297: Innovation in the Pandemic: A Pilot Study of Improvised CPAP for COVID-19 Respiratory Failure

INTRODUCTION: Increasing functional residual capacity with positive pressure improves ventilation perfusion mismatch and relieves hypoxemia However, most noninvasive interfaces must be tightly applied for prolonged periods and may lead to facial decubiti and delirium High flow O2 systems with wide bore nasal cannulas are not widely available in resource-limited settings We hypothesized that applying CPAP through a high flow nasal cannula (improvised nasal CPAP or iCPAP) may provide sufficient alveolar recruitment and greater patient comfort The objective of this study was to describe the use of a respiratory support strategy using iCPAP in COVID-19 respiratory failure METHODS: In a prospective, observational design, hemodynamically stable, non-hypercapnic patients, admitted with COVID-19 hypoxic respiratory failure and on iCPAP, were included For iCPAP ResMed LumisTM 150 VPAP ST-A was used with Respicare High-flow nasal cannula iCPAP failure was characterized as advancement to conventional BiPAP or endotracheal intubation Data on ROX index, iCPAP failure, organ failure, survival to discharge was collected RESULTS: 113 patients were admitted;improvised CPAP (iCPAP) was applied in 45 (49%), 6 (13%) alternated between bilevel positive pressure and iCPAP Mean APACHE II score was 19 3 ±4 7, 37 (82%) were male, mean age was 52±12 9 years Systemic steroids were given to 45 (100%), 41 (91%) patients underwent awake proning, 8 (18%) and 7 (16%) were given concurrent tocilizumab and convalescent plasma Mean PaO2/FiO2 ratio at admission was 129 ±63 (range 43,380) Mean CPAP applied was 12 ±2 7 cmH20 For patients receiving concurrent BiPAP, mean IPAP was 15 ±3 7 cmH20, EPAP 11 ±4 cmH20 were applied Mean ROX index 2 hours after iCPAP application was 10 9 ±5 5 iCPAP failure occurred in 3 (6 6%) patients;1 required intubation and invasive ventilation and 2 were advanced to only BiPAP support 39 (87%) patients were discharged alive from the ICU;2 (4 4%) developed acute kidney injury, 5 (11%) had deranged liver function tests One patient (2%) had a cerebrovascular event 1 patient (2%) had a cardiac arrest CONCLUSIONS: iCPAP application in carefully selected patients and with intensive care monitoring, may represent a valid form of respiratory support in COVID-19 hypoxic respiratory failure