The Accuracy of Predicting Obstructive Sleep Apnea During Colonoscopy Under Conscious Sedation

Abstract

Purpose: Obstructive Sleep Apnea (OSA) is a common disorder with increasing prevalence with age. It is estimated that 93% of women and 82% of men with moderate to severe sleep apnea are not diagnosed with the disorder. Compared to normal individuals, patients with OSA are more sensitive to the effects of sedation and the disorder may in fact be unmasked by sedation. The aims of this study were to determine the predictive value of preset clinical signs and symptoms for OSA and to identify patients who may deserve prolonged monitoring after moderate conscious sedation for colonoscopy. Methods: Consecutive patients undergoing outpatient colonoscopy under conscious sedation with meperidine and midazolam were enrolled. Information about body mass index (BMI), neck circumference (NC), day sleepiness (DS), decrease vigilance (DV), mood changes were recorded as well as the oropharyngeal Mallampati score and the Epworth and the Stanford Sleepiness Scales. The endoscopist was blinded to the information prior to colonoscopy. Patients who developed audible snoring for > 10 s while in the left lateral decubitus position whilst maintaining an oxygen hemoglobin saturation of > 90% were suspected of having OSA. The total dose of sedatives used, the duration and difficulty of the procedure were recorded. After informed consent, ambulatory polysomnography was performed on suspected OSA patients and on age-and BMI-matched controls from the study cohort. Results: 131 patients were enrolled and 24 (3.1%) (22 M and 2 F) suspected of having OSA based on the above criterion. These patients had higher BMI and NC, and were more likely to have a medical history of snoring, DS, chocking or gasping sensation, DV, witnessed sleep apnea, personality and mood changes (all P < 0.05). They also had predominance of Grade III/IV Mallampati score and higher Standford and Epworth scale scores (all P < 0.01 compared with controls). There was no difference in the type or dose of sedation or in difficulty of colonoscopy but the total duration of the procedure was significantly higher compared to controls (21.2 ± 5.6 min vs. 18.1 ± 5.6 min; P= 0.01). 19 patients and 12 matched controls consented to ambulatory portable polysomnography. Moderate or severe OSA was identified in 15/19 patients and in 2/12 controls (78.9% vs. 16.7%; P < 0.01). Conclusion: Audible snoring in the left lateral decubitus position during conscious sedation for colonoscopy is highly predictive of the presence of OSA. Colonoscopy time is prolonged in these patients and may also be associated with longer recovery from sedation. Given the serious nature of OSA, such patients need to be carefully identified by the performing endoscopy team and referred for sleep medicine evaluation.