HACOR score to predict NIV failure in patients with COVID-related hypoxemic respiratory failure managed in the ordinary ward and in the critical care setting.

We evaluated the prevalence of non-invasive ventilation (NIV) failure among patients with COVID-19-related pneumonia, managed in the ordinary ward and in the HDU/ICU and we tested the prognostic role of the HACOR score in those different settings. This is a retrospective study, conducted in the University-Hospital Careggi. We included all subjects with COVID-19 and ARF requiring NIV between March 2020 and May 2021, respectively managed in the ordinary ward (G1) and in the critical care setting (G2). Clinical parameters, HACOR and SOFA score were evaluated at Day 0 and after 1, 2 and 5 days of treatment. The primary outcome was NIV failure. 13% G1 patients and 40% G2 patients underwent endotracheal intubation (ETI). NIV was successful in 60% G1 AND 43% G2 patients (p < 0.001). In G1, compared to those with successful NIV, patients who underwent ETI, had a higher HACOR since the baseline evaluation (T0: 6 [5-6] vs 5 [3-6]; T1: 6 [5-6] vs 5 [3-6], all p < 0.05). An HACOR score > 5 was associated with an increased prevalence of ETI independent to an advanced age and a SOFA score > 5 both in G1 (T1: RR 4.87, 95% CI 1.462-16.275; T5: 3.630, 95% CI 0.979-13.462) and G2 (T0: 1.76, 95% CI 0.906-3.422; T1: 3.38, 95% CI 1.386-8.265). Among patients with COVID-related-ARF, NIV could be managed in the ordinary ward in a consistent proportion of patients and, among them, an HACOR score > 5 was independently associated with increased NIV failure from the earliest evaluations.