Quality programs best served with clinical data analytics

Key Points

PAYING FOR PERFORMANCE will cause many health plans to rethink how they contract with provider networks, and experts say there are tremendous changes ahead.

"There are various stages of evolution in different parts of the country. Some of them are focused on the practice, and some of them are building out to a more collaborative community model," says O'Kane.

"We've been talking about this for 20 years," says Andrews. "Clinical data has been hard to get and difficult to work with. It's reminiscent of where biotech and pharma were about 15 years ago, when they started generating a lot of data. Suddenly there is real insight-and everything changes."

The indicators that are in use now have been developed over time largely on the basis of claims data and are still rudimentary.

"We also need to better understand the relationship between quality and cost," he says. "Even if we made people healthier, what was the impact on the economics? Right now it's an assumption that higher quality costs less."

The first order of business is putting the infrastructure in place: extracting the usable data, opening up access to it, and filtering out invalid data. Andrews cautions against taking a narrow view that simply creates an "electronic paper flow."

"There is value to electronic referral, for example, but it's a low-value activity compared with performing population health studies and beginning to understand how we could affect the care of large populations with the chronic conditions that drive much of healthcare cost," Andrews says.

Paul Oates, senior enterprise architect at Cigna, says the industry must be careful in using the existing data. Studies performed in recent years show that improved data and data collection efforts help drive better control of costs-but those results were demonstrated by baselining a historical experience, applying changes to the process of managing care, and then seeing what happens.

"It's a challenge to identify causality in those results-to make sure the results are not due to chance but to something we can all influence, in partnership with physicians, and that we can repeat," says Oates.

He says legacy systems facilitate analysis of claims information, which conveys what occurred for the purposes of payment, but not why. The time lag around claims-based data also can create a lengthy delay between the event and the reporting of it downstream.

His group is restructuring for better analysis of data sources with clinical depth and breadth, such as lab and pharmacy details. Cigna hopes to focus less on data gathering and more on dissemination of actionable information to providers. Within some medical practices, the plan embeds a care coordinator to support members; in others, the practice uses the analytical tools provided by the plan to improve the quality, affordability and experience of care.

"There's a lot of variability; we need to be able to work with what physician practices are most comfortable with-and not all of them are-using electronic health records. Sending the information back out in a way that physicians can use in their workflow, based on their capabilities, is critical," says Oates.