Predicting and reducing cranioplasty infections by clinical, radiographic and operative parameters – A historical cohort study

Cranioplasty is a relatively straightforward and common procedure, yet it carries a substantial rate of infection that causes major morbidity and mortality. The authors’ objective was to assess the effect of various variables on the risk of developing post-cranioplasty infections, and to enable the prediction and reduction of its incidence, contributing to an improved patient-selection. The medical records, microbiologic cultures, imaging studies and operative reports of patients who have undergone cranioplasty between the years 2008–2014 at Sheba Medical Center, a tertiary care teaching hospital in Tel-Hashomer, Israel, were reviewed and evaluated for potential predictive factors of infection. Cox regression was applied for uni- as well as multi-variate analyses, and a Kaplan–Meier curve and Log-Rank test were used to describe the association between neurological deficit prior to operation and occurrence of infection. Eighty-eight patients who had undergone cranioplasties using autologous as well as various artificial materials were included in the study. The overall rate of infection was 13.6%; median time to infection was 30.5 days (interquartile range: 17.35–43.5). Pre-operative degree of neurological disability was the strongest predictor for infection in both uni- and multi-variate analyses (Hazard ratio [HR] = 18.9, 95% confidence interval [CI]: 1.9–187 p = 0.014). Patients admitted due to trauma (HR = 7.04 CI: 0.9–54.6, p = 0.062) and autologous graft material (HR = 2.88, 95% CI: 0.92–9.09, p = 0.07) were associated with a trend toward a higher risk for infection. In conclusion, careful patient selection is a key concept in avoiding harmful post-cranioplasty infections. Modified Rankin Score yields a well-established tool that predicts the risk of infection.