HealthMap is a previous project from the team now studying sleep health.

Boston Children's Hospital is teaming up with Merck to turn Twitter and Facebook into a new source for sleep health data -- data about how many people in a population suffer from insomnia and what they have in common with each other.

"This is sort of our foray to see how we can mine social media to get an understanding of the patient voice and potentially uncover some patterns around sleep deprivations that aren't sort of apparent from more traditional data sources," John Brownstein, associate professor at Boston Children's Hospital and project leader, told MobiHealthNews. "It represents a whole new class of observational data in health research, being able to characterize the burden across this massive sample size in a way that's been never described before."

According to the CDC, more than 25 percent of people in the United States report occasionally not getting enough sleep and almost 10 percent have chronic insomnia.

Brownstein and his team will take publicly available data from Twitter and Facebook, including message content, frequency, user analytics, and demographic information to determine what social media users are likely suffering from sleep deprivation. They'll then compare that group to the larger social media population and look for risk factors for insomnia. Within the group identified as having insomnia, a smaller group will be given the standard clinical evaluation for insomnia in order to validate the social media methodology.

Merck's role in the study is mainly to fund the research. The pharmaceutical company is interested in the outcome, which Brownstein said will likely be both published in a journal and made available via an online tool.

"We are very interested in pushing the boundaries of the science of social media and to see this as an opportunity to better understand the patient voice, in this case, how people share information about sleep problems and their day-to-day impact on quality of life," Dr. Sachin Jain, chief medical information and innovation officer at Merck, said in a statement. "This data source could offer a powerful tool to monitor the sleep health of a city, state or country, and it may offer additional tools in the prevention or treatment of insomnia."

In addition to working at Boston Children's, Brownstein is the founder of HealthMap, another big data project. HealthMap aggregates freely available online data -- from sources like Google News and the World Health Organization -- to produce a near-realtime map of disease outbreaks. Brownstein's research group has also done some crowdsourcing work on health policy, but the insomnia project is their first effort to use social media data to understand a chronic condition.

He says insomnia was chosen because it's pervasive and hard to study in traditional clinical settings, but it's also not stigmatized like some other conditions, so people post freely about their sleep habits on social media. One possible concern is cause and effect -- whether the sample might be biased by people whose sleep deprivation is caused by social media obsession. But the learning algorithms the team is using to analyze social media content should eventually learn to make a distinction.

"Whether it's industry or medical research, the focus is on clinical data streams generally," he said. "But I think there's this black box that is the patient discussion, what it is, and what it can tell us."