Tracheal amylase as a marker for associated ventilator infections in traumatic brain injury patients

Abstract

Introduction: Ventilator associated infections(VAI),tracheobronchitis and ventilator-associated pneumonia occurs in 15-20%. Risk factors for VAI in traumatic brain injury(TBI)patients include, endotracheal intubation, Glasgow coma scale( Objetive: We aim to assess whether the amylase obtained by routinely tracheal aspirate in Intensive Care Unit(ICU) predicts the development of VAI in neurosurgical patients. Methods: We included adult neurosurgical patients, mechanically ventilated without respiratory infection. Normal chest radiograph within 48 hours in ICU were required. Tracheal amylase was measured at 24 and 48 h by routinely sterile aspiration. Patients were followed to determine development of VAI. Clinical scales as CPIS, SOFA, and APACHE II were applied. Results: Sixty patients were enrolled, 27(45%)developed VAI. Clinical scales did not predict nosocomial infection while amylase at 24h and 48h predicted VAI ( P= 0.001, P 0.001 respectively). Area under the curve (AUC) for 24 hour amylase was 0.84±0.05(95% CI0.71-0.96) for a value of 585 IU/L(sensitivity 84 % specificity 94%). 48h amylase AUC was 0.95 ± 0.05(95% CI 0.86- 0.99) for a value of 364 IU/L (sensitivity 92% specificity 100%). Conclusions: Higher levels of tracheal amylase in the first 24 and 48 hours predicted the development of VAI faster than the currently used scores. Measurement of amylase obtained by routinely tracheal aspiration without bronchoscopy might represent an easy and cheap diagnostic tool predicting VAI.