665 - Improving Elective Intubations (EI) in a Level IV NICU: a Quality Improvement project

NEAR4NEOS, a multi-center, prospective registry for advanced airway management in NICUs demonstrated 1) 52% and 78% first (1-ASR) and ≤2 (≤2-ASR) attempt success rate for neonatal EI 2) increased EI-associated adverse events (EI-AE) with increasing EI attempts and 3) decreased EI attempts with the use of sedation and neuromuscular blockade compared to sedation alone or no medication use. A 10-month retrospective review at our level IV NICU (“Baseline”) and a survey of NICU staff showed 1) lack of standardized documentation and tracking of a) EI attempts, b) EI-AE, 2) inconsistent premedication use, 3) poor team role assignments during EI, 4) variable and inaccurate definitions of EI attempt and 5) concern for inaccurate data.

Objective: A key driver diagram and SMART aim were established (Fig.1) to improve the EI process in our NICU.

Design/Methods: PDSA-1 (10/2019) included introduction of a pre-intubation bundle with a) time-out checklist, b) team role assignment with a “recorder”, c) datasheet, d) NoteWriter documentation template and e) education of the correct definition of EI attempt. PDSA-2 (10/2021) introduced a standardized medication algorithm. PDSA-3 (08/2022) introduced a premedication EMR order set. Data were followed in Epoch-1 (8/2019-10/2020), Epoch-2 (11/2020-9/2021), and Epoch-3 (10/2021-10/2022). Statistical process control included p charts to assess proportion of EI 1-ASR and ≤2-ASR using QI Macros. Chi² analysis, Fisher's exact test and Kruskal-Wallis test were used to assess compliance, pre-medication use, and EI-AE across epochs. 

Results: Both 1-ASR and ≤2-ASR decreased enough to change the centerline after the education of EI attempt (Fig 2). 1-ASR and ≤2-ASR steadily improved from Epoch-1 to 3. Median EI attempt rates improved from 2 in Epoch-1 to 1 in Epochs 2 and 3. Rates of hypoxemia and bradycardia during EI improved significantly across epochs (Fig 3. Tab 1). Recorded severe AE were low at baseline and did not change significantly across epochs. The use of paralytic and vagolytic agents improved significantly from baseline (Fig 3 Tab 2). Unit compliance with process measures improved across epochs (Fig 3 Tab 3).

Conclusion(s):

Our findings support the recommendation for a standardized pre-intubation bundle and medication algorithm for non-emergent EI. At our NICU, since the introduction of this QI project, an increasing use of paralytic and vagolytic premedication has corresponded with a decrease in rates of hypoxemia and bradycardia during EI. Further analysis into type of provider and training level is ongoing and will be available at the time of publication.