Infection Associated With Invasive Devices in Pediatric Health Care: A Meta-analysis Free

We prospectively developed and registered a review protocol (PROSPERO registration number: CRD42021254922). The review has been reported in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses¹¹  and the Meta-analysis of Observational Studies in Epidemiology (MOOSE) reporting guidelines.¹² 

A systematic search for published and unpublished cohort studies and clinical trials examining complications of indwelling invasive devices used in pediatric healthcare was initially conducted on the 25th of May 2021, and an updated search was conducted on the 16th of June 2022. The Medical Subject Heading-informed search criteria were developed by a health librarian (Supplemental Table 5; references in online-only supplementary materials) and searched databases for published studies (Medline [via PubMed], CINAHL, Embase, Web of Science, Scopus, Cochrane CENTRAL), clinical trial registries (Clinicaltrials.gov, Australian New Zealand Clinical Trials Registry, International Clinical Trials Registry Platform), and unpublished study databases (ProQuest Dissertations and Theses, Open Access Theses and Dissertations, and MedNar). Reference lists of included articles were also screened. The results of all searches were entered into the Covidence software program¹³  for screening. A large search was performed to identify a range of invasive device outcomes (infection, thrombosis, device performance), with the results for each complication presented separately for clarity and clinical utility.

To be included, primary studies had to include: (1) indwelling invasive devices (ie, a medical device penetrating inside the body, either through an orifice or skin and remaining in place for an extended period (ie, not only during procedures) commonly used in pediatric healthcare,^14,15  (2) pediatric participants admitted to a hospital (postneonatal or maternal discharge to 18 years), (3) reporting device-associated infections, (4) using cohort (retrospective or prospective) or a clinical trial (with an established standard care arm) design. Studies with full-text data, available in English, and published between January 2011 to June 2022 were included. We excluded studies based in neonatal intensive care units and home settings because of separate health service systems and clinical teams.

Cause-specific infections were evaluated, including: (1) tissue infections, (2) ventilator-associated pneumonia, (3) urinary tract infections, (4) other organ infections, and (5) bloodstream infections. The definitions were aligned with international guidelines (Table 1).^16–18 

As per Meta-analysis of Observational Studies in Epidemiology recommendations,¹²  the eligibility criteria were applied to the systematic search results, and all identified references were screened independently by 2 reviewers (M.T., A.H., G.X., and/or A.L.) using a 2-stage approach. The screening of titles and abstracts and the full-text review were performed by 2 authors using Covidence systematic review.¹³  Any disagreement between these investigators was resolved by discussion with the senior author (A.J.U.).

Two authors (A.H., G.X., and/or A.L.) independently extracted relevant data, including author, publication year, country, study design, setting, number of patients and devices, device type, study method, the case of device failure or infection complications, and device days. The extracted data were reviewed by a third investigator (M.T.).

Each article was independently evaluated for methodological quality by 2 reviewers using the Joanna Briggs Institute Critical Appraisal Checklist for systematic reviews of prevalence and incidence.¹⁹  Any disagreement was resolved by a third reviewer (M.T.). Each question was assessed as either a “yes,” “no,” “unclear,” or “not applicable” with the total number of points for each paper calculated. A quality score was assigned to each article, represented as a percentage, which was calculated by dividing the total number of points scored by a paper by the total number of questions for the given tool. Studies were not excluded based on their bias assessment.

Score confidence intervals (CI) with Freeman-Tukey double arcsine transformations were calculated for studies with dichotomous outcomes (infection or no infection), and Poisson confidence intervals and standard errors were calculated for incidence rate (IR) outcomes. Pooled estimates were generated with random-effects meta-analysis and presented with a 95% CI. IR outcomes (continuous data) were pooled by using inverse variance with the DerSimonian and Laird method, per 1000 device days and 95% CI. Heterogeneity between studies was assessed using I-squared statistics, categorized as low (<25%), moderate (25% to 75%), or high (>75%).²⁰  Outcomes described in a single study were reported descriptively. Posthoc subgroup analyses by patient settings (intensive care, oncology and hematology, all pediatrics (general), renal, cardiac, neuro, and gastro-enterology-nutrition and hepatology) were conducted to examine differences in outcomes between varying risk patient populations and to investigate heterogeneity, which are reported in the main findings per each device. Sensitivity analyses were informed by the risk of bias (ROB) and study design. The ROB of the included studies were assessed using the Joanna Briggs Institute tool for prevalence studies (Supplemental Table 6).¹⁹  The defined questions will be scored with 1 for “Yes” and 0 for “No” or “Unclear.” The total score of each article will be calculated by the sum of its points. Based on this tool, studies will be rated as low risk and high risk with scores 0 to 7 and 8 to 9, respectively. Statistical analysis was performed using Stata 15 (Stata Corp, College Station, TX), with a statistical significance of P < .05.

Figure 1 illustrates the study selection, summarized with a Preferred Reporting Items for Systematic Reviews and Meta-Analyses study flow diagram. The initial search identified a total of 3951 potentially relevant studies. After screening the titles and abstracts, 744 full-text articles were assessed for eligibility. Eventually, 116 studies met the inclusion criteria of infection-related complications and were included in the review (Supplemental Table 6).

The summary of the included studies is shown in Table 2. A total of 61 554 devices and 3 632 364 device days were included. Most studies were retrospective (56 studies; 48%) or prospective (45 studies; 39%) cohort studies, based in North America (39 studies; 38%), involving children admitted to the PICU (48 studies; 47%) or across a pediatric hospital (28 studies; 27%). Eight invasive devices were found in the selected studies. The most frequently studied device was the central venous access device (CVAD; 60 studies; 46%), followed by the endotracheal and tracheostomy device (41 studies; 31%). There were 56 studies (48%) that scored over 80% on quality scores (Supplemental Table 6), but common quality issues included unclear outcome definition, conditions not measured in a standard, reliable way for all patients, and unclear response rate or management. The number of studies varied among different invasive device types and outcomes, with laboratory-confirmed bloodstream infection in central venous access devices (CVADs; nontunneled catheters, tunneled cuffed or uncuffed catheters, implanted ports, hemodialysis, and peripherally inserted central catheters) being most reported (n = 50 studies for proportion and n = 39 studies for rates), followed by ventilator-associated pneumonia (VAP) in endotracheal and tracheostomy device studies (n = 34 studies for proportion and n = 16 studies for rates).

Table 3 presented the pooled proportion and IR of different types of infections by device type.

As the most frequently reported invasive devices, 8% of CVADs developed a laboratory-confirmed bloodstream infection (95% CI, 6–11; 50 studies; 46 014 devices), with an IR of 0.96 per 1000 device days (95% CI, 0.78–1.14; 39 studies; 1 770 107 device days). In the case of CVADs, tissue infection was reported in 3% of the CVADs (95% CI, 0–8; 9 studies; 2162 devices) with an IR of 0.18 per 1000 device days (95% CI, 0.00–0.36; 7 studies; 618 374 device days). The Centers for Disease Control and Prevention (CDC) or/and the National Healthcare Safety Network (NHSN) definition for CLABSI was used in 13 (26%) out of 50 studies for proportion, and 15 (38%) out of 39 studies for IR analysis for CVAD-associated BSIs.

Subgroup analyses of proportion and incidence rates (IRs) of infection across device types by settings are summarized in Table 4. CVAD-associated bloodstream infection in PICU was 3% (95% CI, 2–4; 9 studies; 9768 devices), whereas 15% (95% CI, 9–21; 15 studies; 9 953 devices) in oncology and hematology settings and 19% (95% CI, 4 studies, 10–30; 162 devices) in renal settings. However, IR was lower in the oncology and hematology setting (IR 0.79, 95% CI, 0.53–1.04; 12 studies; 1 004 040 device days) compared with PICU (IR7.50, 95% CI, 3.26–11.74; 7 studies; 14 817 devices days). Analyzing by patient setting, the heterogeneity values were either high (>75%) or could not be calculated because of the small number of studies.

PIVCs had a bloodstream infection proportion of 1% (95% CI, 0–5, 105 devices) with IR of 0.14 per 1000 devices days (95% CI, 0.00–0.53; 1 study, 7 071 device days); however, this finding was based on a single study involving 105 devices in all pediatric setting.

VADs had 55% pooled proportion of device-related bloodstream infections (95% CI, 45–65; 3 studies, 104 devices, Supplemental Fig 2) with IR of 41.16 per 1000 device days (95% CI, 20.19–62.13; 1 study, 413 devices). Device-related tissue infection pooled proportion was 38% (95% CI, 15–63, 6 studies, 502 devices) and IR of 2.58 per 1000 device days (95% CI, 1.34–3.81; 1 study, 7376 device days). There were only 2 patient settings (PICU and Cardiac) analyzed in most subgroup analyses, which yielded similar proportions across complications. Analyzing by patient setting, the heterogeneity values were either high (>75%) or could not be calculated because of the small number of studies.

The nasogastric and gastrostomy tubes had a pooled proportion of tissue infection of 6% (95% CI, 5–8, 2 studies, 1396 devices), which were from all pediatrics (4%, 95% CI, 3–5, 1289 devices) and gastro-enterology-nutrition and hepatology settings (64%, 95% CI, 54–72, 107 devices).

When considering peritoneal dialysis catheters, 2 types of infections were reported. Tissue infection had a pooled proportion of 7% (95% CI, 0–23, 4 studies, 1105 devices), whereas organ infection had a pooled proportion of 21% (95% CI, 8–39, 6 studies, 613 devices). Organ infections were similar in proportion for both settings (all pediatrics and renal).

There were only 2 studies that reported other organ infections for external ventricular drainage, which reported a pooled proportion of 6% (95% CI 4–9, 416 devices) with a similar proportion for neurology and PICU setting.

There were 34 studies reporting on VAP in endotracheal and tracheostomy devices with a proportion of 19% (95% CI, 14–24; 34 studies; 11 498 devices) and an IR of 14.08 per 1000 device days (95% CI, 10.57–17.58, 16 studies; 47 321 device days). Most studies (28 out of 34 studies) were in PICU settings. There was a similar proportion of VAP in PICU (18%, 95% CI, 13–23; 28 studies; 10 962 devices), oncology and hematology (17%, 95% CI, 3–56; 1 study; 6 devices) and cardiac settings (18%, 95% CI, 3–41; 292 devices) and neuro setting had a higher proportion (34%, 95% CI, 28–40, 2 studies, 238 devices). The CDC or/and the NHSN definition for VAP was used in 30 (88%) out of 34 studies for proportion and 14 (88%) out of 16 studies for IR analysis.

The pooled proportion of urinary tract infection (UTI) was 4% (95% CI, 1–25, 1 study, 28 devices) with an IR of 2.86 per 1000 device days (95% CI, 1.37–4.36; 7 studies; 1 611 057 device days). CDC or/and the NHSN definition for UTI was used in 1 (100%) out of 1 study for proportion and 7 (100%) out of 7 studies for IR analysis.

Sensitivity analysis for infections by the ROB and study type is reported in Supplemental Table 7 and 8. Generally, lower ROB studies had a higher pooled proportion (except for VADs and PD catheters reporting tissue infection) and IR of tissue infection for CVADs and UTIs for urinary catheters. Prospective study design had a lower pooled proportion compared with retrospective design, except for tissue infection for CVAD and nasogastric tube, VAP, other infection in both peritoneal dialysis catheters and external ventricular drainage devices, and generally higher IR except for VAP and UTI. Heterogeneity was high (>75%) or cannot be calculated because of small number of studies except for VAP with low ROB, which had low score for heterogeneity.

This systematic review provides a comprehensive summary of infection-related complications associated with commonly used pediatric invasive devices. Synthesizing 116 studies across 61 554 devices and 3 632 364 device days, this analysis has demonstrated that hospitalized children remain at considerable risk for many significant HAIs, associated with their essential devices. Some children with complex and chronic health conditions will have greater than 30 invasive devices in the first year of their life alone.²¹  Although varying in severity, infections can change their health trajectory, causing morbidity, mortality, and delayed treatment.^22,23  Our estimates could act as a guide for quality improvement and innovation projects throughout pediatric healthcare that bridge quality, safety, and research agendas.

Compared with adults, pediatric patients are considered at increased risk of many HAIs with vascular access devices (eg, CLABSI) but at lower risk for others devices (eg, VAP, UTI).^24,25  Over the past 20 years, HAIs have been investigated extensively and used as the benchmark of hospital performance internationally.²⁶  However, attention has been primarily placed on specific devices, especially CVADs. The resources and attention dedicated to CLABSI may be working, with our estimate demonstrating reduced proportions and rates per 1000 device days of CVAD-associated BSIs (8%; 95% CI, 6–11; IR 0.93; 95% CI, 0.75–1.17) when compared with the 2015 estimate (10%; 95% CI, 9–12; IR 1.63; 95% CI, 1.40–1.86]).²⁷  The systems developed to support this reduction have been extensive, including continuous surveillance (endorsed by organizations like the Center of Disease Control and Prevention²⁸  and International Nosocomial Infection Control Consortium²⁵ ), clinical practice guidelines,^16,29  and a multitude of bundle- and checklist-based innovations^4,30  to improve clinician compliance to evidence-based practice principles. A federally funded hospital engagement network improvement model, Children’s Hospitals’ Solutions for Patient Safety Hospital Engagement Network, have successfully reduced the CLABSI (coefficient  =  −0.152; 95% CI, −0.213 to −0.019)⁶  and CAUTI rates (2.55 to 0.98 infections per catheter line days)³¹  across up to 128 hospitals in the United States and Canada. The achievements of the Children’s Hospitals’ Solutions for Patient Safety Hospital Engagement Network serve as a promising example of how collaborative initiatives and shared knowledge can drive substantial improvements in healthcare outcomes. These results underscore the importance of ongoing surveillance, continuous learning, and the widespread adoption of best practices to enhance patient safety and minimize healthcare-associated infections.

Our systematic review has also highlighted several inequities. In particular, most studies were condcuted in highly technical, resource rich settings (eg, North America, PICUs). The study also draws attention to multiple understudied devices and outcomes, often not included in traditional, infection-related surveillance reports, for example, PD catheters and NG or G tubes (7%, 95% CI, 0–23 and 6%, 95% CI, 5–8, respectively).²⁵  Additionally, complications associated with the use of external ventricular drainage catheters (3 studies, 489 devices) were understudied despite the relative risk and burdensome sequelae of local organ infection. A rigorous process to ensure equitable reporting across variable populations, settings, devices, and outcomes is necessary to reduce the sequelae of device-associated HAIs across pediatric healthcare. For this to occur, we need consistent, high-quality, and practical definitions, which are essential to facilitate benchmarking and knowledge sharing. Even within CVAD-associated BSI reporting, definition variation remains (eg, clinical symptoms versus laboratory testing) when comparison was made internationally. More than half of the studies did not mention any standard definitions, such as CDC, NHSN, or Infectious Diseases Society of America 2009 guideline.^32,33  Using nonstandardized definitions can impact the reliability and reproducibility of individual studies. This variation is even more extensive in less studied devices, such as PD catheters, venous access devices (VAD)s, external ventricular drainage, and NG and G tubes, where infections may not be routinely collected in surveillance as for CLABSI, CAUTI, and VAP. Nevertheless, these devices can be responsible for infections with the potential to cause detrimental medical outcomes. These findings are consistent with previous studies and highlight the need for continued efforts to prevent, monitor, and manage device-related infections in pediatric patients.^34–36 

Although there were fewer studies, VAD studies had device-related bloodstream infection, tissue infection, and other organ infections reported across 19 studies and 1696 devices. Although this result is limited by the comparatively small sample size and presence of multiple concurrent devices, the overall picture of high risk in a highly vulnerable patient group is clear. Infections in this population can severely affect patient outcomes and should be a priority for technological improvements and innovation.^37,38  Strategies borrowed from other invasive device types may be beneficial, such as incorporating antimicrobial materials, as previously used in CVADs.³⁹ 

This review provides insights into the magnitude of infections associated with invasive device use in pediatric healthcare and provides opportunities for improvements in practice and patient outcomes. However, the review has limitations. Firstly, this systematic review included only the inpatient studies; hence it may not be generalizable to outpatient and community settings. Because of some studies not reporting a breakdown of infections, total device numbers or total device days, some data were not suitable for meta-analysis, which might have excluded some larger surveillance studies.^40,41  Additionally, consistent definitions were absent across all device types, and the ambiguous infectious outcomes were removed. Some studies lacked clarification on whether the complications were because of the device or because of the treatment regimen the patient was on (ie, bloodstream infections in ECMO or venous access device), and these studies were removed from the meta-analysis. This removal of studies may have resulted in estimate imprecision. Secondly, the estimates could also be underestimated for some devices because of lower publication frequency and relying on author-defined outcome events. A future study incorporating individual participant data meta-analysis method⁴²  or weighted analysis using utilization rate could potentially quantify a more reliable estimate. Lastly, overall, the meta-analysis had high heterogeneity across studies and within subgroups. We have attempted to explore potential sources of heterogeneity through sensitivity analyses and subgroup analyses. Although the heterogeneity between subgroups has been reduced by dividing studies based on quality and type, there is still substantial heterogeneity within patient settings. The hospitalized pediatric population is inherently heterogeneous in nature, as it encompasses a wide range of age groups, disease types, and treatment regimens.⁴³  This may contribute to the observed clinical heterogeneity in the meta-analysis. Despite the limitations, this systematic review can provide an opportunity to target innovation, highlighting the priority areas for further investigation.

In conclusion, this systematic review highlights the significant risk of infection-related complications associated with invasive devices in pediatric healthcare. Although BSI was frequently reported in CVAD studies and VAP in endotracheal and tracheal device studies, other infections, such as tissue infection and local organ infection for other devices, were rarely reported. This systematic review has also identified understudied devices and complication outcomes that should be considered in future research. VADs had a high infection proportion and rate, which requires future focused research and reporting regulations, to isolate device-related complications resulting from other concurrent devices. Because of the high rates and potential sequelae of tissue and local organ infection estimated, nasogastric or gastronomy tube and peritoneal dialysis require standard international practice guidelines and compliance initiatives similar to those dedicated to CVADs. The estimates in the meta-analysis could be used for quality improvement in both clinical practice and research areas, for which we recommend using consistent, high-quality, and practical definitions and knowledge sharing. Reducing the sequelae of device-associated HAIs in pediatric healthcare is essential, and the findings of this study highlight the need for ongoing research and innovation in this area.

We would like to acknowledge and thank the health librarian, Mr Dave Honeyman, for his assistance in putting together our search strategy.