Abstract
BACKGROUND: During the COVID-19 pandemic, a significant number of affected individuals experienced acute respiratory failure requiring invasive mechanical ventilation (IMV). Once IMV becomes necessary, substantial efforts are made to conduct a successful ventilatory weaning process and prevent deleterious effects associated with prolonged positive pressure use. Identifying the clinical characteristics of COVID-19 patients who experienced failure in ventilatory weaning can provide insights to optimize ventilatory management. AIM: This study is aimed at identifying predictors of weaning failure (WF) from IMV in patients with COVID-19 and their association with mortality. METHODS: This retrospective cohort study analyzed data obtained from medical records, including personal and clinical information, IMV parameters, ventilatory mechanics, IMV duration, hospital and intensive care unit (ICU) length of stay, and discharge or death date. All COVID-19 patients admitted between March 2020 and July 2022, aged ≥ 18 years, and requiring IMV were included. Patients were categorized into three groups: weaning success (WS), WF, and no criteria for weaning (NCW). Success was defined as orotracheal prosthesis removal without reintubation within 48 h or, for tracheostomized patients, disconnection from mechanical ventilation for ≥ 48 h. RESULTS: A total of 2.198 individuals were included, of whom 528 (age 52 [39-61] years, 195 men, and body mass index 29 [26-34] kg/m(2)) met the criteria for weaning and were analyzed (WS: n = 380, 195 men; WF: n = 148, 89 men). Individuals in the WS group were younger (WS: 52 [39-61] vs. WF: 56 [47-67] years; p < 0.0001) and had shorter ICU stay (WS: 16 [10-27] vs. WF: 29 [14-45] days; p < 0.0001) and IMV duration (WS: 12 [7-20] vs. WF: 25 [14-37] days; p < 0.0001). Independent predictors of WF included age (OR [95% CI]: 1.028 [1.005-1.052]), length of hospital stay (OR [95% CI]: 0.963 [0.937-0.990]), time on IMV (OR [95% CI]: 1.103 [1.059-1.149]), and driving pressure (OR [95% CI]: 3.750 [1.344-10.466]), regardless of gender and comorbidities. Mortality was higher in the WF group (WF: 69% vs. WS: 24%; p < 0.0001). CONCLUSION: Advanced age, the length of hospitalization, prolonged IMV duration, and compromised respiratory mechanics were the predictors of WF. Furthermore, individuals who experienced WF presented a higher mortality rate compared with those who successfully weaned.