Análise quantitativa da pressão de dióxido de carbono expirado por capnografia como índice preditivo de sucesso na retirada da ventilação mecânica invasiva em pediatria

Abstract

Invasive mechanical ventilation (IMV) corresponds to one of the most common intervention in the intensive care unit; approximately 30% of hospitalized patients have difficulty in the removal of ventilatory support requiring IMV for a prolonged period. Currently, the decision to extubate patients is based on clinical experience of the medical staff and in poor criteria, which makes the decision for IMV remotion harder. The analysis of end-tidal carbon dioxide pressure (PetCO2) by capnography is a continuous method, noninvasive and capable to evaluate the ventilation during the entire respiratory period. Although there are numerous indications to measure the PetCO2, the capnograph is not used during the process of weaning from IMV. Thus, the main objective of this study was to examine quantitatively PetCO2 and indexes extracted from the capnogram in neonates and pediatric patients, as a predictor of successful extubation of IMV in the postoperative period of cardiac surgery. Initially, the capnograph was compared to other commercialized equipment and it showed high levels of correlation (r2 = 0.9872 and 0.9513, p < 0.001), and a mean difference obtained by Bland- Altman test close to zero (-0.56 and -1 mmHg), demonstrating its reliability and accuracy application. In the second phase, a sample consisted of 254 patients were evaluated during the pre and pos-extubation from IMV. There were weaning success in 70.47% (179) and failure in 29.53% (75), being necessary return the patient to ventilatory prosthesis before 48 hours after extubation. Comparing the successful group with the group of patients who have failed, a significant statistically difference (p < 0.001) was found for age (15.51 versus 3.16 months), weight (8.17 versus 4.5 kg) and duration of mechanical ventilation (2.34 versus 6.3 days). Considering ventilatory parameters, no significant difference was observed in the values of PetCO2 and arterial carbon dioxide pressure (PaCO2) between the two groups (success and failure). The Receiver Operating Characteristic (ROC) analysis showed that the mean values of PaCO2 and PetCO2 had low sensitivity and specificity to predict extubation failure. However, P(a-et)CO2 showed a significant difference between the success and failure group of extubation (0.42 versus 1.27 mmHg; p < 0.001) in the pre-extubation period; and it presented area under curve (AUC) of 0.707 which indicates acceptable performance to discriminate the extubation outcome. Furthermore, eight indexes extracted from capnogram presented significant difference between groups (p < 0.001), AUC values greater than 0.8 and high sensitivity and specificity to predict extubation failure, indicating excellent discrimination between groups with high clinical significance, these were: expiratory time, respiratory cycle total time, slope 1, slope 2, slope ratio, α angle, ascending angle and inclination angle. The study has shown that the analysis of ventilatory parameters is relevant in the process of removing the patient from IMV, increasing the safety for the multidisciplinary team on the decision to extubate a patient.