Two patients faced a behavioral health crisis this morning, both experiencing severe depression with thoughts of self-harm. The first called their provider at 8 a.m. and secured the last same-day appointment. The second called an hour later, found no openings for five days, and now sits for hours in a crowded emergency room — their only option for immediate care. This scenario plays out daily across the U.S., highlighting a dangerous flaw in behavioral healthcare: Access to urgent behavioral health treatment often depends more on when someone calls than their immediate need.
This misalignment strains our entire healthcare system. More than 1 in 5 U.S. adults live with behavioral health conditions, and behavioral health-related emergency department (ED) visits have doubled from 2011 to 2020. According to the latest CDC research, the numbers sit at around 47 visits per 100 people — over 157 million people annually. Healthcare organizations can no longer sustain a system that distributes care based on timestamps rather than clinical urgency.
Artificial intelligence is rewriting this narrative. By 2025, intelligent scheduling systems will replace our current queue-based approach with smarter resource allocation that helps connect patients with timely care, reducing the need for ED visits.
The true value of AI in behavioral health will not come from automating paperwork or generating clinical notes — it will come from fundamentally reshaping how we allocate and optimize care resources.
Breaking Free from Queue-Based Care
Forward-thinking health systems are already exploring solutions to the prioritization challenge, starting with risk stratification: A way of evaluating each patient’s clinical needs and circumstances to determine the most appropriate timing and level of care. Intermountain Health has begun integrating risk stratification tools into its primary care settings, moving beyond traditional scheduling to better prioritize urgent appointment requests. However, implementing these changes requires careful consideration of clinical workflows and patient trust.
The U.S. Food and Drug Administration (FDA) acknowledged this challenge in 2022, focusing on regulating AI tools used for patient risk assessment and resource distribution. This regulatory attention signals a shift from viewing this technology as merely an administrative tool to recognizing its role in clinical decision-making.
Recent research by Cambridge University supports this direction, reporting that patients increasingly trust machine learning (ML)-based clinical support systems — provided that there is some transparency about how they work. Trust increases by 5% when patients understand how these systems help clinicians make informed decisions. Furthermore, patient skepticism decreases by 3-4% when patients learn how these tools support rather than replace clinical judgment.
Intelligent Care Prioritization in Practice
Risk stratification offers a smart way to connect patients with the appropriate level of care they need, when they need it. Think of it as replacing our current time-stamp system with one that actually reflects clinical priorities. While this might sound obvious, implementing it at scale requires sophisticated technology to process and analyze multiple data points, such as clinical history and current symptoms, to improve urgent care access.
With risk stratification, providers receive AI-supported insights about their patient population to help identify who might need proactive outreach before reaching a crisis point. A patient who usually attends therapy regularly but suddenly misses several sessions, or someone whose medication refill patterns have changed, might signal a need for proactive outreach.
The technology supporting these decisions has matured significantly. Today’s ML systems analyze anonymized patient data to detect subtle patterns that might escape human notice while maintaining strict privacy standards. These systems don’t make clinical decisions; they provide additional context to help clinicians prioritize their time and resources more effectively.
Key Technology Considerations for 2025
Healthcare organizations planning for this transition need a measured, strategic approach. Success with AI-supported risk stratification requires more than simply implementing new technology, demanding careful attention to operations, training, and organizational culture.
Start by examining your current digital infrastructure. With 55% of behavioral health encounters already happening virtually, most organizations have foundational systems in place.
The next step involves mapping which operational processes would benefit most from intelligent prioritization — particularly areas like intake processes and follow-up scheduling where better resource allocation creates immediate impact.
Technology alone cannot drive this transformation. Staff engagement is essential because behavioral health providers must trust these tools before incorporating them into daily patient care decisions. Clinicians who understand and believe in the system’s ability to support their work become its strongest advocates and champions for change.
This engagement requires active participation. Create opportunities for clinical teams to shape implementation through hands-on system training, workflow design input, and regular feedback channels. When providers see these tools enhancing, not replacing, their clinical judgment, they integrate them naturally into patient care.
Success must be measured through concrete operational metrics. Imagine our same two patients in 2025: When two patients call their provider describing severe depression with thoughts of self-harm, the AI-enabled system immediately flags both cases as high-risk for escalation, ensuring same-day appointments regardless of who called first. This type of intelligent triage leads to measurable improvements: reduced ED utilization, shorter time-to-care for patients, and optimized provider capacity. Organizations can track how effectively they provide appropriate care access, documenting cost savings from improved scheduling efficiency and more efficient resource allocation.
Risk stratification in behavioral health offers a practical solution to one of healthcare’s most pressing economic challenges. By improving how we connect patients with timely care, healthcare organizations can reduce costly emergency interventions while improving care delivery. Successful behavioral healthcare organizations in 2025 will be defined not by the number of patients they see, but by their ability to match the right care to the right patient every time.
Image Source: ID 137795256 | Artificial Technology ©
Thekaikoro | Dreamstime.com
Andy Flanagan
As CEO, Andy Flanagan is responsible for Iris Telehealth's strategic direction, operational excellence, and the cultural success of the company. With significant experience in all aspects of our U.S. and global healthcare system, Andy is focused on the success of the patients and clinicians Iris Telehealth serves to improve people’s lives. Andy has worked in some of the largest global companies and led multiple high-growth businesses providing a unique perspective on the behavioral health challenges in our world. Andy holds a Master of Science in Health Informatics from the Feinberg School of Medicine, Northwestern University, and a Bachelor of Science from the University of Nevada, Reno. Andy is a four-time CEO, with his prior experience including founding a SaaS company and holding senior-level positions at Siemens Healthcare, SAP, and Xerox.
