Under the hood of healthcare innovation: Jared Hawkins and the digital phenotype

Jared Hawkins Boston Children's Hospital

What does it take to change healthcare for the better? In the second of a two-part series on digital health innovators at Boston Children’s Hospital, we profile Jared Hawkins, MMSc, PhD. Like Gajen Sunthara, MSc, featured in part one, Hawkins was named among MedTech Boston’s 40 Under 40 Healthcare Innovators for 2017.

Jared Hawkins, director of informatics at Boston Children’s Innovation and Digital Health Accelerator (IDHA), brings a formidable skill set to his work. With a PhD in Immunology from Tufts University School of Medicine and an MMSc in Biomedical Informatics from Harvard Medical School, his background combines biomedical research (immunology, virology, oncology, genomics) with data science, visualization, computational modeling and software development.

His current work spans an equally diverse range of topics, touching on population and public health, patient experience, decision support and pharmacogenomics. A faculty member in the Computational Health Informatics Program, Hawkins is wired into the digital health ecosystem. He serves as a scientific advisor and co-founder of Raiing Medical (home temperature and fertility tracking) and is the head of engineering and co-founder of Circulation (non-emergency medical transportation via Uber).

The common thread: mining health data to improve the quality of patient care, and developing a better understanding of individual and population health.

Constructing a ‘digital phenotype’ from previously inaccessible data

Hawkins’s deep passion is extracting actionable insights from previously invisible and inaccessible healthcare data. That includes nontraditional digital data streams.

Jared Hawkins of Boston Children's Hospital
Hawkins receiving his “40 under 40” award

“Through social media, search engines, online communities, wearable technologies and mobile devices, people are generating an enormous amount of health-related data that can provide deep insight to individual and population health,” he says.

Hawkins refers to these sources of information as the digital phenotype. The digital phenotype encompasses a huge volume of real-time data (consider the 313 million monthly active users on Twitter alone). Those data enable Hawkins to identify and analyze sub-populations over space and time to investigate problems like sleep disorders and infectious disease.

One current project, CrowdClinical, captures another aspect of the digital phenotype —patient experience — by monitoring real-time discussions about health care on social media. Using machine learning and sentiment-analysis techniques, CrowdClinical scrapes data from social media and patient portals to glean insights into what consumers value. The result: real-time patient experience data on a hospital-by-hospital basis — feedback that could ultimately help hospitals make improvements.

Protecting public health

Hawkins’s team is also pursuing digital detective work around foodborne illnesses. His team has developed efficient machine learning methods for analyzing social media data — like tweets from people reporting food poisoning. These methods can differentiate between relevant and irrelevant reports of foodborne illness and track geographical locations of implicated restaurants.

People are generating an enormous amount of health-related data that can provide deep insight to individual and population health. To date, the team has extracted, digested and processed billions of event-based data points. Their efforts could augment foodborne disease surveillance and improve food safety. Public health officials utilizing this tool can view and evaluate reports within their jurisdiction and engage users by responding to their tweets. In the first seven months of a pilot in St. Louis, the platform captured more than 440 tweets about food poisoning, of which 193 were of public health importance.

Most recently, Hawkins’s team showed that a Boston Children’s app called Thermia, coupled with a commercially available wearable thermometer, predicted seasonal influenza outbreaks in China one month faster than national surveillance programs.

“Jared’s ability to steer multiple projects is unlike any I’ve seen before,” says Gaurav Tuli, a data scientist at IDHA who Hawkins has mentored for several years. “It has made a dramatic improvement in the productivity of the whole team. Jared truly encourages creativity and innovation through his work.”