Résumé : Perioperative complications related to obstructive sleep apnea still occur despite the use of partial pressure end-tidal CO2(PETCO2) and pulse oximetry. Airway obstruction can complicate propofol sedation and a novel monitor combining mandibular movement analysis with capnography may facilitate its detection. Patients scheduled for sleep endoscopy were recruited and monitored with standard monitoring, PETCO2, and Jaw Activity (JAWAC) mandibular movement sensors. A post hoc analysis investigated airway obstruction prediction using a Respiratory Effort Sequential Detection Algorithm (RESDA) based on PETCO2 and mandibular movement signals. 21 patients were recruited and 54 episodes of airway obstruction occurred. RESDA detected obstructive apnea [mean ± SD (median)] 29 ± 29 (21) s, p < 0.0001, before PETCO2 alone. This prolonged the time between obstructive apnea detection and decrease to 90% oxygen saturation 64 ± 38 (54) versus 38 ± 20 (35) s, p < 0.0001. It predicted airway obstruction with a sensitivity and specificity of 81% and 93%, respectively. The RESDA algorithm, which is based on the combination of capnography with mandibular movement assessment of respiratory effort, can more rapidly alarm anesthetists of airway obstruction during propofol sedation than PETCO2 alone. However, PETCO2, pulse oximetry, and clinical monitoring are still required. Trial Registry numbers: ClinicalTrial.gov (NCT02909309) https://clinicaltrials.gov/ct2/show/NCT02909309.