How can analytics help triage behavioral health risk?

Triaging behavioral health risk

Many of the behavioral health-related hurdles that healthcare organizations face are related to risk stratification.

As defined by the National Association of Community Health Centers, risk stratification is "the process of assigning a risk status to patients, then using this information to direct care and improve overall health outcomes."

The process is designed to help healthcare providers categorize patients based on care needs and complexity, enabling clinical teams to provide the right level of care services. Risk stratification also plays a key role in population health management and value-based care, as assigning risk categories can help with population-level care model development and individual-level care plan personalization.

To conduct risk stratification, health systems often develop or adopt patient risk scores, but once a patient is assigned a score, engaging that patient effectively can prove difficult.

"This is a national challenge -- how do we engage the consumer in care when they're ready to get care [and] when they're able to get care," Attallah explained. "That's one series of challenges that's consumer-facing, and then, what primary care providers have is relationships that are unique from other subspecialties in that they tend to have longer-lasting relationships."

He indicated that because of this, integrating behavioral health risk stratification into primary care settings presents an opportunity to engage patients earlier. However, primary care providers, like other clinicians, often face constraints on their time.

"They have high relationship, low time, and they are often ill-equipped to offer effective care -- not because they don't want to," Attallah said, noting that this creates a scenario in which primary care clinicians have the desire and unique position to help meet patients' behavioral health needs, but might lack the tools necessary.

"So we need to enable providers to have, from the consumer and provider perspective, a way to have meaningful interactions -- [connecting] that patient to the appropriate type of service or care as opposed to just triaging for high risk," he stated.

He further noted that processes for triaging high-risk patients already exist, but finding solutions that can stratify mild, moderate and high-risk populations effectively is a daunting task.

"Two of our six top diagnoses are behavioral health diagnoses, so this issue is prevalent in terms of what primary care has in front of them," Attallah continued. "So we're trying to create a mechanism for [clinicians] to provide a very specific intervention that will then lead to a better outcome in terms of connecting [patients] to the appropriate support or service."

He underscored that effectively connecting patients to these services is crucial, as patients directed to services that aren't the right fit for them are significantly more likely to disengage from care long term. To help primary care providers better connect their patients with behavioral health interventions, Intermountain Health is leveraging its wealth of data.

Piloting behavioral risk stratification tools

To make use of its clinical data for behavioral health risk stratification, the health system is piloting NeuroFlow's Integrated Care solution. By capturing this data, the tool can help identify and triage risk within primary care workflows, enabling interventions like suicide prevention and crisis outreach.

But Attallah emphasized that the success of an analytics-driven initiative in the context of behavioral health is largely dependent on how these tools are integrated to support provider and patient experience.

"Primary care lives and dies by its ability to have efficient and effective workflows that work for providers," he said, explaining that the first phase of the pilot program is focused on implementing the Integrated Care tool in a way that works for both clinicians and patients.

He noted that the complexities of behavioral healthcare -- such as the need for patients to undergo initial risk screenings and fill out pages of questionnaires and other documentation -- can lead to poor consumer experiences and patient discontinuation of care. On the provider side, workflow inefficiencies and other contributors to clinician burnout can negatively impact satisfaction.

"Data can be amazing, but if the consumer experience and the provider experience is awful, it doesn't matter how good the technology is -- it doesn't get used," Attallah indicated. "So, this first phase is testing, how well does the tool integrate with our system? How seamless is it for consumers? How useful is it, and is it available within the EHR for our providers?"

By focusing on patient and provider experience from the outset, he continued, the partnership is better equipped to effectively utilize the health system's integrated behavioral health data that's been collected over the past two decades.

The fact of the matter is that behavioral health -- from a financial and a payment perspective -- is really under-resourced and underfunded, not just by the health system, but by payers. The whole ecosystem is not designed for sustainability.

As an early adopter of integrated behavioral health approaches, Intermountain Health has a well-established, standardized workflow for screening and risk scoring. However, much of this was a pen-and-paper process, creating a wealth of potentially underutilized historical data.

The second phase of the pilot aims to learn from that information, including insights into how historical and existing risk stratification, referral and clinical pathway approaches impacted the likelihood that patients would receive the appropriate level of care at the right time.

"Our goal, in phase two -- once we can make sure that the systems are seamless in terms of the [consumer and provider] experience -- is to create the data inputs to help us collectively learn," Attallah stated. "We want to take the existing data that's in the record to better predict what patients' needs are, so that their inputs are minimal."

Minimizing the number of questions posed to patients could have a significant positive impact on engagement, he underscored. Further, a seamless system that effectively helps providers parse relevant data enables them to better guide or escalate care.

These components have the potential to bolster patient education -- including insights into which behavioral health interventions have benefited similar patients -- and strengthen the patient-provider relationship, both of which are key to high-quality behavioral healthcare delivery.

"All too often, we're so focused on getting [patients] access to behavioral health that we're spending very little time on access to effective behavioral health, which are two different things," Attallah noted. "[Intermountain leadership] learned very, very quickly that we have to start with the patient, and in this case, the provider at the ground level, doing the work…. So, we plan to ensure that those workflows and designs are such that the consumer feels a seamless experience and the provider is not going to a separate system to gather this information."

Finding the right technology partner

Improving behavioral health risk stratification can be a daunting task, and Attallah emphasized the importance of learning from missteps along the way and choosing partners who can provide the necessary expertise to help overcome these hurdles.

"We have a lot of people that are interested in solving this problem," Attallah noted. "Intermountain is a large integrated health system, but we're not a technology company…. The reality is that there are many people doing the work, and we have to find partners that are aligned in what we're valuing."

For the health system, measurement and user-friendly experience were critical, making finding a partner with expertise in those areas key. He underscored that focusing on how the combination of a health system's clinical expertise with the expertise of an experienced data and technology vendor -- rather than "looking for the next shiny thing" -- is critical to the success of a pilot dealing with the complexities of behavioral health.

The approach allows Intermountain to accelerate the work it's already done in behavioral health risk stratification. This emphasis on collaboration and collective learning makes the first phase of a project particularly important to create a data- and technology-driven foundation for the work ahead, Attallah indicated.

"Intermountain has an incentive to provide proactive care as we start to think of the movement into value-based care," he stated. "We've been on that journey for a very long time, [and] this helps us get there -- what's best for the patient is to identify where their needs are early on, so that they never have to encounter additional unnecessary costs in healthcare."

This strategy is relevant for many health systems, especially as the industry grapples with the importance of addressing the behavioral health crisis that has been ongoing for decades, Attallah continued.

"The fact of the matter is that behavioral health -- from a financial and a payment perspective -- is really under-resourced and underfunded, not just by the health system, but by payers," he said. "The whole ecosystem is not designed for sustainability."

This is beginning to shift, especially as considerations like social determinants of health (SDOH) are becoming higher priorities for healthcare organizations. But moving the needle on progress in behavioral health requires large-scale efforts and ongoing collaboration.

"It's important to have commitment by the broader health system, not just behavioral health within the health system, but the broader health system. That's part one," Attallah explained. "Part two: We need to think about learning from each other and doing more collaboration, rather than competing, in the space. There's plenty of need, and we don't need to compete -- the problems are so complex, we need to be willing to dialogue and network with each other to learn from these incremental learnings that each of us likely has."

Sharing these learnings is crucial for shortening the pipeline between researching risk stratification solutions that help identify appropriate services for various patient populations and implementing those tools in clinical settings, he continued, adding that doing so can help de-stigmatize behavioral health and cement its importance for patients.

"I think we can all catalyze around that, do this work collectively and leverage technologies to help us enable doing that work with speed, which is important given the level of suffering that exists in our communities and across the country," Attallah concluded.

Shania Kennedy has been covering news related to health IT and analytics since 2022.