Impacto da adição da insuflação-exsuflação mecânica na taxa de sucesso de desmame de pacientes traqueostomizados em ventilação mecânica prolongada

Abstract

BACKGROUND: Tracheostomized patients undergoing prolonged mechanical ventilation (MV) may experience weaning failure due to the accumulation of airway secretions. The addition of mechanical insufflation-exsufflation (MI-E) to the weaning protocol may facilitate secretion clearance and improve weaning success rates OBJECTIVE: To evaluate the feasibility of adding MI-E to a weaning protocol in tracheostomized patients undergoing prolonged MV and to evaluate its impact on weaning success. DESIGN: Single-center, open-label, randomized, controlled trial. SETTING: Intensive Care Unit (ICU) at the University Hospital, Federal University of Juiz de Fora, Minas Gerais, Brazil. PATIENTS: Adult tracheostomized patients receiving MV for more than 14 days and meeting criteria for weaning initiation (respiratory stability, adequate oxygenation, hemodynamic stability, absence of infection, and absence of severe metabolic disturbance). INTERVENTIONS: Patients were randomized into two groups: MI-E group (MI-EG) and control group (CG). In both groups, weaning was performed using spontaneous breathing through a T-piece for 12 hours on the first 2 days and 24 hours for an additional 5 days. MI-EG patients received two daily MI-E sessions, each consisting of 5 cycles (insufflation pressure of +40 cm H₂O for 2 seconds and exsufflation pressure of –45 cm H₂O for 3 seconds). Both groups received standard respiratory and motor physiotherapy twice daily. OUTCOMES: The primary outcomes were the feasibility of adding MI-E to the weaning protocol and the weaning success rate. Secondary outcomes included time to successful weaning, ICU, hospital and 60-day mortality rates, and adverse events related to MI-E. RESULTS: A total of 47 patients were randomized, with 25 in the MI-EG and 22 in the CG. Adding MI-E to the weaning protocol was feasible. Adverse events occurred in 14% of MI-E sessions (hypertension [5.9%], tachycardia [5.6%], and desaturation [2.2%]), but they were transient and did not require treatment discontinuation. Weaning success was achieved in 76% of MI-EG patients and 73% of CG patients (relative risk, 1.05; 95% CI, 0.75–1.46). No significant differences were observed between MI-EG and CG in ICU mortality (16% vs. 18%; p = 0.99), hospital mortality (40% vs. 41%; p = 0.99), or 60-day mortality (36% vs. 32%; p = 0.99). CONCLUSIONS: Incorporating mechanical insufflation–exsufflation into the weaning protocol for tracheostomized patients is feasible but did not increase weaning success rates or reduce mortality.