Contemporary Insertion of Central Venous Access Catheters in Pediatric Patients: A Retrospective Record Review Study of 538 Catheters in a Single Center

Abstract

Objective:

To analyze the rate of infections and complications after surgeon-performed, largely ultrasound-guided, central venous catheter (CVC) placement in a pediatric population and to identify patients at high risk of complications.

Methods:

All children aged between 4 months and 19 years with a percutaneous CVC inserted between January 1, 2000, and July 31, 2013, were included. Patient records were reviewed retrospectively for the occurrence of infection and other complications until CVC removal or the last outpatient clinic visit and compared between patient groups and with the recent literature.

Results:

A total of 538 CVCs were placed in 345 patients. Eight patients (1.5%) suffered complications during placement. There were 84 cases of a suspected CVC infection (15.6%). Catheter-related infections with a positive catheter tip culture occurred in 25 cases (4.6%). Older patients (odds ratio [OR] 0.88; 95% confidence interval [CI] 0.78–0.98) or patients with a double-lumen 7 French Bard-Hickman catheter (OR 0.32; 95% CI 0.11–1.00) had a significantly lower risk of infection. Older patients (OR 0.94; 95% CI 0.89–0.99) and patients with beta-thalassemia (OR 0.35; 95% CI 0.17–0.71) also had a significantly lower risk of suspected infection.

Conclusion:

In general, infection rates in our series were similar to those in the recent literature. Younger patients seem to be at higher risk for CVC removal because of infection prior to the end of treatment. Patients with beta-thalassemia or receiving a double-lumen 7F Bard-Hickman catheter had a lower risk of infection.

Children with severe diseases may require repeated intravenous therapy and blood draws. For this purpose, central venous access catheters (CVCs) are placed routinely in this population. Traditionally, CVCs in small children were inserted using either open surgical procedures or non-visualized percutaneous catheter placement. Unfortunately, these approaches are quite invasive and may carry a significant risk of bleeding and infection [1 –3] or a risk of pneumothorax related to multiple puncture attempts [4].

Recently, ultrasound-guided CVC placement using the internal jugular vein in children has been introduced, and initial data suggest that this approach is feasible and safe [5]. However, post-operative central line infections remain a problem [6], leading to CVC removal prior to the end of treatment and often necessitating prolonged hospitalization or multiple operative procedures [7]. Previous studies did not identify patient groups with a higher risk of catheter-related infections. Moreover, these studies did not report the relation between the use of antibiotics and catheter-related infections [1 –3,5]. These matters are discussed in this paper.

At the Leiden University Medical Centre, surgeon-performed ultrasound guidance is used during CVC placement in children with hematologic and oncologic disease. The purpose of the present study was to analyze the complication and infection rates in a population of children at high infection risk to identify subgroups of patients at especially high risk for central line infections.

Patients and Methods

Patients

Between January 1, 2000, and July 31, 2013, 345 patients aged 19 years or younger with a CVC had at least one hospital admission at the Leiden University Medical Centre. In total, 538 CVCs were inserted and/or removed. All CVCs mentioned in this study as being removed were taken out in our hospital. Patients with peripherally inserted central catheters (PICCs) or catheters that were anatomically placed in a vein other than the internal and external jugular vein or the subclavian vein were excluded. Over the years, the number of CVCs inserted by surgeons in our center increased from 1 in 2000 per year to 65 per year in 2012.

Surgical technique and peri-operative management

Antibiotics were administered as prophylaxis. Flucloxacillin was the standard drug, and vancomycin was used when patients already had a CVC in place or were at higher risk for catheter-related infections.

The CVCs were inserted or removed with patients under general anesthesia and arranged in a supine Trendelenburg position. The chest and neck were sterilely exposed. With the neck slightly extended and turned to the non-operative side, the internal jugular vein was identified using a 14.7-MHz ultrasound probe. When it was suitable for cannulation, a small skin incision was made two cm cranially from the planned puncture site through which a 19-gauge hollow needle on a syringe filled with heparin water was advanced into the vein under ultrasound guidance. When the position was confirmed with ultrasound in combination with unobstructed inflow and outflow, the syringe was removed from the needle, and a 0.035-inch (0.89-mm) guidewire was advanced through the jugular vein toward the right atrium. The proper position was confirmed by radiography. The catheters were placed by a surgical resident plus one of four consultant surgeons.

A skin incision was created in the chest. A subcutaneous trajectory was tunneled from the chest to the skin incision in the neck. The pre-flushed catheter was advanced through the tunnel from chest to neck. A dilator was advanced and retrieved over the guidewire, followed by the catheter. The depth of the catheter (just above the right atrium) was confirmed through plain radiography together with proper inflow and outflow, after which, the catheter was flushed with heparin and locked to an intravenous drip. Before the first use, CVC function was checked routinely by the surgeon by drawing blood.

Post-operative management included standardized pain measurement (visual analog scale) by a nurse and, once daily, a check of the CVC area for any signs of hematoma or erythema. Antibiotic (usually flucloxacillin or vancomycin) was administered in case of expanding erythema or other signs of infections: Fever, sepsis, or laboratory results showing increased infection parameters. Any bandages or adhesive plaster were removed within 24 hours after CVC placement.

Definitions

Patient records were reviewed retrospectively to assess central line infections and any other post-operative complications. An infected CVC was defined in two ways. The first, more sensitive, definition was removal because of a suspected infection. Suspicion of CVC infection was based on the occurrence of fever or blood samples showing increased infection parameters (leukocyte count and C-reactive protein). The second, more strict, definition covered removal with a positive culture tip (a catheter tip contaminated with bacteria other than skin flora). This differs slightly from the definition used by the U.S. Centers for Disease Control and Prevention (CDC). According to the CDC, a central line-associated blood stream infection (CLABSI) is defined by a blood culture collected through a vascular catheter that is positive for an organism. Catheter tip cultures are not used for the CLABSI criteria because (according to the CDC) not all laboratories are able to perform quantified catheter tip cultures [8]. As a result, our strict definition is likely to give estimates somewhat lower than the CLABSI criteria and our sensitive definition higher estimates. In addition, we assessed whether the following peri-operative complications occurred: Hemothorax, pneumothorax, arterial cannulation, bleeding, compromised sterility, and epidermolysis. Follow-up was based on the date of CVC removal or on the last visit of the patient to the outpatient clinic, during which it was confirmed that the CVC was still in situ.

The following pre-operative characteristics were collected at the time of insertion: Diagnosis (leukemia, osteosarcoma, beta-thalassemia, other oncologic disease, or other hematologic diseases, medical history regarding previous central CVCs, age at CVC insertion, weight, height, gender, and the purpose of the CVC placement. Patients with a platelet count <75 × 10⁹/L were considered to be suffering from coagulopathy. In addition, the following peri-operative characteristics were collected: The anatomic location, catheter type and insertion technique, duration of procedure, and successful flushing and fixation of the line. Correct placement of the CVC was always verified using radiography. The use and type of prophylactic antibiotics was collected, as well as complication data. Suspected infections caused by CVCs were treated as follows. First, an antibiotic was administered (mostly flucloxacillin or vancomycin, at times adjusted to the culture antibiogram). Second, the CVC was removed, depending on the severity of the infection (sepsis or prolonged fever [>two days] despite antibiotics) and the opportunity for placing a new CVC. Last, a new CVC was placed after a catheter-free interval of 48 hours. Radiographs were performed only when problems were encountered during routine post-operative evaluation (obstruction of the CVC, build-up of fluid beneath the skin, and inability to draw blood from the CVC).

On CVC removal, the date and reason for removal were recorded: End of treatment, suspicion of CVC infection, dislocated CVC, malfunctioning CVC, or other (e.g., when the patient died with the CVC in situ). In addition, the existence of pus and the results of any cultures were recorded.

Statistical analyses

All analyses were performed using SPSS V. 20.0 (IBM, Chicago, IL, USA). The risk of infection was related to the time the CVC remained in situ (number of catheter d). The number of catheter d was calculated based on the date of insertion and removal, or date of last follow-up in case of CVCs still in situ. Every patient referred to us was included in our prospective database. The electronic patient record was used to capture any missing data. For patients transferred to our center with a CVC in situ, if the date of insertion was not available, the date of the first radiograph at our institution confirming the CVC was counted as the date of insertion. As patients may have had multiple CVCs over the period investigated in this study, the CVC was taken as the unit of analysis.

We compared CVCs in patients in whom an infection occurred with CVCs in patients without an infection using the two definitions described above. Independent samples t-test was used for continuous variables and the χ² or Fisher exact test for categorical variables. Groups were compared on the following variables: Gender, age at CVC insertion, body mass index (weight/height²), underlying disease (yes/no), coagulopathy (yes/no), type of CVC, and purpose of insertion. Using sensitivity analyses, different results were generated for patient age. Variables with a significant (p < 0.05) univariable association were subsequently entered into a multivariable logistic regression. Infection (yes/no) was used as the dependent variable according to the two definitions described above.

P values <0.05 were considered statistically significant.

Results

A total of 345 patients received 538 CVCs (Table 1). The average age of the patients at CVC insertion was seven years and nine months (range four months–19 years). A total of 246 Port-a-Caths were inserted that stayed in situ for an average of 474.1 ± 27.6 days (range 1–2,683 d; median 385.5 d). A total of 281 double-lumen 7F Bard-Hickman catheters (hereafter Hickman catheters) were inserted and stayed in situ for an average of 111.5 ± 6.5 days (range 1–857 d; median 70.0 d). An antibiotic was administered as prophylaxis in 395 cases (88.6%), mostly either vancomycin (208 cases; 52.7%) or flucloxacillin (170 cases; 43.0%).

Table 1.

Demographic and Clinical Characteristics of 345 Patients with a Central Venous Catheter at Leiden University Medical Center, January 1, 2000–August 31, 2013

	n
Patients	345
Male (%)	208 (60.3)
Female (%)	137 (39.7)
Average Body Mass Index (SD)	17.7 (3.2)
Average age at insertion, years (SD)	7 years 9 months (5.4)
CVCs (%)	538
Hickman line	281 (52.3)
Port-a-Cath	246 (45.7)
Other	10 (1.9)
Unknown	1 (0.2)
Access site (%)
Right internal jugular vein	262 (48.7)
Left internal jugular vein	102 (19.0)
Right external jugular vein	59 (11.0)
Left external jugular vein	17 (3.2)
Other	35 (6.5)
Unknown	63 (11.7)
Indication for placement^a (%)
Hematology/oncology	229 (42.6%)
Stem cell transplant	218 (40.5%)
Other	23 (4.3%)
Unknown	68 (12.6%)
Underlying disease (%)
Leukemia	238 (44.2)
Osteosarcoma	58 (10.8)
Beta-thalassemia	82 (15.2)
Other oncologic disease	26 (4.8)
Other hematologic disease	126 (22.9)
Unknown	11 (2.0)
Antibiotics (%)
Prophylaxis	395 (73.4)
Vancomycin	208 (38.7)
Flucloxacillin	170 (31.6)
Other	17 (3.2)
No prophylaxis	13 (2.4)
Unknown	130 (24.2)

Definitions for CVC placement:

Hematology/oncology: these CVCs were needed for supportive care in hematology/oncology patients; e.g., for blood draws and to administer medication.

Stem cell transplantation: these CVCs were used for the actual stem cell transplantation. This means that a patient with leukemia who needed a CVC for the stem cell transplantation was registered as “stem cell transplant” instead of “hematology/oncology” as an indication for placement.

Peri-operative complications

Eight (1.5%) peri-operative complications occurred. A pneumothorax was seen three times in three patients. One patient underwent successful drainage, and the Port-a-Cath remained in situ. The other pneumothoraces prevented reinsertion in the jugular vein, and therefore, a CVC was placed in the femoral vein in one patient, and a PICC line was inserted in the elbow in the other. Two arterial cannulations complicated the procedure. Because of this, one CVC needed a second attempt to cannulate the jugular vein. The other arterial puncture resulted in the CVC tip entering the pleural cavity as proved by radiography followed by an immediate removal of the line (this resulted in one of the pneumothoraces, treated with an intercostal drain). The remaining three peri-operative complications consisted of epidermolysis, bleeding in the neck, and compromised sterility. After stitching the skin, compression of the bleeding, and disinfecting again, these CVCs were placed (Table 2).

Table 2.

Central Venous Catheter Data Including Complications

	n (%)
Total CVCs	538
Removed	426 (79.2)
Still in situ	112 (20.8)
Complication
Pneumothorax	3 (0.56)
Arterial cannulation	2 (0.37)
Epidermolysis	1 (0.19)
Bleeding in neck	1 (0.19)
Compromised sterility	1 (0.19)
Reason for CVC removal
End of treatment	245 (45.5)
Suspicion of a line infection	84 (15.6)
Suspected CVC infection with Post-operative tip culture	25 (4.6)
Dislocation	4 (0.7)
Malfunction	23 (4.3)
Other	70 (13.0)
Average dwell time	147 d (range 15–472 d)

Risk of CVC Infection

The 538 CVCs on average stayed in situ for 147 days (range 15–472 d). In total, 426 CVCs had been removed (79.2%) and 112 (20.8%) were still in situ at the time we performed this study. The CVCs were removed most often because of the end of treatment (245 cases; 57.5%), followed by suspicion of a CVC infection (84 cases; 15.6%), malfunctioning CVC (23 cases; 5.4%), or dislocation (4 cases; 0.9%). Catheter-related infections with a positive catheter tip culture occurred in 25 cases (4.6%). Given the total number of 149,190 catheter d, this means a suspected infection rate of 0.56/1,000 catheter d and a rate of infections with a positive catheter tip culture of 0.16/1,000 catheter d.

Port-a-Caths remained in situ for a total of 116,630 days with a suspected infection rate of 0.27/1,000 catheter d and a proved infection rate of 0.06/1,000 catheter d. Hickman catheter patency equaled a total of 31,344 days with a suspected infection rate of 1.63/1,000 catheter d and a proved infection rate of 0.57/1,000 catheter d.

Table 3 shows the differences between patients with a CVC removed because of a suspected CVC infection and patients without a CVC infection. In univariable analyses, patients with a CVC removal because of suspected CVC infection were significantly younger than patients with a patent CVC (6.0 vs. 8.1 y). Similar significant results were found comparing only patients aged 1–19 years in the sensitivity analyses (7.2 vs. 8.5 y). Furthermore, beta-thalassemia and other hematologic diseases were more frequently the underlying disease in patients with a CVC removal secondary to infection compared with patients with a patent CVC. Flucloxacillin was more frequently given and vancomycin was less frequently given to patients in whom the CVC was removed because of a suspected infection (Table 3). When entered into multivariable analyses, only age at insertion and beta-thalassemia as underlying disease remained as independent significant predictors of risk of suspected infection. Infection risk decreased by 6% for every year of increase in age (odds ratio [OR] 0.94; 95% confidence interval [CI] 0.89–0.99]; OR 0.94; 95% CI 0.89–1.00 in the sensitivity analyses) and was 65% lower for patients with beta-thalassemia (OR 0.35; 95% CI 0.17–0.71).

Table 3.

Differences between Infected and Non-Infected Central Venous Catheters (CVCS), According to Two Definitions

CVC removal because of suspected infection vs. no removal for this reason
	Removal for suspected infection (n = 84) (63.1% placed with ultrasound)	No removal for suspected infection (n = 454) (64.3% placed with ultrasound)	Test of difference
Patient age at insertion (y) (SD)	6.0 (5.6)	8.1 (5.3)	t = 3.31; p < 0.01
% male	64.3	59.7	χ² = 0.63; p = 0.43
Body Mass Index (kg/m²) (SD)	17.6 (3.2)	17.7 (3.2)	t = 0.26; p = 0.79
Percent underlying disease (yes/no)
Beta-thalassemia	28.0	13.3	χ² = 11.53; p < 0.01
Leukemia	6.1	11.9	χ² = 2.39; p = 0.12
Osteosarcoma	6.1	11.9	χ² = 2.39; p = 0.12
Other hematologic diseases	34.1	21.3	χ² = 6.39; p = 0.01
Other oncologic diseases	4.9	4.9	χ² = 0.00; p = 1.00
Percent coagulopathy^a (yes/no)	26.4	36.3	χ² = 2.63; p = 0.11
Type of CVC (%)
PAC line	38.1	47.2	χ² = 2.39; p = 0.12
Hickman line	60.7	50.7	χ² = 2.81; p = 0.09
Purpose of line (%)
Stem cell transplant	47.4	46.2	χ² = 0.04; p = 0.85
Hematology/oncology	46.1	49.2	χ² = 0.26; p = 0.61
Other	6.6	4.6	χ² = 0.55; p = 0.40
Antibiotic (%)	(17 cases missing)	(124 cases missing)
Flucloxacillin	53.7	40.6	χ² = 3.92; p = 0.05
Vancomycin	38.8	55.8	χ² = 6.42; p = 0.01
Other (e.g., teicoplanin, ceftazidime)	7.5	3.6	X² = 1.99; p = 0.18

CVC removal because of positive tip culture vs. no removal for this reason
	Removal because of positive catheter tip culture (n = 25) (64.0% placed with ultrasound)	No removal for positive tip culture (n = 513) (64.1% placed with ultrasound)	Test of difference
Patient age at insertion (y) (SD)	5.1 (4.8)	7.9 (5.4)	t = 2.03; p = 0.01
% males	60.0	60.4	χ² = 0.00; p = 0.97
Body Mass Index (kg/m²) (SD)	16.3 (2.2)	17.8 (3.2)	t = 1.95; p = 0.05
Percent underlying disease (yes/no)
Beta-thalassemia	29.2	14.9	χ² = 3.54; p = 0.08
Leukemia	4.2	11.3	χ² = 1.20; p = 0.50
Osteosarcoma	4.2	11.3	χ² = 1.20; p = 0.50
Other hematologic diseases	29.2	23.1	χ² = 0.48; p = 0.47
Other oncologic diseases	4.2	5.0	χ² = 0.03; p = 1.00
Percent coagulopathy^a (yes/no)	33.3	34.8	χ² = 0.02; p = 0.89
Type of CVC (%)
PAC line	28.0	46.7	χ² = 3.35; p = 0.07
Hickman line	72.0	51.4	χ² = 4.07; p = 0.04
Purpose of CVC (%)
Stem cell transplant	54.5	46.0	χ² = 0.62; p = 0.43
Hematology/oncology	31.8	49.6	χ² = 2.64; p = 0.10
Other	13.6	4.5	χ² = 3.79; p = 0.09
Antibiotic (%)	(9 cases missing)	(214 cases missing)
Flucloxacillin	62.5	39.5	χ² = 3.34; p = 0.07
Vancomycin	31.3	55.5	χ² = 3.60; p = 0.06
Other (e.g., teicoplanin, ceftazidime)	6.3	5.0	χ² = 0.05; p = 0.58

Multivariable analysis, including all variables with p < 0.05
Removal because of suspected infection vs. no removal for this reason
	p	Odds ratio (95% confidence interval)
Patient age at insertion (y)	0.03	0.94 (0.89–0.99)
Underlying disease
Beta-thalassemia	<0.01	0.35 (0.17–0.71)
Other hematologic diseases	0.13	1.72 (0.86–3.43)
Antibiotic
Flucloxacillin	0.56	1.37 (0.43–4.37)
Vancomycin	0.12	2.55 (0.80–8.18)

Removal because of positive catheter tip culture vs. no removal for this reason
	p	Odds ratio (95% confidence interval)
Patient age at insertion, years	0.02	0.88 (0.78–0.98)
Body Mass Index	0.18	0.87 (0.70–1.07)
Hickman line	0.05	0.32 (0.11–1.00)

Note: An odds ratio of 0.88 (patient age) means that the chance for a removal for a positive catheter tip culture decreases by 12% for every year the patient gets older. Therefore, younger patients are more at risk.

Defined as a platelet count less than 75 × 10^9/L.

CVC = central venous catheter; PAC = Port-a-Cath; SD = standard deviation.

Patients having CVC removal and a positive catheter tip culture were significantly younger than patients with a patent CVC (5.1 vs. 7.9 years; Table 3) and had a lower body mass index (16.3 vs. 17.8 kg/m²). This did not apply to the sensitivity analyses. Furthermore, in the group in which CVCs were removed with a positive catheter tip culture, there was a larger percentage of patients receiving Hickman catheters (72.0% vs. 51.4%). When entered into multivariable analyses, only age at insertion and use of a Hickman catheter remained as independent significant predictors of infection with a positive catheter tip culture. Infection risk decreased by 12% for every year of increase in age (OR 0.88; 95% CI 0.78–0.98 and was 68% lower for patients with a Hickman catheter (OR 0.32; 95% CI 0.11–1.00]). Both the group with and that without infection had similar percentages of ultrasound-placed CVCs, so this technique cannot explain differences in the infection risk. In our hospital, 82.3% of the CVCs were placed with ultrasound guidance. Finally, Table 4 shows the culture data of CVCs removed with a positive tip.

Table 4.

Culture Data of Central Venous Catheters Removed with a Positive Tip

Organism	n
Coagulase-negative staphylococci	4
Pseudomonas aeruginosa	3
Staphylococcus aureus	8
S. haemolyticus	1
Enterobacter cloacae	1
Raoultella ornithinolytica	1
Ochrobactrum anthropi	1
Nocardia nova	1
Candida glabrata	1
Gram negative bacilli	1
S. epidermidis	3

Discussion

This study was designed to explore the infection and complication rates in a large cohort of pediatric patients receiving CVCs and compared it with contemporary studies on a similar subject. Compared with other studies [5,9 –14], these were the major differences.

First, the peri-operative complication rate of 1.5% in our study is somewhat lower than most values reported in the literature, which range from 2.4% to 4.6% [5, 11, 12] but equal to the rate reported by Malbezin et al. [9], a study of 5,434 pediatric CVC insertion procedures. The percentages of ultrasound-guided CVC placements in these four studies [5,9,11,12] were 100%, 100%, 31.5%, and 1%. Concordantly, the three cases of pneumothorax resulted in a similar rate of 0.6% compared with the latter study [9], which is slightly lower than the rate of 1%–2% generally reported in the literature [13 –15].

Second, depending on which definition is used for an infected CVC, there are different comparisons with other studies. Counting infected CVC tips, the infection rate of 4.6% in the current study is similar to those reported by other investigators [7,13]. However, removals because of suspected infection combined with the total number of catheter d provide rates much lower than in other studies. The catheter-related infection rate (removals because of suspected infection) for Port-a-Caths is 0.27/1,000 catheter d and 1.63/1,000 catheter d for Hickman catheters. Other published studies report rates between three and five infections/1,000 catheter d [5,16,17]. The numbers from the current study may even be an overestimation, as the catheter d excluded the catheter d at other institutions prior to admission or after discharge. Moreover, our population is at higher risk for infection, as many of the underlying diseases and treatments are associated with immunocompromise.

Third, in the current study, catheters suspected to be infected and catheters removed with a subsequent positive catheter tip culture were used to explore whether certain patient groups were at higher risk for a catheter-related infection. Interestingly, it appeared that younger patients were at higher risk. This is a phenomenon previously observed in epidural as well as CVCs [18,19,20], the latter of which might have been attributed to the fact that older children could have larger and more stable vessels that may better tolerate catheters [21]. Importantly, increased protection from a suspected catheter-related infection in older patients could not be secondary to the immunologic immaturity in the very young, as the sensitivity analyses in patients aged 1–19 years showed similar results.

Although defining a CVC infection by a positive catheter tip culture would be scientifically sound, this does not resemble the clinical situation. In this study, outcomes for proved and suspected infection were calculated to investigate whether there is a discrepancy between the two. In our analysis, all major findings in the proved-infection group were in the same direction as those in the suspicion-of-infection group even though a difference was not always statistically significant because of the smaller numbers of infections. This similarity of findings is important, as age, type of catheter, and the underlying disease may serve as additional criteria for treatment when hesitating about a possible catheter infection based on clinical signs.

Antibiotics were administered as prophylaxis by the following protocol. Flucloxacillin was the standard regimen. Patients with a CVC in situ or those considered at higher risk of developing an infection received vancomycin. In retrospect, patients in whom a CVC was removed had more often received flucloxacillin prophylaxis than patients without CVC removal secondary to a suspicion of line infection. Furthermore, in seven of the 13 cases that had received flucloxacillin prophylaxis and later presented with positive catheter tip cultures, the bacterium was resistant to flucloxacillin. Although this effect was not statistically significant in multivariable analysis, these findings may warrant further investigation.

There are some limitations of this study. First, it is possible that minor complications were not marked in the medical records, resulting in a lower complication rate than in other studies [22,23]. For instance, arterial punctures that were quickly noticed and resolved by cannulating the correct vein were not routinely reported as a complication. However, as these did not influence the post-operative recovery, one might argue about whether this should be considered a complication. Second, although more than seven attempts in the same position was considered a failure in other studies [24,25], it is not possible to compare this with the current data, as the number of cannulation attempts was not clear from the records available. Third, although this study provides accurate data on the infection profile of CVCs needing surgical involvement, it cannot be extrapolated from our data how many infections were treated successfully by antibiotics alone. Finally, this study is obviously limited by its retrospective nature. However, this design did permit the identification of the patient groups that are more at risk for CVC removal prior to the end of treatment. Thus, these findings will facilitate formalizing peri-operative CVC care with special attention to high-risk patient groups.

Taken together, this large series of surgeon-performed, largely ultrasound-guided CVC placements has shown that younger patients are at higher risk for CVC removal because of a suspected or confirmed infection prior to the end of treatment. Other factors lowering the infection risk are beta-thalassemia and the use of a Hickman catheter. These findings will be of value in optimizing the peri-operative care of young patients in need of CVC-based treatments and will help in increasing CVC patency.

Footnotes

Acknowledgment

The authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest (such as honoraria; educational grants; participation in speakers' bureaus; membership, employment, consultancies, stock ownership, or other equity interest; and expert testimony or patent-licensing arrangements) or non-financial interest (such as personal or professional relationships, affiliations, knowledge or beliefs) in the subject matter or materials discussed in this manuscript.

Author Disclosure Statement

This research was carried out without funding.

References

Uderzo

, D′Angelo

, Rizzari

, et al. Central venous catheter-related complications after bone marrow transplantation in children with hematological malignancies. Bone Marrow Transplant, 1992; 9:113–117.

Rizzari

, Palamone

, Corbetta

, et al. Central venous catheter-related infections in pediatric hematology–oncology patients: Role of home and hospital management. Pediatr Hematol Oncol, 1992; 9:115–123.

Hruszkewycz

, Holtrop

, Batton

, et al. Complications associated with central venous catheters inserted in critically ill neonates. Infect Control Hosp Epidemiol, 1991; 12:544–548.

Laméris

, Post

, Zonderland

, et al. Percutaneous placement of Hickman catheters: Comparison of sonographically guided and blind techniques. AJR Am J Roentgenol, 1990; 155:1097–1099.

Arul

, Lewis

, Bromley

, et al. Ultrasound-guided percutaneous insertion of Hickman lines in children: Prospective study of 500 consecutive procedures. J Pediatr Surg, 2009; 44:1371–1376.

Huang

, Chen

, Abdullah

, et al. Strategies for the prevention of central venous catheter infections: An American Pediatric Surgical Association Outcomes and Clinical Trials Committee systematic review. J Pediatr Surg, 2011:46:2000–2011.

Avanzini

, Guida

, Conte

, et al. Shifting from open surgical cut down to ultrasound-guided percutaneous central venous catheterization in children: Learning curve and related complications. Pediatr Surg Int, 2010; 26:819–824.

U.S. Centers for Disease Control and Prevention. Frequently asked questions: Bloodstream infection event (central line-associated bloodstream infection and non-central line-associated bloodstream infection). www.cdc.gov/nhsn/pdfs/faqs/psc/faqs_-bsi.pdf (last accessed April 12, 2015).

Malbezin

, Gauss

, Smith

, et al. A review of 5434 percutaneous pediatric central venous catheters inserted by anesthesiologists. Paediatr Anaesth, 2013; 23:974–979.

10.

Gaballah

, Krishnamurthy

, Keller

, et al. US-guided placement ant tip position confirmation for lower-extremity central venous access in neonates and infants with comparison versus conventional insertion. J Vasc Interv Radiol, 2014; 25:548–555.

11.

Alderson

, Burrows

, Stemp

, et al. Use of ultrasound to evaluate internal jugular vein anatomy and to facilitate central venous cannulation in paediatric patients. Br J Anaesth, 1993; 70:145–148.

12.

Leyvi

, Taylor

, Reith

, Wasnick

. Utility of ultrasound-guided central venous cannulation in pediatric surgical patients: A clinical series. Paediatr Anaesth, 2005; 15:953–958.

13.

Casado-Flores

, Barja

, Martino

, et al. Complications of central venous catheterization in critically ill children. Pediatr Crit Care Med, 2001; 2:57–62.

14.

Costello

, Clapper

, Wypij

, et al. Minimizing complications associated with percutaneous central venous catheter placement in children: Recent advances. Pediatr Crit Care Med, 2013; 14:273–283.

15.

Bruzoni

, Slater

, Wall

, et al. A prospective randomized trial of ultrasound- vs landmark-guided central venous access in the pediatric population. J Am Coll Surg, 2013; 216:939–943.

16.

Basford

, Poenaru

, Silva

, et al. Comparison of delayed complications of central venous catheters placed surgically or radiologically in pediatric oncology patients. J Pediatr Surg, 2003; 38:788–792.

17.

Hall

, Hartley

, Ade-AjayI

, et al. Bacterial contamination of central venous catheters during insertion: A double blind randomised controlled trial. Pediatr Surg Int, 2005; 21:507–511.

18.

Mulloy

, Jadavji

, Russell

, et al. Tunneled central venous catheter sepsis: Risk factors in a pediatric hospital. J Parenter Enteral Nutr, 1991; 15:460–463.

19.

Kost-Byerly

, Tobin

, Greenberg

, et al. Bacterial colonization and infection rate of continuous epidural catheters in children. Anesth Analg, 1998; 86:712–716.

20.

Advani

, Reich

, Sengupta

, et al. Central line-associated bloodstream infection in hospitalized children with peripherally inserted central venous catheters: Extending risk analyses outside the intensive care unit. Clin Infect Dis, 2011; 52:1108–1115.

21.

Jumani

, Advani

, Reich

, et al. Risk factors for peripherally inserted central venous catheter complications in children. JAMA Pediatr, 2013; 167:429–435.

22.

Verghese

, McGill

, Patel

, et al. Comparison of three techniques for internal jugular vein cannulation in infants. Paediatr Anaesth, 2000; 10:505–511.

23.

Janik

, Conlon

, Janik

, et al. Percutaneous central access in patients younger than 5 years: Size does matter. J Pediatr Surg, 2004; 39:1252–1256.

24.

Chuan

, Wei

, Yu

, et al. A randomized-controlled study of ultrasound prelocation vs anatomical landmark-guided cannulation of the internal jugular vein in infants and children. Paediatr Anaesth, 2005; 15:733–738.

25.

Verghese

, McGill

, Patel

, et al. Ultrasound-guided internal jugular venous cannulation in infants: A prospective comparison with the traditional palpation method. Anesthesiology, 1999; 91:71–77.