For health systems under tremendous operational pressure, metrics like high behavioral health discharge rates or reduced inpatient admissions can look like an attractive shorthand for evaluating acute care psychiatry partners. Many vendors lean on these numbers in their marketing—implying that a lower admission rate automatically signals better clinical quality or stronger ROI.
But the reality is far more complex.
These numbers do not exist in a vacuum. And selecting partners based on discharge rate alone—or using them to calculate ROI—can be misleading, and in some cases, harmful if the data lacks clinical and contextual interpretation.
After more than two decades of providing virtual psychiatric assessment across emergency departments, inpatient units, and other acute settings, I’ve seen how easily “good” numbers can obscure the truth.
When I see aggregate discharge rates being touted without context, I have the same reaction every time: they only tell a fraction of the story.
Every hospital or ED operates within its own ecosystem—shaped by patient mix, acuity patterns, care delivery workflows, resource availability, and the community it serves. These factors meaningfully shape clinical presentations and influence the likelihood of a safe discharge.
At Array, our six-month adjusted discharge rate across acute partners is 39%. But the range tells the real story:
%
Underserved Urban ED Discharge Rate
%
Well-Resourced Suburban ED Discharge Rate
- An urban ED serving an economically challenged, underserved population averages 26% discharge.
- A psychiatric emergency screening program within a well-resourced, commercially insured suburban system averages 94% discharge.
Both are appropriate. Both reflect the realities of their settings. And both look “good” or “bad” only if you strip away the context that actually matters.
This is why simplistic comparisons—or cherry-picking the lowest admission rate from a vendor’s entire customer base—are not just unhelpful. They’re misleading.
Context changes everything.
A discharge is never “just a discharge.” It is the outcome of a complex clinical picture—one shaped by medical, psychiatric, and psychosocial variables that differ dramatically across settings.
Some acute environments care for patients with higher clinical severity or multiple social determinants of health that limit access to outpatient follow-up. Others evaluate populations with lower acuity and stronger community resources. These differences directly influence disposition.
Data science helps quantify these realities.
Psychotic disorders, for example, are a well-established risk factor for suicide. A major meta-analysis of 50+ studies shows individuals with psychotic disorders have 40% higher odds of suicide than those without. That level of risk meaningfully shifts the threshold for safe discharge.
%
Before we collaborate with any new partner, we start with discovery.
We need to understand:
- How does the organization define consult triggers?
- What is the typical patient mix and acuity?
- What operational pressures shape decision-making?
- What community resources exist—or don’t exist—to support discharge?
Using national evidence and our internal data, we model what that specific organization can realistically expect from a high-quality psychiatric assessment service.
This isn’t marketing spin. It’s responsible clinical partnership.
Our models rely on factors hospital leaders already know—diagnostic mix, available resources, workflows. We simply apply clinical and data-science precision to help partners understand what’s achievable in their setting.
A high discharge rate is not a sign of a good partner.
A clinically appropriate discharge rate is.
Acute psychiatric assessment requires a careful balance:
- Protecting patient safety
- Reducing unnecessary disruption
- Preserving inpatient access for those who truly need it
- Supporting equitable treatment
- Thoughtfully assessing factors that increase the likelihood of recidivism, including clinical severity, social complexity, and barriers to follow-up care
- Using resources responsibly
This balance becomes impossible when partners chase the lowest admission rate or build ROI projections off numbers pulled from a population that looks nothing like their own.
In today’s data-driven environment, health systems deserve more than appealing metrics. They deserve partners who:
- Interpret data through a clinical lens
- Understand population-level risk factors
- Model realistic expectations based on actual patient mix
- Support safe and equitable decision-making
- Deliver consistent, evidence-based care across settings
Numbers matter. But numbers without context can lead organizations in the wrong direction.
The real ROI of high-quality acute psychiatric assessment isn’t measured by the lowest admission rate—it’s measured by getting patients the right care, at the right time, in the right setting.
That’s the standard health systems should expect.
