Payment Models that Support Medical Home and ACO Principles: Maryland's Experience
Payment Models that Support Medical Home and ACO Principles: Maryland's Experience
AHRQ's Health Care Innovations Exchange held a Web event titled Payment Models That Support Medical Home and Accountable Care Organization Principles: Maryland's Experience on April 25, 2013.
This was the first Web event in a three-part series designed to share novel experiences and lessons learned in putting accountable care organization (ACO) and patient-centered medical home (PCMH) principles into practice.
Judi Consalvo, Program Analyst at AHRQ Center for Outcomes and Evidence
Ben Steffen, MA, Executive Director, Maryland Health Care Commission
Mr. Steffen serves as the Executive Director of the Maryland Health Care Commission, an independent regulatory agency whose mission is to plan for health system needs, promote informed decision-making, increase accountability, and improve access to health care and health care coverage in Maryland. Previously, he served as the Director of the Commission’s Center for Information Services and Analysis overseeing analytic and operational responsibilities for health care practitioner initiatives in the state including development of an All Payer Data Base and the Patient Centered Medical Home Program. Mr. Steffen is board member for the Maryland Health Benefit Exchange, established in 2011 by the Maryland Legislature to implement the insurance coverage expansion under the Patient Protection and Affordable Care Act. He also serves on the Maryland Health Insurance Program’s board of Directors, Maryland’s highly successful high risk insurance pool that serves individuals that are unable to purchase insurance in the current Maryland individual market. Before joining the MHCC, he served as a budget analyst in the Health, Housing, and Income Security Division of the Congressional Budget Office. In this role, Mr Steffen participated in activities that ultimately led to the Medicare Prospective Payment System.
Craig Jones, MD, Executive Director, Vermont Blueprint for Health
Dr. Jones is the Executive Director of the Vermont Blueprint for Health, a program established by the State of Vermont intended to guide statewide transformation of health care delivery including: universal coverage; multi-insurer payment reform; a novel delivery system built on a foundation of patient- centered medical homes and community health teams; a focus on prevention across the continuum of public health and health care delivery; a statewide health information infrastructure; and a multi-dimensional evaluation framework for ongoing improvement with quality and cost effectiveness. Dr. Jones also serves on the Institute of Medicine's Consensus Committee on the Learning Healthcare System in America and the Roundtable on Value and Science Driven Healthcare. Previously, he was an Assistant Professor in the Department of Pediatrics at the Keck School of Medicine at the University of Southern California, and Director of the Division of Allergy/Immunology and Director of the Allergy/Immunology Residency Training Program in the Department of Pediatrics at the Los Angeles County and University of Southern California Medical Center.
Meredith B. Rosenthal, PhD, Professor of Health Economics and Policy, Harvard School of Public Health
Dr. Rosenthal is professor of Health Economics and Policy in the Department of Health Policy and Management at the Harvard School of Public Health. Her research examines the design and impact of market-based health policy mechanisms, with a particular focus on the use of financial incentives to alter consumer and provider behavior. She is currently working on a body of research that examines alternative models for reforming physician and hospital payment. Specific empirical projects include evaluations of several patient-centered medical home pilots, pay-for-performance initiatives, and an episode-based payment system.
What is the Health Care Innovations Exchange?
- Publicly accessible, searchable database of health policy and service delivery innovations
- Searchable QualityTools
- Successes and attempts
- Innovators' stories and lessons learned
- Expert commentaries
- Learning and networking opportunities
- New content posted to the Web site every two weeks
Innovations Exchange Web Event Series
Archived Event Materials
- Available within two weeks under Events & Podcasts http://www.innovations.ahrq.gov
- Thursday May 9, 2013 1-2pm ET
A Close Look at Care Coordination within Patient-Centered Medical Homes: West Virginia's Experience
- Wednesday June 5, 2013 1-2pm ET
Building Health Information Exchanges to Support Accountable Care Organizations and Medical Homes: Delaware's Experience
Motivation: Goals of New Accountability Contracts
- When we seek root causes of quality gaps and cost problems, fragmentation rears its head
- Payment reform can reduce fragmentation by making a single entity accountable for all care
- Incentives and performance measurement are the key levers
- Patient-centered medical homes are one such concept and a building block for others including accountable care organizations
Patient-Centered Medical Home Basics
- Joint Principles: physician-directed care; whole person orientation; coordinated and integrated care; quality and safety; enhanced access; payment system that rewards value (i.e., not resource-based relative value scale)
- National Committee for Quality Assurance has a measurement tool that has de facto become another definition
- Broadly, a set of structures, processes that improve access and reliability of care with a focus on individual patient needs and payment to support all of the above
Initiatives Are Proliferating
- Private/public Patient-Centered Medical Home pilots have proliferated across the country
- All major national carriers are sponsoring some kind of pilot or initiative
- Two Medicare demonstrations
- Numerous existing and emerging Medicaid and other State –sponsored initiatives
- Hoped for effects: improved access and quality of care (population health); improved care coordination and aggressive management of high-risk patients will equate to cost savings
Payment Incentives to Support Medical Homes
- Fee for service is incompatible with medical home concepts: huddles, between visit monitoring, care coordination, and support for self management are not reimbursable
- Payers may add a care management fee – per member per month – to cover these costs
- Such mixed payment may also soften productivity incentives
- Pay for performance or shared savings used to get practices focused on quality, downstream costs
Contracting Challenges with Medical Homes
- Multi-payer environment may make it hard for practices to fully step off hamster wheel
- Small primary care practices (arguably the place we want transformation the most) not good candidates for high-powered incentives
- Shared savings subject to enormous random variation with small numbers of patients
- Need to guard against possible unintended consequences of patient access problems, provider financial losses
Zoom Out to Accountable Care Organizations
- Regardless of how successful medical homes are, primary care cannot fix fragmented care alone
- Building medical neighborhoods and entities large enough to manage total costs (i.e., Accountable Care Organizations) is required
- At a minimum payers should provide incentives for hospitals and specialists to work with medical homes (e.g., BlueCross BlueShield Michigan)
Shared Risk Arrangements
Graph showing the risk sharing arrangement to/from ACO. X axis shows the total spending for ACO patients. Y axis shows the maximum shared savings or loss. Maximum shared savings is 7.5%. Maximum shared loss in Year 1 is 5 %; Year 2 is 7.5% and Year 3 is 10%.
Key Takeaway Points
- Integrated health care delivery requires payment approaches with greater accountability for total costs and outcomes
- Policy initiatives are simultaneously working to encourage implementation of specific clinic models to manage populations and complementary payment mechanisms
- A spectrum of mixed payment and risk sharing approaches are available
- Key issues of balancing appropriate risk, incentives against potential unintended consequences
Maryland Program History
Studies in 2009 showed
- Tools to enhance primary care are limited in Maryland law
- Higher payment for primary care alone would be inadequate
Legislation in 2010 established
- Authority of the state to launch a multi-payer PCMH pilot
- Exemption for a cost-based incentive payment tied to PCMH
- Authority for carriers to establish single carrier PCMH programs with incentive-based reward structure (shared savings) and data sharing
Maryland Health Care Commission
- Convene stakeholders to form multi-payer Patient-Centered Medical Home (PCMH) program: state action exemption to Federal anti-trust
- Develop standards and approval process for single payer PCMH programs (2 programs recognized as of March 2013)
- Participation in multi-payer: 5 commercial and 6 Medicaid managed care organizations
Overview: Multi-Payer Pilot
- Pilots sites included 52 participating practices: including 7 solo physician; 1 nurse practitioner-led; 18 small (2-5 practitioners); 18 medium (6-10 practitioners); 8 large (11+ practitioners); 2 federally qualified health centers
- Practices are broadly dispersed across Maryland
- 330 providers including physicians and nurse practitioners
- Participation agreement binds providers and payers
MultiPayer PCMH Program
Flow chart of the multi-payer PCMH program. The chart shows that the Maryland Health Care Commission administers the program which has three components 1) Practice Transformation with the Maryland Learning Collaborative 2) Innovative Payment 3) Program Evaluation.
What We Have Accomplished
Judi Consalvo: Good afternoon. On behalf of the Agency for Healthcare Research and Quality, I'd like to welcome you to our Web event titled “Payment Models that Support Medical Home and Accountable Care Organization Principles: Maryland's Experience.” I'm Judi Consalvo, and I'm with AHRQ's Center for Outcomes and Evidence. We're very excited about today's topic and glad to see that you share our enthusiasm. We have over 800 registered for this event today. Next slide, please.
Before we begin, I would like to introduce you to our webcast console. The console can be resized to fit your entire browser window. All of the components on the console can be resized, moved, and minimized into the menu dock at the bottom of the console. If the slides are too small, click on the lower right-hand corner of the slide window and drag your mast down to make it larger. Twitter functionality is available in the console for today's webcast. Feel free to participate using the hashtag AHRQIX. We are pleased to offer closed captioning on this web seminar. To access the closed captioning, please click on the link called Closed Captioning that is on the lower right-hand side of your screen view. After you click the link, a new window will display the captioning.
I would also like to remind you that if you experience any technical problems, you may click on the question mark button at the bottom of the screen to access the help guide or click on the Q&A button at the bottom of your screen to contact us with your questions. Our technical staff will work with you to resolve any issues. The last 10 to 15 minutes of this Web seminar is reserved for a discussion based on questions that you submit. Questions may be submitted at any time during the presentation. Simply click on the Q&A button at the bottom of your screen, and then type your question into the Q&A box and select Submit. We welcome your questions and comments on the upcoming presentation and look forward to an engaging dialogue that will promote the spread of health care innovation. Next slide, please.
Today's slides are available by clicking on the widget at the bottom of your screen that says Download Slides. This will generate a PDF–a version of the presentation that you can download and can save, as desired. Next slide, please.
The presenters that you will hear from today are innovators from AHRQ's Health Care Innovations Exchange. For those of you who are new to the Innovations Exchange, I'll take just a minute to give you an overview. AHRQ created the Exchange to speed the implementation of new and better ways of delivering health care. It offers busy health professionals and researchers a variety of opportunities to share, learn about, and ultimately adopt evidence-based innovations and tools suitable for a range of health care settings and population. The Exchange Web site includes a searchable database of QualityTools and service delivery innovations, as well as policy innovations, such as the ones that you will hear about today.
The Exchange also contains both successes and attempts, innovators' stories and lessons learned, and expert commentaries. To assist you in implementing these innovations, AHRQ also supports learning and networking opportunities, such as Web seminars, tweet chats, and podcasts. We post new content to the Web site every 2 weeks on a range of topics, and hope that you will sign up to stay connected with us, if you have not already done so. Next slide, please.
We have a number of upcoming Web events to share innovation, health strategies, and promote the spread of innovations. Our next learning and networking event is our upcoming Web event on May 9th from 1:00 to 2:00 p.m. Eastern time called “A Close Look at Care Coordination: West Virginia's and Arkansas' Experience.” To register for this event or receive more information, please visit our Web site at www.innovations.ahrq.gov. The Web site also holds an archive of our past Web events, podcasts, and tweet chats, and we also invite you to take a look and download materials that may be useful to you in your practice. Next slide, please.
Let's turn to our agenda for today. It is my pleasure to introduce our moderator Dr. Meredith Rosenthal. Dr. Rosenthal is Professor of Health Economics and Policy in the Department of Health Policy and Management at the Harvard School of Public Health. Her research examines the design and impact of market-based health policy mechanisms, with a particular focus on the use of financial incentives to alter consumer and provider behavior. Meredith?
Meredith Rosenthal: Thank you, Judi. I'm pleased to be moderating this Web seminar to help promote the spread of health policy innovation. I'm going to start us off by providing some context for today's presenters. It hardly needs to be said what the motivation bringing us here together is, but clearly underlying concerns about the affordability of health care and uneven quality are at the heart of what we're talking about today. When we look at these problems, as complex as they are, a common theme running throughout is the fragmented health care delivery system, in particular, no individual provider has the right set of tools and incentives to worry about the whole of health care and think holistically about managing patients for the best outcomes at the lowest cost.
The accountability contracting issues we're going to talk about today are a way of using performance measurement and incentives to assign accountability for a population to a provider or a set of providers. For the most part, we'll hear today about a patient-centered medical home initiative–a primary care-based accountability model–but we will also think of these as building blocks and take the discussion towards accountable care organizations which may include a range of other providers.
Just to get us on the same page about patient-centered medical homes, you'd have to be living under a rock to have not heard a lot about these over the least number of years. Of course, this concept has been around for decades, but for a little more than 5 years, patient-centered medical homes have been looked at as a potential source of solution to the health care fragmentation problem and really as a contracting vehicle for improving health care delivery. The joint principles issues in 2007, by a group of specialty societies really are the common thinking around the higher level concept of the patient-centered medical home that many of these initiatives are working towards. The national committee for quality assurance has come up with a measurement tool that allows payers and others to recognize patient-centered medical homes based on objective and concrete-structures and processes. Largely, what we're talking about here is a primary care-based model that allows for proactive, accessible care focused on individual patient needs and, again, provides a way of thinking about a new payment model that would allow those medical homes to treat patients more flexibly, more holistically, and not just through office visits.
Patient-centered medical homes really have taken off in the last number of years. Today we are going to hear about a couple of state-based initiatives, of which there are many. As well, the federal government has gotten into patient-centered medical home pilots and other activities through the Centers for Medicare and Medicaid Services. There are several Medicare demonstrations efforts that Medicaid, in partnership with the federal government, has undertaken in a number of places. These initiatives vary in their emphasis and their focus on pediatric or adult populations, but they are all intended to improve the quality of care and to address affordability concerns and, in many instances, improve access for underserved populations.
Here today we're going to talk about payment incentives around medical homes. One aspect of how policy makers may support medical homes is to think about a payment model that is more in keeping with what medical homes are attempting to achieve. In particular, the payment model that has been favored here is a mixed payment arrangement. Given the problems with fee-for-service and its focus on very narrowly defined billable services, there is an appeal to using essentially a capitation fee–a case management fee–that is a lump sum paid to the practice that gives the practice the flexibility to use those funds for non-visit-based, non-traditional care–to use those funds to allow providers to care for patients in their absence, as it were, for example, by looking at registries, doing outreach. In addition to this mixed payment approach that de-emphasizes fee-for-service, many payers also include pay for performance that targets both quality measures, particularly chronic disease management measures, and also target specific measures of downstream costs, including the cost of hospitalization, emergency department visits and similar types of costs.
There are a number of challenges to contracting with medical homes, and many of them relate to the fact that most primary care practices are quite small, and their patient panels may be quite small. It also can be challenging for payers to operate in this multi-payer environment, when all payers may not be entertaining the same kinds of contracting approaches, and this makes it hard for a practice to fundamentally change what it is doing. It makes it hard for payers to avoid spillover effects and to really get the attention of individual practices, and so you'll hear about some multi-payer efforts. Thinking back about the small practice problem, this can be challenging, not only for setting up all of these contracts, but because there is so much potential for random variation in many of the measures we'd like the practices to be accountable for, including for example, preventable hospitalizations, which may simply be infrequent enough events, so a small practice can't be reliably measured on them.
Then finally, with all kinds of high-powered performance incentives, payers need to consider potential unintended consequences, including the lack of accessibility of practices to certain kinds of high-risk patients that might come with performance incentives or concerns simply about the financial stability of practices. That will lead us to talk about thinking about ways to channel some risk to accountable practices, but guard against some of the most serious kinds of fluctuations.
As I mentioned, in the discussion today, while we'll focus on the experience with patient-centered medical homes, we'll want to talk about moving out from the primary care. Almost surely, primary care-based accountability alone, without requiring changes in accountability for specialists in hospitals, will not be the formula for a long-term success of these initiatives. We should talk about then how to we build shared accountability across unrelated specialty and the hospital practices alongside primary care practices, or think about using these higher level integrated systems as accountable care organizations. There may be any number of intermediate steps. I think there are some good examples in Michigan with Blue Cross Blue Shield–for example, rather than looking for accountable care organizations alone, has begun to give downstream providers–specialists and others–incentives to work more closely with their patient-centered medical homes–something that we've heard from medical homes they are seeking–partners in the fields.
Just to give you an example, I wanted to illustrate here the notion that, even when we think about these large integrated systems acting as accountable care organizations, we can think about a range of risk-sharing arrangements. Accountable care contracting arrangements don't have to be global hard capitation–as we used to call it–but they could involve something like risk corridor. Here is an illustration of an early version of the Medicare shared savings program. On the x-axis you have the total spending for the population. On the y-axis, you have the shared savings–either bonus or penalty. You can see, there is a corridor in which those bonuses and penalties initially shrink, as spending goes up, and then ultimately go negative, but it is capped, and so those flat lines at the top and the bottom indicate this essentially risk protection for the ACO arrangement. We can talk about how much sharing we want to do with accountable care organizations, whether there is a threshold for minimum savings, and all of these contacting issues that need to be worked out in practice. We will hear from Ben and Craig about how those have worked in their states.
Before I pass it off, I just want to note that, as we start thinking about getting more integrated health care delivery, we are going to need to have some flexible payment concepts around very different kinds of delivery models in different contexts. While many payers and policy makers are working on trying to think about what the right delivery model is–is it an integrated system? Is it a network? That will likely vary by state and by population. Likewise, the payment arrangements that compliment those delivery organizations need to be crafted in ways that appropriately handle tradeoffs and risk–protection for high-risk patients–but also do effectively align accountability, as is the goal of these initiatives. It is now my pleasure to introduce Ben Steffen, who is the executive director of the Maryland Health Care Commission. Both in this role and in his previous role, Ben has been highly involved in many of the innovations that Maryland has put in place to improve the access to health care and health care quality in Maryland. Maryland is currently one of five states involved in the Centers for Medicare and Medicaid state innovation models. Ben will share Maryland's experience with the multi-payer program and more. Ben?
Ben Steffen: Thank you, Meredith, and thank you AHRQ and your consultants from Westat for putting this on. Before I begin on our PCMH program, I'm going to give you a very brief background on Maryland. If you're not familiar, we have quite a few very small primary care practices in the state of Maryland, as is often the case where there is a misalignment between carriers and practices. Reimbursement has been an ongoing challenge here in the state, both in the primary care and specialty care study. The state had looked at reimbursement for a number of years of practitioners. As a result of work done in 2009, we had concluded that simply requiring higher payment for primary care would not be adequate for solving the challenges here. Fortunately, work already being done elsewhere suggested that we could try a pilot with an advanced primary care initiative to test some of the ideas we had going forward. Key to those ideas was a concept of a multi-payer initiative. As many of you know, Maryland is the only all-payer hospital rate-setting state, and this concept is popular here. We thought we'd take that into the dimension of primary care.
The state decided in 2010, that the best approach was through legislation. That solved several problems. It allowed us to convene payers together and talk about reimbursement without an action by the state. It is likely that payers would collide with Sherman Antitrust law, if they met together to discuss a reimbursement. We also had to solve several specific provisions in Maryland law that limited our approach to the PCMH program. Specifically, there was a prohibition against cost-based incentive programs in the state. That is practices could not be rewarded for simply reducing spending through an incentive program. Secondly, we also had some very strict limitations on data sharing between carriers back to practices in the state that also would by themselves serve as impediments to the program. We also had as a goal–while we wanted to launch a multi-payer program, we also wanted to have the flexibility for carriers to launch their own programs as well.
To overcome the antitrust provisions, the legislation designated a state agency–the Maryland Health Care Commission–to convene payers, to exercise what is called the state action exemption under federal state anti-trust law over a period of several months. We developed standards and processes for both the multi-payer program and for the single-payer programs, which have also gone ahead in parallel with our multi-payer program in the state. We have two single-payer programs. I'll talk briefly about that later in my presentation. Our multi-payer program includes five large commercial carriers and a sixth Medicaid-managed care–organizations that serve our Medicaid program here in Maryland.
A very brief overview of the program–this is a pilot, so practices actually applied–approximately 200 practices–sought to participate. We selected 52 practices with characteristics that largely mirrored the overall practice population here in Maryland. We have seven solo physician practices, one nurse practitioner-led program, and then the rest of the practices are made up from small to medium to large practices, by our definition of large here in Maryland. Among that group, is two federally qualified healthcare FQHCs. FQHCs presented some particular problems for us, as they are cost-based under the Medicaid program, and we had to fit in our new shared savings algorithm under that arrangement as well. Altogether, 330 providers are included in the program. They include both physicians and nurse practitioners. I think Maryland was one of the first states to actively reach out to nurse practitioners and seek to engage them in the program. Everyone is bound together–both providers and payers–through a participation agreement that spells out the various organization's responsibilities under the program.
The structure of the program is state lead, but payer and provider organizations have significant input into the operation. The three key elements of the program are practice transformation led by a collaborative of Maryland's physicians from the University of Maryland School of Medicine family medicine department, as well as John's Hopkins community physicians, as well as leading edge practices that had already achieved NCQA level III recognition in the state before the program began. Their roles are leading the organizational change, assisting practices in achieving NCQA recognition, assisting in the redefinition of roles within a practice, defining the role of a care manager which is specifically spelled out in our participation agreement, and broadening and reaching out to the community for practices in how they engage with services in their particular localities.
The payment model–Dr. Rosenthal talked a bit about that. We have a hybrid model. It includes an upfront investment payment, as well as a back-end shared savings arrangement. Importantly, the shared savings arrangement is based on the assumption that there would be a recruitment of the investment payments before any shared savings would be distributed.
Lastly, as it's a pilot, we wanted to learn more, even as we went forward, and we had specific requirements and we have a formal evaluation underway. The legislation, interestingly, specified some of the considerations we were to examine as part of that evaluation. In particular, the impact on both patient and provider satisfaction here in the state and the ability of the PCMH model to reduce health disparities.
Briefly, what have we accomplished? In the multi-payer program, approximately a quarter of a million folks are engaged. They are the privately insured and Medicaid patients. When we say engaged, we mean that those patients have a payer behind them. That is the practices receive a fixed upfront payment as part of the initiative. There are, of course, within those practices, other patients who receive the benefits of a PCMH practice, but have no payer incentive payments as part of that access. We've moved all 52 practices through NCQA recognition, with two-thirds of them already at level II or level III already.
We've also worked with practices, and they have successfully reported quality information from their EHRs for both calendar year 2011 and 2012. We've spent a significant amount of time aligning the 24-odd quality measures that we use under our program with other initiatives that were underway in the state, particularly the National Coordinator's–ONC–meaningful use measures. As part of this presentation today, we did crosswalk our measures with those specified and under CMS' ACO initiatives. There are eight measures that are collected and required by CMS, as part of ACO requirements. When we actually cross-walked our multi-payer measures with the ACO measures and with the private carrier measures that are used in the state, we find that fortunately, however, there are only two measures that are used consistently across all programs. We have some work to do here.
The major focus of my remaining time will be on the payment model. As I mentioned, it is a hybrid model. It still remains anchored in fee-for-service payments. It is important to note that the fee-for-service payments are different across the various carriers. That remains so. Medicaid MCOs have different payment levels from the private carriers. Each of the private carriers vary somewhat. Those did not change. What we did, however, was harmonize–what we call–the fixed transformation payments–the upfront payments that go to practices for hiring care coordinators, for adopting certain non-face-to-face visit activities that are thought to benefit there. The payment levels were spelled out in the participation agreement and did not differ by the complexity of a patient. They differed, however, by the level of NCQA recognition that a practice had achieved and, as a practice, moved through from I to II to III, in terms of NCQA recognition, in consideration of the fact that, as a result of meeting the higher levels of recognition, you could do more. They also receive a higher fixed transformation payment.
At the conclusion of each year, we construct a shared savings calculation of our practices–our benchmark–against a historical or baseline total cost of care measure. If they are able to demonstrate that the total cost of care for patients attributed in both years is lower, they would share in that savings. In order to be eligible for shared savings, the practices are required to first of all report on the clinical quality measures and to meet our utilization measures. In the first year, the quality measures only are reported. In the third year, we will be setting up benchmark thresholds for performance on those clinical measures.
A bit more about shared savings–we define total cost of care to include all services. Of course, those who are familiar with both the PCMH and the ACO models, when we talk about total cost of share, we're including services provided within and beyond the primary care setting. In Maryland, we elected to have a somewhat bifurcated approach for the private carriers. We elected not to include pharmacy because we had some self-insured participants who had carved out the pharmacy benefits, and for simply a complexity issue with patient IDs, we elected on the private side not to include pharmacy. On the Medicaid side, we are including our pharmacy. We've recently learned that TRICARE was participating, and they too will be including pharmacy in their shared savings initiative.
We set a benchmark against which practices were compared. It was for each practice–its own total cost of care for patients in the program in the baseline year. That was then inflated forward to account for overall cost growth in the state–7.4% was the figure we had for 2010 to '11, which was a pretty significant growth in cost–I think I would note–compared to some of the figures we've seen nationally, which may account for some of the shared savings we got initially. If practices were able to demonstrate that their costs were below the budget, then they participated in shared savings. The practices earned from 30 to 50 % of those savings, depending the number of quality metrics they reported in the first and in the second year. By the third year, they will be not only held accountable for reporting on those metrics from your EHRs, but also for meeting the quality thresholds that we set in the program.
For those who are interested in a few more details on the shared savings program, I already noted that in order to be considered, a patient has to be attributed in both the baseline and the comparison year. We also adjusted the calculations to exclude certain high costs of patients–those that would die, the cost of trauma care, or patients that had suffered traumatic injury–were also excluded. We capped the patient cost at $75,000 in both the base year and the comparison year. We looked at more sophisticated statistical approaches, a trend at the second standard deviation or the 95th percentile, and we actually elected to use a dollar figure, because it was easier to understand. I'll note here–the third point–that we used our own all-payer claims database to both attribute patients and to conduct the shared savings calculations. We are one of only a few states that are using all-payer claim data to actually distribute our payments. I can tell you that once you move from simply counting, for example, colonoscopies to looking at how reimbursement might be redistributed, everyone pays a lot more attention to what is in that APCD.
Measures on how we looked and how we completed the shared savings–moving from top to bottom, practices first had to demonstrate they could report quality measures if they did that, and you can see that 49 of the 50 organizations reported quality metrics in the first year. Moving then to the second column, of those 49, 23 did generate some shared savings. It wasn't significant, but we were surprised, even in the first year, that savings were produced. Then in the bottom row, you can see that 23 were eligible for some shared savings. Surprisingly perhaps–not surprisingly to others–only one of those practices was among our pediatric subset. We are one of the few states that do include pediatricians in the initiatives. A quick question on why did practices generate savings. They, first of all, had to generate and prove they could submit quality data. We saw–and these were simply by various correlations–a moderate relationship between reduced hospital days and lower total cost of care–about a .3 correlation–a weaker relationship between ED visits and lower costs. Then we simply didn't have enough cases to look at whether there was a relationship between reduced readmissions. As Dr. Rosenthal observed, some of these measures are very difficult with small practices as we face. We also observed a random variation in our small practices that drove some of the savings. Applying the same principles as used in the ACO model, we elected to trim off savings rewards or top them off at 10%. That is practices could make no more than a 10% improvement for purposes of shared savings. If they reduced cost by 4%, it was cut back to 10%, because we thought random variation contributed to those gains.
You can see the impact as you look at this simple scatter plot. Note that there were 11 practices that had their shared savings payments reduced as a result of this rule. The rule was actually permitted under our participation agreement, and I think we thought, when we looked at the first year results, that we were very fortunate to include this.
As I conclude my presentation, a few considerations on a multi-payer program, why we think it is important. This is a term that practice has really presented to us. They like the payer agnostic model. It allows for a consistent shared savings program. It allows to us move towards a consistent metrics. A large multi-payer program potentially can be linked with state improvement goals, and it can broaden participation in the market to smaller carriers by themselves who could launch a program. It is also is important that they talk about how it can help sustain practice transformation. We think external practice support for transformation is critical. How that is funded is really a consideration I wish we had thought more about going forward.
Care, coordination, and management is key. Certainly, the third rail of any advanced primary care initiative, and what we've learned–we end up having embedded care managers, but we think, going forward, a combination of provider base and community-based support may work best. The role of electronic health technology cannot be overstated, and we're fortunate here that we have a very close linkage with our health information exchange, and we're working with them to support the diffusion of information through the HIE to practices. Practices have argued very strongly for standardized data feeds, but have not yet made it happen.
Our evolving efforts that we'll focus, as we move forward from the pilot to something broader, is to continue to enhance primary care functions, to engage in coordination with the community and link providers in a more defined way, to begin to develop this concept of a community health workforce, which means establishing one in many Maryland communities, as one does not exist, and then taking advantage of it and linking data systems. Where we are going then–and, as Dr. Rosenthal noted–we are one of about 20 states that have received planning funds–is to take our multi-payer program up here–330 practices–combine it with our largest single carrier program from Blue Cross Blue Shield, which involves about 2,500 providers. Our Cigna program and other programs that will soon be approved and bring it together through the planning process to increase the number of transformed practices, to more fully engage payers, to engage our communities in a more defined way, standardized metrics, and then to link all of these too with the broader population goals that we in the state have defined in collaboration with the residents, providers, and payers that are participating into this broadly integrated community medical home model going forward. I will stop here and pass this back to Dr. Rosenthal.
Meredith Rosenthal: Great. Thanks, Ben. Now, it is my pleasure to introduce Dr. Craig Jones, Executive Director of Vermont Blueprint for Health. As a veteran innovator involved in Vermont's statewide transformation of health care delivery, Craig has agreed to provide his reflection and perspectives on this topic. Craig?
Craig Jones: Meredith, thank you very much, and I want to thank AHRQ and their team for the opportunity to be a part of this and basically add on a little to the great information that Ben has provided us. We've had a lot of opportunity through the AHRQ Innovations Exchange to present the details of our program and the characteristics of it, so today I'm really mostly going to focus on how we've approached our payment models and what it is intended to achieve. It offers a nice comparator to the Maryland experience, and I think the beauty of this is, is that we're beginning to get a set of examples across the country of the way these hybrid or composite payment models are beginning to emerge. I'll go ahead and just give a quick outline of the program, and then focus more on the payment models and some of the observations.
Our Blueprint model was really called on to develop a fairly comprehensive approach to delivery system reforms, and it was–what I would consider–visionary legislation or guiding legislation that was certainly looking at primary care but looking beyond, and as has been alluded to, to almost community networks of support services. These are the elements that are in play right now, and the Vermont model is expanding statewide. First off is, of course, the patient-centered medical home or what is called the advance primary care practice here. We've expanded to 110 patient-centered medical homes in all health service areas of the state. In addition, we've put in place what we call community health teams, and this was an interesting construct. The intent here was how can you provide the multi-disciplinary support that you really need–the team-based services you really need–to patients and families in all of these different types of primary care settings? Even if primary care works and transforms and improves, how are you going to get those teams available? In our case, the community health team was the answer that we played with and attempted to expand, and it seems to be working well. These are basically multi-disciplinary teams that are put in place by the administrative area in each health service area. They work embedded in the practices. They are effectively an asset or a resource as part of the practice, but they move between several practices, and so they are a fluid entity, if you will–a core resource. We almost view them as a utility to the medical home settings, and with time, they become a key and integral part of those patient-centered medical homes.
Now, we have all of our insurers supporting these core community health teams at a ratio of about 1 fulltime equivalent for every 4,000 in the medical home setting, and that expands and is added to, as the number of medical homes increases. As in Ben's case and in Maryland's case, our medical homes go through NCQA recognition, and with the official recognition, they get the support of the community health teams. In addition, we've seen several of the insurers expand on this and add additional care support resources that are more targeted towards more intensive groups. Medicaid has added on care coordinators to these teams to target their top high risk population. Through our multi-payer advanced primary care demonstration, we not only have Medicare supporting the medical home, payments, and the community health team payments, but they also are supporting what we call SASH teams. These are support and services at home. These are teams based in publicly subsidized housing around the state as a node, if you will. They literally are providing in-home support to high risk Medicare beneficiaries and forming teams. This is really getting at the heart more of holistic health and human services, problems people have with limitations and activities and social and economic needs. The vision here was to try and put in place an advanced primary care transformation, but supported by community networks of teams that can really get toward a more holistic approach to health and human services. Now, this, in our case, is supported through insurer payment reforms. All of our major insurers pay each of the medical homes, based on their NCQA score, and I'll show a bit more about that in the next slide. We consider that, I guess, an upfront or a quality payment, based on how they score. All of the insurers are also sharing the cost of the community health teams and effectively putting in place a utility. There are no prior authorizations, co-pays–no barriers to patients–and so those two payments at this stage form the hybrid that is helping to roll out the Blueprint model.
I want to expand on another key aspect of this, and there is a lot of attention being paid to this in multi-state collaboratives we work with–the other supports that you really need to make a delivery system reform come alive and to really build these community networks or neighborhoods or however you want to view them. We have a tremendous amount of effort going into our health information infrastructure–a statewide health information exchange–but also a central clinical registry being fed from electronic medical record systems, hospital systems, and the network of community users–tobacco succession counselors, health educators, SASH teams–contributing into the centralized clinical registry with the goal of producing what we call a community view or an integrated health record–a real complete health and human record, so that these teams can deliver the best services possible.
In addition, we have an intense investment in development of data systems for evaluation and reporting. As with Maryland, an all-payer claims database that we are using–certainly, to study the impact of the program, to produce practice profiles–so practices can see how they are performing comparatively, and also to use the output for predictive modeling and looking towards the next steps in health reform and payment reform. There are a number of other data systems that are going in place as well, the central clinical registry being a good example as a growing resource. I will just say the build-out technically of this is maybe one of the most complex aspects–getting systems to exchange information or even contribute information that is normalized with good quality has turned out to be a really complicated task, but that's one we're working hard on.
Another key aspect, in addition to the medical homes and community health teams, a network statewide of what we call community self management programs–things like healthier living workshops for patients with chronic disease, healthier living workshops for pain control, healthy and living workshops for diabetes, tobacco cessation workshops, wellness recovery and action planning workshops. There is a growing network that we have of these statewide self-management programs. In the end, I think it is all oriented toward a continuum of services extending from the medical home to the team-based support across communities, linkages with other services, and a strong emphasis on a move toward enhanced self-management ultimately.
Lastly, what we call learning health system activities, and this is essential. Can you really drive transformation in the primary care practice or even across broader settings without the right supports? The health information and the evaluation systems are critical, but we believe there are other aspects to it, such as practice coaches or facilitators, the shared learning forums that they and others interact in; and so we have an array of activities that we put under that category–learning health system activities–and we're hoping that the data and the information generated out of the IT systems comes to play as actionable knowledge through these learning health system activities. That is the bullet list, if you will, of the components that are expanding statewide now. As I said, we're at about 110 medical homes, about 420,000 or 430,000 people with access to those medical homes and community health teams and about 550 primary care providers including a good array of advanced nurse practitioners that are involved in the program.
Let me switch over now to this slide and really focus on the way payment reforms or investments are being brought to play, as we expand the Blueprint. This is a simple way of displaying the payment strings right now. First of all, we have all of our major insurers, including Medicare and Medicaid, paying the medical home a per-person per-month payment based on their NCQA score. This is on top of fee-for-service. It is similar to Maryland. Fee-for-service is still in place, and this, in some sense, can be viewed as a quality payment to help promote or drive patient-centered medical home transformation. It is not a large amount of money, and frankly, not enough, and we are reexamining that, in terms of what primary care really needs, but interestingly, as a targeted payment string, it has been effective in helping to engage primary care practices and drive transformation. It ranges from about $1.50 to $2.50 per person per month, based on the NCQA score.
The second is all insurers, as I mentioned, sharing the cost of the community health teams, effectively putting in place a utility and minimizing barriers for patients and families. These two payments effectively look like one for quality, the other to build a utility and, again, on top of fee-for-service. Now, you notice the remaining arrows there all come from our Blueprint budget, which is state budget. We have an appropriation each year, and we view this as an investment in helping to build the infrastructure that is necessary for a real integrated health services system across the state. We have all of the independent practices, organizations, hospitals, a community of providers that everyone else does. The question is–the challenge is–can we invest in a way that makes them operate with system-ness, even where there really isn't the single system?
The Blueprint grants support project management in each health service area. They are charged with bringing together the primary care providers, the hospital teams, the non-medical providers, mental health, area agencies on aging, social services, public health district office directors, the SASH teams, and on a regular basis, having all of these people meet in an integrated health services workgroup, plan what their community health team staffing will be and continuously work to improve operations, coordinated services, and transitions. It is an incredible role for these project managers. They do an amazing job.
In addition, our grants invest in practice facilitators in each area, and their job is several, but one is, primarily at this point, to help practices prepare to be scored as medical homes and really go through the transformation process–helping practices, so they are not overwhelmed or certainly assisted with this difficult transformation effort. They will stay in place and be focused on continuous ongoing improvement, as practices are scored and become active medical homes. The Blueprint grants also support the self-management workshop network around the state, and this is a growing network that is taking on a critical role as a resource to communities and teams.
Our budget also supports the central clinical registry and data quality. Really, data quality has become a key issue, where we literally work side by side with core data dictionaries to improve the data capture in practice settings by health teams, and make it so that it truly is normalized and can be brought into a single centralized registry. This is where we are working to create the integrated health record views, so teams can really operate across settings. Our contracts also support a range of evaluation analytics and reporting, and ultimately we hope, moving toward predictive modeling, as we link clinical data, claims data, other sources. That is really one of our next steps we are going to be focusing on potentially. This is, from a payment or a financial point of view, a look at the two payment reforms that are driving this stage of transformation, and then the investments that the state is making in helping to develop the infrastructure. Frankly, and this is viewed almost as building a foundation across Vermont, one in which next waves of payment reforms and financial reforms, such as ACOs, can build on top of, and with the hope that we optimize delivery systems and health service operations, so that these higher level payments reforms can have a chance to be successful.
Let me quickly talk about just a couple of considerations and comments. Then you've heard, I think, this echoed throughout the discussion–is higher payment enough? In our experience, we don't believe so. We think that there needs to be a steady investment in the infrastructure and some of the components I outlined for you, for real transformation and, particularly even just within the practice studies–the need for practice facilitators to help prepare and undergo the transformation. We have a University of Vermont team that scores every practice against the NCQA standards, so it is independent and objective. Those two resources alone–the facilitators and the scoring team–are an example of an infrastructure development that has really helped practices transform and begin to operate in this community network. We believe that payment isn't enough, and in fact, there are many of our practices that tell us the investment in the community health team, which doesn't go directly to them as money, is one of the things they value the most. The question about to what degree can primary care organize holistic team services–they can, to a degree and particularly within their walls, do some of that, but to really have multidisciplinary payment, it is critical that we find a way to invest in these teams and get more holistic services.
Let me go ahead and closeout with this slide, and I think this is really the essence of today's discussion. We've heard about a couple of blends of payment that are being taken initially to drive transformation, but the question really is what payment or blend of payments will lead to the most effective and sustainable change? We've heard about these upfront or capacity-like payments–quality payments–really investments, and then presented a view, for example, of how shared savings is coming into play, but we're really at the front edge of this. The question is can we come to a balance that will achieve high quality and savings, so that we don't fall into the dynamic of the 90s, where savings became the sole major driver and gatekeeping was the theme of the day.
With that, let me go ahead and finish up with just a slide that is showing a couple of key advancements that will be coming into play here in Vermont in the next year–things we're excited about. I won't go into each of these. There are some new payment reforms coming in. Particularly, I'll focus your attention on the fifth bullet point. What is expanding is a statewide model for addiction and co-occurring mental health disorders and investment in more community health team staff to help patients treated with support and home therapy, and for mental health disorders in the medical home setting and also specialized centers. It is our first go at integrating specialized centers for the treatment of people with addiction disorders in mental health. New investments, all driven by infrastructure investment, and then we'll move toward payment-based on outcomes and some important societal outcomes in this case, such as reductions in incarcerations or sustained employment. These are just examples, and again, we hope this is the beginnings of a foundation for the next waves of payment reform and financial reform. With that, I'm going to close it out and turn it back over to Meredith. Again, I want to thank AHRQ for the opportunity to be part of this.
Meredith Rosenthal: Great. Thank you, Craig, and thank you both for your insightful presentations. I want to start the question and answer. We have a lot of questions that have come in, but I wanted to start with a question for both of you to try and zoom out from the really important work that both of your states have undertaken, really on the ground, in primary care. Can you talk, I guess first Ben, a little bit about the receptiveness in Maryland of specialists in hospitals to these changes and how Maryland is planning to engage them? Craig alluded to sort of gatekeeper history that we've all been through and wanting to think about this in a way that brings communities and providers together, rather than drives them apart. Ben?
Ben Steffen: Sure, thanks. I didn't mention in my presentation that we have nine ACOs in the state that have been designated by CMS. Unlike other states, because of the way that the ACOs were defined in the Accountable Care Act, Maryland hospitals cannot serve as ACOs, so these are all practice-led programs. In terms of engaging specialists, of course, I think that provides one opportunity limited to Medicare, although all of the payers are looking at that option. I think, generally, the engagement of specialists is a big challenge under healthcare reform today. If you think of sort of the organizational change and the three principles, everyone wants to know is where are we going, what is in it for me, and how can I help? I think we are working with specialists to help figure that out. I think the concept of the medical home binding specialists to these new models is extremely important. We are not there yet. I think the next generation through this CMS planning grant will provide us some opportunities to look at better integration with specialists, as well as integration with hospital services. Our hospital rate study system provides us with some important advantages there to more fully engage hospitals.
Meredith Rosenthal: Great. Thanks. Craig, anything to add to that, in terms of how Vermont is trying to bring the rest of the delivery system into this opportunity?
Craig Jones: I think it is similar to how Ben described. There are ACOs forming and really working on the design of their standards and their measures. I think that is a place in which the linkage of interest–shared interest, if you will–becoming specialists in primary care are very likely to occur. I think the hospitals are also looking at this as a way to help them move into a new business model almost–transform. The goal of these community networks and good primary care services is effectively to reduce unnecessary acute care. That is one of the key outcomes, so they are having to figure how to evolve and to work in a new world, in a new environment. I think the ACO is offering them that opportunity to look at a steady shift in their business model. I would say that is one of the key things that hospitals will have to do.
Meredith Rosenthal: Judi, do we do one more?
Judi Consalvo: No. I think we're coming close to the top of the hour, and I wanted to thank you all tremendously for your participation today, to our presenters with sharing this information and to Meredith and also to our audience. I want to draw your attention to the last slide. If you want to learn more about the Maryland program and the Vermont program, you can go to those. You can click onto those names on this slide, and it will take you directly there, as well as there is a video series on the Vermont Blueprint, which would be very helpful. If you have some additional comments and questions, you will see when you go to these different profiles from Maryland and from Vermont, there is a comment section, and you could submit your questions, your comments, and that will go directly to our innovators who were on this presentation today. Again, you can follow us on Twitter, and you could send us emails to email@example.com. As always, we invite you to explore our Web site and to join us for future presentations. Again, thank you again to everyone.