This blog post is the third in a series focused on how provider organizations, payers, and private equity investors are preparing to respond to disruption, overcome challenges, and lead innovation in the behavioral healthcare space.
Amid growing evidence that integrating primary care and behavioral health can improve health outcomes and reduce cost, payers are changing the way they approach behavioral health and substance use disorder (SUD) services. Instead of carving out these services to be managed by third-party entities, health plans are considering models that integrate behavioral health with physical healthcare.
Caring for complex conditions and behavioral health comorbidities can be costly—studies have demonstrated that health plan members with chronic conditions and behavioral health comorbidities experience annual costs approximately two to four times greater than members with the same chronic conditions and no behavioral health comorbidities. Managing care in a more integrated manner can lead to cost savings, quality improvement, and enhanced experiences for both members and providers.
To implement and operate an effective primary care–behavioral health integrated model, payers must develop a comprehensive strategy that includes investments in primary care support; cost-effectiveness measurement; and data collection, management, and analysis.
Primary Care Support
Developing a successful primary care–behavioral health integrated model begins with payers effectively supporting providers and offering opportunities for primary care providers (PCPs) to enhance their existing knowledge, as well as augmenting care teams with behavioral health expert resources. Primary care is often the principal treatment setting for patients with behavioral health conditions; consider that 70% of all antidepressant prescriptions are currently written by PCPs. While PCPs can successfully manage many patients with behavioral health diagnoses, some are not entirely comfortable or may lack the experience to diagnose and treat behavioral health and/or SUD conditions.
PCPs can benefit from payers coordinating support in the following areas:
- Managing high-risk medical patients with low-intensity behavioral health conditions
- Prescribing behavioral health medications
- Administering medication-assisted therapy
- Treating patients requiring behavioral health urgent care
After a patient is identified as having a behavioral health comorbidity, the PCP faces the challenge of providing care management and ensuring access to resources that align with patient need. Payers should allow providers the flexibility to explore a range of strategies for integrating support into their existing practices, such as colocation with behavioral health specialists or building collaborative care teams that include community health workers who help coordinate care and resources via telehealth.
In order to achieve physical health and behavioral health integration objectives, health plans must reevaluate how behavioral health is addressed in the primary care setting and review care management processes to ensure resources are appropriately allocated, roles are clearly defined, and PCPs are effectively supported.
Cost-Effectiveness Measurement
Investing in behavioral health for patients with chronic conditions and effectively integrating these services can reduce the total cost of care. Estimates indicate that participating in collaborative care can lead to cost savings of approximately 5% to 10% of the total cost of care. The IMPACT study by the University of Washington’s AIMS Center analyzed healthcare spend on older adults with depression who received collaborative care services compared to usual care; results indicated that every $1.00 spent on collaborative care saves $6.50 in healthcare costs.
The opportunity to manage costs and improve quality and outcomes will vary dramatically depending on the line of business, which chronic conditions are present, and which care management processes are employed.
- Line of Business: Prevalence of behavioral health conditions differ by line of business. A primary care–behavioral health integrated care model for one population (e.g., Medicaid) will not yield comparable results for another population (e.g., commercial large group members).
- Chronic Conditions: Certain conditions are more likely to be associated with behavioral health diagnoses, and an integrated model can allocate resources accordingly. For example, 1 in 4 cancer patients may have a major depressive episode, compared to 1 in 13 individuals within the general population.
- Care Management: Payers need to identify opportunities, develop care programs around the identified opportunities, and measure performance to assess the impact of care management.
To date, only a limited number of longitudinal studies have been conducted across large populations and segmented by line of business to demonstrate the cost-effectiveness of increased investment in behavioral health services. Even fewer studies have measured the level of patient engagement at all stages of the care management process to determine the effects of care management on outcomes and cost. Because of the relatively short history of this model of care, health plans should be prepared to test various programs and develop robust performance measurement programs. These analytics will be useful to validate the expansion of behavioral health services and determine which care management programs are most effective.
Data Requirements and Management Processes
To successfully manage physical and behavioral health integration, collecting, managing, and utilizing data is essential.
- Data Collection: The data required to identify risk, manage patient care, and measure performance for behavioral health is more expansive than the data required for physical health. Independent data elements must be collected, including traditional claims, clinical data, engagement information, and social determinants of health data. Furthermore, information exchange between many behavioral health professionals is not robust, and longitudinal monitoring can be limited, especially in SUD treatment. While health plans will depend on providers for much of this information, data gathering can be streamlined if payers create and share well-defined processes and formats for extracting and reporting clinical data.
- Analytics Databases: The lack of structure in many sources of behavioral health data (e.g., text notes may require natural language processing) may lead to challenges developing a database that can integrate multiple data elements. It is necessary to analyze the various data elements together and to identify statistically significant relationships in order to provide insight on the most effective practices.
- Risk Algorithms: Development of risk algorithms can create opportunities for timely interventions and resource maximization. These algorithms help by proactively identifying high- or rising-risk populations, including SUD, serious and persistent mental illness, behavioral and physical health comorbidities, and high-risk behaviors (e.g., medication nonadherence, accidental injury, increased ED visits).
- Actionable Information: Data must be analyzed and returned to providers, care management teams, and members as actionable information. For example, tracking effective member communication strategies and identifying effective outreach channels can lead to improved patient engagement, but this information must first be clearly synthesized and disseminated for provider use.
Many small and mid-market insurers will be challenged to afford or prioritize investing in data collection and management due to lack of scale and, therefore, must be more creative in how they address data collection and analysis.
When strategically executed as part of a transition to a primary care–behavioral health integrated model, we expect these investments in primary care support, cost-effectiveness measurement, and data management to lead to cost savings, enhanced patient experiences, and improved patient outcomes. However, success requires careful planning and execution, as the unique requirements of each market combined with your organization’s capabilities can dramatically affect the size, scope, and financial and clinical performance of the model you look to implement.
Published August 19, 2020