How government data is changing the way we understand health care

Physician utilization data from CMS gives developers the raw material to build tools that allow for the comparison of doctors on a number of criteria.

Shutterstock image: medical professional interacting with a futuristic interface.

Government data is changing what we know about the work doctors do, and is giving developers the ability to transform that data into useful tools.

For the second year running, the government has released information on how doctors and other providers are charging Medicare. The physician utilization data from the Centers for Medicare and Medicaid Services, covering $90 billion in payments across 950,000 providers, gives developers the raw material to build tools that allow for the comparison of doctors on a number of criteria, such as services delivered, charges submitted, and more.

The news was announced at the sixth annual Health Datapalooza conference in Washington, D.C., on June 1.

The data, said Niall Brennan, CMS chief data officer and director of the Office of Enterprise and Data Analytics, consists of 10 million distinct observations, and builds on data released in 2014, to allow for comparisons over time. In addition, CMS released a dataset on prescriptions made by providers using Medicare Part D at the end of April that allows for the comparison of health care provider by prescribing patterns.

"Now you can actually see every piece of care and every drug they prescribe. Is it perfect? No. Is it better than where we've been before? Absolutely," Brennan said speaking on a panel at the conference.

This push for price and prescription transparency, part of the larger Obama administration Open Data policy, is designed to pull back the curtain for consumers on what health care delivery really costs, and gives developers and researchers the raw material to build tools to make the data more usable by ordinary health care consumers.

"I'm ready to declare progress, but not victory," Brennan said. "I think a lot of people think transparency is easy. You just kind of push the big release-data button and it gushes forth and it's done," he said.

CMS was only "gingerly dipping its toes" into the open data world when the first Health Datapalooza took place. But that changed in 2013 with the release of the "chargemaster" list that revealed what hospitals were charging for common inpatient procedures, and the rates at which Medicare paid claims. The data revealed wide disparities in charges for procedures, even within the same metropolitan areas, and garnered pop-culture currency with a prominent mention on "The Daily Show."

Brennan said he encountered some skepticism and fear inside CMS, as the agency moved to release more and more data. "We've deliberately adopted an incremental strategy where the initial data releases were pretty modest, but we had to almost reassure people that the world wasn't going to end if dataset x-y-z came out," he said.

Brennan said that while some data releases made big news, others flew under the radar.

"It's a very market-driven process. We've released data that we thought was going to have a wow factor, and people have yawned, and we've released data where we thought people would yawn, and they've gone wow," Brennan said.

The unpredictable response to data releases has spurred even more releases.

"You almost have to err on the side of openness, because we don't necessarily know the value that others may derive from the datasets, especially when they're combined with other data," Brennan said.