Electronic health records such as those maintained by hospitals and doctors’ offices don’t have a lot of value for healthcare researchers, which means new sources of health data are needed to improve quality in the U.S. healthcare system.

And moving doctors away from paper patient records to EHRs won’t do a lot to reduce costs either, said Robert Litan, vice president for research and policy with the Kauffman Foundation. Litan was speaking at a national conference on Value-driven Engineering in Akron, Ohio, sponsored by the Austen BioInnovation Institute in Akron.

“[EHRs] may lead to faster and better diagnoses,” Litan said. “It will not necessarily save money, but that’s the way it’s been sold.”

Specifically, Litan listed two reasons why EHRs don’t help researchers much. First EHR data is spotty and episodic since most of it is obtained only when patients visit their doctors, and some patients go years between visits.

Second, sometimes EHR data simply isn’t reliable. Doctors are under lots of pressure to get paid, so what they include in patient records is often aimed in that direction and isn’t entirely accurate. “There’s tremendous incentive to upcode,” Litan said.

So if EHRs don’t hold the data that researchers need, then who or what does?

The public, according to Litan. In other words, go to the patients and ask them for their data.

For inspiration, look toward the “Army of Women,” Litan suggested. “Google it, because it’s the future,” he said.

Dedicated to fighting breast cancer, the group solicits health data from women on a voluntary basis through questionnaires, the submission of lab results and more. So far, about 365,000 women have signed up for the project, about 30 percent of whom have breast cancer, Litan said.

Armed with all this data, researchers hope to be able to follow women over time to determine what factors contribute to breast cancer.

“This is revolutionary,” Litan said.

Copyright 2013 MedCity News. All rights reserved. This material may not be published, broadcast, rewritten or redistributed.