Predictive Factors for Failure of Non-invasive Ventilation and High-flow Oxygen Therapy in COVID-19 Patients: A Retrospective Observational Study.

COVID-19 is an infectious disease caused by SARS-CoV2 (severe acute respiratory syndrome by the novo coronavirus). During the progression of COVID-19 infection, some patients may require noninvasive ventilation (NIV) or high-flow nasal catheter oxygen therapy (HFNC). However, its use has proved controversial in this population. Objective: To describe, retrospectively, possible predictor variables related to the use of NIV and HFNC, in order to associate them with their failure and consequent orotracheal intubation. Materials and Methods: observational, retrospective study, according to STROBE model, from the analysis of medical records, between March 2020 and July 2021. Study approved by a co-substantiated committee of D'or research and teaching institute (CAAE: 52534221.5.0000.5249). Inclusion criteria: over 18 years of age, diagnosis of positive COVID-19, hospitalized in the ICU of a tertiary Hospital located in Rio de Janeiro, admitted between March 2020 and July 2021. Exclusion criteria: hospital stay time of less than 3 days; patients who evolved to IOT in less than 48 hours; patients whose therapy (NIV, CNAF or NIV + CNAF) lasted less than 48 hours; patients whose medical records do not have the variables for the study. Primary outcome: incidence of orotracheal intubation. Secondary outcomes: time for orotracheal intubation, length of stay in the ICU and hospital and in-hospital mortality. Statistical analysis: There was no sample calculation due to the exploratory, descriptive and retrospective nature of this study. The Mann-Whitney U test will be applied to assess differences in icu length of stay. Survival time as well as hospital length of stay until the event will be analyzed with Kaplan-Meier estimates; the log-rank test will be used for group comparison. The p-< 0.05 will be considered statistically significant. Preliminary results: 82.6% of the medical records completed phase 1 of data collection. Of these, 64.4% used CNAF, NIV or both. 20% of the medical records passed through phase 1 completed for phase 2 collection, with characterization of the population. The absence of heart disease and dementia, as well as the acute onset of symptoms (less than or equal to 10 days) and age (between 40 and 79 years) showed a statistically significant trend.