Oxygenation Characteristics and Impact of Automated Closed-Loop Oxygen Administration in COVID-19 Patients: A Single ICU, Prospective, Observational Study

Background: An association between dyspnoea, hypoxaemia, and in-hospital mortality has been identified during the COVID-19 pandemic. The study aimed to depict the oxygenation characteristics of spontaneously breathing patients using automated closed-loop oxygen administration. Methods: This monocentre, prospective, observational study was derived from a data warehousing ICU project, Brest University Hospital, France. All spontaneously breathing patients were admitted for acute respiratory distress related to COVID-19 and required O2 administration. Automated O2 was administered using the FreeO2 system (Oxynov Inc., Quebec, QC, Canada), which was set to maintain SpO2 between 90% and 94%. Continuous oximetry recordings were performed at 1 Hz. The primary outcome measure was the percentage of time spent within the SpO2 target. Secondary outcomes were outcome parameters, severity scores (EWSO2, qCSI), time within hyperoxaemia (SpO2 > 96%), hypoxaemia (SpO2 2 L/min in < 2 min) and SpO2 events. Findings: Fifty-nine patients were treated using FreeO2, of whom 21 after weaning from mechanical ventilation, 7 after high-flow nasal canula, and 31 as a first-line therapy. O2 was required for a median of 1.4 days. On treatment initiation, patients were hypoxaemic (estimated PaO2 /FIO2 = 221 [178–258]) and considered at a low to intermediate risk for deterioration (EWSO2 = 8.5 [7.9–10.4]; qCSI = 5.4 [4.4–6.2]). Time within range was high (86.3±10.6%; 90.0 [86.7–91.5]), with short time within hypoxaemia (9.4±9.7%; 6.1 [4.9–8.6]) and hyperoxaemia (2.8±6.2%; 0.7 [0.3–1.4]) ranges. Interpretation: The use of automated closed-loop oxygen administration is efficient and feasible, and it provides valuable information concerning the status of COVID-19 patients. Trial Registration: ReaSTOC study, ClinicalTrials.gov identifier: NCT02893462). Funding Statement: This study was funded by Brest University Hospital. No funding was received from Oxynov Inc. for this study, except in the form of technical expertise for clinical score implementation (qCSI and EWSO2) and interpretation. Declaration of Interests: Erwan L’HER and Francois LELLOUCHE are the inventors of the FreeO2 device and shareholders of Oxynov Inc., which produces the system. Erwan L’HER is a consultant for GE Healthcare, Smiths Medical, and Sedana. Quang-Thang NGUYEN is the research engineer from Oxynov Inc. that is responsible for the development of FreeO2. Victoire PATEAU is the research assistant from Brest ICU, and her salary is funded by Oxynov Inc. Severine ANSART has no conflict of interest. Ethics Approval Statement: Our local ethics committee (Comite d’Ethique du CHRU de Brest) approved the data collection and waived written informed consent according to French legislation.