Risk of Postoperative Complications in Patients with Obstructive Sleep Apnea following Skull Base Surgery OTOLARYNGOLOGY-HEAD AND NECK SURGERY Huyett, P., Soose, R. J., Schell, A. E., Fernandez-Miranda, J. C., Gardner, P. A., Snyderman, C. H., Wang, E. W. 2018; 158 (6): 1140–47

Abstract

Objectives Obstructive sleep apnea (OSA) presents several challenges in skull base surgery, including increased intracranial pressure, worsened OSA with nasal packing, and avoidance of positive airway pressure (PAP) therapy postoperatively. The objective of this study was to examine the risk of postoperative complications in a skull base population with OSA in which PAP therapy is withheld. Study Design Retrospective cohort study. Setting Tertiary care hospital. Subjects and Methods Medical records of 414 adult patients undergoing anterior skull base procedures between January 1, 2014, and January 7, 2017, were retrospectively reviewed. Revision surgeries, skull base infections, sinus surgery, and orbital cases were excluded. Results Fifty-four (13.0%) patients with a diagnosis of OSA were identified. While the known patients with OSA were more likely to require postoperative supplemental oxygen (odds ratio [OR], 4.29; 95% confidence interval [CI], 2.38-7.75; P < .001), there was no increased risk of serious respiratory events or cerebrospinal fluid leak (CSF). To address the likely underdiagnosis of OSA in this cohort, subgroup analyses were performed of patients at high risk for OSA (body mass index >30 kg/m2 and hypertension) and demonstrated an increased risk of serious respiratory events (OR, 4.41; 95% CI, 1.24-15.7; P = .034) and CSF leak (13.6% vs 4.7%; P = .018). Conclusions Skull base patients with known OSA can be successfully managed with diligent care in the perioperative period when PAP therapy is withheld. However, OSA is likely underdiagnosed in the skull base population, and patients at high risk for undiagnosed OSA may be at the greatest risk for respiratory complications and CSF leak. Increased presurgical awareness and implementation of a perioperative management algorithm is needed.

View details for PubMedID 29688821