One of the trickiest pieces in the puzzle of medical home financial sustainability is fully understanding PCMH ROI. Case in point: recent reports claim that Community Care of North Carolina (CCNC) avoided $984 million in Medicaid health spending over four years by placing 1.1 million Medicaid enrollees in medical homes. This headline begs two questions:
How did they do that?
The success is chalked up to the medical home model, but savvy providers—who know how diverse models called “medical home” can be—know that more information is needed to explain exactly what this particular medical home is or what it does.
Did CCNC really save $984M?
Again, experienced medical home observers know that in the world of medical home, “savings” are often very challenging to quantify. Savings figures are always modeled out compared to a control group or what would have been spent—they are almost never an absolute decline in spending compared to the past. So it's always important to look at where the numbers are coming from.
Here, a quick look into Community Care of North Carolina’s particular version of the medical home, the operational changes that resulted substantial savings, and the context of how estimates of CCNC’s savings have changed over the years.
Details of the CCNC model
The main source of avoided cost in a medical home model is almost always its success in reducing downstream ED and acute care utilization. However, while medical homes usually use common general principles to target those goals (care coordination, specially supporting chronically ill patients, etc.), the way those principles are applied always varies.
From a practitioner perspective, the first noteworthy aspect of the CCNC model is that it is not a mainstream, PCP office-based PCMH as we tend to think of it. Instead, it is a set of centralized care coordination services that sit at the network level, acting as "glue" across the continuum for Medicaid patients, and support diverse PCP sites.
To create the network model, the state has divided 4,500 PCPs, 1.1M patients, and 100 counties into 14 regions. Each network is paid $3 per member, per month (PMPM), in part to offset centralized staffing costs.
An average staffing model for a given network includes:
- Medical directors: 1.4
- Local case managers: 42.8
- Pharmacists: 1.8
- Psychiatrists: 1
Source: “2011 Overview”, Community Care of North Carolina presentation, November 8, 2011
CCNC primary care sites include a mix of small, large, independent, and hospital-affiliated groups. It is interesting to note that the sites themselves may or may not have a medical home infrastructure in place, provide any medical home services, or be recognized as PCMHs.
In this model, the PCPs receive a $2.50 PMPM, which helps ensure that practices remain open to the Medicaid population. Physicians identify the patients they perceive as high-risk and in need of additional support up to the case management infrastructure for additional resources resources within the network.
Though the services are centralized and uniform across each network, different networks have evolved different internal structures based on who/what is in the network. For example, depending on whether the region includes a dominant health system or many fragmented providers, the case managers may be deployed differently based on the delivery context. The networks also develop and disseminate best practices internally, leading to the formation of slightly different delivery models.
How does CCNC avoid costs?
CCNC’s success probably stems from many different factors that are hard to boil down. However, for the sake of creating a quick overview, here are a few key features that we notice CCNC having in common with other medical home models that have demonstrated strong results in preventing health spending:
Solid risk segmentation/case management: As a general rule, the more targeted any medical home initiative is toward meeting the needs of patients considered to be a high-risk for ED use or hospital admission, the greater the up-front payer-level ROI will be.
The CCNC model identifies and supports high-risk patients in the population by ensuring that primary care practices remain open to Medicaid patients (via the PMPM to physicians) and using those practices as an intake system for flagging patients who need more support and referring them into care management at the network level. These two elements together represent a very scalable and systematic approach for managing a large population effectively.
Initiatives that target quality and costly utilization: Key examples include major pushes around childhood asthma, diabetes management, pharmacy management, and, most recently, palliative care.
Best practices shared across each network: The network model helps otherwise disconnected providers to come together, creating economies of effort and knowledge regarding key levers for improving population health.
Cross-continuum "glue": There are pros and cons of centralizing (or "outsourcing") care managers rather than distributing that function across provider PCP sites. A likely benefit of the centralized model is that the care managers have a view across the entire system, not just their own provider site.
- Additionally, CCNC benefits from longevity: Though it has been modernized into a PCMH model, CCNC began in 1982 as a rural health initiative aimed at improving primary care access to rural areas. It’s unclear exactly how big a role longevity has played in the savings, but the program enjoys local credibility and momentum for this reason.
CCNC track record attracting more payers
CCNC has become such a high-profile success that employers announced partnerships with the CCNC networks through First in Health, a public-private partnership including CCNC, the State Health Plan of North Carolina, GlazoSmithKline (GSK), KERR DRUG, SAS, and Blue Cross and Blue Shield of North Carolina.
This year, state and GSK employees will be encouraged to participate in CCNC’s medical homes. Co-pays will be waived, and participating physicians will receive an additional PMPM payment for care coordination and chronic care management with these patients.
Double-clicking on PCMH ROI: How much did it save, again?
CCNC is a great case study in medical home ROI, and not only because it has clearly achieved major savings. The organization has had a very interesting history of quantifying its savings, which it generously shares by making different estimates about financial results public. We think that sharing data is, in itself, a great contribution that CCNC is making to the industry as we try to figure out how to effectively model the impact of the PCMH (and population management in general) on health spending.
Over the years, at least three organizations have attempted to model how much money has actually been saved, with each group coming up with a different estimate. Time frames for the studies vary, making the estimates difficult to compare, but consider the following:
- Treo Solutions estimated $1.5B in savings over three years (2007-2009)
- Mercer estimated $708M in savings over five years (2005-2009)
- Milliman, Inc. most recently estimated $984M over four years (2007-2010), with a steady upward step in savings each of those years. Cost reductions are reported to have been $103 million in fiscal year 2007, $204 million in 2008, $295 million in 2009, and $382 million in 2010.
Isolating one year that all three reports have in common, 2007, the estimated savings range from $103M-$273M:
- Treo: $273M
- Mercer: $135-149M
- Milliman: $103M
Source: CCNC "Our Results"
The $170M question: Why the disconnect?
Without fully answering these questions (yet), we note that the different CCNC reports raise many more questions for further research, including:
- Why is there such a major gap in the estimates for a single year, with the gap widening as more years are factored in?
- What lessons can be gained regarding modeling the potential effects of similar (and dissimilar) medical home projects in the future?
The Medical Home Project research team has made "PCMH ROI" one of our key research topics for 2012. So please stay tuned as we continue our work in this area, producing a series of blog posts, white papers, and webconferences related to PCMH ROI.
Subscribe to receive notices of our latest research
Advisory Board members may log in to subscribe to the Medical Home Project blog, the main communication channel of the Project, to keep abreast of the latest medical home developments.
We are planning a webconference focusing on ROI of the PCMH from the perspectives of patients, practices, health systems, and payers, tentatively slated for April. Members who are subscribed to the blog will be invited to join our experts for this presentation when registration becomes available.
Please feel free to reach out to me at firstname.lastname@example.org