To tackle challenges of privacy and information overload, healthcare of the future will use patient data to optimize care.
In October, Harvard University’s HealthMap service identified early signs of an Ebola outbreak in Africa, nine days before the World Health Organization (WHO) announced it.
Using special software to scour thousands of social networks, local news sites, blogs and other social media, HealthMap can detect and track disease outbreaks around the world.
While the WHO announcement actually preceded HealthMap’s findings because the software could not communicate in French to the Guinean government, its overall success shows the huge potential of big data in helping to solve the world’s health problems.
That said, not everyone is embracing big data with open arms.
According to Chris Gough, lead solutions architect at Intel Health & Life Sciences, one of the concerns that comes up frequently is that “big data” is more of a marketing term, basically another expression for business intelligence and analytics that has been around for decades.
However, Gough says the advancements in computing technology are allowing companies to do things that haven’t been practical or even imagined in the past.
According to Gough, about 70 percent of relevant clinical information in electronic health records is in free-form text fields, meaning doctor or nurses notes.
“This unstructured information is typically ignored by traditional reporting tools,” he explains.
“Today, capabilities like natural language processing (NLP) can extract clinically relevant information from these fields.” In other words, a doctor’s scrawl or dictated notes can be transcribed and put into serviceable data.
It was this kind of technology that allowed HealthMap to predict an Ebola outbreak.
But Gough says the culture of the medical industry make clinicians hesitant to adapt from the traditional approach to fixing patients’ health problems.
“Clinicians typically rely on training and experience when making decisions related to the diagnosis and treatment of their patients,” Gough said.
“However, there is really too much information for healthcare professionals to carry around in their heads.”
Gough is confident that if the healthcare industry can use data and analytics to understand patients better, it will be easier to determine the treatment protocol that has the best prognosis for success.
“We believe that the most successful healthcare organizations will make analytics and decision-support tools an accepted part of the patient care workflow,” he said.
“That will be transformational in terms of improving outcomes and controlling costs.” Rock Health, a so-called “seed accelerator” that invests in digital health startups and provides funding and other support to entrepreneurs and engineers, sees great opportunity in the collection of data. But, in an industry like healthcare where long trials and stringent structures rule, embracing the possibilities of mass data sharing has been sluggish.
“The cloud is a tremendous opportunity in healthcare, but the industry has been slow to adopt it,” says CEO Halle Tecco. “It’s partially because of regulation, but it is something that I think all doctors and hospital administrators are aware that it’s a great opportunity.”
One company supporting clinicians in treatment planning is CancerIQ. The digital health company is using a data-driven approach to help oncologists and their patients.
Today, cancer researchers can easily identify patients who may benefit from gene therapy and other treatments under investigation. Patients receive the benefit of having a more personalized treatment plan and more efficient communication with their oncologists.
Another company, NextBio, is a “search engine for the world’s genomic big data.”
Its ever-increasing database includes information on genome mapping, pharmaceuticals, patient data and diseases, and is designed to help researchers develop new treatments less expensively and more efficiently.
Big data is also directly helping patients.
WellDoc is an FDA-approved platform for diabetics that collects data from your smartphone about your blood sugar levels, recent meals and exercise activity to coach you on the proper amount of insulin to take and the most optimal time to administer it.
The platform can also use this data to predict potential hypoglycemic (low blood sugar) reactions and provide information on how to avoid them.
In the field of dermatology, there is digital health startup called Quantified Skin. Its technology analyzes your diet, activities, genetic profile and even a daily selfie to let you know what’s harming and helping your skin.
Intel is also playing a role. According to Gough, Intel’s software engineers are optimizing the most popular life sciences codes associated with genomics analytics to run as quickly and efficiently as possible on underlying hardware components, which he calls “ingredients.”
“We work closely with a large ecosystem of partners to build our ‘ingredients’ into deployable solutions for health and life sciences organizations,” Gough said.
Recent collaborations include working with the OHSU Knight Cancer Institute on cancer research and treatment, and with the Michael J. Fox Foundation on wearable devices and advanced analytics for Parkinson’s research.
Gough is optimistic about the future of big data, but notes the importance of top-down data management.
“If data is spread across the organization and locked away in application-specific or department-specific silos, it will be difficult to succeed with any analytics approach big data or otherwise,” he said.
“Data must be treated as a key asset from the executive level on down in order to truly derive value from it.”
If healthcare organizations can learn to properly manage data, and clinicians can adapt and wholeheartedly embrace big data in their practices, Gough envisions a future where it can indeed help solve the world’s biggest health problems, starting with individual patients.
“Think of a quarterly visit to the doctor’s office to monitor a chronic condition, or monitoring a patient that might be recovering from a recent surgical procedure. If these patients can be evaluated continuously via wearables and other patient-generated data (surveys, vital signs measurement devices, etc.), not only can progress be tracked more effectively, but scarce healthcare resources can be applied where they are needed most.”