A Value-Based Collaborative Care Model by Payers, Providers, and Pharma

When it comes to determining how health care should be delivered and paid for, typically there are two parties at the table: the provider and the payer. In general, the payer is presumed to have the upper hand in the negotiations, given the data they collect and their role in paying providers. While payers have had the financial upper hand, doctors are typically anxious about maintaining their ability to direct patient care, their income and access to networks. At the same time, pharmaceutical companies, pharmacy benefit managers, patients, and other care providers (eg, community-based service providers, physicians addressing comorbidities) not typically included in such discussions. Even with heightened attention to value-based care (VBC), care delivery and reimbursement conversations are still primarily focused on incremental reimbursement and not often a comprehensive model of care.¹ These discussions are often transactional in nature, addressing reimbursement rates on procedures and encounters rather than patient wellness. More broadly, there is little consideration for care coordination or how collectively the provider community can work together to improve all elements of the individual’s health.

It should be no surprise then, that health care delivery has become quite disjointed, with patients bouncing between multiple providers, each delivering only a portion of care with limited connection or communication with other providers also treating the individual.^2,3 This results in care plans created without appropriate knowledge of what other providers may already have done, are currently doing, or are planning to do for these patients, which may result in duplicative testing, patient frustration, and inconsistent or poor clinical and economic outcomes.⁴

Ultimately, this model results in a health care system that expects the patient to be responsible for understanding and navigating how all of their care fits together. Yet, sharing of information with patients is often rare and opaque, and most individuals who are not health care providers feel they must consult “Dr. Google” to understand the completeness of their situation.⁵ Most patients are not comfortable making risk/benefit decisions related to complex biologic treatment, surgeries, or other treatment modalities. Further, patients may receive conflicting information from different care providers, due to the lack of information-sharing among providers—again, a result of the fee-for-service payment architecture. This is particularly worrisome for those with chronic conditions, who require complex treatments and ongoing care from multiple care providers attending to various comorbid symptoms and conditions.

Many health care providers may be nervous about the future of health care, viewing VBC as another way that payers will tell them how to practice. Specifically, providers worry about loss of autonomy when it comes to clinical decision-making, income expectations, and even management of their practice.⁶ If executed thoughtfully, a comprehensive VBC delivery model should produce the opposite outcome, ie, giving providers a prominent voice in designing how they deliver care within these frameworks. By bringing providers into the larger VBC conversation, their clinical expertise can be leveraged to create clinical pathways to guide evidence-based best practice and promote collaboration, which should serve as the backbone of a VBC model (Figure 1). 

With these complex dynamics in mind, a diverse group of stakeholders (the Group) spent much of 2020 meeting to discuss how to build a comprehensive value-based framework for chronic conditions, ie, a model that prioritizes a holistic view of the patient, promotes communication among all those caring for the patient, and proactively confronts common barriers in the way of best practice protocols. As a case example, the Group chose psoriatic arthritis (PsA) to examine and determine potential solutions.

This article discusses how the Group came together and employed shared decision-making, the evolving culture of collaboration, and the resulting plan. The project is an exercise in how to bring diverse, historically adversarial partners leveraging the expertise of all to build something together that is focused on their collective patients/members. Challenges, stumbles, and successes from Year 1 are described, setting the stage for pilot implementation, expected to launch at the beginning of Year 3.

The Group was made up of a payer (3 representatives), multiple specialty care providers (5 representatives), a pharmaceutical life sciences company (9 representatives; the core team for UCB was 5), an analytics company (2 representatives), and one advisor/architect of VBC models.

• CareFirst—payer provides claims data to show longitudinal utilization and cost of care of attributed members; facilitates payment
• Physicians: clinical expertise and building of best practice, clinical algorithms designed to reduce the variations in care and costs of care identified in payer and other data
  • 2 rheumatologists
  • 1 dermatologist
  • 1 cardiologist
• UCB—pharma manufacturer provides expertise in immunology, patient journey data, case studies, information related to medication and other therapeutic protocol adherence (challenges, barriers, etc)
• CrossBridge—analytics organization with expertise in immunology, configuration of clinical algorithms, receives claims, SDOH and EHR data, synthesizes and reports
• BCollaborative— expertise is in building of specialty care value-based models and creating functionally collaborative relationships and models. Provides education and strategic advisement, eg, architecture of specialty care value-based models

The entire Group of 20 members met 3 times virtually throughout 2020. The Group steering committee was made of up representatives from CareFirst, CrossBridge, UCB, and BCollaborative. The steering committee met bi-weekly throughout the year to review data, models, and discuss the input from the larger team. They would review and synthesize the physicians’ comments, needs, concerns, and prepare agendas, pre-reading materials, and presentations for the larger Group meetings. All meetings were moderated by Lili Brillstein, CEO, BCollaborative, an experienced architect of VBC models.

The overarching goals of the group were (1) to gather the full spectrum of stakeholders to assess and address variations in care and costs of care for individuals for chronic inflammatory diseases; (2) explore the current and prospective roles of each stakeholder; and (3) ultimately begin building a value-based model to optimize care for a complex chronic condition.

To identify the most relevant chronic inflammatory condition  to focus on for our value-based framework design, the UCB innovation team partnered with  UCB’s Immunology Medical Affairs leads to assess the current body of evidence, including a retrospective claims analysis, to evaluate co-morbid conditions associated with psoriasis and PsA.⁷ This assessment identified significant increases in cardiovascular disease, depression, obesity, inflammatory bowel disease, and metabolic disease when there was evidence of concomitant joint involvement. The team also evaluated market research and patient journey analyses to understand quality of life (QOL) issues and unmet patient needs in PsA. Finally, they examined interviews with dermatologists and rheumatologists to gain insights on how these two specialties were coordinating care. All of these findings were shared with the larger group to both align on the disease state and to identify potentially modifiable risk factors associated with high cost and poor clinical outcomes. As the group studied the data, it became clear that the prevalence of metabolic disease among this population, which leads to serious cardiac events,required that they prioritize cardiology as an area of focus; thus a cardiologist was also added to the group.⁸

• Before the team met, a packet of pre-reading materials was distributed, including the following:
• an article describing VBC model structures⁹
• an article on developing a model specific to
• chronic conditions.¹⁰
• a real-world evidence analyses on psoriasis and PsA and the drivers and implications of disease progression (compiled by the UCB Health Economics Outcomes Research team)⁷

The first order of business was to quell the expected fear and anxiety and reinforce the shared purpose to care for patients who require expertise from each of the stakeholders across specialties and domains. The Moderator shared her past experiences in building VBC models for specialty care and how the need for respect and listening were key to success. She emphasized the need to engage physicians as the clinical leads (vs being directed by a health plan). Participants expressed anxiety about the model, and the finished product permeated much of the discussion among participants. This was not unexpected, as a paradigm shift may disrupt the comfort of an existing process, particularly if it has been financially rewarding to date. Any proposed change that may impact the status quo understandably can evoke fear and anxiety, often manifesting in a defensive posture, and even anger, at times. Participants were routinely asked for their opinions, perspective and feedback, and there were gentle and frequent reminders of the need to listen to each other. Consistent reminders and prompts from the Moderator on the shared goal fostered greater understanding and empathy among participants, ultimately strengthening relationships. These connections are critically important in ensuring that patients receive well-coordinated, best-in-class care.

The next step was to begin to understand the value that each constituent brings to the table. With payers and providers skeptical about each other’s motives and interests, dispelling preconceptions was an important exercise. By engaging in a “listening session” to discover the patient journey from different perspectives, the team created a heightened awareness of the clinical, emotional, and financial implications experienced by the PsA patient. Each of the partners was provided an opportunity to express their perspectives, desires for the model, and their concerns to provide an opportunity for all to understand the various perspectives, needs and expertise being brought to the table and what needed to be solved for.

Specifically, each participant was asked to share their perspective about patients living with PsA, how they treat them, and what their worries are in terms of patient care and outcomes. As participants shared insights, it became clear that there was tension and misunderstanding among participants. Some of the earliest comments included clear, lingering resentments toward others around the table. For example, it was suggested that some specialists spend more time with patients and others not enough. Some suggested that all of the issues could be solved merely by payer increases to hire more nurse practitioners.

Debate ensued, primarily focused on the pros and cons of adding more people and money to what some would call a broken system. It would take many efforts by the moderator to reset the conversation before the preconceptions and adversarial posture noted at the outset was dispelled, replaced by productive listening and exchange of information. The result: a deeper understanding by all stakeholders of the challenges ahead and urgent need to construct a model that embraces and addresses all elements of the Quadruple Aim: quality, cost, experience, and provider satisfaction.^11,12

This early discussion and exchange of experiences and concerns was essential in order to facilitate a culture change and move beyond old assumptions and enable a focus, not on individual stakeholders, but on the value that each could contribute to the whole. That meant creating an atmosphere of mutual respect through steady moderation of conversations while reminding the Group that no single constituent can solve all the issues; instead, a solution depended on a collective, interdependent set of skills and wills. Further, the team would come to acknowledge that it is okay and in fact preferable to not pre-define the end point. Only through trust, a shared desire to keep the patient at the center of the work, and a willingness to collaborate would the Group achieve success in designing a new framework.

Because the first meeting had focused on open communication, creating connection, and emphasizing participants’ shared purpose, the second meeting began with a more relaxed and focused tone. Stakeholders were able to concentrate on substantive dialogue devoted to understanding the challenges to be addressed as part of an effective VBC model.

Ms Brillstein, as moderator, explained to the group that “Healthcare is not just what health plans have defined as covered benefits. The way care is paid for impacts the way it is delivered. We want to change this and build a model that optimizes patient outcomes. Then we can build a payment model that can support it.”   

To establish a baseline knowledge of VBC models, several presentations were shared to review the various models (eg, Patient-Centered Medical Home, Accountable Care Organization, Episodes of Care, Bundled Payments, and a hybrid-Specialty Care Medical Home). The Group elected to pursue the Specialty Care Medical Home model, as it is designed to be led by the specialist responsible for the primary diagnosis and coordinate and provide care on an ongoing—possibly lifelong—basis, as compared with other models that are more episodic in nature.¹³

Following the initial meeting, participants were asked individually to provide their feedback on the process and the goals of the project. Beyond ensuring unique insights were gleaned from each participant, this information would inform how future meetings would be conducted and how to avoid potential roadblocks. Despite some of the initial tension, the feedback was insightful and practical, and the participants noted that they enjoyed the diversity and dynamics of the group, as well as the opportunity to exchange information and ideas.

It was at the third meeting that CrossBridge was invited to present their tools and capabilities around ingesting claims and other data, configuring the clinical algorithms that the team would define, and the reporting to support a comprehensive collaborative model of care. CrossBridge used a national payer claims data set, which is an aggregation of regional plan data. While not specific to one individual market, the trends and variations are assumed to be directionally the same as any one plan within the region. It also provided a large enough sample size for the group to be able to comfortably rely on the information reflected.

There was some spirited discussion around what the data showed and did not show and how it was presented.  For example, there were questions about why cardiac and gastrointestinal comorbidities were not more clearly reflected in the data and also whether there was a need to study individuals with psoriasis who may not yet have PsA but may be at risk. The team’s cardiologist worked offline with CrossBridge following the meeting to ensure that criteria was properly applied; and the CrossBridge team was asked to reconfigure some of the reports to better address the needs of the physicians.

This is one of the keys to success in these models – listening to each other to understand perspective and what is needed by each to be successful. The value that each stakeholder brings became clear as the group continued to listen to one another (Table 1). As a result of the collaboration among participants, the Group came to the below findings related to the challenges in this population and what elements will be essential to consider and integrate when developing the VBC model. 

The Group discussed the various challenges identified for individuals in this disease population along with the ßrespective details and impact of each. While it is tempting to try to solve for all aspects of care at the outset, it is critical to identify the primary drivers of variations in care and cost of care based on available data and prioritize items that can be most effectively addressed. As the model evolves over time, there will be continual reassessment and review to identify additional variations that may prioritized and addressed through the model.

While goals and metrics will not be determined until the next meeting, the team identified the key challenges facing this population—clinical, humanistic, and economic—laying the groundwork for the drivers that need to be studied.

Clinical Challenges

The key clinical factors impacting care delivery, according to the Group, include delayed identification of patients with PsA, suboptimal patient management, and a lack of care coordination along the patient journey.

Delayed Diagnosis

Some dermatologists do not regularly assess for joint disease in patients with psoriasis. Dermatology training does not focus on screening for potential joint manifestations in patients with psoriasis, thus they may not screen for joint symptoms, or they may not recognize symptoms of joint disease in these patients. In addition, there is a general lack of access to specialists—both rheumatologists and dermatologists—which can result in delays of over 6 months before an initial consultation takes place. Because of these issues, patients can experience a substantial delay in the recognition and diagnosis of PsA (sometimes up to 8-9 years). During this delay, patients can progress and experience permanent joint damage. This is problematic, as Group members noted that patients have better outcomes when PsA is identified and treated earlier.

Suboptimal Patient Management

After diagnosis, patients may still be unaware of the systemic nature of their disease, therefore, they may not share systemic symptoms with their provider. When the primary symptoms improve with treatment, patients may errantly believe they have achieved disease control. If systemic manifestations remain unrecognized, patients are not able to receive the holistic care necessary to properly manage all aspects of the disease. Group members notes that there is a desire to treat patients holistically, but often there is not enough time to do so. Some rheumatologists and dermatologists are only able to spend 15 minutes per patient, which is insufficient time to identify and manage all the manifestations and comorbidities associated with PsA. In some cases, dermatologists delegate the care of patients with psoriasis to nurse practitioners and physician assistants who may have less experience and also lack the time to appropriately manage such patients.

Group members also noted that biologics are underutilized; 90% of biologics are prescribed by 500 of the 10,000 dermatologists in the United States. In 4 states, there are no dermatologists who prescribe biologics. As a result, disease activity is often inadequately controlled and disease progression occurs, leading to reduced function and quality of life.

Lack of Care Coordination

There is a lack of effective communication and coordination of care across specialties. Group members commented that many providers rely on PDFs and faxes to communicate with each other. HIPAA regulations can also limit the ability of providers to share information with one another. There is also no single, shared electronic health record system or central database for patient information; often, each provider utilizes a separate EHR that does not interface or share data with other EHRs. Due to these factors, health care decisions are often made in siloes, as multidisciplinary care can require a substantial amount of time. Specifically, Group members noted that there is a lack of co-management between dermatologists and rheumatologists. Many dermatologists do not coordinate with rheumatologists or vice versa due to the belief that they are able to manage a patient on their own. Because of this fragmented system of care, the burden of care coordination often falls on the patient, who must navigate how to proceed and coordinate their care across settings and specialists.

Psychosocial Challenges

The Group noted that mental health issues are rarely recognized or addressed in patients with PsA, although depression and poor mental health are common in patients with PsA and have an immense impact on their lives and on the manifestation of their physical disease.¹⁴ One rheumatologist participant commented that he was shocked at the degree to which depression inerferes in patients’ daily life, but they never have time to address this dimension of the disease. Behavioral and mental health issues are rarely captured in claims data because they are neither assessed nor documented in patient records. Patients with unaddressed mental health issues are more likely to have lower compliance to medications or adverse behaviors, such as smoking or excess alcohol usage, leading to poor disease control. Patients may also miss appointments and experience a delay in appropriate care.

Overall, social determinants of health outcomes remain unaddressed. Socioeconomic factors and patient behaviors have a significant impact on health outcomes. Lack of consideration of social determinants of health can result in patients who are more likely to have poor adherence to medications and care plans, resulting in poor outcomes (and a higher use of health care resources).

Data generated by pharmaceutical companies regarding humanistic challenges (and their impact on outcomes) is often not leveraged effectively, according to the Group. There are many financial and personnel resources dedicated to understanding a patient’s journey and to studying social determinants of health outcomes, yet such data are rarely leveraged efficiently across stakeholders to improve patient care and outcomes. Thus, patient care is not optimized to address identified determinants of outcomes, leading to poor patient outcomes.

Economic Challenges

The Group determined specific challenges related to access and affordability as well as current reimbursement models and development of new reimbursement models.

Access and Affordability

Lack of access to medications was identified as an issue. For many patients, medications are difficult to obtain due to insurance hurdles (including expensive copays) and denial of prescribed medications. Such challenges are especially evident in patients with Medicare. These complexities lead to inappropriately managed care or care with substantial delays.

Payment policy can also impact provider behavior. Payment and reimbursement practices can have unintended consequences that lead to counterproductive behaviors by health care providers, as operational and financial realities of care cannot be ignored.

Specifically, within the realm of determining reimbursement, it can be difficult for payers to assign attribution for patients who receive care from multiple providers. Patients with PsA often receive care from multiple specialists with no single specialist designated as the primary health care provider, which creates confusion as to whom should be reimbursed for facilitating high-quality care.

Finally, Group members commented that current measures of treatment success (at the level of population health rather than the individual patient) often depend on indirect measures. Indirect, or proxy, measures such as dose escalation or level of opioid use are used to gauge treatment success in patient populations. Members note there is a need for measures of adherence that do not rely on patient-reported data; there is no metric by which to accurately understand the efficacy or success of a treatment/management plan across a patient population.

Developing Optimized Reimbursement Models

Members said that the economic impact of improvements beyond skin/joints are not considered when determining the value of a therapy. While there are many financial and personnel resources dedicated to generating data and real-world evidence regarding the impact of treatments beyond symptom control, such data are rarely leveraged efficiently across stakeholders to inform patient care.

It was also asserted that data-driven systems or models of care do not capture the complexity of care required by patients with diseases such as PsA due to patient complexity and comorbidities. This leads to disjoined or unnecessary care from multiple providers who are functioning independently of one another.

Another challenge is when payers become involved in a claim. Payers tend to get involved only at the point of payment rather than being involved earlier in the care management journey, which may contribute to increased costs for the patient, payers, and the entire health system.

The Group elucidated several considerations when building a VBC model for patients with PsA. They agreed that the model must:

• Be collaboratively constructed
• Incorporate payer and clinical data
• Allow sharing of data with all stakeholders
• Encourage holistic patient care
• Incentivize a shared-care strategy
• Preserve physician autonomy and flexibility,
• when warranted
• Incorporate a compliance threshold with set guidelines
• Include a reimbursement policy based on outcomes.

Further explanation and underlying rationale are provided in Table 2. 

As the team continues to study the data for this population, the first goal for 2021 is to determine the quality metrics that will measure success or failure of the model, then to build the clinical algorithms that will support collaborative, comprehensive care. Some of the quality metrics will be derived from claims data, including rates of emergency room visits and inpatient admissions/readmissions. There will also be data originating from electronic health records that capture more granular clinical elements specific to PsA. Regardless of the data source, ensuring each metric reflects a meaningful outcome of the model remains a principal driver in the selection process.

Once the measures and their cadence have been agreed to, the team will conduct an operational review (ie, clinical workflows, logistics of operations) eg, what will be co-located, what will be virtual; will there be behavioral health screening and if so, where/when does that occur; extended practice hours; etc). This comprises Year 1 of the playbook for this specialty care medical home—a model that we expect to continually review and revise as required.

As of this writing (August 2021), the quality goals are under review by the Group steering committee, and the clinical and operational assessment is well underway to prepare for Year 2.

Authors: Lili Brillstein, MPH¹; Donna Finale²; Michael Udwin, MD³

Affiliations: ¹CEO, BCollaborative
²Head of Portfolio Innovation, UCB
³Medical Director, Practice and Payment Transformation, CareFirst  BlueCross BlueShield

Address correspondence to: Lili Brillstein, MPH
Phone: + 1 (917) 733-5864
Email: lbrillstein@bcollaborative.com

Acknowledgements: The authors would like to thank Drs Herb Baraf, Evan Siegal, Ron Prussick, Joel Brill, Mark Kaufman, and Sara Collins for contributing their expertise to this project.

Disclosures: Lili Brillstein, BCollaborative, served as a paid consultant to UCB for this project. The collaborative project detailed in this article was funded by UCB.