
Managing Hospitalization Risk & Better Home Care Delivery Using LLMs
The first-ever LLM-based smart AI assistant for home-based care.
Project Overview
Updated February 29, 2024
The Problem
The global healthcare industry is facing unprecedented challenges as populations age and present increasingly complex care needs. This strain is particularly evident in the home care industry facing labour shortages and razor-thin profit margins. Managing client records alone can contain a plethora of unstructured information containing important, but seldom leveraged, clinical information. Often occurrences like ‘a mention of pain’ may only be recorded in a note, yet may be a top predictor of a preventable hospitalization.
To remain competitive, flexible and continuing to serve the increasing needs of Canadians, agencies need to manage patient outcomes, optimize their workforce and leverage AI in all aspects of care delivery and clinical evaluation. Large language models (LLMs) offer unparalleled opportunity to query and summarize this data into detailed decision support for caregivers to treat patients more efficiently and spend less time on administration.
How We Are Solving It
This project will leverage the potential of large language models (LLMs) and predictive algorithm for AlayaCare’s flagship AC Cloud product, an extensive health record platform currently used by over 700 home care service providers across North America and Australia. Anticipated features powered by the algorithm will include the first ever LLM-based smart AI assistant for home-based care, offering a broad range of natural language summarization and conversational abilities, risk dashboards, and providing decision support to clinical supervisors and caregivers. Enhanced features will utilize patient clinical data such as comorbidities, diagnoses, previous hospitalizations and falls to automatically identify which patients receiving home care are at risk of hospitalization. As part of managing resources, these features can support modification of homecare provider schedules to mitigate the risk of a patient flagged for needing hospital care.
In collaboration with Acclaim Health, Bien Chez Soi and Polytechnique Montréal; these enhancements to AC Cloud will aim to free up health system capacity and drive more efficient clinical decision making at scale, reducing the rate of adverse events such as falls, emergency room visits and preventable hospitalizations. These features will help AlayaCare further penetrate the $2B global home health care market, representing a significant market share for a Canadian company.
Project Lead
Project Partners
-
"This funding enables us to embrace technology to improve the care Canadians receive at home — something that benefits all of us. By reducing administrative tasks, care workers will be able to increase their caseloads by 10-15%, allowing them to serve more people in need of home care. Additionally, the project aims to reduce the waitlist for chronic care patients receiving home care by 10%. These efficiency gains will not only benefit individuals in need of care but also contribute to lessening the burden on the healthcare system and alleviating overcrowding in hospitals.”
-
“DIGITAL is excited to build upon the over 5 years of investments we’ve made in applied AI with this latest announcement. Our co-investments alongside innovators like AlayaCare, in collaboration with home care providers like Acclaim and Bien Chez Soi, are how we can continue to build on Canada’s reputation as a global center for AI R&D while focusing on advancing global commercial pathways for Canadian companies. This latest announcement with AlayaCare is an example of how Canada can increase equitable access and better health outcomes for Canadians through AI adoption.”