Hypoxemic respiratory failure in patients with acute heart failure.

INTRODUCTION: The incidence of hypoxemia and the indications for oxygen therapy in patients with acute heart failure (AHF) are not well known. METHODS: A large nationwide registry of unselected AHF patients attended in the emergency department (ED) was analysed. Oxygen saturation by pulse-oximetry (SpO2) was measured on ED arrival, and the ratio SpO2/FIO2 was calculated to adjust for the impact of pre-hospital oxygen therapy, stratifying patients into three groups: normal ratio (≥ 452; corresponding approximately to a SpO2>94%, on room air), mild decrease of the ratio (429-451; SpO2~90-94%) and hypoxemia (≤ 428; SpO2~<90%). Unadjusted and adjusted 30-day all-cause mortality was analyzed by category and restricted cubic spline (RCS). RESULTS AND DISCUSSION: Of 18,904 AHF patients, 6949 (36.7%) had normal SpO2/FIO2, 5438 (28.8%) showed mild decrease of the ratio, and 6517 (34.5%) were hypoxemic. Hypoxemia was documented in 40% of hospitalized patients and was observed in all phenotypes, ranging from 18.6% in acute decompensated heart failure, to 100% in acute pulmonary oedema. The severity of most physiological parameters, the level of biomarkers, the rate of comorbidities and the 30-day all-cause mortality risk were inversely related to SpO2/FIO2. Mortality risk progressively increased for SpO2/FIO2<452 but became nearly flat when it was<350 suggesting that the prognosis did not depend only on the severity of respiratory failure. CONCLUSIONS: Impaired oxygenation was observed in nearly two thirds of patients with AHF attended in ED. Hypoxemia occurred in 40% of those hospitalized, affecting all phenotypes and was associated with more severity and worse prognosis. This should be considered when addressing treatment recommendations for these patients.