人工呼吸器付き ICU 患者の目標 O2 を自由化することで、UCHealth で重症患者のケアを最適化

This project will be a multi-hospital, cluster randomized study to implement a targeted oxygen saturation (SpO2) range of 90-96% in adult patients receiving mechanical ventilation (MV) using a multimodal educational intervention strategy. Hospital-level randomization to receive the intervention will occur sequentially for a phased roll-out of multimodal education and real-time informatics tools to enhance adherence of the targeted SpO2 range in eligible mechanically ventilated patients. This phased roll-out is achieved through a stepped wedge approach in a one-way crossover design in which all participating hospital ICUs will ultimately implement the intervention, and the timing of the intervention implementation is randomly ordered. The intervention will occur using 4 clusters that switch from control to intervention at 4 pre-determined time points, 1 for each cluster. For each switch, the design incorporates a 2-week transition period for staff education/implementation at each cluster, during which training will be delivered to each participating hospital to improve implementation. The investigators will provide standardized education and materials to local stakeholders for hospital and ICU-level implementation. During the transition period, the cluster cannot be considered as either receiving the structured usual care intervention (education-based intervention) or the unstructured usual care control (current state, no education-based intervention), and thus patients treated in the hospital during that period will not be enrolled or included in the analysis.

The investigators define a targeted SpO2 range (90-96%), PaO2 (60-100 mmHg) target based on thresholds defined in recent published work that incorporates expert opinions from the field and includes consensus among critical care COGG members that represent UCHealth hospitals throughout the North, South, and Metro regions as well as the site locations at which the intervention is proposed to occur. The oxygen target range will be defined by an oxygen saturation (SpO2) 90-96%, and when arterial blood gases are available, an arterial oxygen pressure (PaO2) 60-100 mmHg. The intervention will start when patients requiring mechanical ventilation arrive to an ICU in a participating hospital and the intervention will continue for any period of invasive mechanical ventilation required during ICU stay. A project goal is to achieve adherence to >95% of eligible patient-hours spent in the target SpO2 range, excluding time on FiO2 30% when above SpO2 96% or time on FiO2 100% when below SpO2 90%. Each hospital site will contribute pre-implementation (control) and post-implementation (intervention) data, with the start of the consensus-based intervention period defined by the randomized timing in the stepped wedge design.

Patients will be ventilated with a standardized UCHealth-wide mechanical ventilation order-set to achieve these SpO2 and PaO2 goals. These order-sets included the ARDSNetwork low- and high-PEEP/FiO2 arms, as well as developed protocols for alternative modes of ventilation, such as APRV which can be used at the provider teams' discretion. Data collection for Phase I began 9/1/2024 and will end 9/15/2025.

Once all hospital clusters have transitioned to the intervention phase, a nested study will commence that involves a patient-level intervention for mechanically ventilated patients experiencing residual occult hypoxemia among participating UCHealth ICUs (Phase II). Data collection for Phase II began 3/25/2026 and will end 7/31/2026. These patients will be identified automatically in the EHR (EPIC) by comparing the SaO2 from an ABG laboratory procedure to contemporaneous SpO2 flowsheet row values and then will be randomized 1:1 to either an active or silent notification delivered in real-time to members of the provider team alerting them to each OH instance, with a 12-hour notification lockout. A unique research study order will be created in EPIC for this nested, Phase II study to facilitate tracking of outcomes for patients randomized to silent and active notification arms in coordination with a health data compass analyst (HDC).

Educational interventions will include:

1. Standardized materials to discuss the rationale and procedural aspects of the intervention, including details on the specific oxygen target-related modifications to the mechanical ventilation order-set prior to starting the intervention phase.
2. EPIC-generated push notification to (1) alert providers in real-time to the presence of OH in their patient with a time-stamp based on SpO2 and SaO2 values compared from vital signs and arterial blood gas oxygen saturation respectively, and (2) inform them of the potential associations of OH with worse outcomes in similar, critically ill patients, and (3) choices for potential changes to management, including an "acknowledge only" option