Fireside Chat: How Digital Technologies Are Transforming Clinical Research

By Benjamin Ross 

March 6, 2020 | ORLANDO—“We’re really heading toward a pathway that’s unsustainable for clinical development,” said Jacob LaPorte, co-founder, VP, and Global Head of BIOME—Novartis’ Digital Innovation Lab. “You look at ROI for R&D and every year it drops, and I think there some real macro-picture issues that are challenging. Solving this problem around clinical development is one of the biggest grand challenges for R&D now.” 

LaPorte was joined by Emmanuel Fombu, VP of Locust Walk, a global life science transaction firm, for a fireside chat during the Summit for Clinical Ops Executives (SCOPE) last month in Orlando. The two discussed how digital technologies can fundamentally transform clinical research.  

“I think the way clinical trials are done today is archaic,” Fombu said. “I think we should have a much bigger mindset when we look at patient populations. The studies don’t stop when a Phase I study stops or a Phase II study stops; it’s a continuous process. You have to continually get information from patients so you can tweak your products and understand your capabilities.” 

Voice-enabled technologies are a new, unique way of capturing this info, says Fombu. Amazon’s Alexa and other voice service devices rely on algorithms that can be used to predict Alzheimer’s, dementia, depression, etc., based on the user’s voice. 

“You have all these devices present in the real-world setting, and you collect datasets, not just to collect data to publish, but to collect data to make the right strategy to improve the life of the patient,” Fombu said. 

“Voice is an interesting one,” LaPorte agreed. “My two-year-old daughter went to visit her cousin, who has an Alexa. And over the course of a couple days, she learned how to say, ‘Alexa, Mickey Mouse Club.’ She couldn’t quite put it all together, but she knew that there was a chance she was going to be able to watch Mickey Mouse Club by having Alexa turn it on. This technology is really starting to embed in everyone’s everyday life, and we can start to collect all this nuanced data as people go about their lives.” 

Fombu agreed, adding that these technologies change the nature of clinical data. 

“What is healthcare data?” Fombu mused. “If we ask the question in general, we start having discussions around EHRs [electronic health records] and blood work. But if you eat a meal today, take a picture of it, and post it on Instagram, is that healthcare data? It is healthcare data. I know exactly what you ate, which is much more useful than me giving you a piece of paper for you to check boxes to tell me what you ate.” 

This new definition of healthcare data means there are new players in the healthcare space, Fombu says. “There’s this new concept of emergent health data where companies like Facebook can predict suicide, depression, alcohol abuse, etc., based on what you write. Even your status tells a lot about your healthcare.” 

These players have a much broader understanding of the user in real time than the typical trial site, says Fombu. Companies like Google and Apple have been collecting health data from their users in real-time, while data collected from patients during their monthly or bi-monthly visits is still just a snapshot of the patient’s health. “In today’s world, the way we do clinical research is like taking a picture; the future is taking a video.” 

That’s not to say these companies are in any way in the business of healthcare, Fombu cautions. “If you say Apple or Amazon did something in healthcare, everyone gets very excited. Forget the fact that they’re in the business of marketing; they’re not in the business of health. Their business is to collect data to sell you products. It’s the interest of those of us in this room to improve healthcare, and I think we need to recognize that.” 

Fombu say the current “static” approach from the healthcare industry goes against what technologies such as artificial intelligence (AI) and machine learning are designed for. Fombu believes companies in other markets have mastered these tools in order to engage with their customers. 

“If you look at Netflix, they started out by sending movies to us in the mail and then we’d send them back; over time, Netflix understood what we liked and would start making recommendations based on our preferences,” Fombu said. “AI is a continuous process; it’s not a static one. You get retrospective data and you look at prospective data, and you gather these data on individuals all the time. We don’t do this in healthcare. We have things labeled by phases, and then we go to the market. So, it’s a period of time now where we need to change our mindset and think about trials as a continuous learning process that doesn’t expire based on the length of a phase within the clinical trial.” 

LaPorte says this approach to healthcare can also impact how patients are recruited for trials. 

“The conventional way of doing a paper-based survey on the quality of life led biostatisticians down a road of saying you need around 5,600 patients [to run a trial],” said LaPorte. “You can now do it with 140 using digital devices, and you can start to do a lot of this outside in the real world.” 

LaPorte says this new path begs the question: “Are we going to start to see a paradigm going forward where people design studies pre-market that are much more focused on getting a few endpoints, and then push more of the expensive and timely research out into the real world?” 

“That’s exactly what’s going to happen,” Fombu said. “You have companies that have limited budgets, but they still need to be able to answer particular questions. Being able to answer that question with a fewer number of patients using these digital resources allows you to add value.”