Implementation of lower oxygenation targets to improve outcome in ICU patients.

Multicenter study to evaluate adherence to lower oxygen targets in ICU patients using a two-phase step-wise implementation strategy. The project will increase the current knowledge as follows:

1. Are the tested implementation strategies successful with respect to adherence to oxygenation targets as recommended?

2. Does introduction of a computerized decision-support module add to the success of implementation?

3. The project will provide new evidence on the effects of implementation of lower oxygenation targets on outcome of patients which may help future large-scale implementation.

Aim of this project is to improve adherence to lower oxygen targets in ICU patients using a two-phase step-wise implementation strategy. RESEARCH QUESTIONs are: What are the effects of a stepwise implementation strategy, including a computerized decision-support module, on guideline adherence, outcome and costs of implementation?

1. April 2012: Start. Retrospective baseline measurement (period: Apr 1st 2011 - Apr 1st 2012);

2. May/June 2012: Introduction of guideline, education, SPC feedback (phase 1);

3. July 1st, 2012: Start data collection first phase of implementation;

4. December 1st, 2013: Phase 2, introduction decision-support application in PDMS. Continuation of measures of implementation;

5. July 1st, 2014: End of study, data analysis;

6. November 1st, 2014: End of project.

The study phases:

1. Retrospective: Baseline measurement over 12-months period. They include in-hospital mortality, ICU-mortality, number of days alive and without artificial ventilation at 28 days after ICU admission, PaO2 values, FiO2 values. All data are collected from existing databases (PDMS and NICE). Severity of illness, using the APACHE IV and SAPS II model, is retrieved for all patients from the National Intensive Care Evaluation (NICE) database;

2. First implementation phase. This phase will consist of providing a written guideline with clear recommendations how to adapt oxygen administration and ventilator settings depending on arterial blood gas measurements. It will include repeated education on the background of lower oxygenation targets (potential oxygen toxicity, international recommendations, present situation with non-adherence to recommendations, safety of lower oxygenation targets) and on actions to be taken to adhere to the guideline. It includes a clear description of preferred PEEP/FiO2 combinations. We will provide continuous process feedback by statistical process control (SPC), and involvement of local leadership to promote a culture that support adherence to the guideline (for more information: see section implementation). After a 3-months wash-in period, data will be collected over the subsequent 6 months. De same data will be collected as in the baseline measurement;

3. Introduction decision-support. The decision-support will be based on exactly the same targets as the guideline from phase 2 and will work in the active, critiquing mode, meaning that it will give decision-support without being asked for but only if the actual situation is not according to the guideline. A pop-up window will appear in the patient data management system if PaO2 is higher than recommended. It will suggest to adapt oxygen administration and/or ventilator settings based on collected arterial blood gas measurements. Actual decisions to change these settings will be made by nurses and physicians in the ICU. Measures from the first phase of implementation will be continued with continuous feedback and repeated education. Data will be collected after a 6-weeks wash-in period. Duration of data-collection is 6 months.

23-aug-2014 changes:
Phase 2: Data was collected over the subsequent 12 months (instead of 6 months).

Phase 3: Addition of "repeated SpO2 measurements" to "A pop-up window will appear in the patient data management system if PaO2 or repeated SpO2 measurements is/are higher than recommended.