Key Takeaways

• Behavioral health nursing turnover reaches 22.8%, the highest of any specialty, yet most facilities measure departures rather than the safety perceptions that predict them months earlier
• Research shows high violence exposure increases turnover intention five-fold, but staff perception of organizational support mediates the relationship, not incident frequency alone
• Healthcare settings measuring safety perception recorded intent-to-leave dropping from 22% to 7%, a 68% reduction in turnover risk captured before any resignations occurred

---

Your vacancy dashboard shows behavioral health nursing turnover at 22.8%. The highest of any specialty in your system [1]. Exit interviews keep surfacing “safety concerns” as a contributing factor. But when you cross-reference incident reports, the numbers look stable.

The disconnect isn’t a data error. It’s a measurement gap.

You’re tracking the aftermath of decisions your staff made months ago. The perception that drove those decisions never appeared on any report you reviewed. Understanding staff safety in psychiatric hospitals means measuring what staff actually feel, not just what gets reported.

The Safety Perception Gap: What Staff Feel vs. What Facilities Measure

The workforce metrics on your dashboard tell a coherent story: turnover rate, time-to-fill, cost-per-hire, engagement composite score. Each metric is reliable. Each is also retrospective. By the time a departure appears in your data, the perception shift that caused it formed weeks or months earlier. For behavioral health Chief Human Resources Officers (CHROs), this lag creates a structural blind spot.

The gap starts with what staff experience versus what gets captured. Research shows that harassment at work is tied to five-fold increased odds of anxiety and nearly six-fold increased odds of burnout among health workers [2]. Those are the conditions that come before resignation. But the incidents driving those conditions rarely surface in your reporting systems.

Eighty-one percent of workplace violence incidents go unreported by healthcare workers who experienced them [3]. Your incident data isn’t stable. It’s silent.

The reasons are consistent: normalization, perceived futility, fear of retaliation [4]. Nurses on acute psychiatric units often treat physical aggression as part of the job description. The incident never makes it into a system anyone reviews. The unreported event still shapes how that nurse answers the question: “Do I stay or do I go?”

The perception of futility deserves particular attention. Among nurses surveyed nationally, 44.8% reported that workplace violence incidents are simply ignored by their employers after being reported [5]. Only 31.7% reported that their employer provides a clear way to report incidents at all [5]. Staff have learned, through repeated experience, that reporting changes nothing. So they stop reporting. Your incident data looks stable while your turnover accelerates.

This creates a compounding problem. Your incident reports understate reality. Your engagement survey buries safety questions in a 50-item instrument analyzed annually. Your exit interviews capture themes but not timing. None of these systems measure the perception that actually predicts whether staff stay or leave.

Metric	What It Captures	When You See It	What It Misses
Turnover rate	Departures after they happen	30–90 days post-decision	The perception shift that preceded the decision by months
Exit interviews	Stated reasons for leaving	At resignation	Staff who stay but disengage; incidents never reported
Incident reports	Recorded events	After filing	81% of incidents that go unreported
Engagement survey	Annual composite score	Once per year	Quarterly or monthly perception changes on specific units
Safety perception baseline	How staff feel about organizational safety response	In real time	Nothing, if you measure it

The bottom row of that table is where the opportunity lives.

How Safety Perception Predicts Intent-to-Stay

The research connecting safety perception to retention isn’t speculative. It’s recorded across multiple large-scale studies. The mechanism is now well understood.

A multi-center study on workplace violence and turnover found that high-level violence exposure increased turnover intention risk by more than five-fold [6]. But the critical insight is what mediates the relationship: perception of organizational support. Staff who experienced support and validation after incidents showed far lower turnover intention than those who perceived indifference [7].

“Intent-to-leave dropped from 22% to 7%, a 68% reduction in turnover risk captured before any resignations occurred.”

The incident didn’t determine whether they stayed. Their perception of how the organization responded did. That’s not a staffing trend. That’s an organizational failure showing up in your vacancy data.

This distinction matters enormously for workforce strategy. It reframes the problem from “reduce violence” (difficult, partially outside your control) to “prove organizational commitment to safety” (achievable, directly within your control). The realization often hits during facility walkthroughs: a visible, fast response to an incident on the unit can do more for retention than a violence prevention program staff never see in action.

The pathway operates through burnout. Workers in low psychosocial safety climate environments had 0.35 times the odds of avoiding burnout compared to those with good psychosocial safety climate [2]. The workplace itself is the risk factor. Not the patients, not the acuity level, not the shift length.

Burnout functions as the proximal antecedent to turnover intent. Organizational factors, including perceived safety, are among the most modifiable predictors [8]. Low psychosocial safety climate creates reduced job control, limited autonomy, insufficient support. Conditions that directly predict increased turnover intention [9].

Sixty percent of nurses have changed or left their job, or considered leaving, due to workplace violence [10]. In behavioral health, where violence exposure rates exceed general healthcare settings, these percentages translate to workforce instability that compounds with every departure.

No one should face violence while trying to help others heal.

Hospitals using ongoing perception measurement have recorded retention shifts within 90 days [11]. That confirms safety climate predicts turnover intention through proven pathways: burnout, reduced organizational support, psychological demand accumulation. Worth noting: 90 days is fast, but it still requires consistent measurement. The signal only becomes useful with repeated data points at the unit level.

The takeaway for your workforce strategy is direct. You don’t need to eliminate workplace violence to stabilize your workforce. You need to change how staff perceive your commitment to their safety. That perception is measurable. And it’s movable.

Measuring Safety Perception: From Baseline to Benchmark

Your employee engagement survey likely includes safety-related questions already. The problem isn’t absence of data. It’s how that data is structured, analyzed, and acted upon.

Safety questions buried in a 50-question instrument, analyzed annually, and reported as a single composite score can’t function as a leading indicator. Leading indicators need frequency, specificity, and actionability.

Validated measurement instruments exist. The Psychosocial Safety Climate Scale (PSC-12) has been validated across four domains with Cronbach’s alpha exceeding 0.84. It shows significant associations with burnout, depression, anxiety, stress, and turnover intention [9]. The Agency for Healthcare Research and Quality’s Surveys on Patient Safety Culture (AHRQ SOPS) contains dimensions including teamwork, communication, and management support that correlate with worker felt safety and retention outcomes [12]. These aren’t experimental tools. They’re established frameworks your team can adopt.

Press Ganey data shows that employees in the top engagement quartile score between the 97th and 99th percentile on safety culture items. Bottom-quartile employees score at the 1st to 2nd percentile [13]. That gap is enormous. Safety perception isn’t a separate construct from engagement. It’s a core component. When safety perception drops, engagement follows. When engagement drops, turnover follows.

Measurement Approach	Frequency	Specificity	Actionability	Leading Indicator Value
Annual engagement survey (composite)	Once per year	Low (blended score)	Low (no unit-level detail)	Minimal
Quarterly safety perception pulse	Every 90 days	Moderate (safety-specific)	Moderate (trend visible)	Moderate
Ongoing perception tracking with unit-level data	Ongoing	High (unit, shift, role)	High (targeted intervention)	High

The CHRO who treats safety perception as a leading indicator pulls safety-specific questions from the broader engagement instrument. They establish a unit-level baseline. They track changes at least quarterly. When perception drops on a specific unit, they investigate before the turnover spike shows up in next quarter’s dashboard.

Behavioral health facilities don’t yet have published industry benchmarks for safety perception scores. AHRQ SOPS provides the established benchmarking framework. Your team can use PSC-12 and SOPS dimensions to establish internal baselines and track directional improvement. The benchmark that matters most is your own trajectory: are perception scores improving, stable, or declining? That directional signal predicts your next quarter’s retention.

What Moves the Needle: Interventions That Shift Perception

Understanding that perception predicts retention creates an obvious question: what changes perception? The research identifies three categories of intervention. The evidence shows they compound when combined.

1. Proven Organizational Responsiveness

Research on transparent, non-punitive reporting systems shows that facilities with visible follow-up can shift staff perception from “reporting is futile” to “reporting leads to change” [14]. This isn’t about policy language. It’s about observable behavior.

When staff activate a call for help and help arrives visibly and quickly, that single experience reshapes their perception of organizational commitment. When they report an incident and see recorded follow-up, the perception of futility breaks. Chief Nursing Officers (CNOs) describe a specific version of this: the difference between a response that arrives in 30 seconds and one that arrives in three minutes registers as organizational indifference. Even when both technically count as “responsive.”

2. Multi-Component Programs

Studies comparing intervention approaches found that multi-component violence prevention programs showed larger and more sustained perception improvements than single-session training [15]. Training alone doesn’t change perception because training addresses individual capability, not organizational response.

This isn’t a checkbox. It’s a commitment. Perception shifts when staff see the system respond, not when they learn a new technique in a classroom.

3. Supervisor Communication

Managers who received coaching on communicating safety as a priority produced significantly improved subordinate perceptions of safety climate [16]. This is a high-leverage mechanism because it’s directly within HR’s control. Ensure your CNO has established expectations for how charge nurses communicate safety commitment on every shift.

The recorded outcomes confirm these mechanisms operate at scale. ROAR customers report up to a 38-point lift in “I feel safe at work” survey responses [11]. Staff satisfaction grew from 57% to 73% within three months. Nearly 80% of team members reported increased confidence in handling safety concerns. Employees feeling “very prepared” to respond to an incident increased from 38% pre-pilot to 76% post-pilot.

A full violence prevention intervention in one study reduced violence incidence from 63.85% to 46.15%, a 30% reduction, while workers’ coping resource scores increased significantly [17]. The objective reduction in incidents matters. But here’s what the research keeps confirming: staff who feel prepared and supported show different retention behavior than staff who feel exposed and abandoned. Regardless of actual incident rates.

Perception over frequency. Every time.

The mechanism is consistent across research and practice. Perception moves when facilities prove, visibly and repeatedly, that staff safety is an operational priority backed by responsive systems. Not a poster on the wall. Not an annual training module. A system that responds when activated.

[soft_cta text="One facility recorded intent-to-leave dropping from 22% to 7% after implementing perception-informed safety strategies. See how leading organizations are turning safety perception into a measurable workforce metric." button="Request a Demo" url="https://roar.the-devoted.dev/request-a-demo/%22]

From Perception Shift to Retention Impact: The Documented Pathway

Before putting safety perception measurement and intervention into practice, 22% of staff at one behavioral health facility said they would consider leaving due to safety concerns. After: 7% [11]. That’s a 68% reduction in turnover risk, captured before any resignations occurred. This isn’t a satisfaction survey improvement. It’s a leading indicator of workforce stability, measured in real time.

The financial translation is direct. Each 1% reduction in nursing turnover translates to about $289,000 in annual savings. The average replacement cost per departing nurse is $61,110 [1]. For a behavioral health system running at 22.8% nursing turnover, even modest perception-driven retention improvements generate substantial returns.

Scenario	Turnover Reduction	Annual Savings (per $289K/point)	Nurses Retained (per $61,110 each)
1-point reduction (22.8% → 21.8%)	1 percentage point	$289,000	~5 nurses
3-point reduction (22.8% → 19.8%)	3 percentage points	$867,000	~14 nurses
5-point reduction (22.8% → 17.8%)	5 percentage points	$1,445,000	~24 nurses

One important qualification: the $289,000 figure is a national average across nursing specialties. Behavioral health replacement costs can run higher due to the smaller candidate pool and longer onboarding periods for psychiatric units. The directional math holds, but your actual per-nurse cost likely exceeds this benchmark.

Every dollar in prevention returns multiples in retention.

U.S. hospitals incurred $18.27 billion in total annual financial costs tied to violence, with 80% representing post-event costs [18]. Technology-enabled safety programs have recorded workers’ compensation claims reductions of 20% to 50% [11]. That connects perception improvement to the financial outcomes you can quantify for conversations with your Chief Financial Officer (CFO). When staff feel safer, they report incidents through proper channels. When incidents are recorded and responded to, workers’ comp claims reflect actual events rather than accumulated, unreported exposure.

The 22% to 7% intent-to-leave reduction represents a before-and-after measurement of the perception-to-retention correlation in a behavioral health setting. The difference between an unmeasured claim and a quantified leading indicator is the difference between a wellness initiative and a workforce strategy.

Building a Safety Perception Strategy for Your Facility

Deloitte identifies workforce safety and psychological well-being as emerging strategic metrics for workforce planning dashboards [19]. Systematic reviews confirm that safety and psychological well-being are explicit components of effective retention programs [20]. Leading healthcare systems are integrating safety perception into the same workforce planning frameworks they use for engagement, compensation, and career development.

Building this capability needs four elements.

Establish a baseline. Pull safety-specific questions from your existing engagement survey or put a validated instrument like the PSC-12 into practice. Measure at the unit level, not just the organizational level. Your overall score may look acceptable while specific units are in crisis. The pattern across behavioral health facilities is remarkably consistent: the organization-wide average masks one or two units where perception has already collapsed and turnover is about to follow. The baseline tells you where those units are.

Track changes with frequency. Annual measurement can’t function as a leading indicator. Quarterly pulse surveys on safety perception, at minimum, provide the trend data that predicts retention shifts. When perception drops on a specific unit between quarters, you have a 90-day window to intervene before turnover shows up.

Correlate with retention intent. Add intent-to-stay questions to your safety perception surveys. The correlation between perception scores and intent-to-leave is the proof point that turns safety perception from a “soft” metric into a workforce planning tool. In behavioral health deployments, teams recorded intent-to-leave reduction from 22% to 7% after safety perception measurement and intervention [11].

Intervene where perception drops. Perception data without intervention is just measurement. When a unit shows declining safety perception scores, work with your Chief Security Officer (CSO) to evaluate response protocols. Work with your CNO to assess whether staff on that unit are experiencing unreported incidents. The intervention is organizational responsiveness: proving, visibly, that the facility acts on what staff report.

Here’s a practical starting point for the next 90 days:

• Pull safety-specific items from your current engagement survey and score them separately by unit. Can you identify which units fall below the organizational average?
• Add two to three intent-to-stay questions to your next pulse survey, tied directly to safety perception. Do you have a way to correlate perception scores with stated retention intent?
• Review your incident reporting workflow. Does your system close the loop visibly enough that the reporting nurse sees what happened after they filed?
• Ask your CNO whether charge nurses on behavioral health units have explicit language for communicating safety commitment at shift handoff. If the answer is “not really,” that’s your highest-leverage coaching opportunity.
• Identify your single highest-turnover behavioral health unit and run a focused safety perception baseline there first. One unit, measured well, proves the model faster than a system-wide rollout.

The ROI formula for your CFO conversation is straightforward. Each percentage point of turnover reduction saves about $289,000 annually [1]. Facilities putting perception-informed safety strategies into practice have recorded 3x ROI within five years and six-figure premium savings on workers’ compensation. The investment case doesn’t need speculation. It needs the same before-and-after measurement discipline you apply to any workforce initiative.

Your turnover dashboard will still show 22.8% tomorrow morning. The exit interviews will still cite safety concerns. But the CHRO who treats safety perception as a leading indicator, measuring baselines, tracking shifts, intervening before intent-to-leave becomes resignation, gains something lagging metrics can’t provide: the ability to see who is considering leaving before the resignation letter arrives. The research connecting staff safety in psychiatric hospitals to retention is clear. The measurement tools are validated. The recorded outcomes exist. The remaining question is whether your team will keep measuring departures after they happen, or start measuring the perceptions that predict them.

[closing_cta eyebrow="MEASURABLE OUTCOMES" headline="Measure the Perception That Predicts Retention" desc="Organizations discovering the gap between what staff feel and what dashboards show are establishing perception baselines, tracking changes over time, and documenting the retention impact that traditional workforce metrics miss. A conversation about measurement methodology is the logical next step." button="Request a Demo" button_url="/request-a-demo/" link="See how one provider achieved a 40% reduction in staff assaults and response times under 2 minutes for 87% of alerts" link_url="https://roar.the-devoted.dev/national-behavioral-healthcare-provider-case-study/%22]

References

1. NSI Nursing Solutions. 2025 NSI National Health Care Retention & RN Staffing Report. https://www.nsinursingsolutions.com/documents/library/nsi_national_health_care_retention_report.pdf
2. CDC MMWR. Vital Signs: Health Worker-Perceived Working Conditions and Symptoms of Poor Mental Health. https://www.cdc.gov/mmwr/volumes/72/wr/mm7244e1.htm
3. AHRQ PSNet. Addressing Workplace Violence and Creating a Safer Workplace. https://psnet.ahrq.gov/perspective/addressing-workplace-violence-and-creating-safer-workplace
4. American Nurses Association. Unreported Workplace Violence: Why Is This So Common? https://www.nursingworld.org/content-hub/resources/workplace/unreported-workplace-violence—why-is-this-so-common/
5. National Nurses United. Workplace Violence Report. https://www.nationalnursesunited.org/sites/default/files/nnu/documents/0224_Workplace_Violence_Report.pdf
6. PMC. Workplace Violence and Turnover Intention. https://pmc.ncbi.nlm.nih.gov/articles/PMC12811911/
7. PMC. Organizational Support and Turnover Intention. https://pmc.ncbi.nlm.nih.gov/articles/PMC7750754/
8. PMC. Burnout and Turnover Intention in Healthcare. https://pmc.ncbi.nlm.nih.gov/articles/PMC11496712/
9. PMC. Psychosocial Safety Climate Scale Validation. https://pmc.ncbi.nlm.nih.gov/articles/PMC11737061/
10. ROAR for Good. An Analysis of Workplace Violence Statistics in Healthcare. https://roar.the-devoted.dev/blog/an-analysis-of-workplace-violence-statistics-in-healthcare/
11. ROAR for Good. Internal data, 2024. Internal data
12. AHRQ PSNet. Culture of Safety. https://psnet.ahrq.gov/primer/culture-safety
13. Press Ganey. Supporting Patient Safety Culture in Healthcare Requires Higher Employee Engagement. https://www.pressganey.com/resources/blog/supporting-patient-safety-culture-in-healthcare-requires-higher-employee-engagement/
14. PMC. Transparent Reporting Systems and Perception Change. https://pmc.ncbi.nlm.nih.gov/articles/PMC11980070/
15. PMC. Multi-Component Violence Prevention Programs. https://pmc.ncbi.nlm.nih.gov/articles/PMC12542813/
16. PMC. Manager Safety Communication and Subordinate Perceptions. https://pmc.ncbi.nlm.nih.gov/articles/PMC9742354/
17. PMC. Comprehensive Violence Prevention Intervention Outcomes. https://pmc.ncbi.nlm.nih.gov/articles/PMC10605776/
18. American Hospital Association. Costs of Violence. https://www.aha.org/costsofviolence
19. Deloitte. Key HR Metrics for CHROs. https://www.deloitte.com/us/en/services/consulting/articles/key-hr-metrics-for-chros.html
20. PMC. Safety and Psychological Well-Being in Retention Programs. https://pmc.ncbi.nlm.nih.gov/articles/PMC10341299/