Clinical practice audit concerning antimicrobial prophylaxis in paediatric neurosurgery: results from a German paediatric oncology unit

Background: Perioperative antimicrobial prophylaxis (PAP) has been identified as an important target for internal audits, concerning the judicious use of antibiotics. Paediatric oncology patients with brain tumours face an increased risk of surgical site infection (SSI) after neurosurgery and receive routine PAP in this setting.

Patients and methods: All patients younger than 18 years admitted to the paediatric oncology centre (POC) with a neurosurgical intervention. Systematic audit of routine clinical data is divided in two groups: retrospective (Jan 01, 2012-March 31, 2014) and prospective (April 01, 2014-March 31, 2015) referring to an internal PAP guideline, invented in Jan. 2014). Surveillance of SSI up to 30 days after the operation with standard criteria (Centres for Disease Control and Prevention, USA).

Results: In total, 53 neurosurgical operations were analysed in 33 paediatric oncology patients. Twelve patients received more than one operation. The detailed analysis of PAP revealed prophylactic cefuroxim doses about 30 mg/kg instead of 50 mg/kg and no repeated dosing in operations lasting longer than 4 h. In addition, Cefotaxim, which is not indicated as PAP in neurosurgery, was used instead of Cefuroxim (or Ampicillin-Sulbactam) in 23 % of all cases in the retrospective and 18 % of all cases in the prospective audit. PAP for more than 3 doses (>24 h) was administered in 66 % in the retrospective group and in 60 % in the prospective group (p = n.s.). In both groups, no SSI was detected.

Discussion: This first comprehensive audit of PAP in paediatric oncology patients undergoing neurosurgery outlines significant opportunities to improve clinical practice in terms of correct dosing, the correct choice of the antibiotic, a correct timing schedule and a shorter duration of PAP. In addition, our results illustrate in detail the challenges in clinical practice when an evidence-based approach to improve a standard workflow has to be implemented.