Rethinking How Hospitals Measure Behavioral Health Success

​Originally published on MedCity News on July 23, 2026.

By Shannon Werb, Chief Executive Officer at Array Behavioral Care

Two patients who present in the emergency department in identical ways can have very different treatment experiences, even if they’re both assessed within the standard window and discharged within the target length-of-stay. Both treatments would appear normal in an operations dashboard. However, one patient could get better, while the other ends up back in the emergency department weeks later, even worse than before.

Most hospitals rely on emergency department behavioral health metrics that can’t tell these two patients apart. Discharge rates and patient length-of-stay are often the default performance indicators, but they don’t accurately reflect whether behavioral health cases received the right care. A high discharge rate is typically read as operational success, but in behavioral health, it can signal the opposite: patients moved through the system before their symptoms were properly addressed. A short length-of-stay carries the same implications. It’s marked as a win operationally, but there is no way to tell if the patient actually got the right level of care for longer-term clinical success.

Between 2017 and 2019, roughly 53 of every 1,000 emergency department visits were related to mental illness, with a potentially significant portion of patient outcomes being monitored by non-optimal metrics. The risks of misdiagnosis with behavioral health are high, with potential increases to patient symptom acuity as well as additional strain on the health system through preventable readmissions. Hospitals that see improved long-term results in behavioral health patients will be those that prioritize metrics aligned with their care pathways.

The one-hour psychiatric consult is a classic example of a non-optimal throughput metric for behavioral health patient treatment. It’s a common hospital performance target measuring how quickly a psychiatric clinician can evaluate and disposition a behavioral health patient in the emergency department – but focusing on speed may reward the wrong behavior. It shouldn’t be about how fast a clinician got there or how long they stayed, but whether the patient was effectively assessed, stabilized, and routed to the right level of care.

There are substantial risks when throughput metrics like turnaround-times, discharge rates, and patient length-of-stay are prioritized for behavioral health cases. The first is risk stratification: a patient with moderate or high acuity can leave the emergency department within the target length-of-stay window without being matched to the level of follow-up care their symptoms warrant.

The second is the readmission that doesn’t read as one. A treated behavioral health patient who returns in a month or two because they’re not seeing results is seen as a new case. Often, there isn’t enough data from a quick previous diagnosis to advance the patient’s treatment path. They’re stuck answering the same questions, and the cycle continues.

Disposition shaped by capacity can become a risk as well when bed space in the emergency department is limited. If a patient’s perceived acuity is lower, they may be expedited to open up space for another patient with symptoms that present as more acute.

Different metrics are available to help hospitals get closer to answering the real question: whether behavioral health patients received the right care. The emergency department’s job isn’t to resolve a behavioral health crisis on its own. It’s to assess and stabilize the patient, then connect them to the right next level of care. So, the metric that matters most at the acute stage is whether that connection actually happens. One established metric already does this: the HEDIS Follow-Up After Emergency Department Visit for Mental Illness (FUM) measure, which tracks the share of behavioral health ED visits with a follow-up visit within 7 and 30 days of discharge. It’s an acute-care metric, but it measures whether the patient reached outpatient care, making it the bridge between the two settings.

Nationally, even under the widely tracked FUM measure, only about half of Medicaid patients receive any follow-up within 30 days of a behavioral health emergency department visit, and far fewer, closer to a third, within 7 days. That number might vary slightly across demographics, but the problem remains the same. High-acuity behavioral health patients early in their treatment path don’t have the established relationships of chronic cases, and are less likely to seek out follow-up care without an active outreach effort from the medical system.

While FUM tells you if the patient reached the next level of care, it doesn’t tell you whether that care worked. That question belongs to the outpatient setting, and it has its own signals. In outpatient behavioral health, symptom trajectory is a useful signal for whether treatment is actually helping. Instruments like the PHQ-9 and GAD-7 symptom scales, for depression and anxiety respectively, read change over time. One score is just a snapshot. Multiple scores across visits, combined with how often a patient returns to the emergency department, give a much stronger read on whether a treatment path is working.

Putting this into practice within emergency departments helps better match clinical resources to patient needs. Master’s-level clinicians can assess patients and conduct follow-ups, psychiatrists can focus on patients who need escalation, and inpatient beds open up for patients who require them.

The downstream impact on patients is tied to what they leave with at discharge, and whether they’re connected to outpatient care. Ensuring proper follow-up is key. A severity score, like a PHQ-9 or GAD-7, gives outpatient clinicians a baseline to work from. Documentation of acute moments in time for the patient can serve as an anchor for any return visits or additional measurements. If actions are taken and scores still don’t improve, the care pathway can be escalated to a higher level of care. If a patient’s scores improve and symptoms decrease, clinicians can step down treatment to a maintenance level.

Dashboards frequently rely on metrics that are easy to measure, but hospitals optimized for throughput metrics like length-of-stay and discharge volume may continue to see patients who leave fast and return faster. Resources will continue to be wasted on preventable readmissions, psychiatrist hours spent re-evaluating patients who weren’t fully assessed the first time, and inpatient beds tied up by patients who didn’t need that level of care.

Transitioning to outcomes-tracked metrics will give hospital leaders a clearer picture of what’s working and where patients are falling through. Hospitals that start measuring acuity-matched disposition, follow-up rates, and symptom improvement will catch the gaps that throughput metrics miss entirely, and stop cycling the same patients back through the emergency department.