RESUMO
BACKGROUND: Unplanned extubation is one of the most common preventable adverse events associated with invasive mechanical ventilation. OBJECTIVE: This research study aimed to develop a predictive model to identify the risk of unplanned extubation in a pediatric intensive care unit (PICU). METHODS: This single-center, observational study was conducted at the PICU of the Hospital de Clínicas. Patients were included based on the following criteria: aged between 28 days and 14 years, intubated, and using invasive mechanical ventilation. RESULTS: Over 2 years, 2,153 observations were made using the Pediatric Unplanned Extubation Risk Score predictive model. Unplanned extubation occurred in 73 of 2,153 observations. A total of 286 children participated in the application of the Risk Score. This predictive model was created to categorize the following significant risk factors: 1) inadequate placement and fixation of the endotracheal tube (odds ratio 2.00 [95%CI,1.16-3.36]), 2) Insufficient level of sedation (odds ratio 3.00 [95%CI,1.57-4.37]), 3) age ≤ 12 months (odds ratio 1.27 [95%CI,1.14-1.41]), 4) presence of airway hypersecretion (odds ratio 11.00 [95%CI,2,58-45.26]) inadequate family orientation and/or nurse to patient ratio (odds ratio 5.00 [95%CI,2.64-7.99]), and 6) weaning period from mechanical ventilation (odds ratio 3.00 [95%CI,1.67-4.79]) and 5 risk enhancement factors. CONCLUSIONS: The scoring system demonstrated effective sensitivity for estimating the risk of UE with the observation of six aspects, which overlap as an isolated risk factor or are associated with a risk enhancement factors.