Better Analytics for Better Mental Healthcare

Individuals with severe mental illness (SMI) represent a critical and complex group within the Medicaid system. While they make up about 20% of Medicaid recipients, they account for nearly 50% of total healthcare spending. Despite this significant investment, their health outcomes are often poor, with life expectancies 10 to 20 years shorter than average.

A major contributing factor is the high rate of physical illness among this population. In fact, advanced conditions like heart disease, chronic kidney disease, and Type II Diabetes are three to four times more common in individuals with SMI. And as mental health declines, physical disease burden commonly increases.

This pattern highlights a fundamental problem: Although physical and mental health are deeply connected, the healthcare system often treats them in isolation.

This article examines the persistent problems in caring for the SMI population, the limitations of current analytical methods, and how a new, integrated approach to data can transform care.

The Challenge of Serving People with Severe Mental Illness

Individuals with SMI face cognitive, emotional, and social impairments that make it difficult to manage their own care. At the same time, physical care providers are often not equipped to handle the unique communication barriers presented by this population. This disconnect perpetuates a cycle of declining health and rising costs, with traditional quality improvement measures failing to make a significant impact.

One of the primary barriers to progress is the use of oversimplified analytics. Many current solutions fail to capture the full complexity of the bio-psycho-social factors that influence health outcomes for the SMI population. Attempting to break down these interconnected problems into simpler components often misses the bigger picture. It’s like trying to fix a complex engine by looking at each part individually without understanding how they work together. For populations with mild health issues, a simplified focus can be effective. But for those with complex needs like SMI, it often leads to treating symptoms rather than root causes. The result is a reactive cycle where resources are spent on isolated issues without achieving systematic improvements in health or cost.

A Transformative Approach: Whole Health Informatics

To truly address the needs of the SMI population, a more sophisticated, programmatic approach is required. This is where “whole health informatics” comes in. This approach integrates several key components to create a comprehensive view of an individual’s health journey. Among the core components of this approach:

• Multifactorial longitudinal analytics (MLAs) are advanced analytics that measure the combined impact of multiple biological, psychological, and social factors on health and cost outcomes over time. MLAs allow agencies to understand what’s driving current outcome, predict downstream effects, and design more effective programs.
• The enriched longitudinal health record (LHR) serves as a single source of truth, combining claims, member, and provider data into a comprehensive health record. This repository provides the data foundation needed for advanced analytics.
• Scalable, on-demand cloud infrastructure provides the necessary computational power to process vast amounts of data efficiently. This allows for large-scale analysis that is both powerful and cost-effective.

By breaking down the silos that traditionally separate business, information, and technology management, this integrated infrastructure provides actionable insights. It empowers Medicaid agencies to design, monitor, and refine policies and programs that lead to sustainably better health at a lower cost.

Case Study: An Innovative Health Home Model

Nearly a decade ago, one state Medicaid agency implemented a specialized health home model (SMI-HH) designed for individuals with SMI. Unlike traditional models, this program featured care coordination delivered by behavioral health professionals in longer, therapy-style sessions.

To evaluate its effectiveness, an analysis was performed of seven years of data from 2.5 million Medicaid members. The analysis examined the health and cost trajectories of individuals referred to the SMI-HH, comparing those who engaged with the program to those who did not. The results were remarkable, demonstrating the significant positive impact of the SMI-HH program:

• Cost and mortality reduction. The program achieved an adjusted year-over-year cost reduction of more than 8% and lowered the adjusted four-year all-cause mortality risk by more than 25%. This strong performance prompted the state to prioritize evaluating other programs using the same advanced analytical approach.
• Effective care coordination. The study confirmed that care coordination is more effective when delivered by behavioral health providers who are trained to engage with the SMI population. In contrast, a primary care–focused model had no meaningful impact on the total cost of care for this group.
• Higher initial costs (which lead to long-term savings). Members who engaged with the program initially incurred higher costs as deferred care was addressed. Over time, however, costs stabilized and then declined as the individuals’ health improved. This highlights the importance of planning for upfront investments to achieve long-term savings.
• Protective role of minor conditions. The analysis revealed that low-severity chronic conditions, such as migraines, can serve as entry points for care. Behavioral health providers can use these complaints to build rapport and uncover more serious underlying health issues.
• Psychiatric complexity as a key driver. The degree of psychiatric complexity was identified as a significant independent driver of cost and mortality risk. This underscores the need for early intervention to prevent the compounding effects of mental and physical health issues.
• Cost neutrality. Over several years, this program has come to approach cost-neutrality.

Addressing Psychiatric Complexity Across Healthcare

This case study has broad implications. It underscores how psychiatric complexity affects the way individuals engage with healthcare services across many different areas. It also demonstrates why an effective analytics platform must account for the distinct needs of the SMI population within various programs. Here are some of the key takeaways:

• Chronic physical illnesses. Individuals with SMI have a higher prevalence and severity of chronic conditions. Programs focused on these conditions need to incorporate targeted strategies for this population.
• Emergency department (ED) use. The SMI population uses the ED more frequently, and analysis shows that about half of this utilization is for necessary medical care. This finding reflects these individuals’ greater physical disease burden and suggests that a blanket approach to reducing ED use would be inappropriate.
• Maternal and child health. Mental illness is a primary driver of adverse maternal outcomes among Medicaid recipients. With postpartum suicide as a leading cause of maternal mortality, targeted mental health support should be a critical component of maternal health programs.
• Justice-involved populations. Individuals leaving the justice system have high rates of SMI and face numerous challenges. Addressing their mental health needs is crucial for enabling successful reentry and reducing recidivism.

A Path to Programmatic Improvement

Caring for the SMI population requires moving beyond one-size-fits-all solutions. Whole-person care begins by understanding how “whole” a person is—cognitively, emotionally, and socially. As a person’s ability to self-organize diminishes, healthcare must become more integrated to meet their needs.

By leveraging a best-in-class whole health informatics infrastructure, Medicaid agencies can gain the insights needed to design effective, data-driven programs. This approach enables agencies to quantify and act on the factors influencing health and cost for individuals with SMI, leading to material health benefits and reduced costs. For the first time, it offers a proven path to systematically address the challenges of psychiatric complexity and improve the lives of our most vulnerable populations.