A55-18 Factors Predicting Unplanned Extubations

Abstract Rationale Unplanned extubations (UE), particularly self-extubations (SE), are associated with acute hypoxemia, aspiration, emergent re-intubation, and increased intensive care unit (ICU) utilization. They often arise from modifiable drivers such as agitation, delirium, inadequate sedation, and workflow gaps. At Newark Beth Israel Medical Center, we noted recurring UE events across multiple ICUs and hypothesized that most of these events would occur during times where there is a higher patient-to-nursing staffing ratio, such as shift change and weekends, and would not be associated with the use of restraints. Methods We performed a retrospective QI review of consecutive UE events at an urban teaching hospital in 3 separate closed adult ICUs, between February 2024 and October 2025. (N = 30 events). Data was collected and recorded into de-identified database, including time, (day of week, timing of day in relation to shift change;), mechanism (self-extubated, accidental dislodgement), presence of physical restraints, sedation (status, type, recent changes), ventilator mode (pressure regulated volume control vs pressure support), and presence of an endotracheal tube (ETT) holder. Outcomes were classified as: required re-intubation, remained extubated without noninvasive support, or remained extubated on noninvasive positive pressure ventilation (NIPPV). Results Conclusion On preliminary analysis, we have not identified any clear associations due to the small sample size. However, most patients did not require re-intubation and remained stable, suggesting they were already near readiness for ventilator liberation at the time of UE. We found that the majority of UE (73%) occurred on weekdays, and less often within an hour of shift change (80%) than at other times (20%). These events occurred despite the use of ETT holders (96.7%) and 2-point mechanical restraints (70%). While a larger sample size is needed to delineate statistically significant factors associated with UE, our initial data does not support our hypothesis that these events occur during periods of lower staffing ratios. This data may support an opportunity to improve our process of timely extubation. Our future QI aim will be to partner with nursing and respiratory leadership to standardize paired daily Spontaneous Awakening Trials (SAT) and Spontaneous Breathing Trials (SBT), and standardize timely extubation criteria. We will also implement a delirium bundle with early mobilization and family engagement to reduce mechanical restraint use. We will continue to monitor UE rate per 100 ventilator days, re-intubation within 48 hours after UE, mechanical restraint usage rate, and completion of sedation documentation as balancing and outcome measures. This abstract is funded by: None