Using AI to Support Healthcare Amid a Mental Health Provider Shortage

THE CHALLENGES OF ADDRESSING A MENTAL HEALTH PROVIDER SHORTAGE

When considering the challenges posed by a mental health provider shortage, it’s crucial to understand that these shortages exist across various levels of behavioral health expertise, Malloy said. Psychiatric providers, licensed clinicians, and support staff are all key to the provision of behavioral health services, and there are shortages in each group.

However, for an organization like Futures, Malloy noted that working to recruit and retain these staff is a top priority.

“So being in a 24/7 facility, we can't have a shortage," she said. "We have certain licenses that we abide by [and] we have to have a certain amount of nurses, a certain amount of clinicians on staff.”

These staffing requirements can pose a challenge of their own, though, as the requirements are often vague, and determining whether the ratio of staff to patients is effective for delivering high-quality care depends on the severity and acuity of Futures' cases at a given time, Malloy explained.

Because of this, and due to variations in staff availability and demand for services, the organization sometimes needs to overcompensate in terms of staffing and how clients are assessed.

The nature of the care provided at Futures requires an 'all hands on deck' approach, meaning that the facility generally needs to be ‘overstaffed’ in order to be adequately prepared for unexpected events, like a staff member leaving. This was something that became increasingly common during the pandemic and the rise of telehealth, she stated.

In doing so, Futures has been able to prevent care gaps resulting from slowdowns in recruiting, which, according to Malloy, could take months now instead of weeks.

Wood echoed these sentiments, indicating that regional variations in staff availability are also a major hurdle for behavioral health facilities.

The lack of available mental healthcare providers differs from state to state and city to city. While Asheville, North Carolina, where Oasis is located, is fortunate to have many providers, other rural locations in HCANA’s network, like Chattanooga, Knoxville, and certain areas in Tennessee, face limitations in provider availability, he explained.

To attract providers, HCANA is advertising job postings and working to upgrade its care teams. But that approach can only go so far, and the network’s current staff also need further support.

Supporting staff across the organization is a major goal for both HCANA and Futures. To achieve that aim, the two organizations are leveraging an artificial intelligence (AI)-based mental health screening tool known as Aiberry.

HOW AI CAN STREAMLINE PROCESSES ACROSS MENTAL HEALTH ORGANIZATIONS

The AI tool is designed to provide mental health insights by analyzing visual, audio, and language features pulled from screening conversations between individuals and their mental health providers. HCANA and Futures are both using the tool throughout each patient’s care journey, from intake to at-home follow-up.

Malloy explained that mental health and addiction treatment require ongoing support to help patients maintain accountability and develop adaptive coping skills. As part of this process, tracking an individual's progress and identifying patterns that contribute to their mental health is critical.

“If I come in and my depression is way up here, what [the tool] helps us do is look at what's happening while [the individual is] in with us,” Malloy said. “Did they just have a family session? Did they just talk about their trauma? Did they not sleep last night? We can take that, look, and pinpoint a timeline of exactly when they're spiking.”

These insights are invaluable because not everyone is self-aware enough to identify and verbalize the factors that negatively impact their mental health, she added.

The tool also allows Futures to better gauge the efficacy of treatment interventions in near real-time because it can capture these nuances, whereas traditional methods may take months to determine whether a treatment is working well.

This capability is important when navigating a mental health provider shortage, as it enables the organization to better allocate staff and optimize resources for interventions that are working and for individuals with potentially unmet needs that require additional support.

Wood noted that at HCANA, the tool has been implemented not only for intake but also to streamline individuals’ weekly check-ins prior to seeing a therapist. These check-ins take approximately ten to 15 minutes, which helps carve out more time for therapy sessions that have lengthened to an hour to an hour and a half.

Recently, HCANA has implemented another capability of the tool, allowing individuals to download the technology to use at home for follow-up. Wood further indicated that the tool could be used to generate more outcomes data for therapist review from these follow-ups, which would benefit aftercare planning.

Moving forward, the organization hopes to integrate the AI technology with other tools and processes to improve care. For instance, integrating the solution into EMR systems would allow HCANA to automate uploads of assessment outcomes into client files, eliminating the need for manual data entry.

USING TECH TO SUPPORT ENGAGEMENT, RISK ASSESSMENT, AND SAFETY

As experts have debated what role AI could play in mental healthcare and how digital solutions may help tackle provider shortages, one major question is always present: how will these technologies impact patient experience and safety?

Malloy and Wood were initially concerned about how individuals would react to the AI tool, citing difficulties in getting some people to complete traditional assessments.

In Malloy’s experience, standard pen-and-paper assessment methods are taxing for individuals and often require incentives like gift cards to encourage participation. Certain groups, like veterans, first responders, firefighters, EMTs, and police officers, can also be resistant to assessment or prone to minimizing their struggles.

She explained that people can feel shame around the assessment process, particularly when it requires them to write about their experiences, as many individuals subconsciously seem to believe that writing things down makes them true.

Using Aiberry helps remove this barrier, leading to increased engagement. Malloy indicated that patients at Futures are more proactive and motivated to complete their next assessment with the tool. Their enthusiasm appears to stem from a sense of empowerment they get from using the tool.

Wood describes a similar experience at HCANA.

“The clients, if I had to throw a percentage out there, 95 percent of them are very open to [Aiberry’s tool] and actually enjoy using it,” Wood said, noting that HCANA had more initial hesitation from staff and clinicians when the tool was first deployed.

“The most challenging thing for us has been getting our longer-term therapists, so therapists that have been practicing 20, 30 years as a master's level clinician to say, ‘I'm okay with bringing technology into my session,’ because they are the master's level professional, and they've worked very hard to get to where they're at,” he added.

However, once staff were given more guidance and training on the tool, HCANA saw improved acceptance and willingness to adopt the technology.

The tool also provides risk assessment capabilities and additional insights that can enhance patient safety.

Malloy explained that in the type of work done at facilities like Futures, sometimes the outcome is death. Suicide is a major cause of morbidity for individuals receiving mental healthcare, and assessing whether a patient is at risk for suicide is a significant priority for mental health organizations.

Futures uses the AI tool in combination with other solutions, such as the Columbia Suicide Severity Rating Scale (C-SSRS) and the Suicide Assessment Five-Step Evaluation and Triage (SAFE-T) assessment, to evaluate patient risk throughout treatment and ensure patient safety.

Thus, leveraging an AI tool as part of a mental health organization’s overall strategy can not only help tackle challenges like a provider shortage but also help empower individuals to be more involved in their mental healthcare.

“I've always struggled with the way [mental health is] assessed,” Malloy said. “I think that sometimes you get it so wrong, and when you get it wrong, it changes the path of someone's life. [The tool] gives so much strength to the individual; it really does empower them to be a part of the process… It's collaborative, and it needs to be.”

“It's just about taking that stressor off that person and getting really to the nitty gritty of what's going on, so we can get into the solution,” she continued. “Because mental health isn't a one size fits all. We have to understand that, and sometimes we have to think outside of the box and do what's necessary for that person.”