Abstract
BACKGROUND/AIM: This study analyzed clinical factors impacting the survival of COVID-19 patients with acute respiratory distress síndrome, or ARDS (CARDS) to ICU compared to non-COVID-19 ARDS patients. METHODS: Clinical variables from 1,008 CARDS cases and 332 ARDS cases were computed using learning algorithms. The multivariable Cox proportional hazards regression models with the enter method evaluated risk factors and ICU mortality relationships. The survival analysis was completed with Kaplan-Meier and the log-rank tests. RESULTS: A Random Forest model revealed that mechanical ventilation-related factors, oxygenation, blood pH, superinfection, shock, and ICU length of stay have the greatest effects on ICU survival. According to a multivariate Cox model, reintubation and a high-flow nasal cannula were essential for survival in CARDS patients during the ICU stay. The length of stay in the ICU diminishes in patients older than 45 years, regardless of the source of ARDS. CONCLUSION: This study gives recommendations for the respiratory care of ARDS in COVID-19 patients.