This technology is an artificial intelligence software that extracts clinician expertise from nurse documentation to predict clinical decisions.
Unmet Need: Clinical decision support systems demonstrating improved patient outcomes
In hospital settings, early identification of patients at risk for deterioration could prevent avoidable adverse outcomes, such as mortality and sepsis. Clinical decision support systems and patient surveillance algorithms can assist clinicians in diagnostics, patient monitoring, and treatment pathways to identify at-risk patients. However, these systems typically rely heavily on physiologic factors, such as labs and vital signs, that are often late indicators of deterioration. Current systems can also lead to alert fatigue, have inefficient processing times, and provide biased outputs.
The Technology: Machine-learning-based clinical decision support system leveraging clinician expertise
This software is a machine-learning-based predictive model that utilizes clinician documentation along with patient clinical, physiological, and contextual data to predict the clinical trajectory of the patient and provide care recommendations. This approach harnesses clinician expertise in understanding subtle, yet observable clinical changes that are not well-captured in physiological data and electronic health records, which may lead to earlier identification of deterioration compared to existing metrics. This technology also incorporates bias mitigation for equitable care and includes a front-facing app for clinicians, patients, and caregivers to track patient progress over time.
This technology was validated in a clinical trial, which demonstrated that the software decreased patient mortality, sepsis, length of stay, transfers, and readmissions.
Applications:
- Early warning system for hospitals
- Diagnostic tool
- Medical monitoring interface
- Healthcare equity and bias mitigation method
Advantages:
- Leverages nursing and clinician expertise
- Enables earlier identification of patient deterioration
- Displays trends over time for clinicians, patients, and caregivers
- Incorporates bias mitigation
- Promotes equitable care
Lead Inventor:
Sarah Collins Rosetti, RN, Ph.D., FACMI, FAMIA
Patent Information:
Patent Pending
Related Publications:
Hobensack M, Withall J, Douthit B, Cato K, Dykes P, Cho S, Lowenthal G, Ivory C, Yen PY, Rossetti S. “Identifying Barriers to The Implementation of Communicating Narrative Concerns Entered by Registered Nurses, An Early Warning System SmartApp.” Appl. Clin. Inform. 2024 Mar; 15(2): 295-305.
Rossetti SC, Dykes PC, Knaplund C, Cho S, Withall J, Lowenthal G, Albers D, Lee R, Jia H, Bakken S, Kang MJ. “Multisite Pragmatic Cluster-Randomized Controlled Trial of the CONCERN Early Warning System.” medRxiv. 2024 Jun.
Hobensack M, Withall J, Cato K, Dykes P, Lowenthal G, Cho S, Ivory C, Yen PY, Rossetti S. “Understanding the Technical Implementation of a Clinical Decision Support SmartApp: A Qualitative Analysis. Stud Health Technol Inform. 2024 Jan; 310: 1382-3.
Rossetti, S.C., Knaplund, C., Albers, D., Dykes, P.C., Kang, M.J., Korach, T.Z., Zhou, L., Schnock, K., Garcia, J., Schwartz, J. and Fu, L.H., 2021. “Healthcare process modeling to phenotype clinician behaviors for exploiting the signal gain of clinical expertise (HPM-ExpertSignals): development and evaluation of a conceptual framework.” J Am Med Inform Assoc. 2021 Jun; 28(6): 1242-1251.
Tech Ventures Reference:
IR CU2480, CU24273, CU24275
Licensing Contact: Sara Gusik
