The Cost of Sustaining a Patient-Centered Medical Home: Experience From 2 States

Thanks to Dr. Kralewski for his insightful comments. The practices were a convenience sample of NCQA PCMH Level 3 recognized practices (Colorado) and practices with a decade of experience implementing a local PCMH model (Utah). They were chosen to represent a range of practice types, including FQHCs, private practices, and university-owned clinics. We did not address substitution effects, but only estimated time per NCQA PCMH element and standard at one point in time. Substitution effects could be a useful component of future work.

Our "cost per clinician" calculations included PAs, NPs, and physicians as "clinicians," but we recognize that PA/NP roles vary by practice. Some serve as physician "extenders," while others have their own panel of patients. We agree, as stated in the discussion, that future research should examine cost variation by practice type and staffing to identify the most efficient practice models. While space did not permit including it in the published paper, we are currently preparing such an analysis to report separately. This analysis may provide insights into costs of different NP/PA roles.

Our data did not permit calculation of revenue or net financial performance of the practices. However, our estimates of potential revenue from alternative payment models suggested at least the possibility for financial margin from implementing PCMH. These may be fruitful topics for future work, which also should include a larger sample, ideally selected to represent the distribution of types and locations of practices implementing PCMH.

Competing interests: None declared

This manuscript provides important information needed to advance the debate over the costs associated with the implementation of medical practice improvement initiatives. It would be helpful to know more about how the "high-functioning" practices were selected and whether financial performance was included and how substitution was accounted for when new hires replaced some work done by physicians and nurses but overall it is a good addition to what is known about the costs of sustaining these innovations.

My main concerns are that the authors didn't go far enough in the analysis of their data set. Two additional opportunities are evident. First there is considerable variation in the costs associated with maintaining the standards and it appears that some practices are able to accomplish this at a much lower cost than others. How this is achieved is an important issue and should be explored. If some practices have found a better way to implement and sustain these standards it would greatly enhance the adoption rate by others.

Second the authors should re-analyze the costs per FTE provider and PMPM when the contributions of the nurse practitioner are included. In some of our studies the NPs had panels of 1500 to 1800 patients. I think that the cost picture would look quite different if they are included in the analysis. In fact if the costs drop to around three or four dollars per patient visit it would be in the range of the differences expected in clinic payment rates from the health insurance plans. Could it be that some of these practices are in fact making money by adopting these standards? Some data on net revenue and net revenue after operating costs in these practices would be helpful.

Competing interests: None declared

On behalf of the Patient-Centered Primary Care Collaborative (PCPCC), I want to commend and thank the authors for this valuable contribution to the literature. As we work to transform the health care system for the benefit of multiple stakeholders, including primary care providers and the patients and families they serve, we need to better understand the details regarding the cost of sustaining the patient-centered medical home (PCMH) in primary care practices. This study provides a good step toward quantifying the cost.

As the authors noted, the cost for sustaining PCMH functions in these 20 practices in 2 states averaged $4 per member per month (PMPM), which is consistent with other reports. Although PCMH transformation and sustainability can be expensive and time consuming, we need more research that describes the cost to practices relative to the total cost of care savings for payers and purchasers. We are hopeful that federal and multi- payer initiatives will provide us with this further research. The CMS Multi-payer Advanced Primary Care Program (MAPCP) and the Comprehensive Primary Care (CPC) initiative, which include PMPM payments to select practices delivering advanced primary care, will allow researchers to examine total cost of care savings relative to PMPM investment. This will be a critical next step to fully valuing primary care and assessing the capacity for upstream funding to impact downstream costs.

In the meantime, the PCMH model continues to demonstrate savings, but primary care teams often do not have access to total cost of care data. Our Annual Review of Evidence report demonstrates impressive cost savings in peer-reviewed studies, state or multi-payer evaluations, and industry reports. In addition, our Primary Care Innovations and PCMH Map demonstrates cost, quality, and health outcomes from initiatives across the country. The evidence continues to build - but comparing apples to apples across PCMH initiatives remains a challenge.

As others have noted in the comments, this study further demonstrates the limitations of fee-for-service (FFS) payment in terms of PCMH sustainability. All the practices studied were paid according to the antiquated and wholly inadequate FFS model. It is no secret that practices operating under FFS payment are not adequately paid for PCMH functions (such as communication with patients via phone and email, group visits, population health and quality improvement strategies, etc.) and instead are incentivized to increase the volume of services provided. As cited in a recent JAMA Internal Medicine study, practices that are incentivized to focus on cost containment and rewarded with shared saving were able to hold down costs AND improve quality of care. Practices in this study generally had no financial incentive to focus on value versus volume.

Primary care not only needs and deserves a pay raise, but we must also shift payment to focus on value. The current RVU-driven FFS payment system undervalues primary care and underscores our strong push for risk- adjusted comprehensive primary care payment. As we migrate toward this model of payment, practices like the ones referenced in Utah and Colorado will be incentivized to provide team-based high quality care to their patients with fewer FFS billing, coding, and administrative hassles through prospective alternative payment models that reflect a greater investment in primary care while controlling overall total cost of care.

When we look at the larger picture, this study also speaks to the need for clear ways to define, measure, and value the PCMH. The authors acknowledge that none of the 20 practices had fully implemented all the PCMH functions at the time of the study, and PCMH services were defined subjectively. It will become increasingly important to standardize what PCMH is for the purposes of value-based purchasing in light of impending Medicare and private sector payment reform.

We must also remember that the value of PCMH has different meaning to patients, payers, and health care providers. The authors quantify the cost of sustaining a PCMH but acknowledge their major goal is to improve patient care and health outcomes, built on professional values and team- based care. At the end of the day, measuring the value of advanced primary care for those who pay for these services - patients, families and consumers as well as health plans and employers - is imperative.

The Patient-Centered Primary Care Collaborative strongly advocates for payment and policies that advance primary care in meaningful ways. This study supports our efforts by giving us valuable insight into the ongoing cost of medical home sustainability. While there is a need for further research and this study's small size limit the results, the study's finding that a $4 per member per month (PMPM) payment is no surprise - and actually supports the need for an even greater investment in primary care given the documented savings to total cost of care.

Competing interests: None declared

This research is a significant contribution to primary care in the US. Sitting in a payer owned desk chair right now typing this I can tell you that payers seem to have realized the initial cost of PCMH recognition but are putting a lot of pressure on us to pull back on the supplements PCPs once they are recognized. There is a reluctance to pay for the ongoing costs of sustaining a PCMH operation, in part because those costs are so difficult to measure and identify. Specialists threaten to withhold services if they are not paid but our PCPs will not discharge patients from their practice simply because the patient's insurance is not supporting their services. The conundrum is that the family physician's ongoing loyalty to their patient panel precludes them from acting in a way that supports only the physician's self interest. Many specialties don't have a stable panel of patients so withholding services in a financial dispute is much more abstract, yet it is the rare specialist who would abandon a patient under their care. Now we can at least say that these are not expenses due to practice inefficiency but rather established, recognized and quantified expenses that are essential to giving the best care we can to our patients. Thank you!

Competing interests: None declared

Thanks to Drs. Wergin and Brady for their thoughtful comments. Dr. Wergin places our findings into appropriate context toward payment for value from advanced primary care. Dr. Brady's estimate of national cost to sustain PCMH may be in the right ballpark. However, even so it could be a great investment. The US total spending on health care is approaching $3 trillion (http://www.chcf.org/publications/2014/07/health-care-costs-101). A reduction in cost of just 1% would mean savings of $30 billion. Even if Dr. Brady is correct that the ongoing investment in PCMH totalled $10b, net savings would still be double that - even more if the higher percentage savings documented in PCMH studies were realized across the board.

Competing interests: None declared

Thank you to these authors for their hard work on this important analysis. Adding in Ho and Antonucci's article on the cost of initial certification (http://www.annfammed.org/content/13/3/269) being around $50,000+, we now have a clearer picture of the cost of PCMH transformation both to practices and to our society. To independent practices, spending $100,000 is financially unsustainable without a huge boost in reimbursement. To society, if one assumes there are 100,000 practicing primary care providers each paying $100,000/year to maintain certification, the end result is a staggering $10 billion/year in added costs. Unless the improvements in quality can produce at least that much in savings, and I have yet to see any studies which give me hope that it will, the net result is higher costs and more consolidation.

Competing interests: None declared