Leveraging Data Science for Rapid Response to Mental  Health Crises

August 11, 2022

The first thing you hear when calling the doctor's office is: “if this is an emergency, please hang up, and call 911.” These words are ubiquitous within the healthcare community, yet, little is known about the care journey for patients experiencing a mental health emergency. Most mental health practices do not track the number of patients who call experiencing a crisis, and for patients that end up calling 911, there is little data shared about the wait times for addressing these emergencies. Recognizing the disjointed options that exist for addressing mental health crises, Cerebral decided to develop a data-driven approach to improving clinical safety.

On a typical day, patients send anywhere from 5,000-10,000 messages to Cerebral via our proprietary electronic medical record (EMR) message system. Our message system gives patients the opportunity to connect with us on a wide variety of topics ranging from appointment rescheduling to questions about treatment plans, medications and more. It also serves as an avenue of contact for patients who are experiencing a crisis.

Each message is reviewed and addressed by the patient support team. While this approach is best suited for non-emergencies, more serious mental health crises, such as suicidal ideation, need to be responded to as soon as possible. With this as our guide, we developed ways to improve our emergency response time. 

Using machine learning-enabled tools to improve mental healthcare delivery

Machine learning (ML) and artificial intelligence (AI) are only possible at scale. Because of Cerebral’s scale, we are uniquely positioned to develop and implement cutting edge ML/AI tools that will improve clinical outcomes. All messages sent to via Cerebral’s proprietary EMR message system are and will continue to be viewed by Cerebral’s professional staff. With our new ML/AI tool, we now have the ability to respond to patients in crisis in near real-time.


With a cross-functional team of clinicians, crisis specialists, and data scientists, we leveraged millions of patient messages to build Crisis Message Detector 1 (CMD-1). CMD-1 is a tool that shares messages from patients experiencing a mental health crisis to our crisis response specialists. The model is designed to alert the Crisis Response Team of messages that may indicate suicidal ideation, homicidal ideation, non-suicidal self-injury, or domestic violence, among other emergency situations. Once alerted, the Crisis Response team reaches out to patients directly to assess the level of immediate risk and mobilize emergency contacts and/or local responders as deemed necessary.

CMD-1 allows us to significantly reduce the amount of time a patient experiencing a mental health crisis must wait before receiving support. During a week-long pilot, CMD-1 screened over 60,000 EMR messages and flagged over 500 potential crises. The model successfully detected over 99% of all crisis messages, and as a result, crisis specialists were able to respond to patients in less than 9 minutes on average.

We have officially launched CMD-1 at scale, allowing our crisis specialists to support patients when they are at their most vulnerable. This is just the beginning for machine-learning-enabled chat message triage. In the coming months, we will apply this technology to improve response times for medication concerns, scheduling issues, and support requests across the board.

Special thanks to the Cerebral employees and advisors who participated in the design, build and testing of our CMD-1 System: Tyler Heist, Madeline Holmes, Matthew Rubashkin, Sid Salvi, Shaked Azzam, Dr. David Mou, Dr. Matthew Nock and the entire Cerebral Clinical Quality and Safety teams!

Cerebral Inc. picture

Cerebral Inc.

Cerebral logoCerebral logoCerebral logo
If you're in emotional distress, here are some resources for immediate help:
Call 911
National Suicide Prevention Hotline:
Call 988
Crisis Text Line:
Text Home to 741-741
Let’s stay in touch
Google Play ButtonApp Store Button