The road to RHIOS: Health care IT networks on the horizon

The availability of health data has implications for individual patients, health-care systems and policymakers, yet despite advances in information management, patient health records to a large extent are still scattered and difficult to retrieve. A research team at the W. P. Carey School of Business created a database of patient information that aggregates information about disease and treatment from physicians, clinics and hospitals without compromising confidentiality. The database has been useful to health-care planners and public health researchers — and is a model of local innovation as the nation moves toward the regional health information sources.

In the nine months since he was named national health care IT czar, Dr. David Brailer has talked himself hoarse urging hospitals to computerize patient medical records, then link electronically with one another, local clinics and doctors' offices.

If American health care honchos listen to Brailer, we'll be able to find out our latest cholesterol numbers or Pap test results as quickly as checking e-mail. Even better, since they'll be in an electronic rather than a paper file, our records will follow easily as we switch doctors or insurance plans, move out of state or even land in an ambulance while vacationing.

Unfortunately, most communities are light-years away from being organized into the "regional health information organizations" (RHIOs) — local electronic networks — that will help turn Brailer's vision into reality.

AZHQ launched in Yuma

But an ambitious database project conceived by William G. Johnson, economics professor in the W. P. Carey School of Business health management and policy program, may jump-start local health care providers. Called "Arizona HealthQuery," the project may eventually bring "inter-operability" to Arizonans far ahead of other states.

The database already contains 100 million medical records from health facilities in Yuma — the location of the original project where the project's efficacy was proven — and from a dozen Phoenix area hospitals, several clinics, urgent-care centers and physician practices, AHCCCS (Arizona's Medicaid system) and county health departments.

Every quarter, Arizona HealthQuery participants — called "data partners" — pony up disks or tapes with updated patient-care records. Using software designed for HealthQuery, Johnson's research team adds the new records to the database, then generates reports forecasting or tracking health trends. Thanks to the continually updated medical records, the trending information is recent and fresh.

"Say a hospital wants to know how many people under age 18 were treated for diabetes in the last year. We provide that information — not just on their patients but for all medical records at other hospitals in the database," explains Mary Rimsza, a physician and co-investigator with Johnson. "There's an 'a-ha!' moment when hospital administrators realize they can use it to also get XYZ — some special project — done."

Data partners can't access the firewall-protected database themselves, but get a slew of reports at no cost, and can request customized reports for a fee. They also retain ownership of their own medical records, although Johnson and Rimsza may use the database for their community-wide research, which so far has included studies on childhood asthma, hospital readmissions, uninsured care, mental illness, diabetes and thyroid disease.

"We see patients when they're very sick or having surgery. Typically, we don't know what happens to them otherwise — if they seek home health care, or end up in another emergency room, or are seeing their regular doctor for something," explains Wendy Lyons, vice president of Scottsdale Healthcare (two hospitals, 570 beds). "Arizona HealthQuery lets us track patients across their lifetimes with all health encounters, or at least all health encounters with others in the database."

Database is growing

While it's free to participants, someone must foot the bill. Sponsors - Arizona State University, Flinn Foundation and St. Luke's Health Initiative in Phoenix — have poured $1.1 million into HealthQuery. But Johnson says the database is growing so quickly due to new data partners that he and Rimza are trolling for more grant money.

The HealthQuery is still far from being a RHIO, though, according to Dr. John Hlamka, chief information officer at CareGroup Healthcare System (six hospitals, 700 beds) based in Boston, Mass., and CEO of Ma-Share, a RHIO founded in 2003. But Arizona HealthQuery founders are on the right track. "A RHIO should start with things like an administrative data — an aggregation of databases that can be used by the community. That's easier than sharing clinical data, and brings the community together," Hlamka says.

"However, once you've done that — especially once trust among hospitals and other providers is established — you can evolve into a true, real-time, transactional data exchange of clinical data, as Brailer envisions," Hlamka adds.

Johnson wasn't trying to build a RHIO five years ago, when he initially asked a rural hospital, 281-bed Yuma Regional Medical Center, to share its medical records. To protect patient privacy, identifying data is removed (name, social security number) and replaced by a master-patient index number. The ASU-Yuma Regional collaboration planted seeds that blossomed into Arizona HealthQuery in 2004. As more hospitals, clinics and physician practices join, the database deepens, spawning more and richer research opportunities.

New federal laws guarding patient privacy (HIPAA) mandate "business-associate agreements" between entities sharing medical records — in this case, HealthQuery and its data partners. The agreements dictate tight security practices intended to prevent accidental or criminal disclosure of confidential patient information. There's no chance of hackers intercepting the private information during electronic transmission, Johnson notes, because HealthQuery staff collect the tapes and disks through "face-to-face hand-offs."

Rimsza and Johnson are the first to admit their model isn't complete, or perfect. In fact, the researchers recently learned about the RHIO model of data sharing while pitching HealthQuery to hospitals. They would be asked, "Is this a RHIO? Because we need to join or form a RHIO."

Something similar is happening at the University of Mississippi Medical Center (611 beds) in Jackson, Miss., where chief information officer Mike Smith says, "A year ago we didn't know we were working on a RHIO — I think a lot of people don't. Our emphasis was on putting together a public health surveillance project."

It began when UMMC officials decided to electronically pull real-time data on infectious disease from the hospital's departments — admissions, ER, OR, pharmacy and lab — and stream it to the state health department. To get it done, Smith bought and implemented clinical decision support system software. Later, he plans to hook up with other hospitals in Mississippi, eventually forming a full-fledged RHIO.

Goal is real-time patient data

Most of the RHIOs up and running have a crucial difference from HealthQuery; they operate without a central, shared database instead creating an "electronic pointer" that forms a bridge between existing networks of member hospitals, clinics and physician practices.

This way, members can actually find a particular patient's specific, private medical records. This is the other functionality Brailer is pushing, because an inter-operable society saves millions by reducing duplicate tests and cutting hospital labor costs. There's also an overriding humanitarian aim: saving lives.

How? With the right software attached, electronic medical-record systems digest streams of real-time patient data — noting the 20 mg. dose of Demerol going into your IV, or warning the nurse not to give you that same dose if your records indicate an allergy to Demerol. Your doctor can read your liver biopsy report from his laptop instead of waiting till he gets to the hospital, and ask a nephrologist in New York for a second opinion on your electronically-stored CAT scan while still at your bedside.

Forming RHIOs to create a national medical record system within 10 years, as President Bush has ordered, will cost at least $276 billion, according to a recent study published in "Health Affairs: The Policy Journal of the Health Sphere." It also will require an unlikely level of collaboration from competing hospitals, docs and insurers.

Running the system will cost $16.5 billion annually. Then there's the issue of keeping confidential medical information private at the same time that everyone from patients to insurance-claim clerks are tapping into highly sensitive databanks.

Some industry insiders — like Santa Fe, N.M., consultant Vince Ciotti — say the RHIO concept is unrealistic given marketplace pressure, to say nothing of inept data handlers or thieving hackers. "I wouldn't want my hospital's data out there publicly, so that others could observe how my physicians practice, how my case mix is managed, how many appendectomies are done, Ciotti says. "It makes no sense from a competitive business sense."

But Scottsdale Healthcare's Lyons points out that she and other data partners get only anonymous, aggregate data, such as the number of appendectomies citywide, from the HealthQuery database. "I can't ask for the number of appendectomies at St. Luke's across town, but I can get the number of appendectomies done at all other hospitals" lumped together, she explains.

Is it safe to share confidential patient records with a non-medical entity such as HealthQuery, a product of academic researchers? Yes, says Peggy Stanborn, vice president of business development at St. Joseph's Hospital and Medical Center (520 beds) in Phoenix. "I'm actually very confident about this. I've worked with Dr. Johnson on other projects over the years, and know the efforts they go to for data security and integrity — they do the best that it's possible to do."