Does It Work?

Two peer-reviewed studies provide the primary independent research on crisis stabilization centers as of 2025: Burns et al. (2025) on emergency department diversion, and the Arizona State University / Connections Health Solutions study (2024) on patient reutilization. Program-reported data from Arizona’s AHCCCS system provides cost and impact figures cited in legislative contexts.

What Independent Research Has Found

Emergency Department Diversion

A 2025 study by researchers at Indiana University’s Richard M. Fairbanks School of Public Health examined longitudinal data from 2016 to 2021 across 1,002 zip codes in Arizona, Florida, Kentucky, Maryland, and Wisconsin. Funded by the Agency for Healthcare Research and Quality, the study found that availability of walk-in crisis stabilization services was associated with statistically significant reductions in mental-health-related emergency department utilization. The study used a multi-state panel dataset with controlled research design and was published in Health Services Research.

The study’s limitation is important to name: it examined availability of services, not quality of services. A crisis stabilization center in a zip code is associated with lower ED utilization whether that center is excellent or mediocre. The research confirms the diversion effect; it does not confirm that all centers produce equivalent outcomes.

Patient Reutilization in Arizona

A 2024 study by researchers from Arizona State University and Connections Health Solutions, using linked Medicaid claims data and electronic health records, examined how patients flow through Arizona’s crisis system. The researchers found that 57.4% of individuals entered Arizona’s crisis system through mobile crisis or crisis facilities rather than emergency departments, and that 63.3% of patients who received care at the Tucson Crisis Response Center did not reutilize crisis or emergency services within 30 days.

The 63.3% non-reutilization figure is meaningful: roughly two-thirds of patients who received care at this facility did not return to any crisis or emergency service within a month. The limitation: Connections Health Solutions, the facility operator, was a co-author of this study, which creates a conflict of interest that readers should weigh. The study used linked claims and electronic health records rather than self-report, which adds methodological rigor. But independent replication at other sites and with other operators has not yet been published.

What the Independent Research Does Not Show

Neither of these studies demonstrates long-term revolving-door reduction. The 63.3% non-reutilization figure covers 30 days; no multi-year independent study has tracked whether crisis stabilization reduces the long-term pattern of repeated emergency encounters. The Indiana University study shows diversion from ERs but does not address whether people who divert to crisis centers fare better over time than those who receive ER care. These are the research questions the field has not yet answered with independent rigor.

What Program-Reported Data Shows

Beyond independent research, a substantial body of program-reported data describes operational outcomes. These figures reflect what specific programs report about their own performance. They are not independently verified.

Arizona / Maricopa County: Cost and System Impact

Arizona’s “Crisis Now” model, operating in Maricopa County for over 30 years, reports the most frequently cited cost and impact data. Program-reported outputs from the Arizona Health Care Cost Containment System (AHCCCS) estimate that the state’s $100 million investment in the full crisis continuum (phone line, mobile teams, and stabilization facilities) produced the following outcomes compared with communities without such a system:

Reduced potential state acute care inpatient expense by $260 million
Avoided $37 million in hospital psychiatric boarding costs
Freed the equivalent of 37 full-time police officers from mental health transportation duty
Reduced cumulative psychiatric boarding time by the equivalent of 45 years

These figures are model estimates from program administrators, not independently audited results. The $260 million reduction is a projected cost avoidance (the cost of inpatient care that did not occur), not a direct accounting of dollars spent and saved. Finance-minded budget directors will distinguish this from an audited return-on-investment calculation. The estimates are consistent with the causal story (people stabilized in crisis centers do not require inpatient admission), but they should be presented as program-reported projections rather than verified outcomes.

Nashville Return-to-Jail Rate

Travis County, Texas Judge Andy Brown cited a Nashville, Tennessee figure to the Texas Tribune: 80% of people who received care at a crisis stabilization center did not return to jail. This figure comes through a second-hand source (a judge citing another city’s experience) and reflects one program’s outcomes with a specific population. It illustrates the potential of the model but should not be presented as the typical result across all centers.

Polk County, Iowa: First-Month Volume

Polk County, Iowa’s crisis stabilization center admitted more than 30 people in its first 30 days, which the local editorial board noted “exceeded operator expectations.” This is a simple volume figure that demonstrates demand exists; it does not speak to outcomes.

Phoenix Crisis Response Center: Volume and Outcomes

Connections Health Solutions, which operates the Phoenix Crisis Response Center (described by Connections as the nation’s largest dedicated psychiatric emergency center), reports serving more than 25,000 individuals annually and achieving “significant reduction of hospital admissions in Medicaid and behavioral health populations.” These are operator-reported figures, consistent with the 2024 Arizona State University study but not independently verified at the claimed volume.

Fulton County Atlanta: First-Six-Month Volume

Fulton County’s behavioral health crisis center served more than 500 patients in its first six months of operation following its October 2024 opening, according to LaTrina Foster, director of Fulton County’s Department of Behavioral Health and Developmental Disabilities. This is program-reported data from a center that had been open less than a year at the time of reporting.

Cost Comparisons

Arizona’s AHCCCS program-reported data estimates the per-boarding cost in emergency departments at $2,264 per person, drawn from independent hospital cost research. No independent multi-site cost comparison of crisis stabilization stays versus ER stays has been published as of 2025.

Memphis’s center was built with a dedicated first-responder drop-off area specifically designed to keep officer transfer times short — the Daily Memphian documented the 15–30 minute standard against the several hours an ER transfer typically requires. The time comparison reflects a cost-per-officer-hour dimension that police departments find meaningful even when dollar figures are not precise.

The Maricopa $37 million in avoided boarding costs translates to roughly $2,264 per avoided boarding episode, a figure drawn from independent hospital cost research that AHCCCS applied to its own volume data. This methodology is defensible as an estimate, though it combines third-party cost data with program volume claims.

The broadest cost claim, that Maricopa’s $100 million investment saved $260 million in reduced inpatient expense, should be presented as a program-reported projection, not a verified fiscal impact. This is the figure most likely to be challenged in a legislative hearing, and the distinction between “projected cost avoidance” and “documented savings” matters for credibility.

Law Enforcement Outcomes

Police departments in multiple documented jurisdictions have reported operational benefits after crisis stabilization centers opened.

Clarksville, Tennessee Police Chief Ty Burdine: the facility “will help streamline the process of connecting individuals to appropriate care while allowing officers to return to other duties more quickly.” Hawaii’s Dr. Chad Koyanagi: before the center opened, “a police officer could end up waiting hours with a patient” but now “can be back to the beat in like five or 10 minutes.” These are officer experience reports rather than measured outcomes.

What We Don’t Know

The honest accounting of evidence gaps:

Long-term revolving door reduction has not been demonstrated with independent multi-year research. The 30-day reutilization data from Arizona is the best available figure, and it covers a single operator in a mature system. Whether crisis stabilization reduces the 6-month, 12-month, and 5-year pattern of repeated emergency encounters remains unstudied with independent rigor.

Whether outcomes generalize across facility quality is unknown. The Indiana University study finds a diversion effect for crisis stabilization centers as a category. It does not measure whether the effect differs between high-quality and low-quality programs, between well-funded and underfunded centers, or between centers serving urban and rural populations.

Optimal length of stay, staffing ratios, and care model specifics have not been compared across sites in independent research. The Indiana University study confirms the diversion effect as a category finding; it does not identify which design features drive better outcomes within that category.

Youth-specific outcomes are particularly thin — this subpopulation is the least studied in the field. The youth-focused centers that opened between 2021 and 2025 are recent enough that longitudinal outcome data does not yet exist. Johnson County, Kansas opened a 10-bed youth center in spring 2025; Alaska’s Juneau youth center opened in late 2023. Neither has published outcome data. Whether outcomes demonstrated in adult-focused programs generalize to youth-specific centers has not been independently studied.

Generalizability of the Indiana University finding. The Burns et al. study examined availability of walk-in crisis stabilization services across zip codes, not the quality or specific design of those services. A center built in a zip code but operating at minimal capacity, with limited hours or understaffed clinical positions, may not produce the same ED diversion effect as a well-staffed, fully integrated facility. The study confirms the category effect; it does not isolate which design features drive the effect within the category.

Bottom Line

The Burns et al. 2025 study found a statistically significant association between crisis stabilization center availability and reduced emergency department utilization across 1,002 zip codes in five states — independent peer-reviewed evidence on the model’s core claim. Program-reported data from Arizona’s AHCCCS system projects $260 million in reduced inpatient expense from a $100 million investment; that figure is a model estimate from program administrators, not an independently audited fiscal impact. Long-term revolving-door reduction, quality variation across centers, and youth-specific outcomes have not been independently studied as of 2025.