This must be “Big Data Week” here at Healthcare Informatics. We have had a guest blog on the topic by Bob Kelley, senior vice president for analytics at Truven Health Analytics, and a thoughtful piece by Editor in Chief Mark Hagland. I’d like to add to the mix with some reporting on a nice presentation I saw this week by Cynthia Burghard, research director for accountable care at IDC Health Insights.
Mark’s blog noted that healthcare CIOs, CMIOs and other IT leaders could easily list numerous roadblocks in their way on the journey towards the full leveraging of big data tools. Most hospitals, medical groups, and health systems lack the data warehouses and data marts, the robust data analytics and business intelligence capabilities, and above all, the interoperability, needed to turbocharge their efforts, he noted.
IDC’s Burghard noted that so far hospitals are not investing in big data and analytics, with 78 percent of providers surveyed last year investing less than $1 million to date, as opposed to 50 percent of payers reporting spending more than $1 million. Provider investment is lower because hospitals are bogged down with their EHR implementations. “The well for technology investment is a bit dry,” she said.
When asked what they would like to do with big data, providers say they would like to analyze their own operations and also have an interest in machine or device data.
One big challenge, Burghard says, is deciding which data is relevant to the business or clinical question you are asking. Another challenge, especially outside of academic medical center settings, is the lack of staff with the right mix of analytical skills. Burghard recalled a case study involving an integrated delivery network in the Midwest that worked with IBM to understand drivers of congestive heart failure. They determined drug and alcohol use and living situation were key drivers. But when she asked them what they were going to do with that information, there was silence on the other end of the phone. “There is a lack of understanding of how to operationalize big data,” she said. “Once I get that information, what do I do with it? Most healthcare organizations haven’t faced those challenges.”
A few vendors, including IBM, Oracle and Dell, can offer healthcare providers a full big data solution. But because investing in enterprise big data tools is expensive and the staffing is hard to find, we are seeing the emergence of a new business model: analytics as a service.
Vendors in this space include Apixio, Explorys and Humedica. “This is going to be the way healthcare will consume big data,” Burghard said. “It is just too complicated for the average healthcare organization to invest in.”
Burghard closed with several recommendations, including:
• Recognize the value of untapped data assets in supporting fact-based decisions by physicians, case managers, pharmaceuticals, patients and caregivers.
• Understandthe implications of operating without critical information and build use cases to address the challenges. “The goals of health reform and accountable care cannot be realized without access to big data technologies,” she said. We are used to not having good data for decisionmaking, and have adapted, she added. But look at what technology you have and the staff investments required. What would it take to make that leap?
• Consider a shared services model for technology deployment with non-competing partners. The cost of big data technology is outside the financial resources of many healthcare providers, so new business models will need to be developed.