On Nov. 5, consultant leaders at the New York-based Deloitte presented what the consulting firm called one in of that organization’s “Dbriefs Health Sciences series,” under the title, “The Big Data Revolution: Unlocking the Power of Health Care Analytics.”
Harry Greenspun, M.D., senior advisor in the Deloitte Center for Health Solutions at Deloitte LLP, moderated a (webinar-based) panel with three other consultants. Those three others were David Biel, principal and national technology lead in the Life Sciences & Health Care Practice at Deloitte; Brett Davis, principal in Deloitte Consulting LLP and general manager in Deloitte Health Informatics LLC; and Dan Housman, director and chief technology officer at Recombinant By Deloitte, Deloitte Consulting.
Among the topics the Deloitte consultants discussed were the top trends shaping healthcare right now; “enterprise data challenges: asking the hard questions”; and “making analytics a game-changer.”
After an initial framing of some of the issues, the consultants discussed some of the new technologies, including patient-monitoring technologies, rapidly coming into healthcare right now. In that context, Davis said, “I think we’re just at the forefront [beginning] of this national dialogue. This data has so much potential to unlock new innovation, in terms of insights about patient populations, and what’s working and why, but there’s going to be a healthy tension around how this data is going to be used, in terms of privacy.”
In one of a series of instant online polls of the webinar audience, the Deloitte team asked audience members which pressures they felt were most significant for them in their organizations. The results of that first poll: police changes, 34.0 percent; an “economy that is stretching thinner every day,” 21.4 percent; “the need to stay relevant in an increasingly competitive market,” 20.0 percent.
Greenspun asked Davis, “How will organizations monetize this” investment in infrastructure and other technologies? Davis responded, “If you’re a provider moving into risk-based contracts around, say, congestive heart failure, being able to monetize this will be around figuring out what’s happening around your population, and then, say, targeting it down to males between 45 and 60 who are on particular drugs, etc. That’s the level of granularity we’ll be able to get to. The monetization comes from the insights from the data” that will help providers figure out how to manage particular populations, he added.
What kinds of convergences are taking place around incentivization and around technology development right now? Greenspun asked the other panelists. Per the broad convergence taking place, Biel said that “It’s interesting to see it happen,” he said. “Some of it is regulatory-driven, with the accountable care push happening. We see a lot of providers getting into accountable care,” he noted. “And we’re certainly seeing a large number of payers buying up providers, so they’ll have a certain amount of a captive audience, if you will, and from a data perspective, that’s driving the need to pull together financial and clinical data, so that you can really examine a cohort of a population and get to analytics that are meaningful.”
Housman noted that because of the convergence of incentives right now, a lot of activity is taking place, both from a business standpoint and from an IT standpoint. Still, he said, “A lot of groups are just moving through that first stage of their development of analytics and being able to build predictive models in terms of things like getting people on particular drugs, etc. In an advanced organization, they can start to analyze whole care pathways,” he added. “But to do that, they’ll need hospital data, ambulatory data, out to what’s going on in patients’ homes, and what are their environmental factors that are relevant? So all this is going on, but it’s early days, with providers and payers sorting through things they didn’t have before.”
“We now have the ability to collect all sorts of information, and I daresay,” Greenspun said, “dirtier data.”
“Yes,” replied Housman, “and one of the challenges we face is sorting signal from noise. The good news is, in a lot of other industries, we’ve developed a lot of technologies for doing that, and the good news is that it’s now beginning to happen in healthcare. These new sources will present all sorts of new opportunities and challenges.”
Looking at the gaps
In another instant audience poll, the Deloitte consultants asked audience members to cite their biggest data gaps: were they with operational data, clinical data, financial data, translational data? The results were as follows: operational data, 34.5 percent; clinical, 25.0 percent; financial, 20.9 percent; and translational, 19.4 percent.
Greenspun shared with his panelists that there remains “a need for a common pool of data” in patient care organizations.” Housman agreed, saying, “Yes, that goes back to the issue of bringing data together and reconciling it. Some healthcare organizations have more aligned objectives in one area versus another, depending on whether you’re a health plan or provider; so your recognition comes out of what you’re trying to achieve, year to year.”
“Making the data usable has never been easy,” Greenspun noted. “Yes,” Housman agreed, “and it requires building advanced models to look like. What is a contract going to look like if we have a complicated risk model between a provider and a payer? And that’s a difficult thing, right? The clinical model starts moving into the financial model, but it builds into… framing the questions in conjunction with stakeholders like physicians and patients who aren’t in the c-suite, to drive the answers to those questions.”
And Davis added, “The operative word for me is ‘enterprise.’ Particularly for organizations that have just gotten done investing tens, sometimes hundreds, of millions, in electronic medical records, it’s a real challenge to start creating some of these data entities in a one-off sort of way,” he noted. “So healthcare organizations need to start thinking through some of these models in a comprehensive way, so we don’t just end up with a lot more new silos.”