The patient-centered medical home (PCMH) is not a pill. It would be much easier to evaluate this primary care reform if it were.
In this issue of JAMA Internal Medicine, Rosenthal et al1 report their evaluation of the pilot program of a patient-centered medical home model of primary care at 5 practices supported by a consortium of 3 Rhode Island health insurance plans. Findings on patient outcomes were mixed. During the 2-year postintervention period, patients in the pilot practices had 11.6% fewer emergency department visits that were potentially avoidable compared with matched control patients in other practices, with a nonsignificant decrease of 5.2% in total emergency department visits. A similar trend was observed for hospital admissions, with nonsignificant decreases of 15.1% for potentially avoidable admissions and 2.9% for total admissions. No significant differences were found in diabetes care and cancer screening, although colon cancer screening rates increased by a sizable but not significant 37% during the postintervention period.
Primary care advocates may conclude from this study that a significant reduction in potentially avoidable emergency department visits and strong trends in favorable outcomes for avoidable hospitalizations and colon cancer screening add to the evidence that supports redesigning primary care. Skeptics may conclude that the lack of significant differences in quality measures and overall use is evidence of the lack of effectiveness. Thus, it is important to place in context evaluations of the patient-centered home model.
If the patient-centered medical home were a pharmaceutical product, research to justify Food and Drug Administration approval would need to demonstrate safety and therapeutic benefit. By this standard, the substantial efficacy of the patient-centered medical home for colon cancer screening and the lack of reporting of any patients in the 5 practices who had bad care experiences would support approving the patient-centered medical home as an efficacious and safe product. This conclusion would be bolstered by a recently published systematic review of patient-centered medical home evaluations that found “a small positive effect on patient experiences and small to moderate positive effects on the delivery of preventive care services.”2(p175) As is the case for new medications, expectations that health plans pay for practice redesign would follow soon after regulatory approval.
No such luck for the patient-centered medical home. It is not enough for it to be nonharmful and to demonstrate some degree of efficacy. In many quarters, the patient-centered medical home is judged on whether it is a “home run” for the ills of the US health system3—specifically, a 3-run homer achieving the triple aims of better health, better patient experience, and lower costs. Reducing overall costs is a paramount aim for the Centers for Medicare and Medicare Services and other payers. The ability to generalize payment reform for primary care beyond the current pilot and demonstration projects largely depends on the ability of the Centers for Medicare and Medicare Services to convince the actuaries at the Office of Management and Budget that the reforms cost the federal government no more—and ideally less—than would otherwise have been spent.
Clamor for the reforms that are part of a patient-centered medical home reflects the broad recognition that health systems do not function effectively or efficiently without a strong foundation of primary care and alarm that this foundation is crumbling in the United States. Revitalizing primary care requires an infusion of resources and changes to care delivery.4 Of course, primary care is deprived of resources and undervalued and consumes only approximately 6% of total health care spending in the United States. So, it is legitimate to question the reasonableness of making an investment in the revitalization and retooling of primary care contingent on promptly “bending the cost curve” for the remaining 94% of health care spending.
Those who evaluate patient-centered medical home interventions contend with finding sensible methods for studying complex changes in health care delivery in the dynamic setting of community practice.5 A pharmaceutical product can be manufactured with uniform specifications and delivered in a standardized manner. The patient-centered medical home, however, is a multifaceted intervention. It involves changes in the organization, structure, processes, culture, and financial model of a practice. The diverse components include redefining roles, responsibilities, and tasks for everyone from the receptionists to the physicians; reengineering patient scheduling templates; performing proactive outreach to high-risk patients; and adopting payment models that blend fee for service, capitation, and pay for performance. Practice transformation has more in common with continuous quality improvement than a rigid clinical trial protocol. There are inevitable compromises between fidelity to the prescribed patient-centered medical home model and adapting the model to the particular circumstances and context of different practices. Things get messy.
To evaluate practice interventions, such as the patient-centered medical home, researchers often must capitalize on natural experiments over which they have little control. A consequence of evaluating a natural experiment is that studies are frequently underpowered to detect clinically important effects, as was the case in Rhode Island. Quantitative meta-analyses that integrate findings across studies may compensate for the limited power of individual studies, but heterogeneity among studies of patient-centered medical homes often precludes such an approach.2 Moreover, in the study by Rosenthal et al, 2 years of follow-up may have been too short for full implementation of the intervention and for evaluating the intervention’s effects on outcomes. No doubt Rosenthal et al would have preferred to study more than 5 practices and to continue the evaluation beyond 2 years of follow-up.
When all the available information is imperfect, organizations must make strategic decisions. Decision makers use a collage of scientific evidence, case studies, and their own hunches. For example, the Department of Veterans Affairs (VA) has a remarkable trajectory of quality improvement during the past 20 years. A critical ingredient has been the VA’s investment in strengthening primary care and implementing its own patient-centered medical home model known as the Patient Aligned Care Team.6 In 1994, when the directive committing the national VA system to primary care transformation was issued, leaders of the department did not wait before acting for a pooled P < .05 from a meta-analysis of patient-centered medical home evaluations. The sponsors of the Rhode Island patient-centered medical home initiative likewise seem to consider the external evaluation by Rosenthal et al and their own internal tracking data sufficiently compelling to commit to supporting 20 additional practices per year for the next several years.7
Research on primary care transformation in the United States is important, using methods of evaluative and implementation science suitable to the complexity of the topic. The information generated provides critical formative feedback to those implementing patient-centered medical home models and an overall sense of whether the movement to reform primary care is headed in the right direction. Policymakers, however, will do the public a disservice if they wait for incontrovertible scientific evidence that the patient-centered medical home is a magic triple aim pill with a large and immediate financial return on investment. The patient-centered medical home is neither a pill nor a wonder drug. On the basis of the available evidence, insurers, the states, and the federal government should act to revitalize the indispensable primary care foundation of our health system.
Corresponding Author: Kevin Grumbach, MD, Center for Excellence in Primary Care, Department of Family and Community Medicine, University of California, San Francisco General Hospital, 1001 Potrero Ave, Ward 83, San Francisco, CA 94110 (firstname.lastname@example.org).
Published Online: September 9, 2013. doi:10.1001/jamainternmed.2013.7652.
Conflict of Interest Disclosures: None reported.
Grumbach K. The Patient-Centered Medical Home Is Not a PillImplications for Evaluating Primary Care Reforms. JAMA Intern Med. 2013;173(20):1913-1914. doi:10.1001/jamainternmed.2013.7652