Transforming healthcare training: How Fraser Health is using AI

Modernizing a health authority’s clinical system is a massive undertaking. Not only are there millions of patient files, but tens of thousands of staff and medical staff to train in order to become proficient on the new system. The efforts are worth it, however, as the systems provide streamlined workflows and improved clinical decision support, resulting in better patient care.

In 2022, Fraser Health adopted a new electronic health record system to serve over two million people in 20 communities in British Columbia. With tens of thousands of clinicians learning the new system, they also required a quick way to get questions answered and troubleshoot issues. Otherwise, staff might need to manually sort through more than 350 reference and resource documents or flood Fraser Health’s IT help desk with calls.

The Digital Patient and Provider Experience Team knew artificial intelligence (AI) could help solve the problem.

They turned to Amazon Web Services’ (AWS) robust set of AI and machine learning (ML) solutions to create a powerful virtual AI assistant that could sift through vast amounts of information and provide instant answers. Within four months, the virtual AI assistant was running, adhering to Fraser Health’s rigorous privacy and security requirements.

The virtual assistant uses several AWS AI and ML services to ingest, analyze and respond to staff members’ queries by presenting them with the information they needed to complete their tasks.
For example, Amazon Lex creates a conversational interface where users can write their questions, Amazon SageMaker builds, trains, and hosts machine learning models, and Amazon Kendra enables intelligent searches across different content repositories.

What once took an average of 15 minutes to find now takes mere seconds, significantly reducing troubleshooting and boosting time spent on patient care across the network.

“Since day one we’ve heard from users how much time they’re saving by using this system, and how much it’s improved their workflow,” said Alim Sunderji, executive director, Digital Technology Services at Fraser Health.

Fraser Health staff and medical staff have welcomed the virtual assistant. As users are onboarded to the new clinical system, at least 300 so far, they are praising its efficiency and convenience, highlighting its ability to expedite problem-solving processes effectively and enabling them to stay focused on patient care. Overall, the impact has been overwhelmingly positive.

“It was nothing short of remarkable how quickly we were able to get the Virtual AI Assistant implemented,” Sunderji continued.

Today, thousands of healthcare and life science customers globally turn to AWS, with a growing number adopting AI and generative AI solutions, to transform back-office functions, such as billing and chart coding, to frontline patient care, such as helping doctors take notes during appointments.

“Organizations are turning to AWS to develop chatbots and AI solutions to handle challenges in staff engagement, development and upskilling, to assist with large scale transformations or day-to-day operational support,” said Elizabeth Keller, Healthcare Business Development Lead, AWS Canada. “Fraser Health’s virtual assistant is a great example of how AWS’ AL and ML services can positively impact workflow and patient experiences.”

Spurred by this success, Fraser Health is already looking towards the future, developing a second-generation virtual assistant that will incorporate generative artificial intelligence and a Large Language Model (LLM) to continuously improve on the tool, like answering strategic queries, better anticipating follow-up questions, and engaging users in a more conversational manner to further enhance its utility and user experience.

“We find that 99.9 percent of users are engaging with the Virtual AI in English by typing, but what’s amazing about this system is that it’s equipped with Amazon Translate to provide high-quality translations in a multitude of languages if required, and can also respond to user voice prompting,” said Mike Chiu, director, Technology and Architecture at Fraser Health. “As we expand out across the health authority, we anticipate diverse user needs, so we’ve built the system to anticipate a multitude of user needs and experience expectations.”

Fraser Health’s adoption of chatbot technology marks a significant milestone in the evolution of healthcare training and efficiency. By harnessing AI’s power, they have not only streamlined processes but also set a new standard for innovation in the industry. As the healthcare landscape continues to evolve, embracing such transformative technologies will be essential for staying ahead of the curve and delivering optimal patient care.