Preoperative chemotherapy and corticosteroids: Independent predictors of cranial surgical-site infections

Bryan A. Lieber, Geoffrey Appelboom, Blake E. Taylor*, Franklin D. Lowy, Eliza M. Bruce, Adam M. Sonabend, Christopher Kellner, E. Sander Connolly, Jeffrey N. Bruce

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

36 Scopus citations

Abstract

Objective Preoperative corticosteroids and chemotherapy are frequently prescribed for patients undergoing cranial neurosurgery but may pose a risk of postoperative infection. Postoperative surgical-site infections (SSIs) have significant morbidity and mortality, dramatically increase the length and cost of hospitalization, and are a major cause of 30-day readmission. In patients undergoing cranial neurosurgery, there is a lack of data on the role of patient-specific risk factors in the development of SSIs. The authors of this study sought to determine whether chemotherapy and prolonged steroid use before surgery increase the risk of an SSI at postoperative Day 30. METHODS Using the national prospectively collected American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) database for 2006-2012, the authors calculated the rates of superficial, deep-incisional, and organ-space SSIs at postoperative Day 30 for neurosurgery patients who had undergone chemotherapy or had significant steroid use within 30 days before undergoing cranial surgery. Trauma patients, patients younger than 18 years, and patients with a preoperative infection were excluded. Univariate analysis was performed for 25 variables considered risk factors for superficial and organ-space SSIs. To identify independent predictors of SSIs, the authors then conducted a multivariate analysis in which they controlled for duration of operation, wound class, white blood cell count, and other potential confounders that were significant on the univariate analysis. RESULTS A total of 8215 patients who had undergone cranial surgery were identified. There were 158 SSIs at 30 days (frequency 1.92%), of which 52 were superficial, 27 were deep-incisional, and 79 were organ-space infections. Preoperative chemotherapy was an independent predictor of organ-space SSIs in the multivariate model (OR 5.20, 95% CI 2.33-11.62, p < 0.0001), as was corticosteroid use (OR 1.86, 95% CI 1.03-3.37, p = 0.04), but neither was a predictor of superficial or deep-incisional SSIs. Other independent predictors of organ-space SSIs were longer duration of operation (OR 1.16), wound class of ≥ 2 (clean-contaminated and further contaminated) (OR 3.17), and morbid obesity (body mass index ≥ 40 kg/m2) (OR 3.05). Among superficial SSIs, wound class of 3 (contaminated) (OR 6.89), operative duration (OR 1.13), and infratentorial surgical approach (OR 2.20) were predictors. CONCLUSIO NS Preoperative chemotherapy and corticosteroid use are independent predictors of organ-space SSIs, even when data are controlled for leukopenia. This indicates that the disease process in organ-space SSIs may differ from that in superficial SSIs. In effect, this study provides one of the largest analyses of risk factors for SSIs after cranial surgery. The results suggest that, in certain circumstances, modulation of preoperative chemotherapy or steroid regimens may reduce the risk of organ-space SSIs and should be considered in the preoperative care of this population. Future studies are needed to determine optimal timing and dosing of these medications.

Original languageEnglish (US)
Pages (from-to)187-195
Number of pages9
JournalJournal of Neurosurgery
Volume125
Issue number1
DOIs
StatePublished - Jul 1 2016

Keywords

  • 30-day readmission
  • Chemotherapy
  • NSQIP
  • Predictors of infection
  • Steroids
  • Surgical-site infection

ASJC Scopus subject areas

  • Surgery
  • General Medicine
  • Clinical Neurology

Fingerprint

Dive into the research topics of 'Preoperative chemotherapy and corticosteroids: Independent predictors of cranial surgical-site infections'. Together they form a unique fingerprint.

Cite this