“Early warning scores are tools used by hospital care teams to recognize the early signs of clinical deterioration in order to initiate early intervention and management,” said Dr. Andrea Waddell, Medical Director Quality Standards and Clinical Informatics. “Knowing ahead of time that a patient may be at risk of harm can help us develop intervention strategies such as increased nursing attention and adjustments to their plan of care.”

Data from the Canadian Institute for Health Information in 2021-22 revealed that 1 in 17 hospital stays involved unintended harm, with nearly 50% of them being preventable. Waypoint’s Dr. Waddell is also the Regional Clinical Co-Lead for Mental Health and Addictions at Ontario Health’s Mental Health and Addictions Centre of Excellence. She and her team of researchers are seeking to change this statistic to develop an early
warning score to help stop the risk of harm before it occurs.

Artificial intelligence, incorporating machine learning algorithms and natural language processing, has revolutionized various sectors and mental health care is no exception. AI can analyze vast amounts of data, identify patterns, and generate valuable insights to improve patient outcomes. When applied to mental health, AI has the potential to enhance early detection, personalize treatment plans, and reduce the burden on healthcare providers.

Early warning scores have been widely adopted in many acute medical settings, but this methodology has not been applied in mental health settings. The development of an EWS system involves the continuous monitoring and analysis of various patient-specific factors to assess the risk of deterioration. By combining historical data, real-time monitoring, and AI algorithms, a predictive model can be created to identify subtle changes that may indicate an impending crisis. Ideally alerting care providers up to 72 hours in advance so meaningful interventions can be put in place.

Waypoint and its expert staff care for some of the province’s most severely ill patients. The hospital has a 20-bed acute mental health program, has submitted a proposal to the Ministry of Health to add an additional 20-bed unit, and is shifting the culture intentionally to become a learning health system; making the hospital uniquely positioned to build this early warning model. Leveraging existing frameworks, expert opinion, and literature, the hospital is proposing variables for an EWS and testing a machine-learning model on 2022 patient data. Frontline clinicians, patients, and families will provide input at every step to guide the selection of the final algorithm. Once finalized, the EWS will be piloted in some Waypoint units using a rapid-cycle quality improvement model.

“Early Intervention and timely detection of deteriorating mental health conditions is really about advancing person-centred care,” said Dr. Nadiya Sunderji, President and CEO. “Artificial intelligence enables personalized care plans tailored to individual patients’ needs, taking into account their specific risk factors, treatment history, and response patterns.”

Artificial intelligence unlocks tremendous potential in developing Early Warning Score systems for mental health patients, providing healthcare professionals with tools to detect deteriorating conditions at an early stage. By leveraging AI’s capabilities in data integration, predictive analytics, natural language processing, and remote monitoring, these systems can enhance patient care, improve outcomes, and reduce the burden on mental health services. AI-driven solutions hold the key to revolutionizing mental health care for a brighter and healthier future.