Date of Completion


Embargo Period



Jane Ungemack, Greg Rhee, Kellie Randall

Field of Study

Public Health


Master of Public Health

Open Access

Campus Access


Introduction: Despite an increasing trend for children to visit the emergency department (ED) for behavioral health reasons, this is not the ideal treatment setting. One alternative service in Connecticut is Mobile Crisis Intervention Services, which has been shown to be effective in diverting behavioral health cases from the ED. Mobile Crisis provides children in crisis with a response in the community from a clinician, who can de-escalate the crisis and link the child to services. It is important to look at how this service is being used, and one such element is what factors are related to children who have multiple episodes with Mobile Crisis.

Research Questions: 1. Are different sociodemographic groups more or less likely to return to Mobile Crisis? 2. What presenting problems are associated with repeat Mobile Crisis episodes? 3. Is the child’s clinical acuity and history of mental health care associated with the number of Mobile Crisis episodes? 4. How are the characteristics of the initial Mobile Crisis episode, and the services provided during that episode, related to the subsequent number of episodes (e.g. source of referral to Mobile Crisis, length of service, follow-up care, and referral to long-term care)?

Methods: This study employed a retrospective cohort design, analyzing data for children who had their first Mobile Crisis episode between 2015 and 2017. Data for these children were looked at for two years following their first episode, to determine how many episodes each child had during that time. In addition to descriptive analysis of the sample, binary logistic regression and zero-inflated binomial regression were used to look at the relationships between various factors and the outcome variable.

Results: Findings were largely consistent across the two regression models, indicating stronger support for the results. There were several notable findings. Certain presenting problems – disruptive behavior and harm/risk of harm to self – are associated with an increased number of Mobile Crisis episodes. Additionally, children with more acute symptoms at intake and a history of behavioral health services were more likely to return to Mobile Crisis. Children were most likely to have multiple episodes if they were referred by their family rather than a third party. There were also several associations found for demographic groups, with females, younger children, and children living in certain regions of the state being more likely to have multiple episodes.

Conclusions: This research has important implications for Mobile Crisis and for behavioral health services as a whole. Though additional research is needed, these findings begin to provide a framework for how this service is used, and may be able to inform practice. Investigation of this service is not only useful for Mobile Crisis in Connecticut, but for mobile crisis response and other crisis management systems in other states.

Major Advisor

Jane Ungemack