Predicting and Responding to Clinical Deterioration in Hospitalized Patients by Using Artificial Intelligence:...

Area of Science:

• Healthcare Informatics
• Clinical Decision Support Systems
• Patient Safety

Background:

• Early identification of clinical deterioration is crucial for improving patient outcomes but remains challenging in busy hospitals.
• Artificial intelligence (AI) offers potential solutions for predicting clinical deterioration through predictive models.
• Existing hospital systems face challenges in effectively integrating AI for real-time clinical decision support.

Purpose of the Study:

• To assess if an AI-enabled work system, using the SEIPS 2.0 model, improves clinical outcomes.
• To describe the implementation of the clinical deterioration index (CDI) predictive model and associated workflows.
• To define emergent properties of the AI system that influence clinical outcomes.

Main Methods:

• A mixed-methods approach informed by the SEIPS 2.0 model will be employed.
• A modified stepped wedge design over 11 months, with three stages: baseline, physician-only AI prediction, and multidisciplinary team AI prediction.
• Quantitative data from electronic health records and qualitative interviews with multidisciplinary teams will be collected and analyzed.

Main Results:

• Pilot study initiated in December 2020.
• Full study results are anticipated by mid-2022.
• The study is designed to capture both process and outcome data related to AI implementation.

Conclusions:

• This protocol outlines an evaluation approach for a complex, AI-driven intervention.
• Assessing both processes and outcomes is vital for understanding AI's impact on clinical deterioration management.
• The study emphasizes the importance of a multifaceted evaluation for AI-enabled healthcare systems.