Cefazolin versus vancomycin for neurosurgical operative prophylaxis – A single institution retrospective cohort study

Objectives: Cefazolin and vancomycin are common choices for neurosurgical antimicrobial prophylaxis. Cefazolin is typically first-line due to its lower toxicity profile and specificity for gram-positives such as skin commensals, while vancomycin is often reserved for patients with cephalosporin or penicillin allergies. However, one randomized clinical trial demonstrated superiority of vancomycin for cerebrospinal fluid (CSF) shunt insertions at a hospital with a high prevalence of methicillin-resistance Staphylococcus aureus (MRSA). We aimed to evaluate the association of prophylaxis choice and incidence of surgical site infection (SSI) at our own institution. Patients and methods: This was a retrospective cohort study of patients who underwent a neurosurgical operation from January 2013 to April 2016 at one particular hospital belonging to our institution. We included patients who received either only cefazolin or only vancomycin as their pre-incisional prophylaxis. Vancomycin was substituted for cefazolin in patients with known penicillin or cephalosporin allergy. Procedures requiring multiple attending surgeons were excluded. We defined a SSI as a confirmed culture isolated from the wound, implant (if pertinent), or CSF (if pertinent) within a year of surgery. Multivariable logistic regression was performed with consideration of antibiotic, operation performed, wound class, and procedure length. Results: A total of 859 operations met study criteria; 664 patients received Cefazolin, and 195 received Vancomycin. We identified 22 SSIs, with 14 in the cefazolin (2.2%) and 8 in the vancomycin (4.1%) group. Upon logistic regression, there was no significant association of vancomycin substitution with incidence of SSIs between the two groups (odds ratio, 1.59; 95% CI, 0.42–6.00, p = .49). In the cefazolin group, 8/14 cultures were positive for S. aureus compared to 1/8 of the vancomycin group. Conclusions: There was no significant difference in neurosurgical site infection incidence when vancomycin prophylaxis was substituted for cefazolin. S. aureus was isolated from patients who received cefazolin at a higher rate although this was not statistically significant. At our institution, S. aureus makes up 36% of isolated organisms from inpatient and intensive care units. Institutions should consider their own investigations into local antibiograms, SSI rates, and choice of prophylaxis.