Countless entrepreneurs around the world have launched digital health companies in the last decade and more than 100,000 mobile apps are currently on the market—all with the goal of making us healthier. But how many of these companies have concrete evidence to support their health claims?

Not many, says Sean Duffy, co-founder and CEO of Omada Health, a digital health company that develops programs for people at risk for chronic disease. “It’s still rare,” he adds.

In fact, there’s such a lack of clinical and economic evidence on the effectiveness of digital health products and services that a new startup was created just to focus on the problem. Evidation Health works to parse the marketing claims of digital health companies, which are rarely proven using scientific evidence, says CEO Deborah Kilpatrick.

“I wouldn’t couch it in the context of, ‘these companies are making claims they can’t back up,’” she says. “They just haven’t [proven them]. Whether or not they can remains to be seen.”

Kilpatrick blames little time, knowledge and money spent on evidence in the digital health world for the lack of proof. “In all fairness,” she says, “digital health as a sector is still fairly young.”

Many digital health companies are just now reaching maturity—they’re angel funded or in Series A or B rounds—and the sector is still figuring out what it means to prove effectiveness, Kilpatrick says. The entrepreneurs building these companies frequently come from the consumer sector, rather than healthcare, she says, so while they understand the importance of evidence, they don’t know how to generate it.

Clinical trials—the standard in proving the effectiveness of health interventions—can be prohibitively expensive. And these tests usually require a partner with ample resources and access to patients, such as a hospital. Healthcare systems are often reluctant to partner with small companies that don’t have evidence, and these companies can’t get the evidence without a partner. Another hitch: hospitals are accustomed to being paid to host clinical trials, but startups want hospitals to pay them to pilot their solutions.

Market potential in the consumer tech world is based on customer adoption, Kilpatrick says. But in healthcare, often slow-moving hospital systems, insurance payers, and pharmaceutical companies drive adoption. “It’s a much more complex and larger process of adoption,” she says.

Several factors play a role in a hospital’s adoption of digital health technology, says Dr. Richard Milani, chief clinical transformation officer at Ochsner Health System. Healthcare needs disruption, he says, but it must be done carefully.

“You want to be comfortable that the technology you’re utilizing is reliable and accurate,” Milani says. “That’s a genuine concern. If you have some scientific data or peer-reviewed data to back it up, that certainly makes it a little easier.”

Evidation Health, which is backed by GE Ventures and Stanford Health Care, brings together digital health companies and healthcare systems. Evidation designs and executes outcome studies based on medical, behavioral and contextual data. The studies range from 72-hour investigation into Fitbit published by Buzzfeed to a long-term look into whether a medication adherence app can improve a hospital’s clinical outcomes.

“This allows us to redefine for clients and partners the way they think about their business on the basis of these outcomes,” Kilpatrick says.

Duffy’s Omada Health is one notable example of a digital health company that has published its outcomes data in respected peer-reviewed journals. Omada offers an employer-sponsored 16-week online digital health program to promote behavioral change in people at risk for chronic disease. It was founded in 2011 in response to the growing lack of evidence in the digital health sector, Duffy says. “We had the intent to publish from day one,” he says. Omada’s first peer-reviewed study wasn’t released until the company had a year’s worth of data. Now, Duffy claims Omada is the only company in its space to publish two-year data from a digital program.

There’s no doubt, Duffy says, that producing evidence-based studies is logistically complex. The cost of an entry-level study is likely to reach into the hundreds of thousands, though more robust research can tip into the millions. But, he says, it’s the right thing to do. After all, we wouldn’t subject patients to a new pharmaceutical or surgery without the evidence to back it up.

“I don’t know why digital should be the exception,” Duffy says. “That’s the direction I think digital health needs to take to gain increasing trust.”

Based on the number of queries he gets from entrepreneurs who want to generate evidence to back up their products and services, Duffy says digital health companies are beginning to realize that the best way to commercialize their tech is to prove it works. “In the next four or five years,” he says, “it will become less rare.”

Sign up for Data Sheet, Fortune’s daily newsletter about the business of technology.

Clarification: An earlier version of this article claimed that a health study was made by Buzzfeed. This was not the case, the study was completed independently by Evidation Health for Buzzfeed.