Respiratory Support and Clinical Outcomes in Critically Ill Patients with COVID-19 in Intensive Care Unit: A Retrospective Study.

Background: Appropriate respiratory support is crucial for improving the clinical outcomes of critically ill patients infected with the SARS-CoV-2 virus. This study aimed to investigate the different modalities of respiratory support and clinical outcomes in patients with COVID-19 in intensive care units (ICUs).

Materials and methods: In a retrospective study, we enrolled 290 critically ill COVID-19 patients who were admitted to the ICUs of four hospitals in Mazandaran, northern Iran. Data were extracted from the medical records of all included patients, from December 2019 to July 2021. Patients' demographic data, symptoms, laboratory findings, comorbidities, treatment, and clinical outcomes were collected.

Results: 46.55% of patients died. Patients with ≥2 comorbidities had significantly increased odds of death (OR=5.88, 95%CI: 1.97-17.52, P=0.001) as compared with patients with no comorbidities. Respiratory support methods such as face mask (survived=37, deceased=18, P=0.022), a non-rebreather mask (survived=39, deceased=12, P<0.001), and synchronized intermittent mandatory ventilation (SIMV) (survived=103, deceased=110, P=0.004) were associated with in-hospital mortality. Duration of respiratory support in nasal cannula (survived=3, deceased=2, P<0.001), face mask (survived=3, deceased=2, P<0.001), a non-rebreather mask (survived=3, deceased=2, P=0.033), mechanical ventilation (survived=5, deceased=6, P<0.019), continuous positive airway pressure (CPAP) (survived=3, deceased=2, P<0.017), and SIMV (survived=4, deceased=5, P=0.001) methods were associated with higher in-hospital mortality.

Conclusion: Special attention should be paid to COVID-19 patients with more than two comorbidities. As a specific point of interest, SIMV may increase the in-hospital mortality rate of critically ill patients with COVID-19 connected to mechanical ventilation and be associated with adverse outcomes.