The big data disruption: How big data analytics can impact health care

October 17, 2018

Originally published by Crain's Detroit Business 

Banks, grocery stores and tech companies like Amazon and Apple gather vast amounts of data from consumers to personalize experiences, become more efficient and strategize new services, product offerings and partnerships. Yet retailers, financial institutions and IT firms comprise only a portion of the $28.65 billion global market for so-called “big data.”

In health care, big data is being used to predict epidemics, improve care, prevent unnecessary diseases and deaths, motivate patient fitness and lower costs. According to BIS Research, a global technology research firm, the growing use of wearable devices, at-home medical testing services, electronic health records, and other applications and innovations will cause big data in health care to surge to a $68.75 billion global market by 2025 from an estimated $14.25 billion in 2017.

For the health care industry, big data is comparatively new. Experts offer varying definitions that boil down to this: Big data is disparate pieces of information in and outside of healthcare organizations that reflect individual or group behavior and can be leveraged to provide the best care available now and in the future.

Dominick Pallone, executive director of Lansing-based Michigan Association of Health Plans, which represents 13 health providers, said big data is instrumental as the health industry moves from a pay-per-volume to a pay-per-value type of service. A smooth transition means getting the right data to the right people at the right time to improve outcomes and lower costs.

“We recognize this is where the future is going and are excited about where this will take us,” Pallone said. “We’re starting to coordinate data, ensuring our care managers are actively engaging with our physician partners and the physicians and clinicians have at their fingertips the information they need to provide appropriate levels of care to the people. Without unlocking big data, we won’t be able to make that transition smoothly.”

To be effective, Pallone and other experts contend that big data needs buy-in not just from health care systems and insurers but from government, social organizations, employers and consumers.

Silos breed the need

When thinking of silos, one might conjure images of farmland with barns and tall circular structures protecting vast amounts of grain from outside contaminants. In this respect, silos are good not just for farmers but also for the communities they serve.

In health care, silos occur when information in each clinical, specialty, pharmacy, service, insurance and governmental health department is separated. These silos benefit no one. Instead, they generate communication gaps and thwart positive outcomes.

Health care silos tend to be barriers to finding answers for critical questions like:

“What can we do to improve employee health?

“How can we reduce hospital readmissions?

“Where is the opioid problem highest in my community?

“Why haven’t doctors been able to treat my shortness of breath?”

Tim Pletcher, executive director of Michigan Health Information Network (MiHIN) Shared Services in East Lansing, explained how silos negatively affect health care:

“When a patient changes health plans, the new health plan does not know the information history the other health plan had. So, they often think the beneficiary needs a bunch of new preventative services that might have been done before. However, if the new health plan has no record of them being done, then the government might rate the health plan lower,” Pletcher said.

“Therefore, the health plan is obligated to notify the doctor that the patient needs the service. But if the doctor has been seeing the patient and knows the service was performed, then the notification from the health plan is just annoying versus helpful.”

Silos in the health care industry prevent physicians, pharmaceutical companies, manufacturers and payers from accessing and interpreting important data sets, instead encouraging each group to make decisions based upon a part of the information rather than the whole. As a result, their short-term fixes don’t improve the sustainability of operations or resolve root problems, reports HealthITAnalytics.com. The report adds that “for patients, this is really bad news as it results in delays in diagnosis and delays in access to treatments and appropriate care.”

Silos aren’t just prevalent at hospital systems and providers, Pletcher said, but they happen at employers and originate at the federal and state level, at groups such as the Michigan Department of Health and Human Services, Michigan Automated Prescription System (MAPS), Supplemental Nutrition Assistance Program (SNAP) and foster care programs.

“These different groups are not sharing data,” Pletcher said. “(Those silos) need to go away if we’re really going to harness the benefits of big data for dealing with problems, which are not all medical … but some are social determinants-related.”

Harnessing the power of big data

As assistant vice president of Translational Science and Clinical Research Innovation at Wayne State University in Detroit and associate chair of Clinical Research for WSU’s Department of Emergency Medicine, Dr. Phillip Levy and his team develop, design and conduct studies on the determinants of health and diseases in the Detroit area based on massive sets of data.

By gathering, sharing and analyzing data, Levy and his team can aggregate information to get a bigger picture of what is happening at the community level.

Among the team’s projects is gathering data from emergency departments, EMS runs, medical examiners and others about the opioid epidemic, where overdoses occur and the type of drugs used.

“How you really leverage big data, to me, to drive health outcomes is by sharing it,” Levy said. “You can see on the map where the overdoses are coming from. … It’s suburban. They’re coming into the city to do this. It’s obviously in different areas; but as you pull in these sources, you put it together into one common resource. All of the sudden now, you can get the big picture without just talking about little pockets of overdoses.”

Having such information allows communities to close the loop by installing Narcan boxes in the targeted areas or by directing police to find the dealers who are selling the drugs that are causing the overdoses.

That same data can help employers as well.

While employers don’t know which employees are on opioids, big data can tell them if they have a plant that is in a community with an opioid problem, explained Elaine Coffman, senior vice president of Health and Benefits at Troy-based Marsh & McLennan Agency. “Ten years ago, you couldn’t do that,” she said.

Levy’s team also is mapping blood pressure data in the city to identify areas with concentrations of uncontrolled hypertension.

“Hypertension is the single most important variable that contributes to health-disparity outcomes in the city of Detroit,” Levy said. “People are more likely to die of heart disease in Detroit than they are anywhere else in the state and far more than in other similar regions of the country.

“A lot of that chases upstream to diabetes, but if you’re able to figure out where the uncontrolled blood pressures are by mapping it out at census-tract levels, then you’re able to look at (medical) codes to figure out how common are things like dialysis that may go with uncontrolled blood pressure. Then, you know specifically where you want to target.”

Essentially, one of the key benefits of using big data in health care is the ability to take something from a concern to a crisis, said Jane Harper, director of Privacy and Security Risk Management at Henry Ford Health System in Detroit.

“We’re able to say, ‘This is not just an issue. This is not just an ad-hoc problem. This is a crisis,’” Harper said. “And this allows the government to put in controls, like MAPS and other controlling preventative options, to try to redirect us down the correct path.”

Coffman said employers can also use big data to determine if they have workers with uncontrolled hypertension.

“If you’ve got biometric screening data that says you’ve got all these people who have high blood pressure, and you run that data up against your claim data, you can actually see your zero-dollar people. Meaning, one of the biggest risks for an employer are the people who have the hypertension who are not doing anything about it,” Coffman said.

Organizations like hers share claims files with health coaches to help close the gap in care by tracking that employee and helping the employer and health plan determine if they have the right things in place to provide the care that’s needed.

Detroit-based Health Alliance Plan of Michigan is putting big data to use with a medical management therapy program, said Annette Marcath, HAP’s chief information officer and vice president of Information Technology.

“We’re using claims data, pharmacy data and a number of different factors and taking individuals that are on 20 or 25 different medications and working with the physicians and the information they have to get the patients to a point where they’re not taking that many (medications) and thereby improving their health and their outcome,” she said.

On the medical care management side, Marcath said HAP is doing more to ensure clinicians have the information they need to interact directly with patients and providers to improve results.

How big data might help me

Think of a digital e-reader. E-reader services suggest books consumers will like based on data such as how much they like or dislike the books they read, their age, gender and location.

With big data, researchers can find ways to provide targeted care to patients — like how an e-reader works. By merging data from various diagnostic tests with a patient’s medical history and social determinants, providers are working to create targeted treatment and preventive health strategies.

In February, Grand Rapids-based provider Priority Health launched Wellbeing Hub, which analyzes risk, claims and self-reporting data to develop and provide tools tailored to specific health and wellbeing needs, said Jeffrey Smith, director of Information Services at Priority Health.

The hub focuses on chronic disease management and targeted programs. For instance, members may be offered a weight management, smoking cessation or medication therapy management program based on their health status and personal interests. Members can opt to receive this personalized information via text, email or phone.

Similarly, Marcath said HAP uses data to encourage walking and other preventive health activities and provides bonus points to benefit members financially.

“It’s taking some of that … non-traditional data and using it to influence even the premiums that a member may have and their copays,” Marcath said.

Experts contend big data in health care should evolve to the point of having a smartphone-linked heart-rate or personal blood pressure monitor improve patient care.

“Often, our phones know more about ourselves than anybody we ask, even with the amount of information we’re gathering,” Smith said.

Nonetheless, WSU’s Levy said the dialogue between patient and provider is not often patient-driven. So, when patients come to the emergency room for concerns like shortness of breath, physicians routinely focus on possible heart failure when the cause might be anxiety related to social isolation or depression.

Added Marcath, “You have a patient coming in that has all this information on the phone that the physician is seeing for the first time. How are they going to act upon that at that point of care?”

She said a doctor can tell a patient her blood pressure or heart rate is high today, but the patient could say, “I can tell you from my phone that it’s been good.”

How big data can reduce overall costs

Big data also can help employers manage insurance cost trends and provide insight into benefits or resources they can provide employees, said Marsh & McLennan’s Coffman.

Keeping this in mind, employers are investing in a variety of tools that can increase employee use of telemedicine and employee assistance programs (EAP), which can help lower costs.

“What many employers are investing in is a communication hub where you’ve got one interface that will get people to the right place at the right time, because it’s so complicated,” Coffman said.

This helps employees get a handle on and plan for all aspects of an illness or a procedure, from the emotional component to the care and financial components. And by using real-time data analytics with technology hubs, employers can make sure workers get the care they need and, therefore, remain productive.

“With claim information, they can tell if a person may be struggling with low-back issues. So, these aren’t just intake hubs, they’re outbound hubs where someone is getting on the phone or sending a text message and saying, ‘Hey, did you know you’ve got this expert medical opinion service available? If you’re considering back surgery, you can have this sent to the best spinal care institute in the country for review,’” Coffman said.

Fundamentally, Pletcher said, “The biggest opportunity for big data is to help bridge the gap between clinical and administrative boundaries — what’s paid for, how it’s paid for, what’s wrong, what you need. Because those things matter to patients.”

Harper added that big data is an opportunity to invest in and improve not just the overall well-being of an employer’s greatest resource — their people — but also to invest in a reduction of their bottom line cost for health care so that we finally build a model that we can sustain and allow improvements in care for everyone.