By 2020, the healthcare system is projected to store 50 times as much information, 25,000 petabytes, meaning machine intelligence will be essential to complement human intelligence to make sense of it all.
Currently, startups are catering mostly to providers, rather than directly aiding patients with care and health decisions. According to the survey conducted by the accelerator, solely enterprise-focused startups outnumber patient-focused startups at nearly a 10-to-1 ratio.
In the future, though, the report speculates that such patient-focused startups can help patients make better wellness and health decisions, as well as aiding in triage.
The variety of data should also increase. At the moment, startups are focused on using clinical data (71%), as well as claims (42%) and patient-generated data (42%). One expert quoted in the report, Dr. Sam Ho, the chief medical officer of UnitedHealthcare, thinks that transactional, clinical and lifestyle information will become more readily available and important for startups in that space.
Personalized, predictive analytics face some particular challenges when applied to healthcare, however. Analytics used in, for example, a Google search can be quickly assessed: the search is conducted, completed and used within the space of seconds. By comparison, analytics used in healthcare may take years to assess.
More certainty in the regulatory space would help the industry, the report argues. This spring, the Food and Drug Administration, the Office the National Coordinator for Health Information Technology and the Federal Communications Commission unveiled a draft framework to regulate the health IT industry. As a part of that framework, the agencies promised more detail into the regulation of clinical-decision-support software—the category into which much of personalized, predictive analytics would fall.
“Most” clinical decision support software, per the framework, would fall under ONC's purview, with high-risk software being overseen by FDA. But further details distinguishing what software would be regulated by which agency—and how each agency would regulate the software under its jurisdiction—have not emerged, and industry observers have been frustrated with the lack of regulatory clarity on the technology.
Follow Darius Tahir on Twitter: @dariustahir