Administration of Prophylactic Antibiotics for Central Venous Port Placement in High-risk Pediatric Patients Is Not Associated with Decreased Infections

Background: For children in need of chronic intravenous therapies (chemotherapy, antibiotics, parenteral nutrition, etc.), central venous port placement (CVPP) provides constant venous access that is both reliable and concealable. The most common complications after CVPP are bloodstream and/or port-site infections (BPIs). Evidence to support the use of antibiotic prophylaxis to prevent microbial inoculation of the bloodstream and seeding of the implanted port is insufficient to establish its universal use. Previous studies have demonstrated higher infection rates in patients who are immunocompromised. We hypothesized whether prophylactic antibiotics would lower infection rates among high-risk patients. Defined risk factors included malignancy, thrombocytopenia, neutropenia (absolute neutrophil count, ANC <500/uL), fever within 24 hours of CVPP, and age under 5 years. Our objective was to assess the incidence of and risk factors for BPIs within 30 days of CVPP in a large cohort of pediatric patients to inform the use of perioperative antibiotic prophylaxis.

Methods: Retrospective chart review was conducted of patients under 21 years old who underwent CVPP at Children’s Hospital of Philadelphia between 1/1/2014 and 12/31/2018. Primary outcome was BPI within 30 days of CVPP. Secondary patient characteristics were collected from the medical record. This study was deemed exempt from Institutional Review Board approval (IRB#19-016413).

Results: Over the five-year study period, 1104 patients underwent 1241 CVPP procedures. Mean age was 7.7±5.5 years, and 617 (56%) patients were male. Most patients had underlying malignancy (n=914, 83%; 57% solid tumors, 43% liquid tumors), and chemotherapy was the most common indication for CVPP (n=1061, 86%). CVPs were most often placed in the internal jugular vein (n=995, 80%) and the subclavian vein (n=207, 17%), and harbored a single lumen (n=1192, 96%). CVP replacement was required at least once in 114 (10%) patients. BPI was the indication for CVP replacement in 39 of 114 (34%) patients. Prophylactic antibiotics were administered in 324 (26%) cases. BPI occurred within 30 days of placement in 52 of 1241 (4%) cases, mostly in patients with cancer (87%), and in a bimodal age distribution with peaks at 2 to 4 years old and 12 to 13 years old. Prophylactic antibiotic administration did not yield a significant difference in the rate of BPIs among patients with cancer (p>0.99), patients less than 5 years old (p=0.8), patients who were febrile within 24 hours of CVPP (p=0.39), or patients with neutropenia or thrombocytopenia within 72 hours of CVPP (p=0.58, 0.99, respectively; Table 1).

Conclusion: For patients with presumed risk factors for BPI (malignancy, immunosuppression, a preoperative fever, or young age), the use of prophylactic antibiotics did not appear to be significantly protective against BPI within 30 days of CVPP.