Meet the Mother of All Patient Registries

The Agency for Healthcare Research and Quality is building a massive patient database to help healthcare providers identify effectiveness projects years before researchers get around to publishing their analyses.

When news of this federal project first popped into my inbox last week, it sounded like government redundancy at its best:  The Registry of Patient Registries, or RoPR, for short, was now available online, it said.

Taxpayer dollars at work, creating the mother of all patient databases? Now that's an ambitious effort.

But what the Agency for Healthcare Research and Quality is trying to develop, I now see, looks like a terrific idea. It has rich potential to eventually help providers more quickly identify effectiveness findings for myriad therapies and procedures around the world, years before researchers and peer-reviewed medical journals get around to publishing their analyses.

In an interview last week, I asked Elise Berliner, AHRQ's Technology Assessment Program Director and RoPR's project manager, to explain what RoPR is and what she thinks it can eventually be.

"Right now, we really don't know what the registries are, and who the owners of those registries are. There are many that you've heard about, but lots and lots that you haven't," she explains.  AHRQ wants to make a list.

Berliner says she thinks that worldwide, there are potentially "thousands" of registries that might be included, though only 63 are loaded into RoPR to date.

One registry included in the first rollout looks at catheter infection and rehospitalization rates for people who use parenteral nutrition at home with five years of follow up. Organized by the American Society for Parenteral and Enteral Nutrition, it has obvious implications for healthcare costs.

Another registry collects data from 5,000 rheumatoid arthritis sufferers to see which medications do a better job. This registry, run by Consortium of Rheumatology Researchers of North America, is enrolling 3,000 people in an effort to learn why patients start or stop drugs, and is monitoring changes in their ability to perform daily activities.

The Society of Thoracic Surgeons and the American College of Surgery, which keep the largest databases of heart procedures in the nation, are running the Transcatheter Valve Therapy Registry. This registry is gathering information on valve replacement procedures performed on 15,000 patients to examine safety and "real-world outcomes" such as major cardiovascular events, life-threatening bleeding, mortality, or other adverse events within 30 days. 

These registries, of course, only exist via the voluntary efforts of physician and hospital participants, who usually pay a fee to enroll. Nearly all of what those who run these registries might learn about quality of care gets reported back to the hospitals and doctors involved.  The information is not used by payers for pay for reporting or for pay for performance programs.

And almost none of the results of the registries are public, at least not yet.

However, these registries—especially those involving surgical procedures—are starting to get noticed by lawmakers as a potential way stakeholders can more quickly assess best practices and value. 

As part of the New Year's American Taxpayer Relief Act, Congress directed the Secretary of the U.S. Department of Health and Human Services to look at the potential of registries to become a platform for quality measurement, comparison, and transparency. One might presume that in time, the strategies that appear to produce superior outcomes may become the standard, with consequences for those who don't measure up.

Is that what AHRQ is trying to accomplish with its Registry of Patient Registries?

Berliner says she's not the right person to answer that question. But, she adds, "my understanding is similar to yours. What we're doing is the first step of what you said. You're right. There are a lot of these registries that already exist. You have to know who has the data…We're just trying to connect people have the answers with the people who have the data."

Berliner says RoPR has four goals:

1. Assist registry sponsors to rapidly record in RoPR relevant information on outcomes and common data elements
2. Reduce redundancy of registries
3. Increase the value of new registries that use the standardized data elements, and
4. Identify registries capturing relevant or similar information for collaboration or linkage.

It's hard to say how quickly the RoPR project will catch on, or if it ever will.  The American College of Surgeons, which has more than 500 hospitals and hundreds of data points in its National Surgical Quality Improvement Project wouldn't say.  But an ACS spokesman says the ACS is in the process of entering NSQIP data into the RoPR database.

One tremendous potential benefit from such a registry is to tear down the Tower of Babel that various researchers are using to measure their outcomes today, and move toward a common language or format. "Today, it seems, everybody has their own definitions," Berliner says. "[There are] cases in which people use different words to describe the same thing, and use the same words to describe different things."

At a recent meeting of Academy Health, Berliner presented slides in which she gave examples.

The European Society of Cardiology's CARDS study (Cardiology Audit and Registration Data Standards) defines bleeding during a hospital stay as an event in which the patient's condition "required close monitoring and/or specific treatment, e.g. transfusion of blood or blood products."

But the American College of Cardiology's National Cardiovascular Data Registry of  Percutaneous Coronary Intervention (PCI) defines bleeding during a hospital stay as a "suspected or confirmed bleeding event" that was associated with a hemoglobin drop of 3 g/DL or a transfusion of whole blood or packed red cells.

And different organizations use different biomarker thresholds to define a heart attack after a PCI. Some registries measure outcomes by readmission rates, but differ in the timeframe, with one using 45 days and another only 30 after hospital discharge.

Currently, most, if not all, of the 63 registries in RoPR have links in the National Institutes of Health's www.clinicaltrials.gov.  But trying to find registries there would be like looking for a needle in a haystack, Berliner explains. That's why it's important to separate out those registries in a separate system.

"Our system is kind of confusing," she laughs, "but only because we think it ultimately will be less confusing." Berliner believes strongly that the RoPR database must have a curator to help organize the listings. "The idea is to take registries in the U.S., like those run by Kaiser Permanente, and those from other countries like Sweden and Australia that have really good national registries, and pull the data together."

The RoPR project was enabled by a $5 million, three-year grant from the American Recovery and Reinvestment Act of 2009, which also included a lot of stakeholder engagement efforts and produced a registry handbook for physicians and others who want to create their own registries.

The next three years will be funded for an undetermined amount through the Patient Centered Outcomes Research Institute (PCORI), Berliner says.

Now, the trick is to get more people to register their patient registries, Berliner laughs. "We're trying to reach out," she says. "Hopefully, you will write an article, and that will make some people want to register."