Practice-based Research Networks (PBRNs) in the United States: Growing and Still Going After All These Years

What Is the Same?

Practice-based research networks are alive and well and continue to do important work to understand and to improve primary care practice and its intersections with specialty care, public health, and our communities. There is evidence to support this assertion. In 1994 we could identify 28 PBRNs in the United States. In 2003 there were 111 PBRNs registered on the Agency for Healthcare Research and Quality (AHRQ) website, and now there are 169. These networks are conducting research studies and publishing their work. Although the list is likely incomplete, the AHRQ PBRN Research Bibliography lists more than 100 publications from PBRNs every year since 2005, peaking at 220 in 2010 (http://pbrn.ahrq.gov/pbrn-literature).

There is still great need to understand the epidemiology of primary care practice and person-oriented outcomes of care. The studies in this issue of JABFM display the continuing importance of descriptive research using the naturalistic, observational opportunities inherent in frontline practice derived from knowing the patients and their problems and observing how health is won and lost. The new national Swiss network is a reminder that sentinel surveillance of primary care practice remains a critical function for PBRNs.⁵

There remains a tension concerning generalizability of results from PBRN studies about phenomena as they occur in practice, with built in selection and observer bias. PBRNs study health care as it unfolds daily in real-world practices in diverse communities. PBRNs thrive on their own selection biases—it is part of what makes them (and their pragmatic trials) important. Traditional randomized trials also have difficulty with generalizability because trials often exclude most of the population for which results are intended. PBRNs continue to wrestle with translation of knowledge acquired elsewhere to the frontlines of health care. People still wait for care that matches their situations.

What makes a PBRN a PBRN is the same now as it was in the beginning. The essence of PBRNs is that the relationships formed among the investigators, network administrators and practitioners and their office staffs are expected to endure over time and independent of the topic of interest, with active participation of practice members at all stages of study development, implementation and reporting.

Many of the conditions PBRNs study have stayed amazingly consistent and constant for more than 30 years. This is no surprise because the primary care problems of the people served do not change much over time. Primary-care practices are continually confronted with acute illness, chronic disease, mental health issues, and the social determinants of illness. Kudos to Shared Networks of Collaborative Ambulatory Practices and Partners (SNOCAP), who demonstrated that a PBRN can improve control of diabetes and hypertension.⁶

Some studies in this issue of JABFM are not from PBRNs, but they required practice-based research to get to results critically important to practice and policy.^7,8 As has been the case for centuries, not all primary-care research requires a PBRN. PBRNs are still essential laboratories for primary care research but not the exclusive venue.

Unfortunately, PBRN research and primary-care research in general remain fragile and woefully underfunded. Nearly every year, including 2015, the U.S. Congress threatens to shut down or severely limit funding of the AHRQ, the only federal funding source that has taken an enduring active interest and leadership role in PBRN research. Intrepid PBRNs and primary care researchers march on, doing important, desperately needed work on shoestring budgets.

What's New?

Two events have profoundly affected the research agenda of PBRNs: publication of the companion reports, To Err Is Human⁹ and Crossing the Quality Chasm¹⁰ in 1999 and 2001, and the passage of the Patient Protection and Affordable Care Act (ACA) in 2010. Before these reports, practice systems improvement was not so high on the agenda of PBRNs, but it has come to the forefront as the transformation of U.S. primary care practices has proceeded to become safer, more timely, more patient centered, more effective and efficient, and more equitable.

The ACA has increased demand for robust primary care and simultaneously accelerated the consolidation of large health systems and the employment of primary care clinicians by these large health care systems. These have been a double-edged sword for PBRNs. Time that once could be spent on practice-based research has often been trumped by the immediate needs of increased patient-care demands. Decision making about PBRN studies changed, including a loss of autonomy at the local practice level that can create another drag on PBRN research. It can be more difficult to convince large health care organizations rather than local, small- or medium-sized practices of the value of participating in studies to understand and to improve primary care practice. And, there are often challenges in decision making about studies and implementation of studies secondary to consolidating entities, electronic health records (EHR) vagaries, and big systems with a spectrum of possibly conflicting goals.

In contrast, if a large health care organization “gets it” and values practice-based investigations, additional resources may be available as part of core business to support projects seen as quality and practice improvement. Of course, large health care organizations have many patients and hundreds of thousands of individuals' health records in electronic format, providing an ideal opportunity to study the epidemiology and outcomes of primary care practice. Once the full potential emerges for recording, linking, and aggregating data on individuals, their signs and symptoms and conditions, their communities, care processes, and outcomes, EHRs and associated data warehouses will provide powerful tools to accelerate advanced PBRN research. Emerging vanguard national networks such as the Practice Partners Research Network, DartNet, the Health Care Systems Research Network, and others are developing and using new tools to extract and aggregate data residing in electronic health records to study primary care, the largest platform of health care delivery on a grand scale.

Another new development is the use of increasingly sophisticated and varied scientific methods by PBRNs. Analyzing and making sense of large, practice-based data sets requires the always evolving analytic methods to cope with bias and error. Large, pragmatic trials with few exclusions and straightforward outcomes can be accomplished at lower cost in large, electronically linked networks, leading to greater external validity.

Regardless of large system or small practice, implementing and sustaining high-quality care for all patients is now at the center of PBRNs' research efforts. Many are shifting efforts from a single disease focus per study to practice systems transformation to improve care across a wide range of conditions. An example is the Fernald et al study,¹¹ which shows PBRN persistence through a 3-stage project yielding a toolkit to improve testing: to move the dial on hypertension and diabetes measures. This same study also reminds us how important it is to map practice processes and how difficult it can be to complete practice-based research. Thirty-one practices were recruited, 24 consented, 22 actually participated, 20 did something, and only 4 practices were able to complete the 6-week intervention for quality improvement. This reported experience provides additional evidence for the need for a “primary care extension service” to support practice transformation, as proposed by Grumbach and Mold.¹²

A welcome new trend on the PBRN scene is the inclusion of patients as research partners, pioneered by Jack Westfall and colleagues¹³ at the High Plains Research Network. This network's patient advisory council formulates questions with investigators, designs studies together, interprets results together, and sculpts dissemination of results so they get to who needs to know them. This is a truly patient-centered approach. In this issue, the study by Goodman et al¹⁴ to validate a simple literacy measure in primary care is evidence of the recognition that patients must be partners in their care.

The Future: Speculations

Concluding Comments

Once again we see the stunning complexity of primary care and the meagerness of the infrastructures to discover and support it. The resilience, adaptability, and continued expansion of PBRNs is on display in this issue of JABFM. This is good news for those seeking transformation of health care such that it actually matches patients' goals and needs and improves both individual and population health. It is obvious that PBRNs remain an important part of the U.S. research landscape. Those clinicians, patients, and investigators working within them have established a way of life that is adaptive, underappreciated, and rich with opportunity. Surely the time when proper financial support for practice-based research becomes part of practice business models and a priority for the medical research enterprise is closer than ever. It cannot come too soon.