Propeller Health is working with Boehringer Ingelheim.

Pharma's investment in the digital health space is getting more and more real and substantive as companies move beyond the pilot stage and start to launch real digital health interventions, from Teva buying up smart inhaler company Gecko to Otsuka gearing up to launch a commercial drug integrated with a Proteus sensor.

But digital innovation in big pharma is still in many ways an uphill battle. At Health 2.0 this week in Santa Clara, speakers from four big pharma companies spoke about why it's so hard to innovate and how digital is changing the way they think about clinical trials and outcomes.

"There’s an enormous number of paradoxes and contradictions in pharma," moderator Monique Levy, vice president at Decision Resources Group, said. "You’ll see great vision, but then budgets get cut. You’ll see some kind of fascination with what’s happening with Google and moonshot thinking, but then huge risk aversion to putting any kind of substantial budget or resources into innovating or understanding digital."

Eddie Chen, global head of search and evaluation for Sanofi, believes there's a fundamental culture clash to be overcome in bringing digital innovation to pharma.

"We as an industry have tried to refine a development process over time, which has been fairly effective for the industry," he said. "When you try to apply that, it doesn’t square quite well with the rapid, iterative development world we see on the technology side. That’s where we have the fundamental gap, which is one of cultures and in terms of what can be done in what period of time. We’ve yet to find that delicate balance that can optimize that."

But Larry Brooks, director of business innovation at Boehringer Ingelheim, believes that gap also houses the opportunity for digital health, which allows pharma companies to think about their products and their business in a new way.

"We don’t touch the product through the life cycle of its patent, and every other industry does continually iterate," he said. "Think about cars. Every single year, they come out with a new car and iterate it. This is where digital health has a real opportunity. Even if you come out with a highly novel therapeutic, it will have a fast follower, but if you can innovate the entire solution and address the endpoints, you can bring something that’s truly unique but also meets a need."

Sandra Nichols, senior director of the digital innovation group at AstraZeneca, said that an organization's commitment to innovation needs to be backed up by funding and longterm thinking.

"There’s this tension between delivering on this year and setting ourselves up for the future," she said. "And with the squeeze that we’re under, this year is so important that people have trouble investing for something that’s not being delivered this year. Not only all the way up the chain do we need commitment to innovation, we also need to increase funding for it for more than one year, much as we do for drug development."

Everyone in healthcare is more focused on outcomes than ever before, panelists said, but the outcomes they're focused on are also changing. Evaluating digital interventions isn't just about measuring clinical efficacy, but also about measuring engagement, adherence, and new measures of quality of life.

"I think the alignment around outcomes is driving more collaboration than historically we’ve seen," Nichols said. We’re entering this era where if you don’t have an outcome, you’re not gonna have an income. So I think it’s pulling us together in a way that’s unique. We are beginning to look at how can we impact an outcome in ways we didn’t before. ... What we’re learning is the sooner you can get out there and do a beta test, the better learnings you’re actually going to have. Let people experience [your intervention] and see what the outcome is."

And digital health tools, with their unprecedented ability for day-to-day real world monitoring, can lend themselves to the development of new, meaningful outcomes measures -- not just things like wearables and sensors, but also patient communities like PatientsLikeMe.

"How we contextualize what’s happening with the patient is very different [than it used to be]," said Jessica Burton, and outcomes research scientist at Genentech. "You have these new endpoints, where the primary endpoint is progression-free survival. Well what does that mean to a patient? What does that actually provide in terms of benefit? You can use these new ways to look at a patient experience and really show, without progression, this is what happens to a patient. You can see impact on a daily basis,  and with that better data  you can have that audit trail you need when you’re in pharma."

On the other hand, as patient's attitudes and capabilities change, it becomes harder to support traditional clinical trials. If two study participants are also talking about their experimental drugs on PatientsLikeMe, that could spell trouble for the trial.

"Patients want information, they tell us all the time, and if you don’t give it to them they’ll find it," Burton said. "But the way in which they’re seeking information leads to complications of ‘are you unblinding?’ I mean if they’re finding out that they have side effects and this group has this and that group has that, how do you account for that in your analysis? We have to grow to accommodate these new types of data."

All of this is prompting the beginnings of a sea change in how pharma thinks about clinical trials, but the challenge will be in getting the FDA to accept new study designs and new outcomes measures. In the meantime, pharma companies that can manage the innovation hurdles continue to explore new ways to learn from patients, even as they keep running traditional trials.

"To the extent that our trials going forward can be a more accurate reflection of how people live day to day, real world evidence is a great strength, with technology as the key enabler," Sanofi's Chen said. "[When it comes to] the manner in which pharma trials are traditionally executed, more and more there are questions about trying to force-fit that, especially around solutions that involve digital."