UCSF

Objective

Crisis services are undergoing an unprecedented expansion in the United States, but research is lacking on crisis system design. This study describes how individuals flow through a well-established crisis system and examines factors associated with reutilization of such services.

Methods

This cross-sectional study used Medicaid claims to construct episodes describing the flow of individuals through mobile crisis, specialized crisis facility, emergency department, and inpatient services. Claims data were merged with electronic health record (EHR) data for the subset of individuals receiving care at a crisis response center. A generalized estimating equation was used to calculate adjusted odds ratios for demographic, clinical, and operational factors associated with reutilization of services within 30 days of an episode's end point.

Results

Of 41,026 episodes, most (57.4%) began with mobile crisis services or a specialized crisis facility rather than the emergency department. Of the subset (N=9,202 episodes) with merged EHR data, most episodes (63.3%) were not followed by reutilization. Factors associated with increased odds of 30-day reutilization included Black race, homelessness, stimulant use, psychosis, and episodes beginning with mobile crisis services or ending with inpatient care. Decreased odds were associated with depression, trauma, and involuntary legal status. Most (59.3%) episodes beginning with an involuntary legal status ended with a voluntary status.

Conclusions

Crisis systems can serve a large proportion of individuals experiencing psychiatric emergencies and divert them from more restrictive and costly levels of care. Understanding demographic, clinical, and operational factors associated with 30-day reutilization may aid in the design and implementation of crisis systems.