Editor’s Note: Healthcare Informatics has compiled together a five-story Special Report section on data analytics for its July/August print issue. This story, and one that can be read here about a fast analytics team at the University of Michigan Health System, are part of that special section.
More and more, across healthcare organizations nationwide, data and analytics tools are being seen as a means to improve efficiency and quality. Yet, according to one survey from KPMG LLP, a New York City-based audit, tax and advisory firm, only a small fraction of those in the industry are using these capabilities to their fullest potential.
The March 2015 survey of nearly 300 respondents who identified themselves as being employed by providers, payers, or life sciences companies, found that only 10 percent are using advanced tools for data collection with analytics and predictive capabilities. Twenty-one percent indicated that they are still only “planning their journey.” Of the other respondents, 16 percent said they are using data in strategic decision making, while 28 percent are relying on data warehouses to track key performance indicators.
Indeed, as the provider community continues to prepare for the shift to value-based care and being at risk for various patient populations, it’s as clear as ever that sophisticated analytics tools will be a necessity going forward—even if adoption levels are still currently low. Speaking to the survey’s results, Bharat Rao, Ph.D., KPMG LLP’s national leader for healthcare and life sciences data analytics, says that many organizations are not where they need to be in leveraging this technology, and that providers need to employ new approaches to examining healthcare data to uncover patterns about cost and quality.
Dr. Rao, who has more than 60 patents tied into the realm of healthcare informatics and analytics, personally looks at analytics on a “full stage,” that moves from descriptive (what happened and why it happened) to predictive (what will happen), and then to prescriptive (what I should do about it). “I will say that we have gotten pretty good at descriptive analytics,” Rao attests. “There are tools out there that do a good job of painting a picture of the near past. It’s no longer acceptable to take 45 days to get quality measure reports back out,” he says, offering an example of improvement in that area.
Bharat Rao, Ph.D.
However, Rao points to both the huge gap and potential for predictive analytics, and he notes that prescriptive analytics is something that doesn’t happen in healthcare today. “Readmissions tools do a reasonable job, but there is a big gap there. One thing that has changed over the last decade is the recognition by provider organizations that analytics is not a nice shiny toy, but something that has become increasingly important for them to survive,” Rao says.
Diving Into Uncharted Waters
It was a few years ago when David Seo, M.D., current University of Miami (UM) Health System associate vice president, information technology for clinical applications, and chief medical informatics officer (CMIO) of the Miller School of Medicine, and other health IT leaders at the health system began to truly understand the evolution of where healthcare was going. “Patient-centered medical homes and ACOs [accountable care organizations] were the trends under the main idea of managing risk,” Dr. Seo says. “I was getting multiple calls and visits from vendors offering analytics solutions, one after the other, and what became clear was they were not offering a true full suite of what a health system needs to manage risk. Our own EHR [electronic health record] vendor talked to us, but even what they could provide was limited.”
Seo says that UM Health System was looking for company that had a long track record of understanding data analytics and security, so it ended up choosing Lockheed Martin, a Bethesda, Md.-based global security and aerospace company with involvement in healthcare analytics. “We knew were headed towards a clinically integrated network and other things of that nature,” Seo says, noting the need to establish a data environment, implement big data analytics and predictive modeling tools, and start to stratify patient data and conduct risk assessments.
David Seo, M.D.
Seo agrees with Rao in that predictive analytics in healthcare “is still very much in its infancy no matter who you talk to.” Indeed, aside from the basics such as readmissions, true predictive analytics has not come to fruition, he notes. To this end, University of Miami Health System started out with a diabetes risk model, and clinician leaders have shown that the model can fit within providers’ workflows, Seo says. He adds that the risk model can be ordered through the organization’s order entry system, or it can have a patient ask to run that risk model themselves in test environments. “The risk model returns a score, so you understand your risk of developing diabetes over the next five years, for example. And now we are engaged with our clinical staff to [look at] things such as what is the threshold we would set to apply an intervention, for instance,” Seo says.
Seo further emphasizes the importance of the health system’s work around different validations, which he says is a necessity before a risk model of this scope goes into production. He explains two key areas around validations. First, the validation of a phenotype or a diagnosis using EHR data needs to be validated for the system. “If I am going to say you do or do not have diabetes for example, that needs to be valid, and you need to understand what the positive predicted value of that phenotype is,” he says.
Second, he says, the diabetes prediction needs to be valid for a specific population. “I like to say that population health will be local, so the diabetes model that we pull from the literature has been validated form a highly specialized population that perhaps is of different racial or ethnic origins from our south Florida population. So what we’re doing is validating the phenotype in our population, and also understanding what the performance of that model is in our population. These are two important steps before going live with this prediction model,” Seo says.
Other healthcare organizations are making their own advancements in the analytics space. Rao points to industry leaders such as Mayo Clinic and Cleveland Clinic, where prescriptive analytics, such as precision medicine, is happening in pockets. “Leading organizations are starting to get ahead of the curve, and are recognizing that healthcare is changing into an at-risk model, so you as an organization will be responsible for care outside your four walls,” Rao says. “You are now responsible for the entire cost of the patient, so you have to track what happens to them,” he adds. As such, the organizations that are at the cutting edge are recognizing that even though today the portion of [value-based] payments is only 5 to 10 percent, it will be 60 percent by 2020, according to some folks, he says. “Providers are gearing up to get the data infrastructure, care coordination tools, analytics tools, and contracting tools to deal with that transition. That’s starting to happen,” Rao says.
Moving forward, Seo stresses that while vendors are now rolling out the tools to make disease management easier, health systems need to re-engineer their operations since it’s not just about looking at the doctor-patient relationship anymore, but rather healthcare leaders have to think about it now in terms of one-to-many simultaneous relationships. “Healthcare organizations have to readjust their care delivery patterns to fit this population health idea,” he says. “It can be hard, and it’s not the way we traditionally practice medicine. Also, some of your population will be managed this way while others won’t be, while finally keeping in mind that you have pressures of new payment models,” Seo says, speaking to all of the challenges health IT leaders now face.
Seo additionally notes two core challenges that can become present in this area of analytics and disease management: provider behavior and patient engagement. Regarding the former, simply turning on alerts in a system, or sending alerts at the point of care, will lead to failure if that’s the objective someone is looking for, he says. “We have engaged with subject matter experts who are M.D.s, as we have been developing our work around diabetes, and they have been involved in a number of our activities and interactions. They have been fully participatory, they have bought in and [been supportive], and if you don’t get that, you won’t get true change in physician behavior,” Seo says. “Doctors can be very good in finding a way around something they don’t agree with, so that’s what you’ll get. Or you will get compliance without commitment to the process. Provider behavior starts with engagement early in the process from all levels.”
Regarding patient engagement, Seo says that in newer accountable care models, there’s accountability for all parties involved—including patients and their families. “You want to give patients a method to interact with their own information. Our mindset has been to give them as much data as we can in a safe and appropriate manner so in their discussions with physicians, they understand what’s going on and can participate in their own healthcare.”
Rao says that from an analytics perspective, the top challenge above all is, that when looking at this transformation of care, there is a large portion of the healthcare provider population which is doing well in the current fee-for-service system, and there isn’t an incentive enough for them to change. “So there is a mindset that says we need to change, and that needs to come,” Rao says. When the feds made the announcement about 60 percent of care being tied to quality, I expected there to have been seismic shockwaves through the community, but people haven’t reacted like that yet. It’s almost as if they think it might not happen or after the election things will change,” he speculates.
Rao further notes the amount of unstructured data that is locked away. “How to you make it accessible for analytics? We collect data and notes every day on patients, and that needs to become actionable,” Rao says. He adds that the good news is that there are tools here to help that are about to become more sophisticated. “People say healthcare data is messy, but nothing was messier than the Internet. Google, Yahoo, Bing and Microsoft have done the greatest job in making that unstructured data useful, and now [the Internet] is the single greatest resource in history of mankind,” Rao says. “It’s about taking that free text and crunching it in a way to find the patterns that make sense. That’s a technology revolution waiting to happen.”
