This Expert Panel is Archived.

This Expert Panel is no longer active as of December 2018. Thanks to those who posted here and made this information available to others visiting the site.

Primary Care Integrated, Complex Care Management

Posted: 11 Aug, 2014   Recommendations: 38   Replies: 74

As we grapple with unsustainable health care spending and an aging and increasingly chronically-ill population in the United States, we need new strategies for organizing and improving the efficiency of health care delivery. One particularly promising strategy is complex care management (CCM) in which specially trained, multidisciplinary teams engage and address the complex needs of a subset of the most complex, high-cost patients. CCM, when designed and implemented appropriately, can improve health outcomes and reduce unnecessary utilization and costs. In fact, CCM initiatives can pay for themselves.

GHDonline is pleased to welcome an exciting group of experts to share their experience in designing and implementing successful complex care management programs:

     • Ken Coburn MD, MPH, co-founder, CEO and Medical Director of Health Quality Partners
     • Steve Counsell MD, Founding Director IU Geriatrics, a John A. Hartford Foundation Center of Excellence in Geriatric Medicine, Indiana University
     • Clemens Hong MD, MPH, Instructor of Medicine, Harvard Medical School and Chief Science and Innovation Officer at Anansi Health,
     • Rebecca Ramsay BSN, MPH, Director of Community Care, CareOregon
     • Jan Van der Mei RN, MS, ACM, Ambulatory Care Management Director, Sutter Health Sacramento-Sierra Region

During our week-long discussion, panelists will address the following questions:

     • How are you approaching the care of the most complex, high-need, high-risk patients in your health delivery context? How have you specifically adapted your approach for your context or selected population?

     • Evidence for complex care management is inconsistent, which may be due to sub-optimal strategies to address key steps in complex care management. What approaches have been most successful for your teams in: selecting and engaging patients, identifying and prioritizing care gaps, and delivering interventions to address patient needs across the biopsychosocial spectrum? What factors are most critical to success in these areas?

     • Care management programs must develop strong relationships with patients, primary care providers, and hospital and community-based providers (e.g. behavioral health, in-home support services, and social services agencies). Who are your critical partners, and what approaches have you used to best engage them?

     • What are the keys to successful implementation of CCM? What challenges have you encountered with implementation? What “do’s” or “don’ts” have you learned?

     • How do you plan to ensure that your work in complex care management is sustainable? What challenges the sustainability of your approach? How are you building the case for sustainability or growth of the model in your context?

This panel is part of our US Communities Initiative, which is supported by the Agency for Healthcare Research and Quality (AHRQ), and aims to foster discussions between health care professionals on evidence-based practices, and translating these practices across disparate settings, to improve health care delivery in underserved populations in the US.

In an effort to understand the impact of our Expert Panels, please take our short (4 question) survey before the discussion begins:

If you work in a complex care management program, or are beginning to implement one, and would be interested in speaking with us about the impact this Expert Panel has on your work, please contact us at:

We’re looking forward to a rich discussion next week – please join the conversation and share your questions or comments for our panelists!



Sudip Bhandari Replied at 5:09 PM, 11 Aug 2014

In preparation for next week's discussion, I wanted to share a couple of resources that might be of interest:

"Toward Increased Adoption of Complex Care Management" (linked below - PDF) is The New England Journal of Medicine’s article co-authored by our panelist Clemens Hong that provides an overview of the importance, opportunities and challenges of complex care management programs.

"Caring for High-Need, High-Cost Patients: What Makes for a Successful Care Management Program?" (linked below - PDF) is a recent Issue Brief (published August 2014) from The Commonwealth Fund, also co-authored by Clemens, which provides an in-depth analysis of the operational procedures and best practices in the field of complex care management.

As always, we look forward to hearing from you - please don't hesitate to share any additional resources you might recommend before we begin our discussion next Monday!

Attached resources:

Jonah Pierce Replied at 7:11 AM, 12 Aug 2014

Looking forward to this.

Maxwell Madzikanga Replied at 7:19 AM, 12 Aug 2014

I am sure it will be a very enriching interaction. Topic very relevant and insightful!

brianne fitzgerald Replied at 9:34 AM, 12 Aug 2014

as someone who has worked in several complex care management organizations in the US I am here to say that boot on the ground implementation is far more difficult that development plans in the office. Having also worked in several developing countries doing the same thing I see it as a far easier implementation due to the fact that there are less specialists, fewer choices for patients and a different all around experience with and about healthcare. So looking forward to this discussion

Jonah Pierce Replied at 11:37 AM, 12 Aug 2014

The second attachment only has the first page. the rest of it is blank.

Emmanuel Byaruhanga Replied at 12:04 AM, 13 Aug 2014

Look forward to this

Sudip Bhandari Replied at 10:34 AM, 13 Aug 2014

Hi Jonah (and anyone else having issues with the PDF!) -  very sorry to hear that you are experiencing difficulty with the second attachment. You may try opening the link in a different web browser, if you have one (Firefox, IE, Safari, etc). You could also try saving the file (right click on the link, "save as") and then open with a PDF viewer. If you're still having issues, please don't hesitate to get in touch directly:

Sudip Bhandari Replied at 1:24 PM, 13 Aug 2014

A reminder that our very short Expert Panel survey is open until Monday (August 18) 9 am Eastern Time- your responses to these 4 questions will help us evaluate the impact of these types of discussions here at GHDonline and provide us with incredibly helpful feedback. Please take a moment to fill out the survey if you haven't already:"

christophe millien Replied at 12:04 AM, 18 Aug 2014

I'm very exited to be part at this discussion. I 'm working in poor country where complex care is's an hornor for me share my little wxperience about that in country Haiti

Steven R. Counsell, MD Replied at 12:47 AM, 18 Aug 2014

Hello - I too am excited to be a part of this discussion! My experience in complex care management stems from my roles in health services research and clinical leadership in Geriatrics in Indianapolis at Indiana University (IU) School of Medicine; Eskenazi Health, a public safety net healthcare system; and IU Health and it's Accountable Care Organization and Medicare Advantage Plan.

We developed and tested at Eskenazi Health a new model of geriatric care management called GRACE - Geriatric Resources for Assessment and Care of Elders - now referred to as GRACE Team Care(tm) ( GRACE continues at Eskenazi Health targeting complex older patients for enrollment at time of hospital discharge. Over the last 3 years, GRACE Team Care has been implemented in the IU Health Medicare Advantage Plan targeting frail elders at high risk for hospital admission. Patients enrolled in GRACE Team Care receive home-based care management by a nurse practitioner and social worker who collaborate with the primary care physician and a geriatrics interdisciplinary team and are guided by 12 care protocols for common geriatric conditions.

Through our GRACE Training and Resource Center, to date, we have provided training and technical assistance for implementation of GRACE Team Care in 9 healthcare systems in 5 states. As with the original trial, replications of the GRACE model are demonstrating similar results including improved quality of care, decreased hospital and ED utilization, and lower costs.

In the coming week I hope to be of help to those thinking about these critical issues. I am honored to be a member of the Expert Panel and look forward to learning how others are targeting Medicare and dually eligible (Medicare and Medicaid) beneficiaries for complex care management interventions and strategies for program implementation and scaling.


Steven R. Counsell, MD
Mary Elizabeth Mitchell Professor
Director, IU Geriatrics
Service Line Leader for Geriatrics, IU Health Physicians
Scientist, IU Center for Aging Research
Indiana University School of Medicine

Dr. Liddy Kiaty - Figueroa Replied at 1:11 AM, 18 Aug 2014

I'm really excited to be part of this discussion. I know will be helpful for the work I'm trying to implement in my country, the Dominican Republic.


Dra. Liddy Kiaty Figueroa
Global Health Delivery Intensive, Harvard School of Public Health

Clemens Hong Replied at 1:25 AM, 18 Aug 2014

Welcome to this online panel discussion! I'm one of the panelists, and I'm honored to be participating. I look forward to learning more about how people are approaching this important work across our global community. I'll be speaking from two different perspectives this week: 1) I led Commonwealth Fund study on primary care-integrated complex care management programs that led to the organization of this online panel, and 2) I've been a part of several complex care management demonstrations for uninsured, Medicaid and dual-eligible patients during my career (e.g. the Transitions Clinic Network - - that improves care for formerly incarcerated individuals returning home from prison). I will tell you more about my current implementation work in a separate post.

I'm excited to be joined on the panel by 4 spectacular panelist with immense experience doing this work. I hope you'll all take advantage of this opportunity to interact with them, as they bring experience working in different US health delivery contexts and caring for different types of vulnerable, high-need, and often very high-cost patients. I know I will definitely take advantage of the opportunity to learn from them and all of your experiences in this space. I look forward to the conversation!

Dr. Liddy Kiaty - Figueroa Replied at 1:46 AM, 18 Aug 2014

Thank you so much! I will use this time to increase my knowledge in my most recent project, the implementation of a telehealth system to provide specialize medical assistance in rural communities, join effort with the Ministry of Health.

Rebecca Ramsay Replied at 2:34 AM, 18 Aug 2014

Hi everyone! I am also a panelist for this week's online dialogue about complex care management strategies for high risk/high cost patients. My experience spans 9 years working to develop collaborative population programs targeting our highest risk/highest cost members in partnership with our delivery system from within a Medicaid and Dual-Eligible managed health plan in Portland, Oregon called CareOregon. Oregon's Medicaid health care system has undergone significant transformation over the past two years with the advent of 16 legislatively mandated regional Coordinated Care Organizations (CCOs) - no longer a Medicaid Managed Care Organization, CareOregon is a member of four regional CCOs and administers a SNP Medicare Plan for dual-eligible benificiaries. For this week's online panel I will be speaking from two perspectives primarily: (1) As a member of the governance and evaluation oversight committees for our largest CCO's CMMI Innovation effort aimed at scaling up and successfully implemeting five metro-area programs that target the CCO's highest cost and most vulnerable members, and (2) As the co-designer of one of those five new programs, The Health Resilience Program, which trains and deploys a non-traditional community outreach workforce embedded in primary care, specialty care, and detox care settings. I am honored to be a panelist and expect to learn a tremendous amount from the other panelists and the wide array of participants - we are still very much on a learning journey with these new programs, and we have much to sort out as evaluation data trickles in and our grant funding sunsets. I will talk more specifically about our experiences tomorrow as the online discussion continues.

Rebecca Ramsay, BSN, MPH
Director of Community Care Programs, CareOregon

Clemens Hong Replied at 4:29 AM, 18 Aug 2014

I want to raise two context-related approaches from my recent experience in two consecutive posts:
1) The role of community health workers (CHWs) in complex care management
2) Healthcare delivery system-embedded vs regional care management models

The role of community health workers (CHWs)

Premise: CHWs could play a key role in complex care management - particularly in resource poor settings and in the care of vulnerable populations (e.g. Medicaid, dual-eligible, and uninsured individuals) who are often hard-to-engage using traditional healthcare delivery approaches and more often face significant challenges from behavioral health problems, trauma, and poor socioeconomic circumstances.

I currently work with Anansi Health, a new non-profit organization. We founded Anansi Health with the mission to help delivery systems develop CHW-integrated models of care delivery for the highest risk subgroup of Medicaid and dual-eligible patients. We believe highly-trained CHWs can play a central role in the care of these incredibly vulnerable and poorly engaged individuals. CHWs work as the lynchpin of Anansi Health's complex care management model with support from multidisciplinary complex care management teams that include nurses, behavioral health focused social workers, and others. CHWs bring immense value to health delivery systems because they are uniquely qualified to engage and accompany patients to better health, and help healthcare delivery teams and community partners provide more patient-centered care. In addition, providing employment for community members in these communities may have downstream impacts on community development and health.

We are currently working to support the Los Angeles (LA) County Department of Health Services in the development and evaluation of CHW-integrated, complex care management model. LA County will embed 25 CHWs into 5 large practices and work closely with nurse care managers, social workers and others to address the needs of the highest risk subset of primary care patients. The goal is to improve care and care experience for this high-risk subset of patients while reducing unnecessary utilization of acute care services and healthcare spending enough to generate a return on investment.

How do people feel about the role of community health workers in complex care management? What is your experience trying to integrate this workforce into more traditional care delivery teams? What have been the benefits/challenges, barriers/facilitators, "dos" and "don'ts"?

Clemens Hong Replied at 4:35 AM, 18 Aug 2014

I want to raise two context-related approaches from my recent experience in two consecutive posts:
1) The role of community health workers (CHWs) in complex care management
2) Healthcare delivery system-embedded vs regional care management models

Healthcare delivery system-embedded vs regional care management models

Premise: We should look to develop more regional population health and complex care management approaches. Small practices rarely have the resources or the expertise to implement their own programs. Large healthcare systems and payers that have the capacity to develop or fund their own complex care management programs, understandably, don't want to investment resources into programs and see that investment to go to patients for whom they are not accountable. Regional care management organizations, if properly structured, bring economies of scale, allow for inclusion of small practices and healthcare systems, and may work beyond this issue of accountability to address the needs of a patients regardless of who insures them or where they receive their care.

Many large healthcare delivery entities that are discussed in the commonwealth fund report run excellent complex care management programs. There are many advantages to building your own program if you have the resources and expertise to do it - not the case for many healthcare entities in the US. Many states are moving towards regional approaches to care, including Oregon, Vermont and North Carolina (included in the commonwealth fund report), and others. There is an opportunity to leverage resources available in these regional centers to support robust complex care management initiatives, as these sites demonstrate. Another example, is Health Quality Partners (HQP), where Ken Coburn (panel member) is the co-founder and CEO. HQP contracts with multiple providers within the region to provide care to their highest risk patients. Through these contracts, HQP brings an evidence-based, high-quality program to participating practices in the region, including many small practices that would not have the resources or the know-how to do it themselves. These examples demonstrate some of the strengths of regional approaches. The challenge are many. The biggest them? Of course, getting people to play together!

So what do people think about strengths and weaknesses of healthcare delivery system-embedded approaches and regional care management? Should we support policies that spur adoption of more regional approaches? Would people play together to do this in your context? Would your organization participate? Should we build on the lessons from HQP and create private entities to do this? Where would that be appropriate and what are the risks and benefits?

Kenneth Coburn, MD, MPH Replied at 2:48 PM, 18 Aug 2014

Hi and best wishes to all. I'm delighted to be on the panel to share the experience of Health Quality Partners (HQP). We currently work with over 110 primary care practices across 7 counties in southeastern Pennsylvania, USA. These practices vary in size and resources - many having only a few providers and limited support staff.

Regional Care Management Model:
We've served as a shared regional resource to unaffiliated practices and practices affiliated with several different health systems, providing population health and complex care management services in the area for over 12 years. Our advanced preventive care model of community-based nurse care management has, to date, served older chronically ill adults and as a regional resource has demonstrated improved health outcomes, reduced acute care utilization, and net savings. More info about our results is available on our website at

Key Elements to Successful Regional Uptake:
One key to recruiting practices has been our model design. It asks relatively little of practices wishing to participate; 1) access to medical records (for review by our nurse care managers) and 2) willingness to take well-prepared and organized calls (SBAR communication format) about patients with a change in status or other safety or medication reconciliation concern. The goal is to detect and address proactively a number of potential emerging threats to health before they cause illness or harm. The longitudinal preventive interventions, monitoring and assessments provided by the HQP program are delivered in close collaboration with primary care providers, but run, for the most part, in parallel with tight coupling back to the primary care providers as patient needs warrant.

Another key has been the development of a management system to ensure reliable delivery of our model, including the rapid identification of significant variations in performance over time or across nurses. This has been essential to the mentorship/management approach we've taken to developing the skills and capabilities of the nurses fully and to identify the need to change protocols when critical care objectives are not being reliably achieved.

Though unwieldy at times, we've also benefited by using a "federated" approach to engaging the larger health systems in our service area. This involves the systems agreeing to share patient data and to lease staff employed by their health system to HQP (we pay back salaries and benefits). In some ways, fundamentally, we've been able to develop a regional model because we've brought the funding source for the program with us - by means of national demonstration project and a health plan contract. We are now working to make the transition entirely to the "free market" with the demonstration project scheduled to end in December 2014.

Benefits of the Regional Approach:
Efficiency - able to serve smaller practices across a region with cross coverage to ensure service reliability. The program also can cover more ground with one lean, centrally shared set of management and IT resources (as a shared center of excellence).
Effectiveness - depending on the context of the embedding practice, in some situations, the regional model may also be more effective; by making staff more mobile and able to go to homes and other locations and the avoidance (intentionally or unintentionally) of staff being co-opted by the practices for other tasks. (Not all docs fully appreciate, understand, or respect the unique roles of the latest generation of nurses or other health professionals in these kinds of models.) In addition, an experienced and skilled management and support team is available to all nurses (not necessarily possible in some embedded models).

Disadvantages of the Regional Approach:
There are occasions when tighter, more readily available communications with practice staff and providers would be beneficial - more efficient and potentially more effective. But, in our experience, this can usually be addressed by developing a good rapport and trust with key practice staff to enhance access and response times. I don't think that routine or standing case reviews or rounds are very efficient or helpful as opposed to a proactive, but still "just in time" sense and respond system.

What are your thoughts and experiences? Any questions about our model? Thoughts on how we can improve?

Rebecca Ramsay Replied at 2:55 PM, 18 Aug 2014

I am responding to the first of Clemens's questions related to the use of CHWs in complex care management models.

In the development of our Health Resilience Program, we initially wrote a relatively vague job description for a new workforce we were calling, at the time, "engagement specialists". We knew we wanted to recruit a distinctly different role that would be seen as less medical, and less hierarchical, to help us develop trusting relationships with the most vulnerable patients. We knew this workforce would need to be comfortable working in community settings, in patients homes, shelters, under bridges.....but to also be able to gain credibility and work well across many traditional health care settings. And we had a hunch (which has now been confirmed) that the root cause of much of the suffering and high cost utilization in our population would be trauma & BH complexity (anxiety, depression, some severe persistent mental illness, and addiction), overlaid with intergenerational poverty and social isolation. We were absolutely shocked that the candidates applying for our new positions were often Master's level social workers. Due to the candidate pool, we ended up hiring quite a number of MSWs,a few LCSWs, and some highly skilled health educators/health coaches. 3 years later we have learned a ton. Our community outreach MSW's(we now call them Health Resilience Specialists) have done a great job with this highly traumatized population (65% of whom have Substance Use disorders, 72% of whom have depression, 23% of whom have complex mental health conditions, and all of whom are living in poverty)...I'm speaking of our patients, not our staff!. Our health educators/health coaches have also done very well, however their comfort level around addictions, trauma, and mental health conditions is lower. All of our staff are trained in trauma informed priciples and motivational interviewing, and they are truly inspiring in their ability to level the playing field, engage and build trusting relationships with our vulnerable patients, and reduce suffering through their compassion, social justice values, and common condition of humanity. BUT, what we believe we are missing is the common "lived' experience that most of our staff do not have relative to their clients.
For us, CHWs might be that partnering workforce but we are leaning more in the direction of a "peer" workforce, which could be slightly distinct. In Oregon, the CHW workforce has been largely defined by cultural identification or neighborhood identification. When we look at the demographics of our high risk/high cost Medicaid and Dual population, there is not a large minority representation (many smaller racial and ethnic groups) and they are geographically dispersed. But, what has really stood out as a common lived experience is trauma, addiction issues, and some mental health conditions (predominently PTSD, anxiety, depression). So, we are on a journey to partner with community agencies that hire and train "peers" who have lived experiences in these domains. We have about 7 peers so far who are working with our Health Resilience team, and they are adding a tremendous amount of value! The facilitators have been: (1) bringing them into our team huddles and community of practice to build a shared knowledge and vision, (2) having the Health Resilience Specialists identify the appropriate and willing clients and introduce them to the peers, (3) helping the peers learn about and navigate the primary care setting through shadowing, etc, (4) and pairing the peers to certain clinics (and by virtue of that, Health Resilience Specialists) to help reduce drive time and strengthen relationships. The barriers have been: (1) most peers cannot work full time, so their hours vary and scheduling can be challenging, (2) lack of knowledge about primary care settings, (3) varied training programs - HIPPA and privacy practices are not always understood.
Hope this helps!

Heidi Behforouz, MD Replied at 5:57 PM, 18 Aug 2014

Thanks, everyone.
First... for all the groundbreaking and HARD work you have done to show us how to better care for high need patients.
And second, for the opportunity to learn from you.
A few questions:
1) If you were to go back to the drawing board...would you do anything differently?
2) What do you see as the potential pitfalls that care systems will face as there is more and more of a push for them to manage their high need/high cost patients?
3) What are the best ways to identify patients who will most benefit from your care management? How do you assess "impactability" as many predictive tools do not function very well in identifying patients who will respond most to your efforts?
4) When do you "give up" on a client? Do you have a protocol to handle patients who don't respond to your efforts? What do these look like?
5) Are any of you in the business of "scaling up" your efforts? Do you think it best to do this by becoming vendors of care management for different communities or by providing TA to systems to develop their own care models?

Eleni Carr Replied at 5:58 PM, 18 Aug 2014

Question: I work in semi-urban area just north of Boston and I oversee complex care management services in about 10 community health centers. The goal of our complex care program is to manage the care of very complex patients - the 5% of patients who consume up to 50% of healthcare resources - patients at high, high risk for ED and inpatient visits, discoordinated care and signficant clinical comorbidities. This target population is small subset of patients who require intensive care coordination and support. The challenges we face is that our care managers receive referrals from clinic providers and staff for more routine patient needs around food, housing, transportation, med issues, medical equipment, immigration problems, etc.... These are not necessarily needs of the most complex patients but a large cadre of our panels - let's say up to 30-40% of our patients. Trouble is, if we focus on these everyday needs of everyone, we can't focus on the more complex care needs of that top 5%. How do other clinics handle this? What suggestions do folks have about how these needs can be met with existing resources? Thanks, Eleni

Charla Parker Replied at 7:16 PM, 18 Aug 2014

It might be helpful to develop criteria for referrals for complex care
management versus referrals for patient care support. Perhaps you have a
resource data ase that can be shared and utilized by a MA/referral clerk or
the individual MA's assigned to the patient care team.

Note: tracking referrals is a requirement for NCQA PCMH recognition, so
perhaps the IT director could be enlisted to help with automating and
tracking referrals for both complex care management and patient care

Charla Parker, MPA, NCQA PCMH CCE
Practice Coach

Charla Parker, M.P.A.

Practice Transformation Facilitator

Clemens Hong Replied at 9:32 PM, 18 Aug 2014

Eleni Carr. Thank you for your comment. From your description, I'm not sure which issue is at play here, but we saw two distinct issues in programs we studied that may relate to the issue you are facing: 1) a lack of understanding of the program and program goals and 2) operational control - who "owns" the complex care management team?

1) Is it possible that the providers and practice leaders either don't truly understand what the program is designed to do or have so many overwhelming, unmet needs in the social service space that they are taking advantage of your resources? One thing that programs did to address this issue is to educate and re-educate provider and practice leaders over and over again about the purpose, goals of the program, etc. Providers frequently do not understand what care managers do or what a care management program is designed to achieve - or they forget. Some programs did quarterly practice "roadshows" where they addressed issues surrounding the program and did this critical re-education. This approach might help you address provider and leader concerns around the referral process while simultaneously improving the quality of referrals. Really clarifying what is an appropriate referral would help, as Charla Parker suggested. You could even create a quality metric around appropriate referrals and drive improvement by giving feedback to practices on the quality of their referrals.

2) Or is it that the operational control of the program held by the practices - so they are doing what they want with the resource to address their unmet needs? Programs that had central operation control of their care management personnel were often better able to ensure that the resources were used to meet program goals or in your case that referrals could be limited to ensure that referrals were appropriate. In contrast, when operational control was held by primary care practices, it was more difficult to protect the program without strong leadership or a practice champion in place to do so. As a result, some programs moved the locus of control from the practice a central organization.

Steven R. Counsell, MD Replied at 10:45 PM, 18 Aug 2014

Heidi, Thanks for the excellent yet challenging questions! I will try to address your questions #3 and #5...

#3) Identifying patients who benefit most and will engage. - In our experience with the GRACE model, having specific enrollment criteria that identify patients at "high risk of hospitalization" combined with close collaboration with the patient's primary care physician in confirming patient appropriateness and enrollment into the program are all key. Simple criteria for "high risk" older adult 65+ include multiple chronic conditions, geriatric condition (e.g., depression, dementia, or falls), functional limitations, lives alone or caregiver burden, and hospital admission in past year. We have found that enrolling patients at transition from hospital or SNF to home is a point in time when the patient more often has a perceived need and accepts help. Since GRACE is a home-based intervention, if the patient does not accept a home visit, this is a natural identification of someone who is likely not to engage with the complex care management program anyway and vice versa.

#5) Scaling up - become vendor or provide TA? - We are doing both in Central Indiana in GRACE Team Care. For example, we have become a "vendor" for IU Health Plans Medicare Advantage program in which the health plan pays for the program which is run by our Geriatrics program. At the Indianapolis VA, however, we provided training and technical assistance in their start-up of GRACE. For health plans and systems outside Indianapolis we are providing training and technical assistance through the GRACE Training and Resource Center. This function is consistent with our mission as an academic medical center to provide inter-professional education in complex care management for frail and high risk elders.
My comments are oversimplifications but hopefully are still of some help.

Thanks again! Steve

Steven R. Counsell, MD
Mary Elizabeth Mitchell Professor
Director, IU Geriatrics
Service Line Leader for Geriatrics, IU Health Physicians
Scientist, IU Center for Aging Research
Indiana University School of Medicine

Rebecca Ramsay Replied at 11:03 PM, 18 Aug 2014

Eleni Carr - Really important questions. To underscore what Clemens has written: Our CCM program is funded and is administerred by the regional Coordinated Care Organization (and funded by the health plans) so we have been able to set the program's primary goals, create data-driven referal criteria, and then educate providers on "why" we have approached the program in this way. It is true that it takes time for providers to truly understand why we are targetting certain patients, and they often want us to be more flexible in our entry criteria because there is so much need for support, especially with the safety net populations - but we have had a lot of success through relationship building and reeducation. Part of our centralized infrastructure design is to dedicate people like me (and a program manager) to work on this collaborative knowledge transfer and create a common vision with our network.

Rebecca Ramsay Replied at 11:25 PM, 18 Aug 2014

Heidi - thanks for asking these questions. I will try to address #3 and #4:

Identifying patients and assessing for impactability:
We have found that a combination of real-time utilization event reports (ED and Hospital) AND provider/community referrals is the best method of identification. The real time event reports give our staff the ability to engage and assist patients when they are in some sort of "crisis" which seems to do two things (1) offers a window of openness to change and to accept support among the patients, and (2) gives our Health Resilience Specialists a more direct window into what triggers these crises and utilization events for their clients. We have a centralized triage coordinator who scrubs these utilization reports looking for those that are more clearly 'avoidable' - this allows us to approximate impactability. The provider and community referrals improve over time as people better understand the program and the target population. I have a bias that many PCPs, especially those with years of experience and with long-standing relationships with their patients, are better predictors of high cost utilization than most predictive models. So, we spend time talking with providers about how they might refer in this way.

The thing I'll say about "impactability", and this also addresses question #4 about "when to give up on patients" is that if you simplify this, there are two factors to consider - the readiness of the patient and the tools in the toolbox of your care managers/care teams. In terms of readiness, we've seen absolute miracles happen - patients who everyone had given up on suddenly become activated and ready to we've learned to plant seeds with patients, not spending a lot of time when they dont seem ready, but always keeping an open door. As the founders of Motivational Interveiwing, Rolnick and Miller, have concluded, the process of change is never static...its a continuum we are all traveling along every minute we are alive, so its really hard to predict when a person will shift. In terms of the tools in the toolbox, early on we learned how prevalent addiction was in our, we made some changes to our training curriculum for all staff, put a staff person in a detox setting, and began partnering in new ways with addiction providers - these programatic additions give us new tools that now change our definition of "impactability".

Kenneth Coburn, MD, MPH Replied at 11:47 PM, 18 Aug 2014

Heidi, in response to your Q's #3,4,5

3 - Based on the subgroup CMS / Mathematica identified as having the biggest reduction in hospitalizations and cost using our model, we are targeting Medicare, over 65 yo's with heart failure, coronary artery disease, COPD, or diabetes (1 or more of these Dx's) and at least 1 hospitalization in the past year (due to any cause). This represents about 16-18% of all Medicare beneficiaries.

4 - We never give up on a participant in our program ... provided they are at least willing to continue to be contacted by us. Our model is very longitudinal. Once meeting eligibility criteria, we don't "discharge" people from the program. Evidence available from our demo suggests that those in the program for extended periods continue to derive benefit. Patience is indeed a virtue that luckily our nurses have in abundance, which often pays off in the end. Patients and/or families do in fact change, if you get to work with them long enough and have a sufficiently solid relationship.

5 - We are in the midst of "scaling up" our efforts. To do so, we're developing an integrated set of capabilities specifically designed to support implementation of our program in a scalable IT platform. The capabilities include; policies and procedures, staff training modules, participant educational materials, data management with decision support, and advanced analytics with statistical process control charts - all available through a single sign on, website, useable on mobile devices (we've been using iPads). In addition, we are offering this along with management consulting, training, and mentoring to help others learn our model, which when combined with the IT platform provides a kind of 'start-up service as a system'.


Kenneth Coburn, MD, MPH Replied at 12:11 AM, 19 Aug 2014

On the observation that "Evidence for complex care management is inconsistent ..." I'd offer the set of design principles that we think are critically important to sustained effectiveness of an advanced preventive care model; person-centeredness, population-relevance, robustness, and reliability. Being explicit in addressing each of these areas has been a great help to us.

Person-centeredness: involves starting with wherever the person is. We use motivational interviewing and the transtheoretical model of behavior change along with compassionate, non-judgmental, steadfast support and engagement.

Population-relevance: involves identifying as many determinants of health for the target population served as possible. Those making the final cut are those for which other smart folks have already demonstrated effective interventions or we think we have a way to address.

Robustness: means including as many different interventions (30-35) to address as many important determinants of health as you possibly can. Robustness also means sticking with patients for the long haul, and interacting with them frequently, mainly in person as a one to one encounter or in group classes.

Reliability: relates to having a management system (we monitor over 200 process measures) that can detect variation from the intended care delivery in order to rapidly and vigorously address these. Another element of reliability is having back-up nurse coverage and the ability to flex coverage or assignments when case loads or patient panel acuity becomes too great for a given nurse. Reliability also means a standardized system of staff training with ongoing refresher and mentoring efforts. Finally, we regularly survey patients about their experience in the program.


A/Prof. Terry HANNAN Replied at 12:34 AM, 19 Aug 2014

This is a very interesting discussion. Just today I received notification of a publication that I believe has relevance to this discussion and it is authored by persons who have contributed to the GHDonline site on a regular basis.

christophe millien Replied at 1:35 AM, 19 Aug 2014

#1: in my country the situation is dificult.we have lot of challenges-
Lake of qualify human ressources. sometime we use task shifting.
Internationals parteners help us by training and by creating appropriate
infrastructure like mirebalais university hospital in a win win
partnership. One way that we use is
To make that is to make the vision and the mission of you program to be
the own vision and mission of all people who involve in .beside that
valorisation of the work of any person who involves in the program must be
well done. We did our work in an integrated context bedside of others
clinical care.

Clemens Hong Replied at 2:46 AM, 19 Aug 2014

It’s clear from Ken Coburn’s comments the intense approach HQP takes to program management and data driven improvement (over 200 process measures!!) to ensure reliable delivery of their model. One thing that is clear from our study of these organizations is that no one has the answer at the beginning. They start with the best approach they can define and continuously improve it. HQP’s approach is exceptional, and I can only imagine how much those metrics and their understanding of the processes that drive their outcomes aid them as they continue to scale their model. One question I have for Ken is how their care managers balance achievement on 200 process metrics with a clear focus on relationship building and engagement? Do you have “relational” quality measures? I would love to hear more!

Attached resource:

Rebecca Ramsay Replied at 2:48 AM, 19 Aug 2014

I wanted to respond to Heidi's questions #2:

#2 Common Pitfalls as Health Systems take on high utilizer population and programs:

I worry that health systems will react from their comfort zones, and that they will try and do more of what they've already tried. What I see most often is a reliance on an overly "medical" lens, and not enough attention paid to social determinants, inc things like social isolation, cultural marginalization, traumatic experiences which influence inforamtion processing, health literacy, and the more commonly talked about behavioral health risk factors. We also need to provide opportunities to view our most vulnerable patients lives on their turf, in their home and community settings - there is so much that influences a person's health and wellbeing that we miss when we only view them in "our" setting. When each of us walks into a traditional health care setting we become a "patient". What we need to learn is how to assess our patients as people, and how to help them live into that identity above and beyond the role of a patient. Going to them is one way to increase the odds of that happening. Using peers, CHWs, and other non-traditional workers can help with this challenge. The other thing I worry about is that programs/new interventions will be started on shoe string budgets that dont allow for the adequate infrastructure to be successful - things like clinical supervision for care management staff, process improvement, data and analytics, mobile technology, program leadership, learning systems, evaluation.

Clemens Hong Replied at 2:58 AM, 19 Aug 2014

Rebecca, I think CareOregon’s health resilience program is incredible in that they work with patients that most programs steer clear of because they wouldn’t know where to begin engaging them in care. The level of trauma and adverse childhood events history is incredibly high. Your increasing use of peers is interesting, and it seems your peers themselves carry a heavy burden of trauma, addiction, and mental health conditions. You cite several barriers in your comments that seem like they could be addressed by hiring them full time as members of your teams. What are the barriers to doing this, from your experience with your 7 peers to date?

Attached resource:

Clemens Hong Replied at 3:19 AM, 19 Aug 2014

On Heidi’s question #3…We are currently completing a study funded by the California Healthcare Foundation on patient selection in complex care management programs due out later this year. We asked 20 programs, including the ones on this panel, this same question. I think the responses from the other panelists have been spot on: 1) use different (e.g. multiple) approaches to triangulate on a high-risk populations (e.g. acute care utilization, markers of disease severity or gaps in care), 2) meet patients at times of high need (e.g. post-hospitalization), 3) ask providers what they think, 4) consider patient readiness (e.g. motivational interviewing) as a key piece of impactability,

I would add two things that I think bring together some of the ideas posed earlier. First, the key to selecting the right populations for a complex care management program is to align the population and intervention to achieve the desired outcomes (includes the outcomes themselves and the time frame in which you hope to achieve them). What is impactable for one program may not be impactable for another - for example, a program with an underdeveloped behavioral health program should probably not select a patient population with a heavy behavioral health burden, whereas one with a robust behavioral health program may choose to focus on these patients. Second, the best practice approach may be to combine quantitative, data-driven risk identification (e.g. predictive models, acute care utilization or disease based entry criteria, etc) with qualitative, clinical assessment by the primary care clinician. This approach takes advantage of the strengths of both approaches. A key point here was raised earlier in this panel discussion – that the primary care clinician MUST understand the program (e.g. goals of the program, what the program does and does not do) to select the right patients.

Attached resource:

Clemens Hong Replied at 4:02 AM, 19 Aug 2014

In my reply to the question of the day on the inconsistent evidence for complex care management, I wanted to provide a construct for discussion. Timothy Ferris (a mentor, a co-author on the commonwealth fund report, and the principle investigator in the Massachusetts General Hospital Medicare Care Management Demonstration that led to 19% savings) uses a diagram adapted from Eisenberg to describe where CCM programs lose their potential to achieve program goals. I've attached the diagram, but I'll include a narrative explanation here using unnecessary hospital admissions as an example:

1) There is a real opportunity to reduce unnecessary hospital admissions - a group of patients that will be admitted unnecessarily to the hospital in the coming year,
2) You must first identify as many of these individuals as possible - if you only identify 50% (and that can be optimistic with current risk prediction algorithms), you've already lost 50% of the opportunity to reduce unnecessary admissions,
3) Then, you must engage those that you identify in your program - if you use a telephonic care management approach (as many have tried), you may engage only a small percentage of those that you identify, leaving very little potential for impact,
4) Once you successfully engage them in the program, you must successfully identify patient needs gaps in the patient's care that your program can address, and
5) Finally, you must reliably deliver the intervention to address the patients needs and thus reduce unnecessary utilization.

We posit that some programs do not adequately address key steps along this general pathway and thus lose their potential to achieve their target outcomes and program goals.

Attached resource:

Steven R. Counsell, MD Replied at 9:28 AM, 19 Aug 2014

In the older adult population, evidence supports that it is the patients having a combination of multiple chronic conditions and functional limitations (basic and instrumental activities of daily living) that have highest health care utilization rates and costs (The Lewin Group, Jan 2010). Since being old with functional limitations is common but not "normal" aging, and since persons with functional limitations frequently depend upon others for assistance, functional limitations identifies older adults likely to benefit from geriatric assessment and interdisciplinary team interventions. In addition, patients with dementia, depression, and falls have high health care utilization rates and costs, and these "geriatric syndromes" typically have multiple contributing factors demanding multifaceted interventions for best outcomes. Older adults with functional limitations and/or geriatric syndromes are thus the "sweet spot" for geriatric care management interventions.

In our experience in GRACE dissemination, we consistently find the following "keys to success" for engaging patients, identifying care gaps, and delivering interventions across the biopsychosocial spectrum:

1. Nurse practitioner (or nurse) and social worker in-home geriatric assessment including exploring a person's health goals and preferences. The in-home assessment not only identifies previously unknown health determinants and conditions but also facilitates relationships and usually tells "the rest of the story" as to a person's health status, medications actually taking, social issues, and self-management abilities.

2. Intervention is focused on geriatric conditions often not given adequate time and attention by the office practice due to limited resources (e.g., depression, dementia, falls, medication management, caregiver issues and social supports, etc.), thus complimenting and not duplicating primary care disease management (e.g., HTN, CHF, DM, etc.).

3. Home-based and proactive care management that integrates with community resources and social services, including for effective transitional care from hospital to home or SNF to home.

4. Longitudinal care that develops relationships with the older adult and caregiver, and primary care provider and practice, to build trust and facilitate health behavior changes and management adjustments over time in a continuous improvement manner.


Elizabeth Glaser Replied at 12:04 PM, 19 Aug 2014

I have two questions:

1.I see that some programs reach out to individuals at high risk of homelessness , such as those recently released from prison, but do any of the CCM programs specifically work with homeless or formerly homeless men and women to keep them in care? For example, Boston Health care the Homeless Program has successfully retained men and women in their primary care clinic for many years, but I am not sure if they espouse a particular model of care. Have any of the CCM programs partnered with homeless programs to improve care to this high care, high cost population?

2. I briefly read the Commonwealth report on cost and utilization, however I did not see mention of a thorough cost-effectiveness analysis of CCM programs; are you in process of doing such an analysis?

Thanks very much for this interesting and relevant panel,


Kenneth Coburn, MD, MPH Replied at 1:06 PM, 19 Aug 2014

In response to Clemens' Post #30 questions ...
How do care managers balance achievement on 200 process metrics with a clear focus on relationship building and engagement?
Building and maintaining an engaged relationship is job #1. A few examples. We look closely at (and have standards for and measure) patterns of patient-nurse contacts; frequency, type, duration, gaps. We measure the degree to which nurses change their pre-booked schedules "on the fly" to accommodate changing patient needs and look at the variation in doing so between our nurses. We do random post-visit calls to patients to assess the degree to which the nurse followed key protocols, including, among other things, person-centered inquiry and medication reconciliation. While a good chunk of the metrics we use are used by all nurse care managers, some are primarily used by nurse supervisors and program administrators. Having statistical control charts makes it easier to identify signal from noise at a glance, reducing the time needed to use these reports. In addition, some measures are more important than others with review cycles and attention paid commensurate with importance.

Relational quality measures? The most consistent source of data on this are mailed surveys. We generally have a response rate of 65-85%. The scoring of predefined questions is helpful, but the gold is in the open comments (and we typically get many!). Nurse care managers are regularly scheduled to have observed visits to patients selected by their nurse supervisor. And we look at variation between nurses in addressing difficult issues such as end of life planning, abuse and neglect, etc. We commonly see statistically significant variation on these areas of assessment and counseling that require additional mentoring by the supervisory nurses.

Janet Van der Mei Replied at 5:56 PM, 19 Aug 2014

Greetings, as a member of this panel I must confess I am a bit intimidated by the expertise of the other panelist and have appreciated their comments. I work for a large healthcare system in Northern California and am the director of our Ambulatory Case Management Programs in one of the five regions of our healthcare system. I am part of a regional case management department which spans the continuum of care. My programs are funded by the six hospitals in our region, our Medical Foundation, as well as an IPA and a Medical Group. The basis of our programs is our Sutter Care Coordination Program comprised of teams which include RN case managers, LCSW’s, Health Care Coordinators and most recently Pharmacists. The teams are physically embedded within the Medical Group/Foundation care centers and support all of our Internal Medicine and Family Practice physicians. Our IPA doctors are supported telephonically. Although our program started as a result of moving into capitated arrangements with our contracted health plans, we now support any high risk patient who has a Sutter PCP.

Identifying the high risk patients has been a work in progress. We have recently developed a program to help identify these patients which we use in conjunction with our PCP's to identify those patients we should proactively engage. We also identify high risk patients at the time of transitions from one setting to the next; hospital, SNF, Home health, using specific criteria. These patients are then followed for at least 30 days post discharge and on-going should continued follow up be indicated based on the case managers assessment and interventions needed.

Jan Van der Mei RN, MS, ACM
Ambulatory Case Management Director
Sutter Health Sacramento Sierra Region

christophe millien Replied at 8:03 PM, 19 Aug 2014

#2 in my country the most critical factor is poverty. This poverty is
organize by that we call structural violene mainly responsible for social
exclusion. The bad ecomic condition will discrease access to the health
system, to basic primary need like food( malnutrition )
housing(promuscuity) education etc. This situation that we call social
exclusion is associeted with the weakness of the health system link to bad
economic condition to make services available and accesible to the
population. Beside that the notion that we call social justice and equity
sometime is not in the Center of all decision making and create more
inequalty between social groups and favor more complex health problem that
the health system can't solve really.

christophe millien Replied at 8:31 PM, 19 Aug 2014

#3 our keys partners are :
1- the medical team.
2 - the finance team.
3- our patients
4- the community health workers.
5- politic men in power and activists
These elements are very importants for manage complex health care program
but all these things must be well defined by good process and good
monitoring and evaluation system. One important point is to avoid to have a
vertical program with little impact on the health care of the population.

#2 in my country the most critical factor is poverty. This poverty is
organize by that we call structural violene mainly responsible for social
exclusion. The bad ecomic condition will discrease access to the health
system, to basic primary need like food( malnutrition )
housing(promuscuity) education etc. This situation that we call social
exclusion is associeted with the weakness of the health system link to bad
economic condition to make services available and accesible to the
population. Beside that the notion that we call social justice and equity
sometime is not in the Center of all decision making and create more
inequalty between social groups and favor more complex health problem that
the health system can't solve really.

Rebecca Ramsay Replied at 8:43 PM, 19 Aug 2014

Clemens raised some astute questions earlier today about including peers on our team of Health Resilience Specialists. I completely agree that we could probably reduce some of our operational barriers by hiring them as part of our team - this was a learning we had in the first two years of our program. We considered doing this, but for a couple of reasons we haven't: (1) there are a number of local community organizations we've recently learned of that have been hiring, training, and deploying peers for many years - it feels more collaborative and perhaps over the long haul, more effective to take advantage of their experience by contracting with them rather than trying to upskill ourselves, (2) We've also learned that the supervision, training, and support peers need is different than what our Master's Level staff need....before we take that leap, we want to test whether we can accomplish the same thing by partnering rather than building ourselves.

Rebecca Ramsay Replied at 10:50 PM, 19 Aug 2014

In response to Elizabeth Glazer's questions in Post #36:
(1) We have many homeless individuals enrolled in our Health Resilience Program. We do not specifically target this population but they make up a significant number of our high cost patients. Prevalence is hard to accurately determine bc we are just adding housing status to our data tracking (since there are no good markers anywhwere else), but our recent data suggests at least 20% have unstable housing. We do have partnerships with housing providers, however there are many, many regulations and policies governing how subsidized housing is distributed so we dont have preferential treatment for our clients unless they risk into priority groups. What we have learned is that we can help an individual recognize the need for housing when they are clearly eligible, but hesitant or fearful, and we can advocate for them when it comes time to determine the appropriate kind of housing, etc. Sometimes we also have to advocate to get the necessary assessments completed.

(2) in terms of cost effectiveness evaluation of our program, we have been noses to the grindstone over the past 8 weeks with our evaluation team trying to develop a rigorous "propensity matched" comparison group to improve the robustness of our pre-post time series study design. We have great utilization trends, but we realize the necessity of this extra rigor. Unfortunately, we have been unsuccessful to date. This is a very complex science, and when you serve the kind of population we do with such high prevalence of social and environmental risks that cannot be captured in claims data (which is what is typically used to create matched comparison groups), the risk of unmeasured bias in the treatment group is high. That appears to be our problem right now. So, we are turning to the literature and have found a few researchers who seem to be grappling with this very challenge and developing some new matching methods that we are about to try.

Elizabeth Glaser Replied at 12:49 AM, 20 Aug 2014

Dear Rebecca,

Thanks for you response - I am very glad to hear that you do try to address housing , either directly or indirectly, as it definitely impacts adherence to medication, attending appointments and general health outcomes. If case mangers and providers are aware of a patient's housing status it can help them to more realistically discuss and trouble shoot issues with diet and medication adherence which may come as a result of unstable housing.

As for propensity score matching - yes, it is a challenge, but is one of the better means to reduce selection bias in a retrospective observational study. I have attached a link to fairly recent paper from academy health about using electronic health data for observational studies which discusses this as a general resource for those who are not aware of this method.


Attached resource:

Clemens Hong Replied at 2:59 AM, 20 Aug 2014

Christophe Millien - thank you for your posts and contribution our discussion. You make many points that are relevant here in the United States. In our work at Anansi Health, we are very much driven by social justice principles. We have tried to take lessons learned from Haiti through Partners in Health and apply it here in the care of the poor. I think we will see much more community engagement in health care delivery and increasing use of community health workers here in the coming decade. This will hopefully lead to a more unified focus around the whole person and around community development and health, rather than solely healthcare delivery. In the US, vertical approaches have also led to significant fragmentation of healthcare delivery. In the complex care management approaches we are discussing on this panel, we are seeing programs built around primary health care, addressing the whole person, and attempting to coordinate care across this fragmented system. For us, here in the US, faced with extraordinary health care spending and waste, successful complex care management may not only address the health needs of our sickest patients and begin to constrain unnecessary spending, but also serve as further impetus for work upstream - perhaps by reinvesting the savings we generate - at least that would be my hope. I know in Haiti, you face different challenges, and I do not understand the context well enough to comment, but please do let us know if there are ways we can help. I'm not sure if there have been facets of this discussion that have been relevant to you and other international audiences. If you or others have thoughts, I, for one, would love to hear them. How does this work translate to Haiti, Rwanda, India, China, Mexico, Navajo Nation, Germany, Sweden, etc?

Clemens Hong Replied at 4:10 AM, 20 Aug 2014

In response to questions by Elizabeth Glaser:

1) Many of the programs that we interviewed in the Commonwealth Fund study interacted with homeless patients, but few had significant resources to address the core issue - homelessness. I think the Boston Healthcare for the Homeless Program functions in many ways as a complex care management program. The Boston Healthcare for the Homeless Program's Street Team and many like it are great examples of complex care management. They have similarities to some "home-based" primary care models, such as the VA model or Commonwealth Care Alliance, but of course they are distinct. I also think housing first approaches and respite programs incorporate many of the activities of complex care management programs, and I have often thought that they could be interesting places to concentrate regional complex care management efforts. Camden Coalition is famous for hot-spotting by mapping patients who recurrently use healthcare services. Could we purposefully concentrate these high-risk, marginally housed individuals, by housing them first together and then building an advanced care model (e.g. with primary care, complex care management, social services, etc) around them? I'm sure there are models that are doing this currently. Are any of you doing this work?

2) I'm not sure which report you are referring to in this question. However, if you are referring to our report on complex care management, we are not planning additional cost-effectiveness analyses of these programs. A robust analysis of cost effectiveness would undoubtedly be a terrific contribution, and I do believe we will see many analyses come out of the demonstrations that are popping up all over the country. Many programs did their own analyses, and unfortunately, many of them did not use methodologies that were rigorous enough to allay fears that the benefits they saw were due to regression to the mean. That said, there are a number of robust analyses - mostly in the grey literature. I let the panelist comment on their own data, however, the evaluation of the Massachusetts General Hospitals Medicare Care Management Demonstration (matched control) and King County Care Partners (randomized control) are two examples that are not represented on this panel.

Also, thank you for sharing the article. I think data challenges abound in control group selection. In LA, where we are planning a rigorous evaluation of a new CHW-integrated complex care management program, we will perform a randomized evaluation as long as the implementation teams are willing - which remains to be seen. I think we sometimes forget that where you have limited resources (e.g. a care management program that can only take care of 500 patients, when there are thousands of high risk patients with need), the most equitable way to allocate the resource is through randomization. Then, you might also have the potential for the strongest data to support program dissemination. Randomization is definitely NOT always the answer, but I think we operationally choose other quasi-experimental designs without fully weighing the potential benefits of randomization.

Clemens Hong Replied at 4:37 AM, 20 Aug 2014

In response to the question of the day (Panel question #3)...The critical partners obviously depend on the patient population. Like so much in complex care management, the trick is to be prepared for anything, but right fit the intervention to the patient and the patient's circumstance. Given the high levels of biopsychosocial complexity in our patients, we need strong relationships with numerous different entities from hospitals and ERs to SNFs to VNAs to community behavioral health programs, to social service entities. It can be overwhelming for programs to develop so many relationships, and as a result many developed these relationships on the fly - e.g. start the work and then build a resource guide as you go and make it available to all complex care management staff. One interesting model, however, is the Vermont Community Health Team approach. They held community team meetings every month, to which all regional healthcare and social service delivery organizations were invited. At these meetings they addressed areas they could work on collectively to improve coordination across the different entities. They often used case examples to drive discussion and development of systematic solutions. Attendance and organization varied from region to region, but those that bought into this process had many success stories.

I do hope that Rebecca Ramsay will have time to share CareOregon's acute care plan approach to addressing recurrent ED utilization in some of their patients. This has led to incredible engagement from emergency departments in their service region.

I'll give just one other example from our commonwealth fund study. Given the importance of hospital-to-home care transitions in many of our programs, a key partner is the hospital. In addition to real-time hospital discharge/admissions notifications seen in multiple programs, many programs still saw a significant gap in information sharing from the inpatient setting to the complex care management team. Several programs placed a care management team member in the hospital, as a hospital liaison. This person would see CCM team patients in the hospital, ensure that information was transferred to the primary care team and CCM team in a timely fashion, and be the eyes and ears of the CCM team. This person got to know hospital care managers and other care team members over time, and strengthened the partnership between the CCM teams and the hospital partner. GRACE also addressed the hospital-to-home problem by leveraging their inpatient geriatric consult team in what they call their ACE to GRACE approach. I'll leave this to Steve Counsell to describe, if he chooses.

Kenneth Coburn, MD, MPH Replied at 6:30 AM, 20 Aug 2014

Today's question: Who are your critical partners, and what approaches have you used to best engage them?

There are many partners to HQP that are important; our own staff and Board of Directors, participants (patients) their families and support network, physician practices, hospital-based health systems, and the many community service providers in our region (meals on wheels, Area Agencies on Aging, home care services, nursing homes, and many more). A new group that is emerging are program design and development partners. Examples include; 1) a software application development partner helping us build a scaleable, industrial strength IT platform specifically engineered to support our care model; 2) health care organizations or collaboratives of providers wanting to engage HQP to facilitate the design of better models of CCM for vulnerable populations using our design principles, framework, and implementation capabilities. We are also working on establishing (none finalized yet) strategic commercial partners who can profit from our work, by replicating it to scale in a business context with royalties flowing back to HQP in order to support our ongoing, non-profit, applied R&D work to continue to design, test, and disseminate better models of CCM.

For all of these partners we try to apply certain principles. Cultivating positive personal relations, rapport, and high quality communications with individuals is important for everyone of these partnerships. It isn't always easy to figure out how to identify who to reach out to, but making the effort to do so has been very helpful. We've found that honest, open communications, transparency, a willingness to go the extra mile, and actively listening to the needs of our partners has been extremely helpful. We also try to enlist our partners to the greater cause of reducing human suffering through support of our work. In the end we've found a good measure of the quality of our partnerships are the degree to which we can collaboratively problem-solve challenges or barriers together. To date, explicit financial incentives have played either no or a very minor role in making our partnerships effective.

Steven R. Counsell, MD Replied at 10:09 AM, 20 Aug 2014

Although not a formal "cost-effectiveness analysis", we performed a rigorous cost analysis as a part of the NIH funded randomized controlled trial of the GRACE model published as follows:

* Counsell SR, Callahan CM, Tu W, et al. Cost analysis of the Geriatric Resources for Assessment and Care of Elders care management intervention.< J Am Geriatr Soc. 2009; 57(8): 1420-1426.
In brief, we found in the high risk subgroup that hospital and ED utilization and costs decreased each year in the intervention group relative to the usual care group with statistically significant cost savings in the intervention group in the third year. This substantiates the value of complex care management interventions in high risk elders and the importance of establishing a longitudinal program since relationships, self-management skills, and medical management adjustments take time to implement and results are often not immediate.

Janet Van der Mei Replied at 3:53 PM, 20 Aug 2014

In response to the question of community based partners that have been critical to our success.
Our most recent success has been our collaboration with our Skilled Nursing Facility (SNF) partners. We had a readmission rate from our SNF’s of over 20%. We began meeting with key SNF partners on a monthly basis. We provided disease specific education for their clinical staff (Heart failure and COPD/Respiratory) we created specific order sets for SNF admissions to ensure critical clinical components of care were addressed after admission. We assist with the discharge arrangements to make sure our patients have home health following their discharge from the SNF if it is indicated. We developed a hand-off process so that when our complex patients are discharged from the SNF we follow telephonically to make sure they have all of their medications and understand how and why they are taking them. We facilitate follow up appointments and transportation to see their PCP. Our SNF readmission rate is below 10%. We have collaborated with our Home Health partners in a similar way.

Clemens Hong Replied at 4:58 PM, 20 Aug 2014

Dr. Coburn, your comment on scaling through for-profit entities is very interesting. Our NEJM perspectives piece touches on things that might lead help accelerate adoption of CCM models, but one thing we do not discuss in detail are the business models best adapted to scale these interventions in different contexts. Is there a reason why you chose not to scale it yourself as a non-profit entity? What are the pros and cons of lending your R&D to for-profit companies? I could see advantages to scaling as a non-profit - the most important being the ability to reinvest the savings to grow and build health delivery infrastructure rather than pay shareholders. Of course, the big advantage to for-profit entities is the availability of investment capital to accelerate the ability to scale. Commonwealth Care Alliance recently received a large social impact bond to fill their reserves so they can grow more rapidly, but I imagine raising capital in the non-profit space is no easy task.

Clemens Hong Replied at 5:11 PM, 20 Aug 2014

Adding to Jan Van der Mei's comment on SNF's as partners, MGH has two interesting approaches to engaging SNFs. The first is the SNF waiver. MGH negotiated a waiver with Medicare that allows the program to bypass the 72 hour hospital stays for SNF admissions. This allowed MGH care managers and primary care providers to decide when a hospital stay was not necessary - allowing for direct admission to SNFs - for example for IV antibiotics - saving significant money in the process. I believe Geisinger also uses utilization management nurses to approve similar direct SNF admissions among people insured by the Geisinger Health Plan. A second initiative is SNF telerounding. MGH care management team members round weekly with SNF staff in skilled nursing facilities where MGH patients are concentrated via teleconference. This provides opportunities for face-to-face engagement and improved care coordination.

Sutter Health's CCM program is also well integrated with their Advanced Illness Management program. Jan, could you and/or other panelists talk a little about how you partner with palliative care and hospice partners to address end of life issues in your patients? I think this is a critical piece.

Marie Connelly Replied at 6:14 PM, 20 Aug 2014

Many thanks to everyone for such insightful questions and comments in our Expert Panel so far. Thank you to each of our panelists as well for being so open and willing to share your work—it's incredible to learn from all of these examples.

For all of our members here: our focus in the discussion tomorrow will be around do's and don'ts. Our panelists will share lessons learned and what they've found to be key elements of successful programs.

First though, it would be helpful to hear from you:

What are the biggest challenges you currently face in your own complex care management programs? If CCM programs are not yet a part of your work, what do you believe are the largest barriers to implementing such programs in your practice or delivery system?

Please share your thoughts—hopefully our panelists will be able to offer some specific suggestions for things to do (or not do!) to address the current challenges you're facing.

Janet Van der Mei Replied at 8:37 PM, 20 Aug 2014

The Advanced Illness Management (AIM) program is basically palliative care. Patients are often identified in the hospital and referred to AIM at the time of discharge. Referrals to AIM also come from the patients PCP as well as our embedded case managers. Patients who are likely to meet AIM eligibility criteria have a high chronic disease burden and are in active decline, the doctor would not be surprised if they died in the next year, and they wish to continue disease modifying therapy but could benefit from a palliative care approach. They are seen by an AIM Home Health RN who establishes a symptom management plan and goals of care. Once they are established, our embedded case managers follow them telephonically following specific protocols. If the patient has the need to be seen in the home due to worsening symptoms, our team is able to request additional home visits by the AIM Home Health RN. One of the goals of AIM is to work with patients over time to enable them to understand and identify their end of life wishes. Moving patients into Hospice in a timely manner is one of the metrics for the AIM program. Many times patients will accept AIM who are really appropriate for hospice. Eventually they often accept hospice.

christophe millien Replied at 10:00 PM, 20 Aug 2014

#4 the key point is to make sure that your leedership make the majority of
your people involve, understand, and take the project as their own
project. By doing that we have a big chance of success. We need to have an
action plan that contain what we would like to do in the time, material and
finance ressources associeted with budget. Objectives must be well difined.
Evaluation with the team to see what we have to improve. That mean
monitoring and evaluation must be etablished correctly. The challenge in
the implementation is about details. Any important detail that you don't
take into account can be a cause of non achievement. Don' t forget that.
Beside that human, material, and finance ressources represente the big
challenge in poor ressource setting for complex care. Don't forget to
indetify the process, data flow and patient your communication plan

Ruth Staus Replied at 11:39 PM, 20 Aug 2014

Seven years ago I started a solo Geriatric Nurse Practitioner clinic that provides free medical case management services to low income elders who are also chronically and persistently mentally ill. I did this at the request of a local state-funded housing agency. The social workers in the high-rises where these elders reside were seeing huge gaps in the coordination of care. Many of the elders that I see have care managers assigned through their supplemental health plans. They tend to manage their care needs over the phone and it is very rare that they ever actually come out to the elder's apartment to see what is really going on. I have other elders who have numerous case workers (I had one lady in her late 80's with 22 professionals involved in her care) but none of them talk to each other much less to the primary care provider. I often end up taking care of issues myself because it takes a lot less time than trying to figure which person is responsible for assisting the elder. Many of my elderly patients hide their illnesses and problems with obtaining medications from their caseworkers who only deal with them over the phone. My clinic is in the high rise where they live so it is pretty easy for me to see that there is a problem and I have worked hard to establish trusting relationships with these elders so they feel comfortable coming to me for help. The lack of training in geriatrics for all levels of providers is also a huge problem. With only 316 geriatricians teaching in US medical schools, 5 medical schools with geriatric divisions, and about 250 doctorally prepared clinically active geriatric nurse practitioner faculty teaching clinical geriatrics in nursing programs ( I am one of them), I am wondering how effective care coordination can be. In the past week I saw an elder who had just been in the hospital for new onset confusion and gait problems. She had about $5,000 of neuro work up with nothing found and was sent back to her apartment. Her B 12 level was low normal but anyone trained in geriatrics knows that a symptomatic elder with a B12 level under 400 needs further testing. She had a Vitamin B12 anemia which I was able to successfully treat. I had a 95 year old who had been discharged from the hospital where she had been hospitalized with acute onset of poly articular joint pain. She was sent back to her apartment on narcotics which were not helping. Based on her history and one lab test I was able to diagnose PMR and successfully treat her with prednisone. Both of these elders had common geriatric medical issues which get missed with frightening regularity. I have seen numerous patients given dementia diagnoses by a primary care provider who did no work up and the elder turns out to have a brain tumor or normal pressure hydrocephalus. I have no idea what the answer is but from my perspective as a clinician in the trenches of primary care geriatrics for the past 30 years, things are not going well for my patients despite truckloads of services and case workers.

Steven R. Counsell, MD Replied at 8:00 AM, 21 Aug 2014

Thanks Ruth for so well articulating the many issues confounding the care of frail older adults in our current health care system. Your dedication to better serving the complex health care needs of this population is palpable. The GRACE model was inspired by what you describe and involves a nurse practitioner and social worker working in collaboration with the primary care physician. The NP and social worker are supported in weekly team conferences by a geriatrician, pharmacist, and mental health liaison who specialize in geriatrics. Together the GRACE team specifically addresses common geriatric conditions and provides care coordination between multiple settings of care and various physicians and other health care providers integrating medical and social care. Frail elders require a distinctly different care management program that helps deploy clinical principles of geriatrics to supplement the primary care team who typically as you describe have little expertise and/or resources to address their unique issues. Thanks again for your insightful comments. Steve

Steven R. Counsell, MD
Mary Elizabeth Mitchell Professor
Director, IU Geriatrics
Service Line Leader for Geriatrics, IU Health Physicians
Scientist, IU Center for Aging Research
Indiana University School of Medicine

Janet Van der Mei Replied at 2:47 PM, 21 Aug 2014

What are the keys to successful implementation of CCM? What challenges have you encountered with implementation? What “do’s” or “don’ts” have you learned?
Successful implementation begins by identifying your population. Who are the patients who over-utilize or under-utilize care and who would we expect to become sick in the future? What social or behavioral issues do they face and what is the best intervention for them? Key to our success is our partnership with our physicians. Having a team of RN’s, Social work, Coaches and Pharmacist all working together with the physician to identify a plan of care. Also important is changing our approach from one of telling patients what they should do to one of coaching and teaching self-management skills to patients and care-givers. Other key issues are hiring the right people, consistent education, training and on-going mentoring. Having evidence based guidelines to direct new case managers to all potential interventions they should consider when working with a complex patient is also important to have in place.
Our challenges were similar to one identified in an earlier post. Physicians have become used to referring all types of patients who are not necessarily the high risk or the rising risk. Changing those referral habits and expectations is difficult. We are providing education on what alternatives are available and how to access them. We explain that we have limited resources and we need to focus on those most in need of our services.
Don't promise what you can't deliver.

Kenneth Coburn, MD, MPH Replied at 2:55 PM, 21 Aug 2014

In response to Clemens' questions in post #50:
Is there a reason why you chose not to scale it yourself as a non-profit entity? What are the pros and cons of lending your R&D to for-profit companies?

These are great questions and we are very much in search of answers ourselves so I can't say we've figured this out. So far, we've not yet discovered a funding source/mechanism to capitalize expansion as a non-profit (but very interested to learn and understand more about the Commonwealth Care Alliance social impact bond you referred to).

In order to efficiently scale our model, I think we need as much creativity and innovation on the "business" side as we do the care delivery side. Though we've been able to keep at it for over 13 years, we have found it incredibly difficult to get funded to undertake meaningful, long-term CCM R&D. We are prepared to be very opportunistic when it comes to generating new and longer-term revenue streams to fund scaling and ongoing R&D - provided that there is commitment to maintaining fidelity to the model and high quality in the delivery of service. For-profit, not-for-profit, doesn't matter. In my experience, those most interested in considering how to advance and scale our work don't come from any one sector, but seem to be rare individuals in leadership roles that have had sufficient experience to understand that a) we need a new mouse trap - can't keep doing the same old stuff, b) it won't be short, cheap, or easy, c) they care more about evidence than marketing messages and have a longer term, "systems" view, and d) they have a deep commitment, personal accountability, and sense of urgency to improve the health of vulnerable populations.

We haven't yet licensed our R&D to for-profit companies, but we are actively looking to do so. Therefore, I can't speak from experience yet about the pros and cons. That said, we are hoping that pro's include 1) the greater diversity, resilience, and longevity of revenues, and 2) greater chance to have more people benefit from the model in more places, more quickly. The con's that we worry about include 1) loss of program fidelity and reproducible effectiveness, 2) being taking advantage by those far more experienced in doing these business deals than we, and 3) if any "exclusivity" is required in these deals, the unintended consequence of stifling rather than promoting use on a broader scale.

We see many potentially tricky trade-offs ahead. Love to hear the experience and thoughts of others.

Rebecca Ramsay Replied at 4:44 PM, 21 Aug 2014

Hello there - busy day yesterday, but I'm here today to address yesterday's question about critical partners:

The list keeps growing. In the development of our Health Resilience Program our initial co-development partners were our primary care clinics because they are the hub for our new workforce. We also worked closely at the time with community mental health partners to understand the often-confusing eligiblity criteria for commmunity mental health services and to come to an understanding that much of the behavioral health challenges our target population was facing were not those traditionally served by MH providers (trauma, anxiety, depression, chronic pain, addiction). The other early critical partners were the hospitals and the EDs, and these relationships were fostered by the set of complementary interventions that were funded by the CMMI grant that also funded the scale-up of the Health Resilience program. In our community, we now have hospital transitions programs that target the highest risk patients and deploy a pharmacist/RN/social work team to help the patient have a safer discharge - these teams can hand off to a Health Resilience Specialist who can help with the ongoing stabilization and connection back to primary care and specialty care. We also have new roles in the ED, so we have similar partnerships around ED admissions for high risk patients. The ED partnership Clemens mentioned actually started as a health plan strategy, but has been enhanced by our Health Resilience Program. IT involves using a care plan template that goes to the primary care provider (often faciliated by the Health Resilience Specialist or a health plan care manager) - the primary care provider writes on the careplan template the care he/she wants for the patient if/when they show up in the ED. This gets transferred to the ED who then puts it into the EMR. The most common instruction relates to "not treating chronic pain complaints with Narcotics" - we have an opitate addiction epidemic in our State. This probably sounds pretty punitive - but, with the number of overdose deaths we've had, its a patient safety issue. And in collaboration with these instructions they can also put a care manager phone number to call. There are other instructions offered as well to help coordinate the primary care plan with the ED treatment plan. We are about to launch an IT platform to support this work so that we are not faxing templates and requiring the manual upload.

At this point our learning has taken us far beyond these more traditional partners. We are now working closely with detox centers, addiction recovery systems. a trauma institute, paliative care programs, peer organizations, community-based acupuncture clinics (alternative to opiates for chronic pain), community centers who can provide low-cost passes to exercise opportunities, food delivery organizations, Aging and Disability Agencies, low income housing providers, shelters.......everything we are learning is driving us toward more and more community engagement.

Clemens Hong Replied at 5:27 PM, 21 Aug 2014

Ruth Staus. Thank you for your comments. I think your comments speak to much of the fragmentation and failures in healthcare delivery in the US. I think complex patients, whether frail, elderly, or otherwise are most impacted by this, and the types of wrap around, primary care-integrated complex care management programs we are discussing on this panel are one part of the solution. I think the integration into primary care will be essential to avoiding the second issue you raise - of multiple care managers working separately with the same patients.

Clemens Hong Replied at 5:37 PM, 21 Aug 2014

Ken, thank you for your incredibly thoughtful response to my question. I do think there is room for both types of models. I would really like to believe that non-profit approaches can scale as rapidly as for-profit when managed appropriately, but I'm sure finding investors is more challenging. Social impact bonds are promising as there really is an inherent return on investment possible with CCM models. I also think we also need greater exploration of different types of government/non-profit partnerships. King County Care Partners, a highly effective program for ABD Medicaid populations in Washington State, closed when Washington State decided to place most of their ABD population into Medicaid managed care programs. The managed care programs were encouraged to partner with King County Care Partners in the care of these patients, but all plans in King County decided to use their own approach, rather than to contract with an outside entity. It remains to be seen whether these managed care plans have equivalent solutions, but from my perspective, it seems a bit of a shame that a rigorously evaluated, and highly effective program was closed rather than scaled across the state. I know you've faced similar challenges with HQP.

Clemens Hong Replied at 5:59 PM, 21 Aug 2014

In response to today's questions: What are the keys to successful implementation of CCM? What challenges have you encountered with implementation? What “do’s” or “don’ts” have you learned?

I will raise three high-level points gleaned from the Commonwealth Fund study interviews to address this question:

1) The key to successful implementation is to align the population, the intervention, and the outcomes of interest. Sounds simple, but I do believe programs frequently don't spend enough time working to ensure this alignment - for example, a program under pressure to achieve an rapid return on investment at 1 year that does not select a population at high enough risk for admissions to achieve the outcomes, or a program that has a high-cost Medicaid or dual eligible population, but does build sufficient behavioral health or social service support capacity into their program. This right fitting is critical, and requires a detailed stakeholder analysis to identify critical outcomes (ones that ensure sustainability of the program), a detailed asset and gap assessment to understand the strengths and weaknesses of the delivery system, and a detailed understanding of the high risk populations served by the system, so you can work to select the right populations for the intervention.

2) A key "do" - Do engage critical stakeholders early and often - starting from program design and approach to patient selection. These stakeholders include health system leaders, program leaders/champions, primary care teams, and other key, traditional and non-traditional, healthcare delivery partners.

3) A key "don't" - Don't assume you have the perfect model. There is no perfect model for your context. You must learn from others and get started with the best model you can, but build in processes to use data and lessons learned to continually improve the model over time.

Steven R. Counsell, MD Replied at 7:16 PM, 21 Aug 2014

Your “high-level” points are excellent Clemens! On a “low-level” we have found the following facilitators of implementation of the GRACE model:

· Health system and physician group that has an “early adopter” mentality

· Strong physician leadership and clinical champion for the CCM program from within the physician organization

· Financial incentives for primary care physicians that align with CCM program goals

· Shared EMR and care management software for ease of communication and coordination

· Dedicated clinical staff for start-up (CCM duties are not an “add on” or redesign of an already full time workload)

· Site visit to health system already experienced in successful implementation of the chosen CCM model


Janet Van der Mei Replied at 8:23 PM, 21 Aug 2014

Will not be available tomorrow so will give my two cents worth on the question this evening:
How do you plan to ensure that your work in complex care management is sustainable? What challenges the sustainability of your approach? How are you building the case for sustainability or growth of the model in your context?
To be sustainable a program must be able to identify its worth both in cost savings and quality of care. Programs must identify metrics that demonstrate outcomes as a result of the various interventions the team provides. One key metric for us is our hospitals readmission rate particularly with the CMS focus on readmission penalties. CMS has TCM codes and will soon have complex case management codes that physicians may use if they meet specific criteria. This provides added revenue to support complex case management programs. We are tracking how successfully these codes are able to be billed. Another area of focus is advance care planning discussions with our targeted population with the goal to have a completed Advance Directive and POLST. We want to ensure patients understand their end of life choices and to be sure they are appropriately documented so patients do not get unwanted treatment. This is another metric we track. We know that end of life interventions can be very costly and devastating to patients and families if what they want is not identified and followed. As we move towards reducing the total cost of care, payer becomes much less of a factor. Demonstrating the benefits of complex case management and the benefits of lowering cost through reduced avoidable admissions, readmissions and ED encounters become more important to health care systems.

Steven R. Counsell, MD Replied at 2:13 AM, 22 Aug 2014

Sustainability and growth of GRACE Team Care in our local programs in Indianapolis and national replication sites are based on demonstrating added value to the healthcare system and/or health plan by improving quality and lowering costs in high risk and complex older adults. The program must be able to demonstrate successful enrollment of high risk / high cost patients and including a caseload of approximately 100 patients for each NP / social worker dyad. Strong acceptance and positive feedback from patients and their caregivers and primary care physicians is critical to sustainability along with objective demonstration of reduced hospitalizations. Return on investment can be most quickly demonstrated by enrolling patients into GRACE Team Care at the time of hospital discharge and documenting reduction in 30-day readmission rates. Longitudinal follow-up subsequently allows demonstration of reduced hospital, SNF, and ED utilization. Even with demonstration of better quality and lower costs, CCM programs like GRACE are under constant pressure to be more efficient and serve larger numbers of patients. Workforce shortage issues and recruitment of nurse practitioners interested in home visits and collaborative care of frail elders can be a challenge as well.

Under a Medicare Advantage Plan, more appropriate and higher risk adjustment scores lead to increases in revenue to the health plan. This is due to greater recognition and better documentation of medical illnesses and conditions such as depression by the GRACE team. Improved performance on quality metrics leading to higher health plan star ratings with associated added revenue also helps drive sustainability and provides rationale for program expansion. Finally, for health plans or physician organizations taking risk under Medicare and Medicaid, GRACE Team Care attention to geriatric syndromes and caregiver issues along with facilitation of more appropriate utilization of home and community based services contributes to greater independence and quality of life and avoidance of nursing home expenditures .

Kenneth Coburn, MD, MPH Replied at 8:33 AM, 22 Aug 2014

We've seen the challenge of sustainability shift in the recent years due to dramatic changes in the marketplace. Sustainability of even highly effective CCM programs is now more difficult because of competition with and hard core marketing by commercial enterprises working to develop new business in this rapidly growing sector of health care. Most effective CCM programs require a sustained commitment of resources, time to ramp up and often grow organically given the need for people to be trained and become experienced in providing care in a new way, setting, etc. So it takes some time to scale up.

Many companies with little or no actual experience delivering any of these programs have substantial investor backing and are hawking their "innovations" and "solutions" in the "pop health" "space" - claiming faster, easier, and immediately scaleable "proven results". Often this is just pure marketing hype. But it plays well to many newbie ACO leaders who lack deep experience in this field and are feeling time pressure to get some kind of care management approach up and running. They want to think that there really is something out there that can be quickly implemented at scale and get results. Add to that, the widespread fascination and tendency to believe that big data and analytics alone will change outcomes. Even when an effective CCM program has much, much more grounded and solid evidence of effectiveness if it's honest about the time to ramp up to scale to get reproducible effectiveness it may not be selected by an ACO, health plan, etc.

One reason (among many) I think the work Clemens and his co-authors did in developing their Commonwealth Fund issue brief on CCM is so important is because it included a Level of Evidence for reported program results. This is rare in this area of work. We need to develop this approach much more. We need to help educate health care leaders and policy makers to realize how important high quality evidence is - just like we expect our clinicians to be able to value and interpret its meaning in clinical research in the care of patients. This is a huge potential opportunity.

Someday, commercial companies (and non-profits) should be required to have the same kinds of large prospective trials demonstrating real efficacy of their CCM program just as pharmaceutical companies need to do now for FDA approval before being allowed to publicly claim "proven results". People could still be completely free to hawk their wares, do small pilots, and even adopt any program they want across big systems without limitations, but to publicly claim efficacy in marketing materials should require some real proof - that everyone agrees is valid and meaningful. This would elevate the entire field and start to drive real thinking about another area of sustainability that is key to continued improvement - sustained support for real, valuable, and valid R&D to advance this work. This area is likely to be one requiring some government support and oversight, but we'd see more private sector investment in real R&D if marketing "proven results" was restricted only to those who can actually demonstrate effectiveness.

Greg Watt Replied at 9:07 AM, 22 Aug 2014

This is a very concrete question: What title is used by each discipline of your Complex Care Management team? Do their business cards refer to them as a Complex Care RN/ SW /CHW /etc? Have any patients had a response (+ or -) to needing a "Complex" CM? We have had some patients who have been put off by that and am seeking to clarify titles.

Greg Watt

Steven R. Counsell, MD Replied at 11:09 AM, 22 Aug 2014

In GRACE Team Care we have used the title of "GRACE Team" or "GRACE Program" in referring to the CCM service, and individual team members go by "GRACE nurse practitioner", "GRACE social worker", etc. We have had very good acceptance by older patients and caregivers in this regard. -- Steve

Kenneth Coburn, MD, MPH Replied at 1:39 PM, 22 Aug 2014

To Greg, Post #67:
The generic term we use at HQP for the service we provide is "advanced preventive care." When advanced preventive care is targeted specifically to higher-risk, higher-cost populations it becomes one flavor of "complex care management". We've just come up with a 'brand' name for our service (my team thought "advanced preventive care" was too dull and uninteresting) - SPERO (Latin for hope). Yet we still default to calling our front line nurses, "nurse care managers" - a term that isn't very good, but we've not yet come up with something better. Any ideas?

Of course none of these terms really gives much specificity about what it includes or how services are delivered. Care management, case management, disease management, complex care management, population health management, etc. ... all have general definitions, many of which are overlapping, but beneath each of these general labels, there are vast differences in program design and effectiveness.

Rebecca Ramsay Replied at 3:40 PM, 22 Aug 2014

Reply to Greg's Question:
This has really been a journey for us - naming our staff. Obviously the term "Health Resilience Specialist" is not necessarily self explanatory. But, it does conjure up a positive image AND it invites questions about what a HRS does, and how they might be able to help. For those of you who may not know, the term "resilience" has important underpinnings in the science of trauma recovery. I totally understand the negativity some of these system-defined role titles have with patients - we've had similar feedback. You might get a kick out of the trajectory of names our program has evolved through over the past 4 years: Started with "high utilizer program", then went to "hot spot program", then "community care program" and now "The Health Resilience ProgramTM". We finally landed on something that feels positive, more descriptive and resonant of our program goals than the others, and unique (although I just found out that one of the ambulatory ICU's being developed in Portland is now called "The Resilience Center". Hmmmm.....).

Rebecca Ramsay Replied at 6:14 PM, 22 Aug 2014

How do you plan to ensure that your work in complex care management is sustainable? What challenges the sustainability of your approach? How are you building the case for sustainability or growth of the model in your context?

I could not agree more with many of the comments already submitted by other panelists, specifically that we need to sustain this work by measuring across many dimensions of cost and quality, paying particular attention to how these programs and new workforce elements are adding value to providers and delivery systems, and recognizing the social return on investment when we can improve the QOL of our societies most vulnerable citizens and bring them back from the marginalization they have experienced. In the highly socially complex population we are serving, where intergenerational poverty and high prevalence of trauma exist, we need new measures of "hope", "resilience", and "wellbeing" that will really capture what we see happening for many of our clients. Eventually these should translate into better health outcomes, but it wont happen over night and we want to be able to measure improvement along the way. We also see a huge need for more palliative care and advanced care planning in our population, and some of the previous comments are pointing me even more in the direction of measuring that as we go.

In summary, what we have been able to demonstrate after about 2 years of work is a significant reduction in ED and hospital admissions (both at about 35%), but we still need a well designed control group to make certain its not all just regression to the mean. I am quite certain there is more going on than just regression, but we need that evidence. When I calculate potential savings using these figures and then assume 50% is due to regression, we still come out positive on the return on investment, so that is a good thing. It means we are picking the right patients and our intervention is not overly expensive. We also have great provider and care team satisfaction which is absolutely critical as health care reform takes hold and Medicaid expansion taxes our safety net (ultimately a good thing, dont get me wrong!). With these outcomes, I have certainty that we will sustain our grant-funded program for at least a couple more years while we continue to refine, learn, and get more robust in our measurement. Beyond that, the outcomes will be more and more important.


Clemens Hong Replied at 6:23 PM, 22 Aug 2014

I thought this was a fantastic conversation this week. Thanks so much to all the other panelists. I learned so much and hope that others found this valuable. I think the other panel members terrifically addressed the issues around sustainability. I do agree that many people are currently "experimenting" on their own to develop models and there are far too many organizations "selling" models with little to no evidence of benefit. The worst part about this is that many sell these models and then spend far too little time building infrastructure and integration necessary to address the issues faced by these incredibly vulnerable patients. So when the contract comes to an end, the organizations are often no better off than before the contract began. I hope we can create more forums like this for exchange of ideas, so folks can really learn from programs like GRACE, HQP, SutterHealth, and CareOregon and not start from a blank slate. It might help the many organizations that are thinking about complex care management to start with more evidence-based designs or make decisions on vendors that will benefit their organizations in the long term. I think the changing health reform environment will create many different opportunities for sustainability in these types of complex care management programs in years to come. In fact, in many ways, the sustainability of health reform itself may very well depend on our collective ability to build these types of programs and take them to scale. I hope this is just the beginning of many conversations. Thanks again.

Isabelle Celentano Replied at 9:10 AM, 25 Aug 2014

Many, many thanks to our exceptional panelists, and all of our community members who participated in this incredibly rich discussion. We greatly appreciate the insights everyone has shared, and look forward to continuing to discuss these important topics in the community.

We will be working on a Discussion Brief to summarize the key points from this Expert Panel, and will share details as soon as that is available on the website.

In the meantime, we would be grateful for your feedback in our short, 9 question follow up survey. These surveys help us understand the impact of our Expert Panel discussions, and your feedback is incredibly valuable to us.

Please take the survey now, by visiting:

Sudip Bhandari Replied at 9:08 AM, 1 Dec 2014

Many thanks again to our panelists and members who participated in this Expert Panel discussion last August!

To help us understand the longer-term impact of these Expert Panels and plan future events, we have created a very short, 5 question, follow-up survey. This survey will only take 2-3 minutes of your time—please take the survey now at:

This Expert Panel is Archived.

This Expert Panel is no longer active as of December 2018. Thanks to those who posted here and made this information available to others visiting the site.