Abstract
Background:
The risk of infection associated with subcutaneous port (SQP) placement in patients with neutropenia remains unclear. We reviewed the rate of early infectious complications (<30 days) following SQP placement in pediatric oncology patients with or without neutropenia [absolute neutrophil count (ANC) <500/mm3].
Methods:
Baseline characteristics and infectious complications were compared between groups using univariate and multivariate analyses.
Results:
A total of 614 SQP were placed in 542 patients. Compared to non-neutropenic patients, those with neutropenia were more likely to have leukemia (n=74, 94% vs n=268, 50%), preoperative fever (n=17, 22% vs n=25, 5%), recent documented infection (n=15, 19% vs n=47, 9%), and were younger (81 vs 109 months) (p values <0.01). After adjusting for fever and underlying-disease, there was a non-significant association between neutropenia and early post-operative infection (OR 2.42, 95% CI 0.82–7.18, p=0.11). Only pre-operative fever was a predictor of infection (OR 6.09, 95% CI 2.08–17.81, p=0.001).
Conclusion:
SQP placement appears safe in most neutropenic patients.
Keywords: Pediatric, Oncology, Neutropenia, Infection, Venous access
INTRODUCTION
Tunneled central venous catheters are essential tools in the management of children with cancer.1 These devices allow long-term vascular access thereby permitting administration of chemotherapy, antibiotics, supportive medications, total parenteral nutrition, and allowing for necessary blood draws during therapy.2 When compared to other tunneled catheters with external components, the totally implantable subcutaneous port (SQP) offers the advantage of being completely covered by skin and soft tissue, and when not accessed, the patient may shower, swim, and engage in most regular activities. While use of these devices is desirable, many pediatric oncology patients present with severe neutropenia at diagnosis or become neutropenic during treatment with cytotoxic chemotherapy. Because of the potential for increased infectious and wound complications in patients with severe neutropenia, many centers do not place SQP in patients with an absolute neutrophil count (ANC) less than 500 per mm3.3 While this is a common practice, data are conflicting regarding the association of neutropenia and complications after SQP placement.2,10–16
Early infection (within 30 days of central venous catheter insertion) is the most notable and frequently reported complication, occurring in 8–22% of pediatric patients with hematologic malignancy.4–6 SQP are associated with the lowest rate of long-term infectious complications among central venous catheters.7 Despite less infectious risk with SQP, the consequences of infection may be more significant due to increased complexity of infection, more difficult device removal, or a surgical wound resulting from removal. Catheter-related or surgical site infections often necessitate removal of the SQP under general anesthesia and may additionally delay therapy. 8,9
Efforts have thus been made to optimize timing for insertion of SQP to minimize the risk of early postoperative infection. There is currently no standard protocol for SQP placement in the setting of neutropenia and therefore practices vary widely between centers. For example, in pediatric hematologic malignancies, SQP insertion may occur at the time of initial diagnosis, beginning of remission induction therapy, or following resolution of neutropenia.2,3,10,11 Waiting for neutropenia to resolve prior to definitive venous access placement may result in a delay in therapy and multiple venipunctures. The literature is inconsistent regarding the association of neutropenia and complications after SQP placement, with some studies documenting increased infectious risk,10−14 and others demonstrating no significant increased risk.2,15,16
The purpose of this study was to examine the association between neutropenia and the development of early post-operative infectious complications (30 days following the initial procedure) in pediatric oncology patients who underwent SQP insertion in a single cancer center. Our secondary objectives were to document the incidence of early post-operative infectious complications and to identify risk factors associated with SQP-related infectious complications. Furthermore, we sought to describe the type of infectious complications that occurred and the associated interventions, such as antibiotics and port removal.
METHODS
Patient cohort
An Institutional Review Board-approved retrospective review was performed. Electronic medical records of all pediatric oncology patients undergoing SQP placement at St Jude Children’s Research Hospital between January 2013 and December 2016 were reviewed. The time period of this study reflects a culture shift at our institution. Prior to 2013, an ANC value of <500/mm3 was considered a contraindication to SQP placement. However, after 2013 SQP were placed regardless of the neutrophil count.
Port placement and post-operative care
All procedures were performed under sterile technique in the operating room or interventional radiology suite. Prophylactic antibiotics (intravenous cefuroxime or clindamycin in the case of penicillin allergy) were administered to all patients within 30 minutes of skin incision. All catheters were inserted percutaneously under general anaesthesia. Ultrasound guidance was used to guide internal jugular vein access, and anatomic landmarks were used to guide the subclavian approach. The choice of vessel accessed was determined by the surgeon, or interventional radiologist placing the line. Subcutaneous ports were either tunnelled and placed on the parasternal chest wall or placed in the subclavicular position. Correct position of the intravenous catheter was confirmed using fluoroscopy during the procedure and a standard chest radiograph was obtained immediately following port placement. The choice of diameter of catheter tubing was determined based on the weight of the patient, favoring 6.6 French catheters for patients < 30 kg.
Post-operative subcutaneous port care and maintenance were performed according to the nursing policy and procedure manual of this institution. While not in use, the subcutaneous port was locked with heparin (100 units/ml). While in use, the Huber needle accessing the port was changed weekly under sterile technique. Chlorhexidine-impregnated sterile line dressings were applied over port access needles as per hospital policy.
Data collection
Electronic medical records of all pediatric oncology patients undergoing subcutaneous port placement were reviewed. We recorded age at insertion, body mass index (BMI), underlying disease (e.g. leukemia, lymphoma, or solid tumor), history of previous central catheter placement, documented preoperative infection (within two weeks of procedure),, date of initiation of chemotherapy, location of port placement, size of catheter tubing, and service placing subcutaneous port (Interventional Radiology or General Surgery), presence of fever within 24 hours of port insertion. Fever was defined as a temperature (measured by temporal scanner) of >38.5 degrees Celsius, in the absence of an isolated pathogen occurring within 24 hours of SQP placement. Most recent laboratory values were also collected prior to port insertion and included: white blood cell count (WBC), platelet count, hemoglobin, glucose, and absolute neutrophil count (ANC). A delay in therapy was noted if a scheduled chemotherapy session was not started; the exact duration of the delay was not consistently available in patient records. Baseline characteristics and infectious complications were compared between neutropenic and non-neutropenic patients.
Neutropenia was defined as ANC<500/mm3. ANC values were recorded as continuous measures but then categorized into a dichotomous variable (<500/mm3 vs. ≥500/mm3) for data analysis. The primary outcome was early infectious complications (e.g. within 30 days of SQP placement). This was defined as: 1) bacteremia (e.g. isolation of a pathogen from a blood culture drawn through the lumen of the SQP 17–19); 2) surgical site infection (e.g. presence of erythema, induration, and/or tenderness or evidence of purulent discharge at the surgical incision); or 3) tract infection (e.g. presence of erythema, induration, and/or tenderness within 2 cm of the port or catheter tubing.20,21
Statistical analysis
Descriptive statistics were calculated for patient characteristics with means and standard deviations for continuous variables and frequency (percentage) for categorical variables. All calculations were done using SAS software version 9.4 (Cary, NC). All tests were two-sided, and the significance level was set at a p-value of less than 0.05. A t-test was performed to compare continuous variables, and a Pearson chi-square test was performed to compare categorical variables between patients with and without neutropenia.
Logistic Regression Analysis
The association between neutropenia and early post-operative infection rate was examined using logistic regression analysis. Five potential confounding variables were pre-selected for analysis: preoperative fever (within 24 hours of port insertion), receipt of preoperative chemotherapy, history of previous central line, underlying disease (leukemia, lymphoma, or solid tumor) and body mass index (BMI). Underlying disease (hematologic malignancy vs. solid tumor) was pre-selected as the only potential effect modifier.
To identify other potential risk factors for the development of early post-operative infectious complications following SQP insertion, a univariate logistic regression analysis was performed. Fourteen pre-selected variables of interest were examined: preoperative fever, underlying disease, location of catheter (internal jugular vein vs. subclavian vein), location of port (chest wall vs. subclavicular), chemotherapy use prior to port insertion, service placing the SQP (interventional radiology vs. surgery), history of previous central line, history of non-central line infection within two weeks of SQP insertion, BMI, age, gender, glucose, hemoglobin, and catheter size (diameter). All variables with a p-value <0.3 in the univariate analyses were entered into the multivariable model and backwards selection was used to identify potential predictors of infection. All variables with a p-value <0.25 were retained in the multivariable model. Odds ratios and 95% confidence intervals were determined.
RESULTS
Patients
In our study period, 614 subcutaneous ports were placed in 542 pediatric oncology patients (Table 1). The majority (79%) of ports were first-time insertions with the remainder being second (16%), third (5%), or fourth (1%) devices. There were 272 (44%) ports placed in children with solid tumors and 342 (56%) were placed in children with a diagnosis of leukemia or lymphoma. The study cohort included 79 ports (13%) placed in neutropenic patients and 535 (87%) placed in those who were non-neutropenic. Statistically significant baseline differences between nonneutropenic and neutropenic patients included age (mean age 109 vs 81 months, p=0.0002), presence of fever within 24 hours of port placement (n=25, 5% vs n=17, 22%, p<0.0001), underlying disease: leukemia/lymphoma (n=268, 50% vs. n=74, 94%), and solid tumor (n=267, 50% vs n=5, 6%), respectively (p<0.0001, Table 1). The following variables were also found to be significantly different between the two groups: chemotherapy prior to SQP insertion (35% vs. 23%, p=0.0359), previous central line insertion (40% vs. 20%, p=0.0024), and hemoglobin concentration (10.9 vs. 9.0 g/dL, p<0.0001; Table 1).
Table 1.
Patient characteristics by ANC.
| ANC ≥= 500/mm3 N = 535 | ANC <500/mm3 N = 79 | OR, 95% CI | p-value | |
|---|---|---|---|---|
| Age (months; mean, SD) | 109.3 (74.7) | 81.0 (58.7) | 0.00021 | |
| Gender | 0.9042 | |||
| Female | 232 (43%) | 37 (47% ) | F:M | |
| Male | 303 (57%) | 42 (53%) | 1.51 (0.72,1.85) | |
| BMI (mean, SD) | 23.0 (70.0) | 18.2 (5.5) | 0.12171 | |
| Disease | <0.00012 | |||
| LL | 268 (50%) | 74 (94%) | LL:ST | |
| ST | 267 (50%) | 5 (6%) | 13.99 (5.57,35.16) | |
| Fever | <0.00012 | |||
| No | 510 (95%) | 62 (78%) | Y:N | |
| Yes | 25 (5%) | 17 (22%) | 5.59 (2.86,10.94) | |
| Chemo Prior* | 0.03592 | |||
| No | 348 (65%) | 61 (77%) | Y:N | |
| Yes | 186 (35%) | 18 (23%) | 0.55 (0.32,0.96) | |
| Previous Catheter | 0.00242 | |||
| No | 323 (60%) | 62 (78%) | Y:N | |
| Yes | 212 (40%) | 17 (22%) | 0.42 (0.24,0.73) | |
| Service Placing | 0.68852 | |||
| IR | 140 (26%) | 19 (24%) | Surgery:IR | |
| Surgery | 395 (72%) | 60 (76%) | 1.12 (0.64,1.94) | |
| Location Catheter | 0.14902 | |||
| IJ | 172 (32%) | 19 (24%) | SCV:IJ | |
| SCV | 363 (68%) | 60 (76%) | 1.50 (0.87,2.58) | |
| Port site* | 0.70612 | |||
| Chest | 265 (50%) | 41 (52%) | CW:SCL | |
| Sub clavicular | 269 (50%) | 38 (48%) | 1.09 (0.68,1.76) | |
| Hemoglobin (g/dL; mean, SD) | 10.97 (2.11) | 9.00 (1.32) | <0.00011 | |
| Post-op Infection | <0.00012 | |||
| No | 524 (98%) | 72 (91%) | Y:N | |
| Yes | 11 (2%) | 7 (9%) | 0.22 (0.08,0.57) | |
Continuous variables are presented as means and standard deviations. Tests used 1TTest; 2Pearson X2 test. BMI (Body Mass Index). LL (Leukemia, Lymphoma). ST (Solid Tumor). IR (Interventional Radiology). IJ (Internal Jugular Vein). SCV (Subclavian Vein). CW:SCL (Chest Wall, Subclavicular)
1 patient with missing data
Infections
There were 18 total post-operative infections (Table 2). In univariate analysis, the rate of early post-op infectious complications was found to be significantly greater in neutropenic patients (n=7, 9%) compared to non-neutropenic patients (n=11, 2%, p<0.0001; Table 1). No patient in our data set had more than one infection. The median time to infection was 14.5 days following subcutaneous port placement (range, 2 – 30 days). Of those with infection, most patients developed bacteremia (n=12, 67%), with the remainder having port site infection (n=4, 22%) or the presence of both port site infection and bacteremia (n=2, 11%). The most common organism isolated was Staphylococcus aureus (n=4, 28%). The majority of patients (n=14, 78%) were neutropenic at the time of the early post-operative infectious complication. Treatment of infectious complications included antibiotics only (n=10, 56%), surgery only (e.g. removal of subcutaneous port, n=2, 11%), and both surgery and antibiotics (n=6, 33%, Table 2). The majority of infectious complications resulted in a delay of therapy (96%).
Table 2.
Early post-operative infectious complications
| N | N (%) | |
|---|---|---|
| Type of infection | 18 | |
| Bacteremia | 12 (67%) | |
| Port Site | 4 (22%) | |
| Tunnel/track | 0 (0%) | |
| Bacteremia and port | 2 (11%) | |
| Organism Isolated | 1 | |
| Capnocytopagea | 1 (7%) | |
| Coag Neg Staph + Candida | 1 (7%) | |
| Escherichia coli | 3 (21%) | |
| Moraxella non-liquefaciens | 1 (7%) | |
| Pseudomonas aeruginosa | 1 (7%) | |
| Rothia mucilaginosa | 1 (7%) | |
| Staphylococcus epidermidis | 1 (7%) | |
| Staphyloccocus aureus | 4 (29%) | |
| Streptococcus viridans | 1 (7%) | |
| ANC at time of infection | 18 | |
| 0 | 10 (56%) | |
| 100 | 2 (11%) | |
| 300 | 1 (6%) | |
| 400 | 1 (6%) | |
| 500 | 1 (6%) | |
| 600 | 1 (6%) | |
| 900 | 1 (6%) | |
| 2500 | 1 (6%) | |
| Intervention | 18 | |
| Antibiotics | 10 (56%) | |
| Surgical removal SQP | 2 (11%) | |
| Antibiotics and surgery | 6 (33%) | |
| Delay in therapy | 18 | |
| Yes | 17 (94%) | |
| No | 1 (6%) | |
Regression Analysis
Baseline characteristics and infectious complications were compared between neutropenic and non-neutropenic patients (Table 3). No interaction was found between neutropenia and underlying disease (p=0.97). The multivariate regression model did find the presence of preoperative fever and underlying disease (solid tumor vs. leukemia/lymphoma) to be significant confounding variables. After adjusting for underlying disease and the presence of fever within 24 hours of port placement, the increased risk of early post-operative infection in the presence of neutropenia was not significant (OR 2.42, 95% CI 0.82–7.18, p=0.11; Table 3).
Table 3.
Association between neutropenia and infection adjusting for fever, and underlying disease.
| Odds Ratio | 95% CI | P-value | |
|---|---|---|---|
| Neutropenia | 2.42 | 0.82–7.18 | 0.1106 |
| Presence of pre-operative fever | 4.72 | 1.55–14.37 | 0.0063 |
| Disease (LL:ST) | 2.33 | 0.61–8.86 | 0.2158 |
LL (Leukemia, Lymphoma). ST (Solid Tumor)
Univariate logistic regression was used to identify other potential risk factors for early postoperative infection following SQP placement (Table 4). Pre-operative fever was found to be associated with increased risk of infection (OR 7.78, 95% CI 2.76–21.93, p=0.0001). A diagnosis of hematologic malignancy (leukemia/lymphoma) was similarly found to be associated with increased risk of infection (OR 3.92, 95% CI 1.12–13.69, p=0.0321). Finally, lower hemoglobin levels were also found to be associated with increased risk of infection, with a 25% decrease in odds of infection for every one unit (g/dL) increase in hemoglobin (OR 0.75, 95% CI 0.58–0.96, p=0.025; Table 4).
Table 4.
Univariable analysis: predictors of infection
| Odds Ratio | 95% CI | P-value | |
|---|---|---|---|
| Fever (Y:N) | 7.78 | 2.76–21.93 | 0.0001 |
| Disease (LL:ST) | 3.92 | 1.12–13.69 | 0.0321 |
| Location Catheter (SCV:IJ) | 1.18 | 0.41–3.36 | 0.7570 |
| Port Site (CW:SCL) | 1.45 | 0.54–3.85 | 0.4588 |
| Chemo Prior (Y:N) | 0.42 | 0.12–1.48 | 0.1780 |
| Service Placing (Sx:IR) | 1.23 | 0.40–3.79 | 0.7185 |
| Previous Catheter (Y:N) | 0.64 | 0.22–1.82 | 0.4003 |
| Preop Infection (Y:N) | 1.11 | 0.25–4.97 | 0.8849 |
| BMI | 1.00 | 1.00–1.00 | 0.6987 |
| Age | 1.00 | 1.00–1.00 | 0.3795 |
| Gender (F:M) | 1.63 | 0.63–4.18 | 0.3124 |
| Glucose | 1.00 | 0.99–1.02 | 0.5379 |
| Hemoglobin | 0.75 | 0.58–0.96 | 0.0251 |
| Catheter Size | 1.08 | 0.79–1.50 | 0.6343 |
LL(Leukemia, Lymphoma). ST (Solid Tumor). IR (Interventional Radiology). IJ (Internal Jugular Vein). SCV (Subclavian Vein). CW:SCL (Chest Wall, Subclavicular). Y (Yes). N (No)
A second multivariate logistic model was constructed to examine the association between these selected risk factors and early post-operative infection followed by a Wald test for the overall effects of ANC on infection. This process identified two significant variables (fever and underlying disease) and thus a final model adjusting for these two variables was created. This final model found the presence of fever within 24 hours of SQP placement was significantly associated with an increased risk of infection (OR 6.09, 95% CI 2.08–17.81, p=0.0010) after adjusting for underlying disease (Table 5).
Table 5.
Multivariable analysis: predictors of infection
| Odds Ratio | 95% CI | p-value | |
|---|---|---|---|
| Disease (LL:ST) | 4.37 | 0.96–19.86 | 0.0564 |
| Presence of pre-operative fever | 6.09 | 2.08–17.81 | 0.0010 |
LL (Leukemia, Lymphoma). ST (Solid Tumor).
DISCUSSION
In this study, which is the largest single-center series examining the topic, we demonstrated no statistically significant relationship between neutropenia and early post-operative infection following SQP placement. Univariate analysis showed that neutropenia, pre-operative fever, underlying diagnosis (leukemia versus solid tumor), previous chemotherapy, and anemia were significantly associated with increased early post-operative SQP infection. However, multivariate analysis showed that only pre-operative fever was independently associated with increased early-postoperative SQP infection. These finding are in keeping with several other studies that examined early post-operative wound infections in the setting of neutropenia following placement of tunneled lines.3,15
In our study period, the incidence of early post-operative infections following SQP placement was 2.9% (18/614). This is comparable to the incidence of infections found in the literature (0.712.6%).9 Treatment of infectious complications required surgical removal of the SQP in just under half of these patients (44%, 8/18). This is in contrast to the literature that has demonstrated antibiotic treatment failure rate ranging from 11% 2 to 14%.9 This difference may be secondary to the higher incidence of local infections (e.g. port site or surgical site) noted in our group of patients (6/18). This may also be secondary to poor wound healing in the context of neutropenia. Almost all of our patients with infections, whether salvaged with antibiotics or not, had a documented delay in therapy (17/18).
Neutropenia is a common occurrence in the setting of cancer, both at the time of diagnosis and throughout therapy. The optimal timing of placement of tunneled central venous lines with regard to preoperative neutropenia has been an area of interest in several studies. Historically, placement of SQP in the setting of severe neutropenia has been discouraged due to the potential increased risk of serious infections, given the crucial role which neutrophils play in the process of wound healing.14 Some studies have demonstrated that neutropenia (defined variably as ANC ≤500–1000/mm3) at the time of tunneled central venous line insertion is associated with an increase in post-operative infection and line removal and therefore recommend delaying central line placement until ANC recovery.25,26 Conversely, other studies have demonstrated that neutropenia is not a significant risk factor for infection following central line placement.3,15 Our study, which is the largest series to date, did not demonstrate a significant association between pre-operative neutropenia and early post-operative SQP infectious complications.
Pre-operative fever was independently associated with early post-operative SQP infection in this study. The importance of pre-operative fever in the absence of an isolated pathogen and the risk of post-operative wound infection following line placement in the pediatric oncology patient remains poorly defined. Fever is common in children with leukemia. It is hypothesized to be secondary to the release of cytokines from leukemic cells, the administration of certain drugs, or even the transfusion of blood products.32 While fever in an oncology patient is often multifactorial, and may not be an indication of acute infection, it has been found to be associated with a higher rate of post-operative infections following SQP insertion.2 Similarly, our study demonstrated a significantly increased risk of infection following SQP placement in patients that had a documented fever in the 24 hours preceding surgery.
Other factors that have been studied with regard to infectious risk and central line placement include site of line placement (internal jugular vein versus subclavian vein), type of catheter used (subcutaneous port versus tunneled central access devices), and the age of patient. 15,28,29,32 In keeping with the literature, we found that the choice of vessel (internal jugular vein versus subclavian vein) did not influence infectious risk. Prior studies found an increased risk of infectious complications following line placement in children less than 10 years of age. However, in our population, age was not a predictive factor for early post-operative infection following SQP placement. Our study did find that for every g/dL increase in hemoglobin, the odds of developing a post-operative wound infection decreased by 0.75. This is not surprising, given the extensive literature demonstrating an association between anemia and increased risks of postoperative complications (including infection). This is thought to be secondary to lower oxygen carrying capacity and the subsequent lower oxygen tension in tissues, resulting in poor wound healing and decreased local immunity.30,31
Historically, tunnelled central venous catheters were placed exclusively by surgeons. Over the past 15 years, however, these lines are now more frequently placed by interventional radiologists at many centers. Our study found that the service placing the SQP (interventional radiology versus surgery) was not a risk factor for early infection. This lack of a difference may be due to a standardized approach within the institution to line care both at the time of insertion and post-operatively.
Our univariate analysis revealed that patients with a diagnosis of leukemia or lymphoma appeared to be almost four times more likely to develop an early post-operative infectious complication as compared to patients diagnosed with a solid tumor. However, this relationship was not statistically significant in multivariate analysis. When compared to solid tumor patients, the hematologic malignancy population typically presents more systemically ill, and often with profound bone marrow suppression. Further, both groups differ significantly with regards to timing of initiation of chemotherapy as well as the types of chemotherapeutic agents used to treat the underlying disease. Finally, the observed increased risk of infection associated with the leukemia/lymphoma group may be exaggerated given our low event rate and resulting wide confidence intervals (only 3 infections occurred in solid tumor patients with neutropenia).
Ultimately, an important finding of this study was the overall low rate of post-operative infection following port placement in a group of both neutropenic and non-neutropenic patients. However, this low event rate also resulted in wide confidence intervals for many of our analyses. Our small event rate also limited our ability to fit multiple variables into our second multivariate regression model due to over-fitting and likely contributed to the wide confidence intervals seen throughout our analyses. A second limitation of this study is the retrospective design which may result in inherent bias. An example of this was our definition of central-line associated blood stream infection. We chose to accept all infections wherein a pathogen was isolated from a culture drawn through the central line. However, a more restrictive definition would have involved comparing cultures between central and peripheral venous samples and documenting a 5:1 ratio in microbe counts. This would have eliminated the potential for contaminated specimens. Not all patients had peripheral and central blood cultures drawn, and therefore this comparison could not be made.
Conclusions
The optimal timing of SQP placement in the pediatric oncology patient remains poorly defined. Our results did not demonstrate an adverse association of neutropenia with early post-operative infection following SQP placement. However, fever within 24 hours prior to port placement was found to carry a six-fold increase in the risk of infection. Given these findings, we believe it may be safe to place SQP in neutropenic patients without fever in last 24 hours. The presence of fever in the immediate pre-operative period, however, should prompt consideration of delaying placement to reduce risk of infection. Given our low event rate, and wide confidence intervals, further studies are required to assess the relationship between neutropenia and modifiable risk factors on early post-operative wound infections following central line placement in pediatric oncology patients.
How this study will improve care:
Our review of subcutaneous infusaport placement in pediatric oncology patients did not reveal a significant association between neutropenia and early post-operative infection. These data suggest that port placement does not need to be delayed until the neutrophil count normalizes.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
REFERENCES
- 1.Kim HJ, Un J, Kim KH et al. Safety and effectiveness of central venous catheterization in patients with cancer: prospective observational study. J Korean Med Sci. 2010;25:1748–1753. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Junqueira BLP, Connolly B, Abla O et al. Severe neutropenia at a time of port insertion is not a risk factor for catheter-associated infections in children with acute lymphoblastic leukemia. Cancer. 2010;116:4368–4375. [DOI] [PubMed] [Google Scholar]
- 3.Gonzalez G, Davidoff AM, Howard SC, Pui et al. Safety of Central Venous Catheter Placement at Diagnosis of Acute Lymphoblastic Leukemia in Children. Pediatr Blood Cancer. 2012;58:498–502. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Cesca E, Dall’igna P, Boscolo-Berto R et al. Impact of severe neutropenia and other risk factors on early removal of implanted central venous catheter (ICVC) in children with hematologic malignancies. J Pediatr Hematol Oncol. 2014;36: 541–544. [DOI] [PubMed] [Google Scholar]
- 5.Worth LJ, Seymour JF, Slavin MA. Infective and thrombotic complications of central venous catheters in patients with hematological malignancy: prospective evaluation of non-tunneled devices. Support Care Cancer. 2009;17: 811–818. [DOI] [PubMed] [Google Scholar]
- 6.Samaras P, Dold S, Braun J. et al. Infectious port complications are more frequent in younger patients with hematologic malignancies than in solid tumor patients. Oncology. 2008; 74:237–244. [DOI] [PubMed] [Google Scholar]
- 7.O’Grady NP, Alexander M, Burns LA et al. Healthcare Infection Control Practices Advisory Committee (HICPAC). Guidelines for the prevention of intravascular catheter-related infections. Clin Infec Dis. 2011;52:e162–e193. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ross MN, Haase GM, Poole MA, et al. Comparison of totally implanted reservoirs with external catheters as venous access devices in pediatric oncologic patients. Surg Gynecol Obstet. 1988;167:141–144. [PubMed] [Google Scholar]
- 9.Allen CR, Holdsworth MT, Johnson CA, et al. Risk determinants for catheter-associated blood stream infections in children and young adults with cancer. Pediatr Blood Cancer. 2008;51:53–58. [DOI] [PubMed] [Google Scholar]
- 10.Guitierrez I, Gollin G. Exclusion of neutropenic children from implanted central venous catheter placement: impact on early catheter removal. J Pediatr Surg. 2010;45:11151119. [DOI] [PubMed] [Google Scholar]
- 11.Elihu A, Gollin G. Complications of Implanted Central Venous Catheters in Neutropenic Children. Am Surg. 2007;73:1079–1082. [PubMed] [Google Scholar]
- 12.Chen IC, Hsu C, Chen YC et al. Predictors of bloodstream infection associated with permanently implantable venous port in solid cancer patients. Ann Oncol. 2012;24:463–468. [DOI] [PubMed] [Google Scholar]
- 13.Kelly MS, Conway M, Wirth KE et al. Microbiology and risk factors for central lineassociated bloodstream infections among pediatric oncology outpatients-a single institution experience of 41 cases. Am J Pediatr Hematol Oncol. 2013;35:e71–e76. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Shaul DB, Scheer B, Rikhsar S et al. Risk factors for early infection of central venous catheters in pediatric patients. J Am Coll Surg. 1998;186:654–658. [DOI] [PubMed] [Google Scholar]
- 15.Hoss DR, Bedros AA, Mesipam A et al. Severe Neutropenia at the time of implantable subcutaneous chest port insertion is not a risk factor for port removal at a tertiary pediatric center. J Vasc Interv Radiol. 2017;29:398–402. [DOI] [PubMed] [Google Scholar]
- 16.Bamba R, Lorenz JM, Lale AJ et al. Clinical predictors of port infections within the first 30 days of placement. J Vasc Interv Radiol. 2014; 25:419–423. [DOI] [PubMed] [Google Scholar]
- 17.Flynn PM, Shenap JL, Barrett FF. Differential quantification with commercial blood culture tube for diagnosis of catheter-related infection. J Clin Microbiol. 1988;26:1045–1046. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Gaur AH, Flynn PM, Giannini MA et al. Difference in time to detection: A simple method to differentiate catheter related from non-catheter related bloodstream infection in immunocompromised pediatric patients. Clin infect Dis. 2003;37:469–475. [DOI] [PubMed] [Google Scholar]
- 19.Gaur AH, Flynn PM, Heine DJ et al. Diagnosis of catheter related bloodstream infections in pediatric oncology patients lacking a peripheral culture, using differential time to detection. Pediatr Infect Dis J. 2005;24:445–449. [DOI] [PubMed] [Google Scholar]
- 20.Carr E, Jayabose S, Stringel G et al. The safety of central line placement prior to treatment of pediatric acute lymphoblastic leukemia. Pediatr Blood cancer. 2006;47:886888. [DOI] [PubMed] [Google Scholar]
- 21.Horan TC, Gaynes RP, Martone WJ, et al. CDC definitions of nosocomial surgical site infections, 1992: a modification of CDC definitions of surgical wound infections. Am J Infect Control. 1992;20:271–274. [DOI] [PubMed] [Google Scholar]
- 22.Pui C-H, Robison LL, Look AT. Acute lymphoblastic leukemia. Lancet. 2008;371:10301043. [DOI] [PubMed] [Google Scholar]
- 23.Wiener ES, McGuire P, Stolar CJ, et al. The CCSG prospective study of venous access devices: An analysis of insertions and causes of removal. J Pediatr Surg. 1992;27:155–164. [DOI] [PubMed] [Google Scholar]
- 24.Schwartzberg LS. Neutropenia: Etiology and Pathogenesis. Clinical Cornerstone. 2006;8:S5–S11. [DOI] [PubMed] [Google Scholar]
- 25.Press OW, Ransey PG, Larson EB, Fefer A et al. Hickman catheter infections in patients with malignancies. Medicine. 1984; 63:189–200. [DOI] [PubMed] [Google Scholar]
- 26.Howell PB, Walters PE, Donowitz GR et al. Risk factors for infection of adult patients with cancer who have tunneled central venous access catheters. Cancer. 1995;75:13671375. [DOI] [PubMed] [Google Scholar]
- 27.Pickering LK, Anderson DC, Choic S, et al. Leukocyte function in children with malignancies. Cancer. 1975;35:1365–1371. [DOI] [PubMed] [Google Scholar]
- 28.Biffi R, Pozzi S, Bonoma G et al. Cost effectiveness of different central venous approaches for port placement and the use in adult oncology patients: evidence from a randomized three-arm trial. Ann Surg Oncol. 2014;21:3725–3731. [DOI] [PubMed] [Google Scholar]
- 29.Zerati AE, Rigueredo TR, deMoraes RD et al. Risk factors for infectious and noninfections complications of totally implantable venous catheters in cancer patients. J Vasc Surg Venous Lymphat Disord. 2016;4:200–205. [DOI] [PubMed] [Google Scholar]
- 30.Cl Wurzel, Halom K, Feldman JG et al. Infection rates of Broviac-Hickman catheters and implantable venous devices. Am J Dis Child. 1988;142:536–540. [DOI] [PubMed] [Google Scholar]
- 31.Musallam KM, Tamim HM, Richards T, et al. Preoperative anaemia and postoperative outcomes in non-cardiac surgery: a retrospective cohort study. Lancet. 2011;378:13961407. [DOI] [PubMed] [Google Scholar]
- 32.Duce L, Cooter ML, Sharon MS et al. Outcomes in patients undergoing cardiac surgery who decline transfusion and received erythropoietin compared to patients who did not: a matched cohort study. Anesth Anal. 2017. [Epup ahead of print]. [DOI] [PubMed] [Google Scholar]
- 33.Pizzo PA, Robichaud KJ, Wesley R, et al. Fever in the pediatric and young adult patient with cancer. A prospective study of 1001 episodes. Medicine. 1982;61:153–165. [DOI] [PubMed] [Google Scholar]