Making sense of the tide of health data inundating the U.S. healthcare community remains a daunting challenge for driving clinical and financial performance.
However, by unlocking the value of petabytes of patient-specific data, health insurance plans will be able to help members to achieve their healthcare goals, argues Keith Dunleavy, M.D., CEO of Bowie, Md.-based data analytics vendor Inovalon.
“Data is driving how we should manage members and how members want to be managed,” Dunleavy said June 12 in a general session at the America’s Health Insurance Plans Institute 2014 in Seattle.
With healthcare data expected to skyrocket from about 500 petabytes today to more than 25,000 petabytes by 2020, the data is becoming more complex and inferring knowledge from heterogeneous sources is becoming more complicated. To take advantage of the data and provide the right intervention to the right patient at the right time, advanced analytics in healthcare (i.e., segmentation and predictive modeling) is the key to taming the tsunami of data gathered from claims, laboratory, pharmacy, medical benefits, demographic information, provider information, and facility information, Dunleavy says. Applying advanced analytics to patients can proactively identify individuals who would benefit from preventative care or lifestyle changes, he adds.
Currently, capitated managed care is composed of four principal elements: Medicare Advantage, managed Medicaid, commercial exchanges under the Accountable Care Act, and accountable care organizations, Dunleavy notes. In turn, these parts of capitated managed care are subject to five market forces: clinical and quality outcomes, disease and comorbidity accuracy;,utilization data, compliance, and consumerism.
Health plans must operate under these market forces to deliver high quality care with accuracy in their risk score, utilization and compliance, Dunleavy points out. Managing chronic diseases in the face of these market forces and regulatory requirements is a very complex process that can only be achieved through the use of predictive analytics to harness big data in order to identify exactly how patients’ complicated medical conditions are progressing, he asserts.
How a member of a health plan thinks about their disease, themselves, the costs, their provider, and where they will get their care is “truly unique to that person” and evolves over time, according to Dunleavy. Only data can answer the central questions of what to focus priorities on and how to interact with a member and the provider environment, he claims.
Many health plans around the country are making progress internally in analyzing this data, “but at the end of the day if all this analysis creates insight that doesn’t result in an action and impact it is worthless,” warns Dunleavy. Data latency and data integrity continue to be major problems hindering the right intervention with the right member at the right point in time with the right content and applying the right resources in an efficient and effective way, he says.
“In healthcare, unfortunately, the data latency—the time it takes to go from a clinical event to awareness and full availability of the resulting data—can sadly be measured in months,” Dunleavy concludes. “There are some health plans with whom we work that are focusing on data latency and have brought down the time it takes for information to go from existing to being known by the right parties and the right systems dramatically, and it makes a difference.”
At the same time, “information that is erroneous is prolific in the healthcare environment,” he says. “Nobody likes to hear that but there’s a lot of dirty data in the healthcare space.”
Originally published by Health Data Management.