754 - QI Initiative to Reduce Invasive Ventilation in Extreme Preterm Infants

Publication Number: 1730.754

Ahmed Ismaiel, Montefiore Medical Center/Albert Einstein College of Medicine, New York, NY, United States; Lan Jiang, Ann & Robert H. Lurie Children's Hospital of Chicago, Chicago, IL, United States; Allie Liebmann, Albert Einstein College of Medicine, Bronx, NY, United States; Johana Abraham, The Children's Hospital at Montefiore, Bergenfield, NJ, United States; Edymel L. Bondal, Montefiore Medical Center (Weiler), YONKERS, NY, United States; Nefertiti Cano, The Children's Hospital at Montefiore, Teaneck, NJ, United States; Sheri L. Nemerofsky, The Children's Hospital at Montefiore, Bronx, NY, United States

Background: Mechanical ventilation is a cornerstone of neonatal intensive care. Prolonged invasive ventilation is associated with ventilator-induced lung injury, which contributes to the development of bronchopulmonary dysplasia (BPD), a major cause of long-term morbidity in preterm infants. Evidence supports the use of noninvasive ventilation and early extubation strategies to reduce the incidence of BPD and mortality.

Objective: QI project was initiated in January 2024 in an urban Level IV NICU to decrease both the rate and duration of intubation by 20% during the first month of life in extreme preterm infants (24-29 weeks) by December 2025.

Design/Methods: A fishbone diagram and a key driver diagram (Fig. 1) were developed to identify possible contributing factors to intubation and prolonged invasive ventilation. Primary outcomes were the rate and total duration of invasive ventilation. Secondary outcomes included the rate of BPD, length of stay, and time to reach full enteral feeds. Process measures tracked days with an extubation readiness score >60, while balancing measures monitored reintubation rates. 4 PDSA cycles were conducted in addition to the interventions. Cycle 1 focused on enhanced delivery room training and bubble CPAP education. Cycle 2 aimed to standardize extubation readiness and reintubation criteria based on blood gas and minimal ventilator settings. Cycle 3 promoted implementation of non-invasive neurally adjusted ventilatory assist as a primary post-extubation support mode. Cycle 4 introduced an extubation readiness calculator to estimate the probability of successful extubation, validated retrospectively to ensure applicability. A process map flyer was implemented and refined with PDSA cycles.

Results: Baseline data included 142 neonates, with 28 excluded (11 died, 9 had surgical intervention, and 8 transferred). The mean intubation rate was 61%, with an average duration of 10.7 days. There were 78 neonates in the intervention phase, and 20 were excluded (5 died, 7 had surgical intervention, and 8 transferred). After the interventions, the mean intubation rate was 41%, representing a 33% reduction (Fig. 2). Run chart analysis demonstrated a consistent shift below the baseline mean, suggesting special cause variation and indicating a sustained process improvement. The average duration of invasive ventilation decreased from 10.7 days to 4.6 days, a 57% reduction (Fig. 3). The rate of failed extubation was 7%.

Conclusion(s): QI interventions resulted in a reduction in the rate and duration of invasive ventilation in extreme preterm infants, demonstrating sustained improvement in respiratory care.

Fig 1: Key Driver Diagram
This diagram illustrates the key drivers and change ideas aimed at decreasing the rate and duration of invasive ventilation in preterm infants

Fig: 2 Rate of intubation in Extreme preterm infants
Fig 2. Control chart demonstrating a progressive decrease in the rate of intubation among extreme preterm infants during the first month of life . Each point represents 10 patients (except point 11 with 12 patients and point 16 with 7 patients). The QI initiative began at point 10.

Fig 3: Duration of invasive ventilation in Extreme preterm infants
Fig 3. Control chart showing a reduction in the duration of invasive ventilation among extreme preterm infants. Each point represents 10 patients per group.