BRIEF REPORTS CENTRAL VENOUS CATHETER-RELATED BLOOD STREAM INFECTIONS IN CHILDREN UNDERGOING HEMATOPOIETIC STEM CELL TRANSPLANT FOR PRIMARY IMMUNODEFICIENCY AND OTHER NONMALIGNANT DISORDERS Theresa S. Cole, BM,* Elizabeth Rogerson,† Jennifer Collins, Csi,‡ Angela Galloway, MB BS,‡ and Julia Clark, BMBS, BMedSci† Abstract: A single-center experience of catheter-related blood stream infections in children undergoing hematopoietic stem cell transplant for primary immunodeficiency is described. The rate of definite central venous catheter infections was 5.31/1000 line days. Staphylococcus epidermidis was the most commonly identified organism. Teicoplanin resistance oc- curred in 17% of S. epidermidis infections. The central catheter was removed in 21% of infections. Key Words: catheter-related blood stream infection, hematopoietic stem cell transplant, primary immunodeficiency, pediatrics Accepted for publication June 14, 2011. From the *Institute of Cellular Medicine, Newcastle University, Newcastle, United Kingdom; †Paediatric Immunology & Infectious Diseases Depart- ment, Newcastle Upon Tyne Hospitals NHS Trust, Newcastle, United Kingdom; and ‡Microbiology Department, Newcastle Upon Tyne Hospitals NHS Trust, Newcastle, United Kingdom. The authors have no conflicts of interest or funding to disclose. Address for correspondence: Theresa S. Cole, BM, Paediatric Immunology & Infectious Diseases Department, Old Children’s Outpatients, Royal Victoria Infirmary, Newcastle, NE1 4LP, United Kingdom. E-mail: theresa.cole@ncl.ac.uk. Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Web site (www.pidj.com). Copyright © 2011 by Lippincott Williams & Wilkins DOI: 10.1097/INF.0b013e31822940a8 L ong-term surgically placed central venous catheters (CVCs) are essential for children undergoing hematopoietic stem cell trans- plant (HSCT). Catheter-related blood stream infections (CRBSI) are a recognized complication. Much of the literature regarding CRBSI relates to children with cancer. There are a limited few studies looking at infections in children with noncancer diagnoses. There are no specific data of children undergoing HSCT for primary immunodeficiency. Risk factors associated with increased rates of CRBSI in children include hospitalization, intensive che- motherapy, neutropenia, noncancer diagnoses, and HSCT. 1–4 These risk factors cannot be avoided in children undergoing HSCT for primary immunodeficiency. Reported infection rates vary widely, influenced by the patients included and definitions used. Published rates of CVC infection in pediatric hematology-oncology patients vary from 2.1 to 7.4/1000 line days, 5,6 with those in children with nonmalignant diagnoses between 3.1 and 7.6/1000 line days. 4,7 It is hypothesized that rates of CRBSI in children undergo- ing HSCT for primary immunodeficiency would be high because of their many risk factors. This study reviewed CRBSI in pediatric patients undergoing HSCT for primary immunodeficiency or other nonmalignant disorders, to establish which organisms were re- sponsible for CRBSI and to evaluate their susceptibilities to the commonly used antibiotics in the HSCT unit. METHODS A retrospective review was carried out for children under- going HSCT under the care of the Pediatric Immunology team at Newcastle General Hospital between January 01, 2000 and De- cember 31, 2005. Children had triple lumen CVCs placed prior to admission. Routine surveillance blood cultures were taken from each lumen of the CVC twice weekly while an inpatient. Children also had blood cultures taken from each lumen if they developed a fever 37.5°C, if the CVC was damaged or if the cap was lost. Demographic data were collected. The number of days each child had a CVC in place post HSCT until discharge from the ward, removal or death was recorded. Positive blood cultures while they were an inpatient were recorded. Maximum tempera- ture, symptoms, and C-reactive protein (CRP) were documented for the days coinciding with positive cultures. The empiric antibiotic policy included teicoplanin and meropenem for the first line, vancomycin and ceftazidime for the second line, and linezolid and piperacillin/tazobactam for the third line. Antibiotic locks were instilled following positive blood cul- tures but were not used routinely as prophylaxis. Vancomycin locks were used for Gram-positive organisms and gentamicin for Gram-negative organisms. Definite infections were defined as at least 1 blood culture of a pathogenic organism (all nonskin or environmental organ- isms); 2 positive blood cultures with the same skin organism (coagulase-negative staphylococci, diphtheroids, and bacillus spe- cies) within a 14-day period; or 2 positive blood cultures with the same skin organism from at least 2 lumens at the same time. With at least one of the following: temperature 38°C, CRP greater than 20 mg/L, or rise in CRP  15 mg/L or suggestive clinical symptoms (hypotension, capillary refill 4 seconds, tachycardia) in the absence of another source of infection. Investigations were performed to exclude other infections according to clinical condition. These included chest radiograph if any respiratory deterioration, urine, stool, and nasopharyngeal secretions or throat cultures as indicated (hematuria, new onset diarrhea, new cough) and blood for herpes virus infections (ade- novirus, cytomegalovirus, epstein-barr virus, and human herpes virus 6). Possible infections were defined as a single blood culture of a skin or environmental organism with at least one of the follow- ing: temperature, CRP of greater than 20 mg/L, or rise in CRP  15 mg/L or clinical symptoms suggestive of infection in the absence of another source of infection. CVCs were accessed by qualified nurses who completed additional training and followed strict guidelines for handling the line. Handling was kept to a minimum; all reasonable measures were taken to prevent the patient handling the line and introducing contamination. Lumen ends were covered with bungs when not in use. Bungs were changed each time the lumen was accessed. A needle-free system was not used but patients were within an HEPA-filtered environment when lines were accessed. Lumens were flushed with saline or dextrose after drug administration and then locked with heparinized saline. CVC exit sites were covered with Mefix dressings, which were changed twice weekly or when wet or loose. CVCs were accessed using aseptic nontouch tech- nique. 8 The line tip was kept sterile by holding the “unclean” catheter and clamps with sterile gauze (in a sterile gloved hand). Lumen ends were cleaned with 70% alcohol on sterile gauze prior to access. The first 1 mL of blood withdrawn from each lumen was inoculated into the blood culture bottle. If an antibiotic lock had been in place, the first 1 mL was discarded and the second milliliter used for blood culture. Blood samples were inoculated into bioMerieux BacT/ALERT pediatric bottles and processed by mi- crobiology following manufacturers’ instructions. Organisms cul- tured from positive bottles were identified via standard laboratory techniques. Antibiotic susceptibility profiles were established us- The Pediatric Infectious Disease Journal • Volume 30, Number 12, December 2011 www.pidj.com | 1