Application of predictive models (machine learning) on clinical deterioration based on IoT in different sceneries of health assistance: proof of concept

• Grant number: 23/00750-1
• Support Opportunities: Regular Research Grants
• Start date: August 01, 2024
• End date: July 31, 2026
• Field of knowledge: Health Sciences - Medicine
• Agreement: MCTI/MC
• Principal Investigator: Adriano José Pereira
• Grantee: Adriano José Pereira
• Host Institution: Instituto Israelita de Ensino e Pesquisa Albert Einstein (IIEPAE). Sociedade Beneficente Israelita Brasileira Albert Einstein (SBIBAE). São Paulo , SP, Brazil
• Associated researchers: Edson Amaro Junior ; Uri Adrian Prync Flato

Abstract

Clinical deterioration is worsening the clinical condition or acute onset of a severe physiological disturbance. Due to that, an effort exists to develop processes to target timely and appropriate care for deteriorating or high-risk patients. Rapid Response Systems (RRS), for instance, have been implemented to intervene and avoid preventable death, cardiac arrest, or transfer to an intensive care unit (ICU). Although the literature points out the positive impact on the clinical outcomes after the RRS, the technologies are being adopted to support this system for precocious deterioration. Several predictive model-based machine learning has been developed to predict patient deterioration. The algorithms use vital signs to extract information or patterns that might be used to predict clinical worsening. Most of these studies were applied in semi-UCI or ICU, where static monitors provide the data. In order to understand and evaluate model performance in different settings A validated algorithm for clinical deterioration will be retrained iin the emergency department (ED) and in home care patients using monitoring system dynamic for data collection, storage and processing. (AU)