Abstract
We have quite a rich issue this month related to practice-based research networks (PBRNs)—reflections on where they have been, where they should go, how they should happen; lessons learned about recruiting physicians and patients and new research methods; and several clinical studies from existing PBRNs. We had an amazing number of manuscripts submitted this year for the PBRN issue; as a result, this is a powerful issue. Some are under revision for future issues of the Journal of the American Board of Family Medicine, just as we have some articles from PBRNs appearing in most issues. PBRNs have deepened the family medicine research tradition. The importance of primary care research to build the evidence base of our clinical practice, plus the useful work building the methods of primary care research, distinguishes the pioneers in PBRNs. PBRNs are Health Improvement Networks and national treasures to be nurtured.
The issue starts with the idea, proposed by Williams and Rhyne et al,1 that the practice-based research networks (PBRNs) have matured and evolved and now really should be thought of as Health Improvement Networks (HITs). This is an idea whose time has come. It is also clearly substantiated by another article from Williams et al2 that shows practice improvement after participation in PBRN studies and can also be seen in other papers in this issue including those by Daly et al3 and Parnes et al.4 This improvement has strong face validity in that the process of doing the research and thinking about the topic should help physicians incorporate their research experiences into evidenced-based practice for the benefit of their patients.
The PBRNs have been advanced through the careful nurturing of the Agency for Health Care Research and Quality (AHRQ). David Meyers, a former student of mine (MAB), writes a tribute to Dr. David Lanier, a former faculty colleague of mine (MAB), who pushed the idea of the need for PBRNs at AHRQ. We are truly blessed with such wonderful colleagues. We also appreciate the fact that AHRQ has hosted PBRN national conferences and provided substantial funding both for infrastructure and for the ensuing studies. Nevertheless, improvement is still possible. Pace et al5 critique the current partnership between AHRQ and primary care researchers with propositions for what could further enhance the future of the field.
Clinical Research From PBRNs
There are several clinical papers from PBRNs; we selected two clusters: pain (n = 2) and skin infections (n = 3). There is one on low back pain6 from a residency research network in Texas and one on chronic pain and the difficulties of its management in practice.7 Perhaps not surprisingly, a substantial minority of primary care clinicians in these studies had decided against prescribing narcotics because of the great difficulties related to their prescribing, use, and misuse. The three on skin infections (from Iowa,3 from North Carolina, Texas, and Colorado and DARTNet4; and again from Texas8) encompass a large number of groups across the country. Different although similar interventions were associated with increase in appropriate choices for methicillin-resistant Staphylococcus aureus (MRSA) skin infections. We also include one from Oklahoma9 on whether older patients' assessment of physician quality care is associated with medical outcomes.
The Methods of PBRN
This section starts with an article on a logic model for PBRN research,10 outlining well the multiple aspects of the assumptions, inputs, activities, outputs, outcomes, and outcome indicators. The process of creating the logic model also helped the PBRN develop through engagement of the multiple involved parties. Devoe and colleagues11 present the amazing work of pulling together a large community health center network that started in Oregon but has expanded and now includes more than 40 safety net organizations serving over 900,000 patients with nearly 8,000,000 distinct visits. Hamilton et al12 highlight a method for creating efficiency in institutional review board approval for card studies commonly used by PBRNs. We have an article on patient recruitment (small dollar incentives attracted more initial interest but not sustainability13) and two on physician recruitment (No surprise: Money works better—as does recruitment by a recognized leader/clinician champion, plus other great hints) from NorTex, a northern Texas primary care PBRN.14,15 These “technique” papers include one on getting an unbiased patient sample.16 Another highlights two quasi-experimental methods—a stepped-wedge design, and wait-list crossover design—that seem particularly useful in PBRN practices.17
This rich—and enriching—issue is our salute to our many PBRN colleagues advancing primary care research in family medicine.