“Risky States” Optimizing ICU Safety Through Patient Engagement, System Science and Information Technology Beth Israel Deaconess Medical Center; MIT; Aptima Abstract Methods “ICU Intensity Index” Dashboard Jason I will be adding a live link under the Intensity Index I may also take the Environmental predictors and the Harm predictors and have them as powerpoint links. We used existing data and expert analysis to identify fundamental risky states (risk events/event drivers) in the ICU environment, system and people in the system that impact the likelihood of risky events or actual harm to occur. We identified corresponding harm outcomes during and following these “risky states” and will use them to inform the development of new holistic mitigation strategies Identified drivers through retrospective review of root cause analyses Identified harms (voluntary reports, billing data, IHI ICU global triggers tool Analyzed 2 years of retrospective data for all ICU patients 2012 – 2014 Objective Acquired and reviewed 2 years of retrospective data Analyzed retrospective data and develop model Designed application to display real time information to ICU staff about risky level Conclusions It was impossible to create a working list of MD’s and RN’s assigned to care for each ICU patient at any point in time from existing electronic data sources It will be important to capture “drivers” that we have not considered in the first analysis of retrospective data Results Team Collaborators Funded by the generosity of the Gordon and Betty Moore Foundation
Moderate Risk State
High Risk State
Add a Patient Concern
Add a Unit Concern