BACKGROUND AND OBJECTIVE:Patients with chronic respiratory failure are increasingly managed with domiciliary non-invasive ventilation (NIV). There may be limited ability to provide NIV titration for these complex patients, and ventilatory requirements and upper airway support needs may change over time. Therefore, an automatically adjusting expiratory positive airway pressure (AutoEPAP) algorithm may offer advantages over manually adjusted EPAP for treating these patients. This study compared 4% oxygen desaturation index (ODI4%) values during the use of an AutoEPAP algorithm versus manual EPAP titration with the intelligent volume-assured pressure support (iVAPS) algorithm. METHODS:This prospective, single-blind, randomized, crossover study was conducted at six US sites. Patients with chronic respiratory failure (neuromuscular disease, chronic obstructive pulmonary disease, obesity hypoventilation and other aetiologies) and an apnoea-hypopnoea index of >5/h who were already established NIV users underwent a single night of NIV with the iVAPS manual EPAP and iVAPS AutoEPAP in the sleep laboratory in random order. RESULTS:A total of 38 patients constituted the study population. Mean ODI4% was statistically non-inferior with AutoEPAP versus manual EPAP (P < 0.0001). There was no difference in the effect on ODI4% across respiratory failure subgroups. Ventilation parameters and gas exchange were similar with either NIV mode, indicating equally effective treatment of respiratory failure. Sleep parameters were improved during AutoEPAP versus manual EPAP. CONCLUSION:A single night of NIV using the iVAPS with AutoEPAP algorithm was non-inferior to a single night of iVAPS with manual EPAP titration in patients with respiratory failure. CLINICAL TRIAL REGISTRATION:NCT02683772 at clinicaltrials.gov.