Connect with us


“Doctors tell patients not to Google stuff – but we do it all the time”

Health Tech World reports on how Medwise is aiming to bring an end to desperate Google searches by clinicians and provide instant, expert answers when they need them.

Avatar photo



Medwise was founded by Keith Tsui, a doctor who originally trained and practised in Hong Kong. During his medical career, he began to encounter a problem that he soon realised was experienced by most other clinicians too.

“I was working in the hospitals in the middle of the night, treating patients and within minutes, I needed to make decisions,” Tsui recalled.

“Oftentimes I would struggle to find the information I needed, like what the best practice was and what the latest guidance or evidence and knowledge were for me to provide the best care to patients.”

Tsui, like many clinicians, would turn to Google for the information “almost out of desperation”.

He added: “doctors tell patients not to Google stuff but we do it all the time. It’s just not the best tool when you’re making life-changing decisions at the point of care.”

Tsui decided to pivot his career into healthcare entrepreneurship to develop a tool that could solve this issue.

In 2020, he launched Medwise, a question-answering platform that uses AI and natural language processing technology to provide clinicians with the most up-to-date information to assist them in their day-to-day decision-making.

With the backing of Innovate UK, Medwise launched a COVID-19 version of the platform last year. A pilot with the Eastern Academic Health Science Network was a success with an independent survey finding that the platform saved GPs up to 2.6 minutes per consultation. Now, the company is releasing a beta version of the full platform before launching it more widely in the next month.

Medwise uses different AI models for different components of the platform which can be broken down into three tasks. The first is classification. In other words, understanding the user’s question, the search and the intent as well as understanding the content.

After classifying the question, it triggers a retrieval task that identifies the segment that is most likely to answer the question. The platform then uses transformer-based neural networks to provide the final answer which may be a word or sentence within the segment.

The information comes only from established and trustworthy sources such as journal articles and NICE guidelines,Tsui said.

“One of the problems with Google is it mixes patient-facing information from, for example, Web MD and Healthline, with more professional information like guidelines,” he said.

“What we’re doing is making sure that the sources that we use are trustworthy guidance that clinicians know that they can rely on.”

As Medwise is still in its early stages of development, the platform will initially be wider than it is deep; however Tsui and his team intend to use a data-driven approach to make it possible for the system to answer complex questions as it becomes more widely used.

“We’re solving an urgent need for clinicians. They are already very time pressured with a lot of demand on the NHS and the primary care system. By providing them with a better information service, they can find information quicker, save time and provide better care.”

What is especially exciting for Tsui is the platform’s potential to collect and understand what information clinicians need. If Medwise is successful, it will be the first company to gather this kind of data in real-time.

“Information is often lost in a Google search or in a WhatsApp chat asking a colleague or flipping through a book,” Tsui said.

“We have the potential to become that unifying platform where we provide information to conditions but on top of that, collecting a real-world understanding of the information needs of clinicians.”

It is very exciting for us to think about how we can use those data to not just help clinicians make better decisions, but also inform what research should be done, what guidelines should be written that will be more suitable and used at the point of care. I think that could dramatically improve the quality of care delivered to patients.”

Medwise got its start on the Novartis BIOME HealthHub, which provides mentoring advice, partnership opportunities and sustainable investments to data-driven start-ups and SMEs in the healthcare sector.

Medwise was able to make connections with stakeholders across the NHS and telemedicine space and formed its partnership with Eastern Academic Health Science Network through the programme. It was this partnership that propelled the company towards the launch of its COVID-19 question answering tool.

“I think the health programme is fantastic. As a clinician running a start up company, it’s vastly different to my previous life as a medical doctor, so that support was really helpful to get us going.

“A lot of the connections with NHS stakeholders, both in primary care and secondary care were able to help us shape the thinking of the product and how we could deliver the greatest impact with what we had and how we should develop a solution, oftentimes with these organisations.

“Having the connections into the ecosystem partners was super helpful for us to better define the value proposition, move quicker and grow.”

Medwise recently closed its pre-seed fundraising round at £588,000.

Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Trending stories