Objective: To determine the incidence and risk factors of accidental extubation (AE) in a tertiary neonatal intensive care unit.
Methods: A prospective cohort study was conducted to determine AE incidence density per 100 patient-days, during a 23-month period, in 222 newborns receiving assisted ventilation (AV). Logistic regression analysis was used to determine risk factors for AE. The presence of a cyclical pattern in extubation rates, according to the variables of interest, was investigated by Cosinor analysis.
Results: The mean AE rate was 5.34/100 patient-days ventilated. AE-associated predictive variables were: subsequent use of the oral and nasal routes during AV [relative risk (RR) = 4.73; 95% confidence interval (95%CI) 1.92-11.60], AV duration (per day, RR = 1.03; 95%CI 1.02-1.04), and number of patient-days ventilated (RR = 1.01; 95%CI 1.01-1.02). According to the adjusted multiple regression analysis, total AV time was the only independent predictor of AE in this sample (RR = 1.02; 95%CI 1.01-1.03). AV time of 10.5 days showed an accuracy of 0.79 (95%CI 0.71-0.87) for the occurrence of AE. Cosinor analysis showed significant periodicity in overall AE rate and in the number of patient-days ventilated. There was a significant correlation between the number of patient-days ventilated and AE frequency.
Conclusion: Mean AE density was 5.34/100 patient-days ventilated. AV duration was the only independent predictor of AE. The best accuracy for AE occurrence was achieved at 10.5 days of AV duration.