Why diversion dashboards fail (and what “good” looks like)
Crisis diversion governance is not a slide deck or a quarterly narrative. It is a live operating system that proves three things at the same time: (1) people are being diverted appropriately (right person, right setting, right time), (2) the diversion pathway is safe and clinically defensible, and (3) the system is not simply shifting cost and risk to other parts of the continuum. The only way to do that at scale is to run diversion as a measurable workflow with thresholds, escalation triggers, and an audit trail.
Well-built dashboards don’t start with “how many diversions did we do?” They start with the failure modes: unsafe diversion, diversion denial that pushes people into ED/jail, delayed step-down that creates boarding, and repeat crisis that signals unmet need. A governance dashboard should make those failure modes visible early enough to intervene, not explain them after harm occurs.
What a diversion governance dashboard must include
A usable dashboard is organized around the actual decision points: referral/triage, clinical clearance, placement, handoff, and follow-up. Each stage needs measures for volume, timeliness, quality/safety, and equity. It also needs “integrity” measures that test whether the data and documentation match what staff say they are doing.
- Front door: 988/911/ED/community referrals; response times; disposition categories; reasons for non-diversion.
- Clinical clearance: risk stratification completion; consult utilization; medical clearance exceptions; involuntary hold triggers.
- Placement: time-to-bed/slot; acceptance/denial reasons by provider; capacity by acuity; boarding time.
- Handoff and follow-up: warm handoff completion; first follow-up contact within target; medication reconciliation status; care plan adherence checks.
- Outcomes: repeat crisis contacts, ED presentations, jail bookings, safety events, grievances, and adverse outcomes by cohort.
Oversight expectations you should design around
Expectation 1: Auditability and medical necessity alignment. Whether the payer is Medicaid fee-for-service, a managed care plan, or a county authority, the system will eventually be asked to demonstrate that diversion decisions were clinically justified and consistent with medical necessity criteria. Your dashboard therefore must link “what we did” to “why it was appropriate,” with documentation completeness measures (risk tool used, clinician role, supervisory sign-off, and disposition rationale).
Expectation 2: Equity and access transparency. State authorities, county boards, and many payers increasingly expect stratified reporting: diversion rates, denial reasons, and safety outcomes by race/ethnicity, age band, geography, language need, and disability status where feasible. Governance that does not stratify can look “good” overall while masking disproportionate diversion denial or unsafe diversion for specific groups.
Operational Example 1: “Disposition Integrity Review” for 988 and mobile crisis
What happens in day-to-day delivery
Each day, a designated quality lead pulls a stratified sample of crisis episodes from the 988/mobile crisis platform and the dispatch log (including “resolved by phone,” “mobile response,” “diverted to crisis receiving,” and “sent to ED/911”). For each sampled episode, the reviewer checks: risk screen completion, documented rationale for disposition, supervisory consult where required, and whether the disposition matches what happened next (e.g., a “diverted” case that still ended up in the ED within 6 hours). Findings are coded into standard categories and fed back in a 15-minute huddle with the shift supervisor and the clinical director.
Why the practice exists (failure mode it addresses)
This practice prevents “paper diversion,” where the record shows diversion but the person still escalates to ED/jail because the plan was not feasible, the follow-up did not occur, or risk was underestimated. It also addresses inconsistent use of disposition categories across staff and shifts, which makes system metrics unreliable and undermines payer or regulator confidence.
What goes wrong if it is absent
Without integrity review, teams can unknowingly drift into unsafe patterns: high “phone resolution” with hidden repeat callers, mobile teams under-documenting clinical rationale, or supervisors not being consulted for higher-risk presentations. When an adverse event occurs, leadership cannot reconstruct why the decision was made, which exposes the program to corrective action, contractual penalties, and reputational damage with ED partners and law enforcement.
What observable outcome it produces
Within weeks, the program should see improved documentation completeness, fewer “unknown” disposition codes, and reduced mismatches between recorded disposition and subsequent utilization. Evidence includes audit scores, declining rates of ED arrival within defined windows after “diversion,” and improved inter-rater reliability in how risk levels and dispositions are assigned across shifts.
Operational Example 2: Bed management governance that prevents “diversion denial by delay”
What happens in day-to-day delivery
A central bed/slot coordinator maintains a live census for crisis receiving, stabilization, respite, and step-down settings, updated at least every two hours. Mobile crisis and ED liaisons submit placement requests through a standardized form that includes acuity, medical clearance status, and key risk factors. The coordinator runs a twice-daily capacity call with providers to reconcile openings, anticipated discharges, staffing shortfalls, and any acceptance constraints. When a placement decision is made, the coordinator logs the acceptance time, transfer time, and handoff confirmation.
Why the practice exists (failure mode it addresses)
The failure mode is “diversion denial by delay”: people who meet diversion criteria are held too long in the ED or in the community because the system cannot quickly match them to an appropriate level of care. Delays create clinical deterioration, increase use of restraints or involuntary holds, and strain ED relationships—then diversion is blamed for outcomes that were actually caused by throughput failure.
What goes wrong if it is absent
Without active capacity governance, placement becomes a series of untracked phone calls. Providers can deny by default (“no beds”) without visibility into discharge barriers, and ED staff lose trust and stop referring to diversion pathways. Leaders then see “low diversion rates” but cannot tell whether the cause is inappropriate referrals, lack of capacity, provider denials, or broken logistics.
What observable outcome it produces
Observable improvements include shorter time-to-placement, fewer ED boarding hours for diversion-eligible people, fewer “no bed” denials that later prove inaccurate, and a clearer capacity plan (e.g., which acuity tiers are consistently saturated). Evidence includes transfer time dashboards, denial reason distributions that shift from vague to specific, and improved compliance with target timeframes in contracts or MOUs.
Operational Example 3: Repeat-utilizer governance tied to post-diversion follow-up
What happens in day-to-day delivery
The system defines a repeat-utilizer cohort (for example: 3+ crisis episodes in 30 days or 5+ in 90 days) and automatically flags those individuals in the crisis platform. A multidisciplinary review panel meets weekly (mobile crisis supervisor, care coordination lead, peer supervisor, and a payer/county liaison where permitted) to review each flagged case. The panel confirms whether the person had a completed safety plan, whether follow-up occurred within the target window, whether medication access was verified, and whether housing or benefits instability was addressed. The panel assigns an accountable owner and sets a next-action date, then tracks closure.
Why the practice exists (failure mode it addresses)
This practice addresses the predictable breakdown where diversion resolves the immediate crisis but fails to stabilize the underlying drivers (medication gaps, missed appointments, housing loss, interpersonal violence, or untreated substance use). Without governance, repeat crises become “normal,” and the system spends resources cycling the same people through short interventions without durable impact.
What goes wrong if it is absent
If repeat-utilizer governance is missing, teams misinterpret high repeat contact rates as “patient noncompliance” rather than system failure to connect services. EDs and law enforcement see the same individuals repeatedly and lose confidence in diversion. Payers may then restrict diversion funding or impose burdensome authorization requirements, further degrading access.
What observable outcome it produces
When implemented well, the system should see reduced repeat episodes for the flagged cohort, higher completion rates for follow-up within defined timeframes, and fewer unplanned escalations after diversion. Evidence includes cohort-level trend lines, closed-loop referral completion metrics, and documented reductions in adverse events or involuntary interventions for the repeat-utilizer group.
Setting thresholds and escalation rules (so the dashboard drives action)
Metrics only matter if they trigger decisions. Effective governance sets explicit thresholds (green/amber/red) and defines who must act. Examples: (1) if “ED arrival within 6 hours after diversion” exceeds a set rate, trigger a chart review sprint; (2) if a provider’s denial rate spikes, require a joint case conference and capacity reconciliation; (3) if stratified equity metrics show widening gaps, trigger targeted staff coaching and workflow redesign.
Build the governance cadence around operational reality: a daily huddle for integrity issues, a weekly performance review for trends and provider denials, and a monthly executive governance meeting for capacity planning and contracting levers. The point is not to generate reports—it is to prevent the predictable failure modes before they become incidents.