Where research and practice meet

June 2, 2006

Practicing physicians are giving feedback to the academics whodevelop clinical guidelines. Will they listen?

At a time when physician adherence to practice guidelines is being examined under a microscope, the ability of physicians in direct patient care to contribute to research on relevant clinical topics is more critical than ever. That's what practice-based research networks (PBRNs) facilitate. PBRNs are designed to pose questions and find answers that are relevant to everyday practice. In this respect, say PBRN advocates, their work can advance quality improvement more effectively than studies conducted in academic settings with narrowly selected patients.

Thousands of primary care physicians in these independent networks are trying to make sure that the quality measures upon which they are judged make sense. As they attempt to find ways to implement clinical guidelines, they're field-testing them and advising the academicians who developed them about unforeseen problems and shortcomings in the protocols.

For example, the American Academy of Pediatrics' PBRN conducted a study on febrile infants to find out how pediatric practice in office settings differed from the guideline for that condition. Later published in JAMA, the prospective cohort study found that pediatricians who simply followed sick children closely for a few days had outcomes that were just as good as those who complied with the guidelines, which stress testing, antibiotics, and hospitalization.

A similar kind of discovery emerged from a study of a congestive heart failure guideline in an upstate New York PBRN, says FP Paul James, professor and head of the department of family medicine at the University of Iowa College of Medicine. James, who conducted the trial, says that the guideline of the Agency for Healthcare Research and Quality (AHRQ) was written for patients with systolic heart failure, but 55 percent of the CHF patients whom the family physicians in the mostly rural network were seeing had diastolic heart failure. That was because these patients were 13 years older, on average, than the patients in the studies that underpinned the protocol, he says. Since his study came out in 1999, notes James, the American College of Cardiology has revised its CHF guidelines to reflect the prevalence of diastolic heart failure.

PBRN studies have revealed other problems with guideline implementation. Researchers interviewed physicians participating in the Oklahoma Physicians Research/Resource Network, for instance, about why they hadn't controlled blood pressure in all of their diabetic patients. They found "many interesting and valid reasons," observes FP James Mold, a professor in the Department of Family and Preventive Medicine, University of Oklahoma Health Sciences Center.

"We've done a fair amount of work to find out what happens when physicians try to follow these guidelines," adds Mold, director of the PBRN. "It might be helpful if the people who are developing future versions of the guidelines knew some of that information so they could revise them appropriately."

Long history, recent expansion

Practice-based research networks, which are found all over the world, first emerged in the US in the 1970s. But it wasn't until 2000, when AHRQ increased its funding to encourage the formation of new networks, that PBRNs really took off.

Today, roughly 10,000 physicians participate in these research networks. AHRQ has counted 111 US networks, including four for nurses. Of the 12,000 clinicians in the 81 networks registered with the AHRQ-funded PBRN Resource Center, 85 to 90 percent are primary care doctors.

PBRNs are either statewide, regional, or national in scope. Examples of the latter include the AAFP National Research Network, which includes more than 300 family physicians in 45 states; the AAP's Pediatric Research in Office Settings (PROS) network, which encompasses around 2,000 physicians and midlevel practitioners; and the Practice Partner Research Network (PPRNet), which embraces 534 clinicians (most of them doctors) in 37 states. The distinguishing feature of PPRNet is that all of its members use the same electronic health record system.