For more than 2 decades, I have worked in a single-payer, nonprofit health-care system run by the Department of Defense. As a primary care physician in this system, I do not have to deal with health insurance billing and do not have to worry about whether or not my patients can afford the medications I provide. When I admit patients to the hospital, I don’t have to worry that they will lose their house or their children will no longer be able to go to college. If my patients need to stay in the hospital a few extra days, I don’t have someone hassling me. When I discharge someone, I know they’ll be able to be seen by an appropriate physician soon. Despite this freedom, my system’s length of stay, per capita health-care costs and readmission rates are much lower than national averages. In my system, the gap between primary care physicians and specialists is small; we are all comfortably middle-class.

While it has not always been so named, our system has successfully deployed a Patient-centered Medical Home (PCMH). As a result, we have been able to improve access and decrease unnecessary utilization. Technology has allowed us to get data to patients quickly, and has permitted patients numerous ways to communicate with their primary care team. We have decreased the costs that fall within primary care’s domain, such as ER visits and subspecialty visits. Both patients and providers have found the PCMH works well.

This issue of JGIM features articles from a recent summit on the PCMH that aimed to identify the research questions that need to be asked as this new system develops. Many important questions were identified in this conference. The answer to the most fundamental question of exactly how a PCMH is defined has still not been answered completely. We are in the midst of a natural experiment, where many different versions of the PCMH are emerging. This provides the opportunity for it to evolve to meet local consumer and market needs, but also the chance to compare different systems and decide which features work best. It is also unclear how to incorporate academic medicine in the PCMH model, both to train providers how to operate in this environment and how to involve learners in the PCMH itself. SGIM is in the process of planning a summit on educational issues and the PCMH.

A fundamental question is whether or not the PCMH will help both to hold down costs and to support better care and improved patient outcomes. I hold great hope for the latter two issues, but little for the former. The forces that are driving increased health-care costs largely fall outside the realm of the PCMH. Primary care is not responsible for the majority of health-care costs, and a new model of primary care health delivery will not fix the other forces that are driving up costs. The PCMH can offer much, but I worry that if we promise too much, PCMH will appear to fail, even if it meets its patient-centered promises.

Even in the system in which I work, we struggle daily with many of the issues raised by this collection of articles on the PCMH. How do we incorporate learners and exactly what are the skills necessary to work in this new environment? How do we increase patient access to information, yet assure privacy? What elements of our PCMH work and which don’t, and what changes do we need to make to continue to improve? How do we continue to change our system and mind-set from one that is centered on the clinician’s needs to one centered on patients? The PCMH is an exciting, visionary concept, and it will be interesting to see how it evolves. Paying attention to and developing systems to measure the consequences and outcomes of the PCMH is critical, and the papers in this issue of JGIM point us in the right direction.