Abstract
During the first two waves of the COVID-19 emergency in Italy, internal medicine high-dependency wards (HDW) have been organized to manage patients with acute respiratory failure (ARF). There is heterogeneous evidence about the feasibility and outcomes of non-invasive respiratory supports (NIRS) in settings outside the intensive care unit (ICU), including in patients deemed not eligible for intubation (i.e., with do-not-intubate, DNI status). Few data are available about the different NIRS modalities applied to ARF patients in the newly assembled internal medicine HDW. The main aim of our study was to describe a real-life experience in this setting of cure, focusing on feasibility and outcomes. We retrospectively collected data from COVID-19 patients with ARF needing NIRS and admitted to internal medicine HDW. Patients were treated with different modalities, that is high-flow nasal cannula (HFNC), continuous positive airway pressure (CPAP), or non-invasive mechanical ventilation (NIMV). Switching among different NIRS during the hospitalization and the success rate (weaning with the same NIRS) or failure (endotracheal intubation-ETI or in-hospital death) were recorded. Three hundred thirty four ARF patients (median age 74 years), of which 158 (54%) had a DNI status, were included. CPAP, NIMV, and HFNC's success rates were 54, 33, and 13%, respectively. Although DNI status was strongly associated with death (Gehan-Breslow-Wilcoxon test p < 0.001), an acceptable success rate was observed in these patients using CPAP (47%). Multivariate regression models showed older age (odds ratio-OR 4.74), chronic ischemic heart disease (OR 2.76), high respiratory rate after 24 h (OR 7.13), and suspected acute respiratory distress syndrome-ARDS (OR 21.1) as predictors of mortality risk or ETI. Our real-life experience shows that NIRS was feasible in internal medicine HDW with an acceptable success rate. Although DNI patients had a worse prognosis, the use of NIRS represented a reasonable chance of treatment.