI 1.9–7.8; I ² = 0.0%, p = 0.9627).

FIGURE 8.

Proportion meta‐analysis for occlusion.

3.4.8. Wound Dehiscence

There were two studies (244 total CVADs) that reported the incidence of wound dehiscence. Overall, 4.4% (95% CI 2.2–7.3; I ² = 0.0%, p = 0.4623) children experienced wound dehiscence (Figure 9).

FIGURE 9.

Proportion meta‐analysis for wound dehiscence.

3.4.9. Accidental Removal

There were six studies (615 total CVADs) that reported the incidence of CVAD accidental removal. Overall, 4.9% (95% CI 2.0–9.0; I ² = 71.4%, p = 0.0037) CVADs were self‐removed (Figure 10). Two studies also reported that the proportion of TCVADs that were self‐removed was 8.9% (95% CI 1.7–20.7; I ² = 0.0%, p = 0.96).

FIGURE 10.

Proportion meta‐analysis for accidental removal.

4. Discussion

Children with leukemia are a unique cohort of patients who are potentially more susceptible to CVAD‐associated complications [8]. Despite the widespread use of CVADs in this population, the nature and frequency of related adverse events remain incompletely characterized in the literature. Existing studies are limited by small sample sizes, heterogeneous methodologies, and a paucity of long‐term follow‐up data. Consequently, the morbidity and mortality associated with CVAD complications in children with leukemia are not well defined, and their clinical relevance remains uncertain. This systematic review and meta‐analysis aimed to synthesize the current evidence regarding the incidence and outcomes of CVAD‐associated complications in pediatric leukemia patients. By consolidating available data, this study seeks to identify knowledge gaps and inform future research priorities, including the need for longitudinal studies that evaluate both short‐ and long‐term sequelae of these complications.

4.1. CVAD Failure

CVAD failure represented a notable complication, with 18.5% of devices failing and being removed prior to completion of planned treatment, though significant heterogeneity was observed in terms of device type and patient population. Device failure not only interrupts essential therapy but also necessitates additional CVAD insertions, each carrying inherent procedural risks and potential for further complications [41, 42]. In the context of leukemia, where uninterrupted vascular access is critical for delivering intensive treatment, this reinforces the imperative for robust preventative strategies. Targeted interventions to reduce device failure, whether through insertion technique, maintenance protocols, or patient‐specific risk stratification, may help preserve treatment continuity and mitigate avoidable harm [43].

4.2. CLABSI

Central line‐associated bloodstream infection (CLABSI) emerged as the most frequent complication, affecting 31.3% of CVADs that developed at least one episode during therapy, though significant heterogeneity was also observed for this complication. This rate exceeds the incidence reported in general pediatric oncology populations, where a recent meta‐analysis found a pooled CLABSI rate of 21.2% [44]. This elevated risk aligns with existing literature comparing hematological and solid malignancies, in which leukemia is consistently identified as a risk factor for increased infection and CLABSI rates, likely attributable to disease‐related immunosuppression and treatment intensity [8, 45]. Notably, neutropenia at the time of CVAD insertion has been independently associated with higher infection risk [21, 28, 34]. Given this vulnerability, our findings support the need for enhanced infection prevention strategies in patients with leukemia. Elective CVAD placement during periods of profound neutropenia should be carefully considered against clinical urgency, with potential benefit from stricter insertion protocols and targeted antimicrobial stewardship. In addition, institutional efforts to improve compliance with preventative measures have shown measurable impact. For example, Berrueco et al. reported that implementing an educational program focused on bundle adherence, staff training, and ward‐level accountability significantly reduced their institution's CLABSI rate [21].

4.3. Local Infection

Local infection remains a clinically relevant complication in pediatric leukemia patients, with 11 studies (n = 1932) reporting a pooled incidence of 5.4% with high heterogeneity. Among TIVADs specifically, four studies reported a pooled local infection rate of 6.0%, suggesting that even fully implanted devices are not exempt from site‐related complications. While TIVADs are generally considered lower risk due to their subcutaneous placement, repeated access and compromised skin healing may still predispose patients to local infection. The high heterogeneity observed between studies likely reflects differences in local infection definitions, surveillance protocols, maintenance practices, and patient characteristics such as neutropenia or compromised skin integrity [46]. This variability underscores the need for standardized definitions and prospective surveillance to accurately quantify risk and improve comparability between centers. Various preventive strategies have shown promise in reducing local infection rates; chlorhexidine‐impregnated dressings and staff education programs have all been associated with lower rates of local infections in pediatric oncology populations [47]. Implementing these evidence‐based measures alongside rigorous monitoring may mitigate local infections, prevent systemic complications, and minimize therapy disruptions.

4.4. CVAD‐Associated VTE

CVAD‐associated VTE remains a significant complication in pediatric leukemia patients, with 11 studies (n = 4631) reporting a pooled incidence of 5.2% despite high heterogeneity. Among PICCs, three studies reported a VTE rate of 4.3%, suggesting comparable risk across device types. These findings align with previous literature indicating that CVADs, regardless of subtype, pose a measurable thrombotic risk in pediatric oncology populations [48, 49].

The pathophysiology of CVAD‐associated VTE in leukemia is multifactorial, involving endothelial injury during insertion, catheter‐related flow disturbance, and a hypercoagulable state induced by malignancy and chemotherapy. While PICCs are often preferred for their ease of insertion, they may be associated with increased thrombotic risk due to smaller vessel caliber and longer catheter length. Tunneled and implanted devices are also at risk of VTE, particularly in patients with prolonged immobility or central venous stasis [49]. Given the potential for serious sequelae, including pulmonary embolism, catheter loss, and interruption of therapy, these findings underscore the need for vigilant surveillance and risk stratification. Strategies such as ultrasound screening for catheter tip positioning and consideration of prophylactic anticoagulation in high‐risk patients may help mitigate VTE incidence.

4.5. Dislodgement/Migration

CVAD dislodgement/migration was reported in six studies (n = 846), with a pooled incidence of 2.4%. Although less frequent than infectious or thrombotic complications, dislodgement remains clinically relevant due to its potential to interrupt therapy, necessitate device replacement, and increase procedural risk. The low heterogeneity across studies suggests consistent reporting and a relatively stable incidence. Dislodgement may result from inadequate securement, patient movement, or external mechanical forces, particularly in younger children or those with high activity levels. External devices such as PICCs and TCVADs are especially vulnerable, while TIVADs, being fully implanted, are generally less prone to migration. Nonetheless, improper placement or suboptimal fixation can still lead to displacement even in implanted devices. Zhang et al. recently highlighted mechanical complications, including dislodgement and migration, as a meaningful contributor to overall CVAD failure in children, reinforcing the importance of securement strategies and device selection tailored to patient activity and treatment duration [50]. Preventative measures such as sutureless securement systems, subcutaneous anchoring devices, and staff education have shown promise in reducing dislodgement rates and warrant further evaluation in high‐risk cohorts [51].

4.6. Breakage/Rupture

Mechanical failure due to catheter breakage or rupture remains a clinically relevant complication, with a pooled incidence of 3.0% (95% CI: 1.9–4.3) across six studies involving 726 CVADs, and minimal heterogeneity (I ² = 0.0%). Zhang et al. reported a notably higher fracture rate (6%) in pediatric cohorts, particularly among Hickman lines, with younger age and lower weight identified as independent risk factors [50]. Although breakage and rupture are less frequently reported than other CVAD complications such as CLABSI, their consequences can be severe, ranging from extravasation and device embolism to urgent replacement procedures and therapy interruption. The observed rate aligns with prior large‐scale analyses in adults, such as Corley et al., who reported catheter fracture among the spectrum of CVAD failures in 10.2% of devices, with fracture contributing to early removal and increased healthcare burden [52]. Device integrity is influenced by multiple factors, including material composition, insertion technique, dwell time, and mechanical stress during flushing or medication administration. Despite the relatively low incidence, the clinical implications of CVAD breakage warrant proactive prevention strategies. These include staff education on flushing pressures, use of pressure‐limiting syringes, and early recognition of device fatigue. Future research should explore device‐specific rupture rates and evaluate the impact of novel materials and securement technologies on mechanical failure.

4.7. Occlusion

We identified an overall CVAD occlusion rate of 5.3% (95% CI 1.5–11.2), with substantial heterogeneity across studies (I ² = 86.2%). The variability likely reflects differences in device type, patient population, flushing protocols, and definitions of occlusion. Notably, the pooled occlusion rate for TIVADs was lower at 4.4% (95% CI 1.9–7.8), with no observed heterogeneity (I ² = 0.0%), suggesting more consistent outcomes in this subgroup. CVAD occlusion remains a clinically significant complication, often leading to therapy interruption, device replacement, and increased healthcare costs. Elliott et al. reported occlusion rates ranging from 14% to 36% in acute care settings, underscoring the persistent challenge of maintaining catheter patency despite advances in device design and care protocols [53]. Occlusion mechanisms are multifactorial, including thrombotic buildup, precipitate formation from incompatible infusates, and mechanical obstruction due to kinking or malposition. Corley et al. identified occlusion as one of the leading causes of CVAD failure in hospitalized adults, with implications for early device removal and reduced treatment efficacy [52]. Their multivariable analysis emphasized the role of flushing technique, catheter material, and dwell time in predicting occlusion risk. Given the moderate incidence and potential consequences, proactive occlusion prevention strategies are essential. These include protocolized flushing regimens, use of locking solutions, and staff education on early recognition and management. Future studies should stratify occlusion risk by device type, patient age, and clinical indication to inform tailored interventions and improve CVAD outcomes.

4.8. Wound Dehiscence

The findings from the meta‐analysis of 11 studies, encompassing 1932 central venous access devices (CVADs), indicate that 5.4% (95% CI 2.5–9.3; I ² = 90.5%, p < 0.0001) of CVADs developed local infections, highlighting a significant clinical concern in the management of these devices. The high heterogeneity (I ² = 90.5%) suggests variability in study populations, methodologies, or infection control practices, which warrant further investigation to identify contributing factors. These results align with previous research by Ullman et al. who noted that local infections (which could lead to wound dehiscence) in CVADs are often underreported but critical due to their potential to progress to systemic infections [4]. The observed infection rates in our systematic review and meta‐analysis emphasize the need for standardized protocols in CVAD insertion and maintenance. Future studies should focus on addressing heterogeneity and exploring preventive strategies to reduce local infection rates [54].

4.9. Accidental Removal

Accidental removal remains a clinically significant complication, particularly in pediatric populations. This meta‐analysis identified a pooled incidence of 4.9% (95% CI 2.0–9.0) across six studies, with notable heterogeneity (I ² = 71.4%). The rate was higher among TCVADs, at 8.9% (95% CI 1.7–20.7), suggesting device type and securement method may influence vulnerability to dislodgement. Accidental removal can result in therapy interruption, additional procedures, and increased infection risk. Curtis et al., in a scoping review of oncology CVADs, highlighted inconsistent terminology and documentation around premature removal, which may obscure true incidence and hinder prevention efforts [55]. Given the moderate incidence and potential for escalation, strategies to reduce accidental removal should include standardized securement protocols, patient and caregiver education, and early identification of high‐risk scenarios. Future research should stratify risk by age, diagnosis, and device type, and evaluate the impact of securement innovations and behavioral interventions in pediatric cohorts.

4.10. Comparison Between Different CVADs

Finally, this meta‐analysis also sought to examine CVAD‐associated complications in relation to device subtype. Of the complications assessed, only device failure and CLABSI had sufficient data for comparison between TCVADs and TIVADs. TCVADs were associated with a markedly higher incidence of both device failure (32.5% vs. 12.1%) and CLABSI (37.8% vs. 24.6%) compared to TIVADs. These findings are consistent with prior meta‐analyses in pediatric oncology and general pediatric populations, which similarly report elevated complication rates among TCVAD [4, 44]. The observed differences may be attributable to structural and anatomical factors. TIVADs benefit from a subcutaneous port protected by overlying skin, whereas, TCVADs have an external exit site that may serve as a conduit for bacterial entry and is more susceptible to mechanical complications such as dislodgement, fracture, or breakage [2]. Despite these insights, the current literature remains limited in its comparative evaluation of CVAD subtypes across the full spectrum of complications in pediatric leukemia patients. Further prospective studies are warranted to delineate the relative risks and therapeutic roles of TCVADs, TIVADs, and PICCs in this high‐risk cohort.

4.11. Limitations

The findings of this systematic review and meta‐analysis should be interpreted with several potential limitations. The available literature was predominantly retrospective in nature, introducing potential risks of information and selection bias. The incidence rate could not be described as complications per 1000 catheter days as many of the included studies did not report this. An agreed standard of reporting would allow for better statistical comparisons between studies and populations. Additionally, considerable heterogeneity was observed across study populations, reflected in the statistical variability of pooled estimates. This likely stems from international inconsistencies in CVADs maintenance standards and evolving clinical practices. While this review synthesizes all available data on CVAD‐associated complications in pediatric leukemia, it is important to acknowledge that vascular access protocols have undergone significant changes over the past decade. As such, the pooled estimates may not fully reflect contemporary complication rates or current standards of care. To more accurately define the burden of CVAD‐related morbidity and mortality in this population, future research should prioritize large‐scale, prospective studies with standardized definitions, consistent follow‐up, and stratification by device type and treatment phase.

5. Conclusions

CVAD‐associated complications are prevalent among pediatric patients with leukemia, highlighting the critical need to optimize preventative strategies and better understand their contribution to treatment‐related morbidity and mortality. In this population, device failure and CLABSI emerged as the most frequent and clinically significant complications. Comparative analysis of device subtypes suggests that TCVADs are associated with higher complication rates than TIVADs, likely reflecting structural and maintenance‐related vulnerabilities. These findings highlight the importance of targeted research to guide device selection and inform best practices. Future large‐scale prospective studies with long‐term follow‐up are essential to validate these observations and support the development of evidence‐based interventions aimed at reducing CVAD‐related harm in this high‐risk cohort.

Funding

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

All data underlying the findings of this systematic review and meta‐analysis were extracted from published articles that are publicly available in indexed databases.