News

Generative AI tool used to support nurse handovers at Unity Health Toronto

Unity Health Toronto is using generative AI to support nurse handovers, starting with a pilot in the orthopaedic surgery unit at St. Joseph’s Health Centre.

The End-of-Shift Note tool, a feature of the Epic EPR, employs generative AI to draft shift summary notes based on nursing documentation, medication data, and lab or imaging results, the healthcare network shares. Outputs are then checked over and approved by nursing staff prior to being added to the patient record.

Early indications suggest that it is helping to improve care team communication and care planning, Unity Health Toronto continues, with ongoing evaluation of impact on care teams and patient outcomes to determine the potential for wider adoption.

Nursing team leader Dan Jereza considers that use of the tool is saving time, and reducing the mental workload involved in pulling information together.

Earlier this year, Unity Health Toronto announced that it was set to offer its support for Sinai Health’s Epic EPR implementation, building on learnings from its own Epic EPR launch in 2024. Taking to LinkedIn, Unity Health Toronto said: “Guided by Unity Health’s award-winning Epic implementation, this partnership will give patients and care teams more seamless access to health information across organisations.”

Just last month, the Public Health Agency of Canada signed a memorandum of collaboration with Vital, a new collaborative health data platform based out of Unity Health Toronto, focusing on disease surveillance, building capacity with real-time data, and developing AI tools for outbreak forecasting and emergency management.

The healthcare network shares that initial areas of focus will be on using hospital data to detect clusters of symptoms for areas such as opioid toxicity and respiratory viruses, building capacity to identify and intervene on public health concerns, and developing AI tools for public health surveillance, emergency management, and the prediction of outbreaks.

Wider trend: Generative AI 

The European Commission has published its Special Eurobarometer on the Digital Decade 2026, revealing the results of a survey conducted between February and March 2026 on thoughts about AI, technology, safety, and digital. Around 40 percent of citizens report using generative AI tools “at least weekly”, with 69 percent of existing users saying their usage has increased over the last year. 55 percent of Europeans indicate digital health technologies as the technology expected to have the most positive impact over the next ten years, above green digital tech (50 percent) and faster internet connectivity (42 percent). 79 percent also note their support for the development of shared digital public services like digital identities and eHealth.

Bradford District Care NHS Foundation Trust has outlined its plans around generative AI, ambient voice technology, automation and digital recruitment as part of its latest board meeting. For the workforce, tech such as generative AI and AVT, as well as workflow automation and the optimisation of clinical systems, will support reduced admin burden and more time for delivering care, the trust explains. For service users, digital will support easier access and more convenient ways of interacting with services, including through a digital front door centred around patient preferences and personalised digital services that will hep access the right support at the right time.

Cognita Imaging, a subsidiary of Mosaic Clinical Technologies, has been awarded a $1.2 million contract by the U.S. Food and Drug Administration (FDA) to test a new approach for evaluating AI-generated radiology reports. The project will see the use of multiple large language models (LLMs) to “work as a jury” in assessing AI-generated reports across around a million patient exams, according to Mosaic. Cognita’s role will be the development and validation of a framework incorporating several LLMs to check human and AI-generated radiology reports, comparing the judgement of the models to produce a more consistent evaluation.