UT San Antonio study finds word choice is linked to depression and anxiety symptoms in 911 dispatchers
The study demonstrates how natural language processing can evaluate the psychological well-being of a critical but historically under-researched workforce.
Emergency call takers and dispatchers serve as the invisible frontline of public safety, navigating life-or-death crises over phone lines and radios without ever seeing the outcomes of their efforts. Now, a new study from UT San Antonio reveals that the specific words these workers use to describe their most stressful calls can provide early, hidden warning signs of depression and anxiety.
Published in the peer-reviewed journal PLOS One, the study demonstrates how natural language processing — a branch of artificial intelligence — can evaluate the psychological well-being of a critical but historically under-researched workforce.
The research team, led by College for Health, Community and Policy psychology professors Vivian Ta-Johnson, PhD, and Sandra B. Morissette, PhD, analyzed written narratives from 106 emergency communications personnel. The study was conducted in collaboration with the San Antonio Police Department and support from Bexar Metro 911.
“This collaboration is critical to advancing science and well-being in 911 telecommunicators,” said Morissette, who took the lead in establishing the San Antonio community partnerships.
A persistent occupational gap
Emergency call takers and dispatchers are routinely subjected to indirect, continuous trauma by managing high-volume calls involving violence, injury and death. Yet, they face a severe occupational gap. Unlike field-based first responders such as police officers or firefighters, dispatch center workers in many jurisdictions are denied official first-responder classification, limiting their access to dedicated wellness-related resources and funding streams.
“They are the first first responders, but they often don’t get the same kind of support,” explained Ta-Johnson.
They also face a unique psychological toll because they lack closure.
“A common theme is the lingering uncertainty following intense and distressing calls,” Ta-Johnson said, reflecting on the narratives she read. “It’s not uncommon for emergency call takers and dispatchers to wonder what happened and whether the person they were talking to got the help they needed. They usually don’t get that information directly because it’s always on to the next call.”
Measuring emotional tone versus intensity
To understand how this chronic stress shapes mental health, researchers asked participants to complete a standard self-report assessment for depression, anxiety and stress before typing out detailed descriptions of a high-stress workplace event.
Ta-Johnson’s team analyzed these testimonies using natural language processing tools that draw from a validated database of roughly 14,000 English words, each rated on several emotional factors. Using these ratings, the researchers quantified two key dimensions of emotional reactivity reflected in participants’ language: valence, which measures the emotional tone from positive to negative, and arousal, which measures the emotional intensity from calm to excited or panicked.
The results showed that narratives containing a higher proportion of negatively valenced words — such as “blood,” “danger,” “hate” and “shoot”— reliably predicted more acute symptoms of depression and anxiety. Follow-up tests confirmed that individuals meeting clinical thresholds for moderate or higher depressive and anxiety symptoms used significantly more negative language than those below the threshold.
While emergency call takers and dispatchers all experience objectively negative scenarios through their calls, the words they choose to describe the calls can shed light on their subjective, emotional interpretation of the calls, Ta-Johnson explained.
“What stood out was the use of more negatively valenced language, which appears to reflect not just what they experienced, but how they are processing those experiences psychologically,” she added.
Surprisingly, the intensity or arousal level of the language did not predict anxiety, depression, or stress symptoms.
“The experiences they described were all pretty emotionally intense, like suicides and medical emergencies,” Ta-Johnson said. “Language reflecting intensity is, to some extent, an unavoidable part of the profession.”
The findings also suggest that positive and negative word choices operate on independent tracks of emotional expression rather than as direct opposites. While highly negative language flagged mental health risks, a lack of positive words did not necessarily mean a call taker or dispatcher was struggling. In chronically high-stress professions, positive expressions are often suppressed or contextually constrained by the nature of the job rather than poor mental health.
Implementing proactive wellness support
The researchers emphasized that these language-based AI tools are not designed to diagnose mental health conditions, monitor employees punitively or replace human therapists. Instead, they hope the findings will lead to simple wellness tools that can be integrated into regular workplace protocols and used to support the employees who opt in.
For example, call takers and dispatchers could complete periodic, brief written reflections that language software could screen to flag individuals who might benefit from extra mental health resources. This unobtrusive strategy could be pivotal for first responders, who frequently hesitate to seek mental health support due to structural workload demands and workplace stigma.
Looking ahead, Ta-Johnson and Morissette plan to explore whether changing how telecommunicators narrate their trauma — such as writing in a more positive or past-tense frame — can influence reactions to trauma over time, including during narrative writing treatment interventions. They are also in the early stages of establishing partnerships with other police departments and telecommunication divisions to replicate and expand this research.
Co-authors on the paper include UT San Antonio psychology doctoral students, Isabella M. Swafford and Paige Lindsey, psychology postdoctoral fellow, Liliane D. Efinger, PhD, and Assistant Professor of Instruction, Janelle Kohler, PhD, alongside Michael Fernandez of the San Antonio Police Department and researchers from the Warriors Research Institute, Baylor Scott & White Health, Texas A&M University and the Baylor College of Medicine.
Journal
PLOS One
Article Title
Linguistic markers of emotional reactivity and their association with anxiety, depression, and stress among emergency call takers and dispatchers
Article Publication Date
8-Jul-2026
No comments:
Post a Comment