The anesthesia airway evaluation: Correlation with sleep endoscopy findings

The anesthesia airway evaluation: Correlation with sleep endoscopy findings:

Publication date: Available online 23 November 2019
Source: American Journal of Otolaryngology
Author(s): Michael Eggerstedt, Matthew J. Urban, Emily Chi, Ethan M. Ritz, Phillip Losavio

Abstract

Purpose

Drug-induced sleep endoscopy (DISE) is a cost-effective, safe, and reliable tool to evaluate obstructive sleep apnea (OSA) patients by revealing upper airway sites, patterns, and severity of obstruction. DISE provides valuable data because reliable evaluation of the OSA airway while awake has remained elusive. Few studies (with mixed results) have analyzed the correlation between pre-operation, awake airway assessments routinely performed by anesthesia and DISE results.

Methods

Preoperative anesthesia evaluation records and subsequent DISE reports were obtained for 99 adult patients undergoing DISE between 2016 and 2018. All patients carried the diagnosis of OSA, based on polysomnography. Anesthesia-collected variables were compared with DISE findings in an effort to determine if commonly-utilized physical exam findings correlated to patterns of upper airway collapse observed on sleep endoscopy.

Results

Most anesthesia preoperative evaluation variables were not found to be predictive of any identifiable patterns of collapse on DISE, including Mallampati score, ability to prognath, and overall airway assessment score. Obesity did not correlate with circumferential collapse at the velopharynx, or to multi-level collapse. Thyromental distance <6.5 cm was found to be statistically correlated to total epiglottic collapse (E = 2+). Friedman tongue position scores were found to be correlated to velopharyngeal collapse (p < 0.05).

Conclusions

Anesthesia airway assessment algorithms and physical exam findings do not correlate well with findings on sleep endoscopy. DISE remains the gold standard for evaluating levels of collapse and operative planning in the OSA population.