Patient Data 'Mega-Registry' Could Bolster Clinical Research

Healthcare Social Media: Time To Get On Board

If the Office of Management and Budget (OMB) approves, the Agency for Healthcare Research and Quality (AHRQ) will soon establish a "registry of patient registries" (RoPR) to aid researchers doing comparative effectiveness research. The registries referenced in RoPR would include those related to clinical trials and might also encompass patient registries created by healthcare systems and accountable care organizations.

AHRQ's proposal, published in the Federal Register, would create a database of the registries used in some observational trials. While many of these registries have already been registered on Clinicaltrials.gov--the federal government's omnibus listing of ongoing studies--RoPR would provide a supplementary listing that people could use to provide additional details about the nature of the registries and the underlying studies.

The main purpose of RoPR is to allow existing data to be reused in ways that may help answer research questions, according to Elise Berliner, director of AHRQ's technology assessment program, in an interview with InformationWeek Healthcare.

[ Is it time to re-engineer your clinical decision support system? See 10 Innovative Clinical Decision Support Programs. ]

For example, she noted, AHRQ is sponsoring an observational study on glaucoma to answer the question: After two different drugs have failed to control the glaucoma, what is the relative efficacy of prescribing a third drug as compared to a procedure? So far, researchers have collected data on 3,000 patients. Other researchers are engaged in glaucoma studies that pose different questions. AHRQ would like to offer its registry to other scientists who could add their data and have a bigger dataset for their own studies.

The patient registries in RoPR would not contain any data themselves. But if a researcher decided they wanted to collaborate with the institution that owned the registry, Berliner said, they could approach the registry owner. If the owner consented, they'd have to sign a use agreement and ensure patient privacy.

"It's useful to have large databases with data collected all using the same definitions," Berliner noted. "We don't know that people will do that, but we're giving them the option to do it."

AHRQ has seen huge growth in the number of patient registries, she said, mainly because drug and device companies are conducting more observational trials. The device makers study how their devices work in clinical practice, and the FDA often requires the pharmaceutical firms to do post-market studies of their drugs.

In addition, she said, hospitals report quality data on surgical procedures to medical societies such as the American College of Cardiology and the Society of Thoracic Surgery, which use the information to create quality improvement registries. While these aren't research registries, stakeholders can use the de-identified data for secondary research.

Meanwhile, healthcare systems and accountable care organizations (ACOs) increasingly use another type of registry for disease management, care coordination, and population health management. This kind of registry may be populated by data from electronic health records and hospital or practice financial and scheduling systems.

Researchers working on comparative effectiveness research projects could also use ACO registries, according to Berliner. "I'd love to have the data generated by all these ACOs because there so many questions we could look at using that," she said.

Unfortunately, there are barriers to using these registries in research. The registry data must be exceptionally "clean," which isn't always the case in healthcare organizations, and it must use the same definitions of medical terms that the clinical trial data does. "One of the things that makes a good registry is having good definitions and really careful data collection," Berliner explained.

In the long run, AHRQ would like to see more data in all kinds of registries generated by EHRs. Today, most of the information in clinical trial registries is collected on paper or manually entered on a website.

Get the new, all-digital Healthcare CIO 25 issue of InformationWeek Healthcare. It's our second annual honor roll of the health IT leaders driving healthcare's transformation. (Free registration required.)