Healthcare Analytics, Population Health Management, Healthcare Big Data

Tools & Strategies News

Smart Big Data is Key to Population Health, Value-Based Care

Population health management and value-based care don't just require big data - they require intelligently targeted insights that support improvements in patient care.

By Jennifer Bresnick

- Big data is everywhere in the healthcare industry, flooding in by the gigabyte each time a patient seeks care from a provider.   From intake to discharge and every step in between, a patient’s data determines what will happen, who will be involved, how much it will cost, and what the patient should do next.

Michael Blackman discusses leveraging smart data for population health management and value-based care.

Before the rise of value-based care payments predicated on positive outcomes, providers were able to open up a patient’s health record when she stepped through the door, add their notes to the file as they treated a specific concern, and stow it in the filing cabinet again without giving much of a thought to what was happening before, after, or even at the same time in a different clinic.

But pay-for-performance reimbursements tied to longer-term wellness are quickly shifting a provider’s responsibility from point-in-time care to overall health – and that trend that is placing new pressures on an organization’s ability to engage in meaningful, intelligent big data analytics.

The move from sick-care to well-care requires the industry to adopt a more preventative model of population health management, says Dr. Michael Blackman, Chief Medical Officer at McKesson Enterprise Information Solutions, and that model must be built upon a firm foundation of smart, targeted, tailored point-of-care insights.

“Value-based care is supposed to get providers to start asking population health management questions proactively,” he said to  “How do we keep people out of the hospital?  How do we keep them at home?  How do we make sure healthy patients stay healthy so we can maximize our resources when we have to care for the people who are sick?”

The process changes that will allow providers to answer these questions – and still make a profit under value-based contracts – aren’t easy to enact.  Financial and clinical success in this new environment will require a greater reliance on tailored health IT tools and system-wide collaboration. 

Value-based care is a data revolution, not just a reimbursement concern

By and large, the healthcare system has struggled to integrate EHRs and other patient management tools into their workflows. Generating the inertia required to move the industry forward into the big data era will require a seismic shift in the way organizations create, analyze, and leverage patient information.

“Fee-for-service has been problematic, because it doesn’t really incentivize a lot of the things we should have been doing with our data for a long time now,” Blackman said.  “But now that we’re looking at value-based care, there’s a lot more pressure to start performing population health management.”

“The only way to do that is to get data in real-time – or as close to real-time as you possibly can,” he continued.  “And in order to do that, you have to take advantage of every interaction with the patient.”

Read: Preventing Big Data Pain Points During a Healthcare Encounter

Insufficient data interoperability and poor care coordination, coupled with the general movement away from the hospital for as many services as possible, have combined to make that visibility into the patient’s healthcare journey a significant challenge.

“Certainly, the core of most health systems today is still the hospital, but more and more work is being done outside that setting,” Blackman pointed out.  “Patients we would have admitted ten years ago are now being routinely treated on an outpatient basis, and the focus is shifting to keeping patients at home, or in the clinic, or in some lower-cost setting that makes it difficult for the hospital to see the patient’s data after they leave the building – or if they never enter the building at all.”

“That makes transitions of care very important, and it means that health systems have to get a tighter grip on their data and maximize their opportunities to use it.”

Cost-cutting efforts may be blurring the traditional boundaries between care settings, but that can actually be a positive development, he added, if providers cultivate the ability to manage patients holistically across those weakening barriers.

“At Kaiser Permanente, for example, if you show up for a dermatology appointment and your electronic health record shows a note that you’re overdue for a mammogram, they address that immediately in the dermatologist’s office,” Blackman said. “By doing that, the patient doesn’t have to make a whole separate effort to schedule their screening and it’s more likely they’ll get it done, even if it’s not strictly under the responsibility of dermatology.”

The difficulty of crafting that type of seamless care delivery system, he noted, is providing the right clinicians with meaningful data at the right time.

“Because there is so much data out there, you need to be able to present the most relevant information to people and make it very clear which data they need to act on,” Blackman said.  “So if you are designing an analytics system or an EHR that includes population health management alerts, you need to pick the things that are important.”

“Where are you having your biggest problems with outcomes?  What areas present an opportunity to reduce costs?  If you’re having trouble with congestive heart failure readmissions, that’s a place to start.  If you’re having trouble with diabetics frequently coming in with extremely elevated blood sugars, or large numbers of people in diabetic ketoacidosis, that’s another good place to start.”

Providers may also wish to employ a more evidence-based approach to developing population health management guidelines, reducing potentially detrimental variations in care, and eliminating unnecessary spending.

“Evidence-based medicine can be a great thing,” said Blackman.  “There are a lot of processes that we know can work, and we really should make sure we adhere to them.  The challenge is putting the information out there so clinicians can see what they need to do and make the right decisions. They need to be reminded to help their patients schedule those mammograms or colonoscopies or eye tests for diabetics.” 

Integration and collaboration will be critical for success

Leveraging data connections to integrate preventative services is especially important in the primary care setting.  Under many value-based care models, the primary care provider is the first and main point of contact for patients – and it may also be the organization that stands to suffer the most financial harm from sub-par results.  

Since the beginning of the big data era, physicians have continually complained that they are being asked to do more with less: documentation burdens, reporting requirements, and expectations for better outcomes have increased even as their time slips away in quarter-hour increments that make it impossible to meet many patients’ complex needs.

Read: Understanding the Value-Based Reimbursement Model Landscape

“In a primary care office, where clinicians are booked solid in 10 or 15 minute increments all day long, there is no way to do everything that needs to be done without a health IT system that supports the provider instead of taking more of their time,” said Blackman.

“If an elderly patient comes in with diabetes, hypertension, and congestive heart failure, that’s not someone you can effectively see in 15 minutes.  So you take longer, as you should.  And then when the next person comes in, and they’re in their 20s with a cold, your first reaction is to say, ‘Oh good, I can catch up, because this won’t take me 15 minutes.’”

“But you miss the fact that this person is overdue for their tetanus update, or that their weight has been creeping up, or that they’re not sleeping well or may be depressed,” he said.  “The challenge is how to put the information out there to the clinician so that they don’t miss these things.  We need to take care of the whole person.”

Frustrated physicians need to take a collaborative approach in order to succeed in this demanding environment.  Team-based care has become a favorite strategy for many organizations – and for good reason, Blackman says. 

“You don’t need a physician to recognize that a person is overdue for their vaccination.  A medical assistant can see that just as easily, and they can make sure it gets addressed without requiring the physician’s time, so he or she can focus on other aspects of care.”

Collaboration outside of the clinic is equally important, he added, especially for the larger-scale population health management and big data analytics that support a truly value-based ecosystem.

“In a lot of places, only the hospitals have the resources to do the kind of tracking and coordination required for really comprehensive population health,” he said.  “It’s beyond the scope of any individual practice, which is part of the reason why we’re in the midst of one of these waves of consolidation.”

“It goes back and forth: hospitals start buying up practices to achieve scale, and then they discover that practices cost money so they try to get rid of them.  And now, because of the push to value-based care, they’re back to trying to control practices again because they need to be in charge of the entire continuum of care.”

“I think we’ll stay in this mode for a long time,” he predicted, due to the fact that value-based reimbursement is likely to retain its stronghold over the industry for the foreseeable future.

To make the investment in widespread change worthwhile, “we need true interoperability,” he asserted.  “More and more often, the hospital and its affiliated practices are using the same electronic health record, so all of the data is there.  But it’s not always communicated well, and the strategies may not be there to make the data work for the patient.  And, of course, as soon as you start seeing providers that aren’t tightly affiliated with that health system, the situation starts breaking down.” 

“The industry needs to work harder to develop interoperability so the provider can know exactly where the patient has been and what happened to them there rather than repeating tests or asking the same questions over and over because they simply can’t access the right records.”

Avoiding the “data dump” to foster meaningful decision-making

Even if healthcare organizations do manage to develop full data transparency across a patient’s longitudinal health record, they must carefully decide how to use it.

“You can’t just dump all the data onto one screen and leave the provider to sift through in during those tight 15 minute visits,” Blackman said.  “There is no possible way a person can look at all of the data and make sense of it in a timeframe like that.”

“The key will be to set up the system in such a way that it highlights the data that’s important – while keeping in mind that there’s no single answer for what is really relevant for the situation at hand.  It’s going to be different for each patient.  It’s going to be different for each provider.  Tailoring those insights will be a major, major project.”

That project will likely have several main components, including provider-generated data from electronic health records, labs, and pharmacies, patient-reported outcomes and patient satisfaction data, genomic data, lifestyle and socioeconomic factors, and patient-generated health data imported from wearables, home monitors, and other Internet of Things devices.

“You need to figure out how to get that information into the organization, whether that’s through the Internet of Things automatically feeding data into the record or through a portal where the patient uploads their data,” said Blackman.

“Now, that amount of data for every patient all the time would overwhelm a primary care practice.  Frankly, it would overwhelm a health system.  So behind the data, we need algorithms to take a first pass at making sense of it and flagging what a human clinician should take a closer look at.”

Read: The Role of Healthcare Data Governance in Big Data Analytics

EHR developers have not always had the greatest success with clinical decision support systems, EHR alarms, and other alerts intended to guide the clinical workflow.  Alarm fatigue has become a serious patient safety issue, not to mention a stressor on clinicians and their job satisfaction.

Big data analytics can help, Blackman argues, if developers forego the one-size-fits-all approach to designing decision support tools and instead embrace a more nuanced foundation for decision-making.

“Algorithms can do a good job if we make them better at tailoring those alerts for individuals, who may each have slightly different ranges of what could be considered normal,” he said.  “In some situations, you might want to look for an absolute value.  In others, you could be more interested in a percentage change over a certain period of time.”

“For a patient with congestive heart failure who was just discharge from the hospital, for example, one of the biggest risks for readmission is when they start to gain weight.  If you had weight readings for a patient every day, and you saw their weight going up, you can predict with pretty good accuracy when that patient is going to show up in the hospital again.  And if you had that data, you may be able to prevent that admission from happening.”

“People’s weights vary tremendously, so it’s no good to trigger an alert based on the fact that the patient is over 175 pounds or something like that,” he explained.  “That could be a normal weight for a lot of patients, so it doesn’t tell me anything.  Instead, you have to look at whatever their weight was at discharge, and whatever their weight was the day after that, and the day after that.  It’s the change that’s important, not the absolute number.”

If developers can design truly sensitive and accurate algorithms that know what’s dangerous for each patient based on the data in their EHR and fire based on deviation from an accepted trajectory, these tools could be much more valuable for making a decision than an alarm that simply tells the clinician that a patient may be overweight.

“Designing those algorithms is the first step,” Blackman said.  “It is hard, but it isn’t too hard for us to figure out.”

“The second step is getting the right equipment to the patients in their homes, making sure it’s connected, and having a system to retrieve and analyze it.  And third step is having the staff with the time and ability to look at those alerts and follow up with the patient to see what the problem is.” 

Historically, payers have not provided reimbursement for those types of proactive population health management tasks, which makes it difficult to justify the time and expense.

“But we are starting to change that with accountable care, where you’re responsible for the entirety of the patient,” he said.  “Under value-based reimbursement models, it starts making financial sense to do those things, and data tools are going to make it possible to do it well.”  


Join 25,000 of your peers

Register for free to get access to all our articles, webcasts, white papers and exclusive interviews.

Our privacy policy

no, thanks