Electronic Health Reporter

Dean Stephens is the CEO of Healthline, a media group and a health information technology company. Here Stephens discusses healthcare analytics and how it’s important to providers and patients; the ever-increasing importance of harvesting useable and life-changing information from unstructured big data;  analytics in population health; the importance of ACOs and the future of Healthline.

Tell me about your background and your role at Healthline.
I grew up in a small, blue-collar town in New England and was fortunate enough to attend an Ivy League college, which was a rare thing for this town. After college and graduate school, I got lucky to land a policy analyst position for the Washington State governor, but in no time, got drafted into management consulting at Deloitte. Much of my consulting time was spent in the healthcare industry learning first-hand how “upside down” the industry was. Thus, I joined other entrepreneurs to re-imagine this muddled industry and joined Healthline as CEO in 2001, not knowing then that I would end up building two companies simultaneously.

What does Healthline do and how has the company evolved?

Healthline’s mission is to make the people of the world healthier through the power of information. And we do this through two business units – our media group, which consists of our consumer health website Healthline.com, and our health information technology group, which includes a range of search and data analytics solutions built on our market-leading medical taxonomy. We are currently working with some of healthcare’s largest brands, including AARP, Aetna, Pfizer, Sanofi, UnitedHealth Group, Microsoft, IBM, GE and Elsevier.

Describe your personal view of analytics and what that means to the rest of us. Why is this important?

Healthcare is the most information-intensive industry on the planet. The number of diseases recognized today and the permutations on the treatment matches to individuals have exploded over the past 20 years. It’s impossible for an individual physician or a large, sophisticated provider or payer institution to deliver effective treatment across all patients without analyzing vast amounts of complicated data. We limped along in the traditional fee-for-service realm. Now as the healthcare market shifts to value-based reimbursement, the value of information and analysis rises dramatically as providers shift from being rewarded for sick care to well care.

How is healthcare data really being used, or should it be used? And to what sort of specific changes should it drive for the patient or at the point of care?

Today, we are perhaps 2 percent of the way to using healthcare data effectively at the population health and patient clinical decision support levels. I hope to see physicians using patient profile information (including genomics) and mapping that to clinical guidelines and evidence-based treatment options in order to assign the right level of resources to improve that patient’s health.

Tell me more about Healthline’s HealthData Engine; how does it work and how is it different from other data analytics solutions in the industry? Who are your direct competitors?

According to IDC Health Insights, more than 80 percent of today’s health data resides in unstructured formats, such as free-text physician notes, patient histories and hospital admission notes – we call this the gold in EHRs. This percentage is expected to grow, not shrink. Many data analytics solutions on the market today are not capable of making sense of all this unstructured data. Our HealthData Engine platform uses a unique combination of Healthline’s medical taxonomy, advanced natural language processing and data analytics to aggregate and normalize the valuable unstructured information hidden in patient records and other disparate sources. It then combines this information with structured data, such as problem lists and medications, to create a more comprehensive view of the patient’s health. By combing through and making sense of the unstructured parts of a patient record, we can effectively identify patients with un-coded predictive clinical factors, including many psycho-social, socioeconomic and environmental factors. As a result, the engine enables clinicians to make better, more informed decisions and improve patient outcomes.

What kinds of healthcare organizations use your technology? How does it benefit them and their constituents?

We primarily target provider and payer organizations. As the healthcare industry continues to move from fee-for-service care to value-based care, provider and payer organizations are increasingly taking on more risk, which means the value of information insights increases. Our data analytics technologies give them more accurate insights into their patient populations and enable them to better manage risk and improve outcomes. In my view, every one of us (as consumers or as patients) should beg our providers to assign us to a risk pool so that we can determine what actions we want or should take. For example, I have atherosclerosis in my family and cancer. I want to know what risk pool I am in. This data is or could be available.

Can you share some examples of how your customers are leveraging the technology?

We are currently working with Village Family Practice, a group practice located in Houston, TX, to pilot our latest application built on the HealthData Engine. This application focuses on coding accuracy and helps provider and payer organizations effectively uncover missed or inaccurate codes from unstructured patient data, proactively close coding gaps prior to claims submission, and better manage risk.

Other organizations like Aetna, UnitedHealth Group, IBM Watson, Microsoft, GE and Elsevier use our medical taxonomy and advanced search technologies to help provide their end users with more accurate, contextual search results. When a user searches for “heart attack,” for example, our technology will intelligently deliver not only educational content about heart attacks, but also tips on lowering cholesterol, a directory of local cardiologists in the area, and coaching programs tailored to that individual’s demographic.

Much has been said and written about population health management. Tell us why big data analytics in particular are vital to successfully managing population health and risk. Where does the concept continue to lack?

Population health management can only exist when there is financial incentive to manage it. Most providers, including doctors, aren’t currently under this incentive to a large degree. Population health management requires an ability to assign every patient to a risk pool or cohort. To be assigned, a patient’s family history, medical history and even socio-environmental factors need to be identified. Most of this information is stored in unstructured text, metadata and semi-structured formats. This is where big data comes in – to take this polyglot of information, convert it into meaningful data, and analyze it against standards, protocols and guidelines. Big data’s biggest gap today is the lack of patient-derived information about his or her history and behaviors. Programs such as Robert Wood Johnson’s funded Open Notes initiative encourages patients and doctors to put this valuable information into the patient’s digital record.

Are providers and practices really using analytics or is this more of a health system conversation? Who’s really “getting it”?

Leading healthcare institutions and physicians are practicing big data analytics today. It is still in more of an experimental mode than common practice, however. Kaiser, Geisinger, Intermountain, and the famous medical and cancer clinics and institutions have been “getting it” for some time. Certain large payers are investing in this, as well, and let’s not overlook big pharma.

What trend do you think will have the biggest impact on the healthcare industry this year?

Accountable care organizations (ACOs) that are now starting to demonstrate a real impact on care quality will have the biggest impact on the healthcare industry this year. Cost is still an issue since these organizations don’t have that great a grasp on capturing all cost inputs. When the ACOs start to put up some reasonable numbers, the industry will start to truly believe that value-based reimbursement is the future.

What’s next for Healthline?

Healthline will continue to focus on further building out our HealthData Engine and developing additional applications on top of that platform. Stay tuned!