The epidemiology and outcomes of adults with acute hypoxaemic respiratory failure in a low-income country in the context of the COVID-19 pandemic: a prospective, observational, multicentre cohort study.

BACKGROUND: Few data regarding the incidence and outcomes of acute hypoxaemic respiratory failure (AHRF) in low- and middle-income countries exist. METHODS: We undertook a prospective, observational multicentre study at 11 Ugandan hospitals (July 2020-April 2021) to determine the prevalence, aetiology and 28-day all-cause mortality of AHRF (acute shortness of breath plus peripheral oxygen saturation <91% while breathing ambient air) in adults (≥18 years) who required unplanned hospitalisation. FINDINGS: 16 747 adults required unplanned hospitalisation during the study period. The median age of study participants was 50 years, and 65.1% were male. The prevalence of AHRF was 4.1%. The predominant causes were pulmonary (46.8%) and extrapulmonary infection (18.3%). Only 38 patients (5.6%) received invasive mechanical ventilation. All-cause mortality 28 days after hospitalisation was 37.9% and associated with the severity of hypoxaemia at presentation (p<0.001). Risk factors for death included oxygen saturation (adjusted relative risk (aRR) 0.96 (95% CI 0.93 to 0.98); p=0.001), the lung injury prediction score (aRR 1.83 (95% CI 1.43 to 2.36); p<0.001), respiratory rate>30 breaths per minute (aRR 2.39 (95% CI 1.34 to 4.26); p=0.003) and age >65 years (aRR 2.09 (95% CI 1.13 to 2.86); p=0.02). INTERPRETATION: In the context of the COVID-19 pandemic, the prevalence of AHRF among adults requiring unplanned hospitalisation in Uganda was comparable with that reported by previous single-centre studies. Pulmonary infection was the most common cause of AHRF. The high 28-day mortality may be explained by the severity of the disease at presentation and the limited access to advanced organ support, including invasive mechanical ventilation.