Data platforms and emerging technology are set to revolutionise life sciences, according to one of the world’s largest providers of clinical trials.
With a well established software-as-a-service platform, Medidata Solutions is now turning its attention to how it can better utilise the trove of data it is collecting.
Medidata pioneered clinical trials as a software service in the 90s and has grown to be one of the market leaders. In 2017 all of the top 15 selling pharmaceutical drugs were developed with Medidata technology, according to the company, which holds the majority of top US pharmaceutical companies as customers.
Today the company is confident its platform, users and data have reached a point where nascent technology like machine learning can be effective.
“With all of these people and all of this data… going through our platform we realised that we should be able to do something,” said Medidata Solutions co-founder and president, Glen de Vries. “We should be able to do something better than just looking at data in a traditional way.”
One early example is the ability to identify biomarkers of patients by collating data on immune diseases, including from non-traditional sources like a patient genes. By analysing a large amount of diverse data the platform was able to “cluster” patients who had better responded to treatment of the rare disease into subsets. Doing so identified a subset of patients where the response rate to medication more than doubled.
“I see that as actually delivering on this idea of precision medicine,” de Vries said.
“We came up with a way to use all this data to do a better job of giving the right patient the right treatment at the right time.”
The key to success with the application of emerging technology like AI and machine learning to data sets is data quality, to go along with quantity, de Vries said.
Medidata has amassed data from over 16,000 studies run on its platforms – representing over half of the current research being funded by the life science industry, according to the company. 4,000 clinical research coordinators and investigators are certified on the Medidata platform in Australia, supporting both global and local trials.
But the data quality — data comes from rigorously regulated scientific trials — is just as important as quantity, a point others often overlook, de Vries said.
“If you look at big data, if you look at artificial intelligence — call it whatever you want or what you want to connect to it — it’s computer science and it’s statistics.”
“Both computer science and statistics suffer from garbage in, garbage out.”
Collecting, “cleaning” and storing the data in an accessible way extends its benefits to others as well, according to de Vries. He explained platforms users’ ability to access historical data of previous trials was cutting down redundant trials and unnecessary control cases.
Importantly, he says, it’s also breaking down the “siloed thinking” of pharmaceutical companies, which in the past have found it challenging to collaborate with one another.
De Vries said Medidata owes much of its current position to its early establishment. Founded in 1999 as an extension of Oceantek, also co-founded by De Vries five years earlier, Medidata was developed as a solution to the lag between getting research findings and the treatment of patients.
“There really wasn’t any platform — I don’t think we were even using the term platform in the world then — to do that,” de Vries explained.
The founders decided to build it themselves, initially with web and mobile applications as cloud wasn’t available at the time. When cloud did become available, it offered a viable solution for Medidata’s technologically conservative clients, mostly pharmaceutical organisations.
“The idea of them running a web-based application out of their own data centre was something they weren’t interested in so we actually wound up, because of our clients, in what would become known as a cloud business model.”