Medical Informatics: Driving data analytics

BOSTON—The healthcare ecosystem sees about $2 trillion in waste per year, said Heather Fraser, global life science and healthcare lead at the IBM Institute for Business Value, speaking at the Medical Informatics World Conference on April 9.

A transition away from such waste is going to take years, she said. However, the ability to capture, share, integrate and analyze data is game-changing in terms of producing outcomes that matter. Benefits of analytics for the healthcare ecosystem include better selection of drug candidates, better selection of appropriate clinical trial candidates and ensuring patient adherence to treatment regimens.  

“From a provider perspective, the No. 1 priority is about trying to prevent readmissions,” Fraser said. But, issues such as a lack of standards and inability to exchange data prevent everyone from getting the maximum value from the information.

Fraser discussed several key results found in IBM’s report, The Value of Analytics in Healthcare. When considering strategy and governance, 63 percent of the 130 healthcare leaders from around the world said they have a current analytic strategy and 35 percent don’t have a strategy but plan to create one. When asked about how they define strategy, 66 percent said they use a descriptive manner, such as reporting and metrics, and only 34 percent look at strategy on a predictive level, Fraser said.

The study indicated that data management was a major concern for healthcare organizations. Claims data in the U.S. are much more comprehensive than in Europe, she said. Just one-third (34 percent) of respondents said they are able to track a service or product from end to end and another 34 percent said they cannot do so. Eighteen percent of providers can look at clinical history through their EHR or similar systems. Eighteen percent of life sciences companies could track a complete clinical trial or other program. “The analytics abilities that were there were confined and siloed to one department or geographic area,” Fraser said.

The study asked about challenges in adoption of analytics. “Data is not the problem—it’s the analytics,” Fraser said. “Only 39 percent were looking at a balance of people with analytics skills who could build insights back into the business on an ongoing basis.” Currently, the distribution of analytics staff is predominantly decentralized but this is expected to switch toward a more hybrid model.

The study found issues that prevent analytics projects from delivering. Respondents weren’t defining key performance indicators at the beginning of projects, she said. “They didn’t know what things business would value most from projects and when they did get results, they hadn’t worked out how to operationalize them. A lot of that came down to the fact that they didn’t have change management elements worked into their business.”

The report looked at ways to put more efficiencies in place to help get the right patient to the right treatment more quickly. Those efficiencies include better access to data to make sure they are structured and reliable to allow users to refine the treatment and care. To build the skills base in analytics, Fraser said several U.K. universities are combining analytics courses with business skills and economics. “People who understand the benefits of analytics can help you get early insights from those projects.”

No one organization has the data, answers and capabilities, she said. “The only way we’re going to benefit the end patient is if organizations partner. That includes data providers and mobile providers to really drive the best opportunities to gather data and drive insights out of that data.”