How to Build a Workplace Mental Health Escalation Protocol in 30 Days

A workplace mental health escalation protocol should not turn supervisors into clinicians. Its purpose is narrower and more defensible: help leaders recognize when a work situation has moved beyond routine support, route the concern to the right owner, protect privacy, and remove job conditions that may be making harm worse.

The common mistake is to treat mental health at work as either an awareness campaign or an HR benefit. ISO 45003:2021 frames psychological health and safety through psychosocial risk management, and the WHO guidelines on mental health at work, published in 2022, include organizational interventions, manager training, worker training, return to work, and employment support. An escalation protocol connects those ideas to daily decisions.

What you need before starting

Before the 30-day build begins, gather the current EAP process, fit-for-work rules, accommodation procedure, incident response plan, privacy policy, HR case-management flow, occupational health contacts, emergency contacts, and any psychosocial risk assessment already completed. If the company has operations in several countries, legal and medical escalation rules must be checked locally before rollout.

Across 25+ years leading EHS in multinational environments, Andreza Araujo has identified a repeated failure pattern: organizations ask frontline leaders to notice human distress, yet they give those leaders no practical route for what happens next. The result is delay, improvisation, over-sharing of sensitive information, or the opposite problem, silence dressed up as respect for privacy.

This guide is written for EHS, HR, operations managers, and site leaders who need a management system, not a counseling script. The protocol must respect clinical boundaries while still treating psychosocial exposure, threatening behavior, fatigue, conflict, overload, isolation, and post-incident distress as work-relevant signals.

Step 1: define what escalation means in your workplace

Write a one-page definition that separates normal support from formal escalation. Normal support may include listening, schedule clarification, workload review, or a manager checking whether the worker knows available resources. Formal escalation begins when there is possible self-harm, threat of harm to others, severe impairment during safety-critical work, repeated distress linked to job demands, workplace violence, bullying, harassment, substance concerns, or a return-to-work situation that needs structured control.

The definition should make one point explicit. Escalation is not diagnosis. A supervisor should never label a worker with a condition, speculate about medication, or ask for medical details. The supervisor should describe observable work impact, immediate safety concerns, and job factors that need review.

Use the EAP, fit-for-work and accommodation comparison to keep routes distinct. EAP helps with voluntary support, fit-for-work checks task safety, and accommodation adjusts work where appropriate. The escalation protocol decides which route is triggered and who owns the next step.

Step 2: map the first three escalation levels

Create three levels that a non-clinical leader can understand. Level 1 is early work stress, such as overload complaints, conflict, fatigue, irritability, reduced concentration, or withdrawal that does not create immediate danger. Level 2 is significant work impact, such as repeated errors, inability to perform safety-critical duties, intense distress after an incident, suspected bullying, or a pattern of absence linked to job strain. Level 3 is urgent risk, including threat of self-harm, threat to others, violent behavior, medical emergency, or a worker who appears unable to stay safe in the current task.

Each level needs a response owner, response time, privacy rule, and documentation rule. Level 1 may sit with the line manager and HR partner. Level 2 may involve HR, EHS, occupational health, and the manager. Level 3 should trigger emergency or crisis procedures immediately, because the workplace protocol cannot replace emergency care.

OSHA's workplace stress resources and CDC/NIOSH stress-at-work material both distinguish poor mental health and workplace stress from clinical care. That distinction matters because the company can manage work conditions and emergency routing, but it should not pretend to provide treatment through line management.

Step 3: assign owners before the first case appears

Build a responsibility table with HR, EHS, occupational health, security, legal, the line manager, senior operations, and external emergency services. For each role, define what they own, what they must not do, and when they must be informed.

EHS should own the connection to work risk: fatigue, workload, shift patterns, violence exposure, post-incident stressors, job design, lone work, safety-critical tasks, and psychosocial hazards. HR should own employment process, manager coaching, case privacy, policy consistency, and accommodation coordination. Occupational health should own medical fitness advice where the company has that resource.

As Andreza Araujo argues in Safety Culture: From Theory to Practice, culture becomes visible through routine decisions, not declared values. A protocol with vague ownership teaches leaders to delay. A protocol with clear ownership makes the next responsible action visible before the situation becomes personal improvisation.

Step 4: create a supervisor decision script

The supervisor script should be short enough to use under pressure. It should include what the leader may say, what the leader should observe, what the leader should not ask, and when the leader must stop the conversation and escalate.

A defensible script starts with observable concern: "I noticed you seem distressed and you are assigned to a safety-critical task. I want to pause the task and make sure we route this correctly." It avoids diagnosis, private medical questions, blame, and promises of confidentiality the company cannot keep when safety is at risk.

Connect the script to fit-for-work review before high-risk tasks. If the person is operating mobile equipment, entering a confined space, performing electrical work, driving, lifting, isolating energy, or supervising hazardous work, the immediate question is whether the task should pause while the correct owner is contacted.

Step 5: build the privacy boundary into the form

The escalation form should capture only what the workplace needs. Include date, location, role, observable concern, safety-critical task involved, work factors reported by the employee, immediate controls applied, owner assigned, referral route, and follow-up date. Do not include diagnosis, therapy details, medication, family history, or personal speculation.

Privacy discipline protects the worker and the company. When a form invites excessive detail, managers often write too much because they are trying to prove care. That care can become harm when sensitive information spreads through an organization that has no need to know it.

Use mental health accommodations as the boundary check. The business needs to know what adjustment is required and how work will be controlled. It usually does not need the private clinical story behind the adjustment.

Step 6: link escalation to psychosocial risk controls

If every escalation ends as an individual case, the organization will miss the prevention opportunity. Add a field that asks whether the concern may be linked to workload, role ambiguity, traumatic exposure, conflict, bullying, poor change management, impossible deadlines, fatigue, isolation, harassment, or violence risk.

This step converts case handling into risk management. One overload concern may be individual. Ten overload concerns from the same department may indicate work design failure. One post-incident distress case may need support. Repeated distress after similar events may indicate weak incident aftercare and weak supervisor preparation.

The WHO/ILO policy brief on mental health at work, published in 2022, calls attention to work organization and workplace action, not only individual resilience. That is the practical test for this protocol: it should route people to support while also forcing the business to ask which work condition needs control.

Step 7: define the 24-hour review rule

Every Level 2 or Level 3 escalation should receive a documented review within 24 hours, even when the immediate crisis has passed. The review should ask whether the worker is safe, whether the task remains paused or modified, whether HR or occupational health has taken ownership, whether any workplace hazard needs control, and whether the manager needs coaching before the next conversation.

This review prevents two common failures. The first is overreaction, where the worker is removed from work without a clear route back. The second is underreaction, where the manager expresses concern but the same workload, conflict, shift pattern, or exposure remains untouched.

Connect the 24-hour review to return to work after mental-health absence when leave is involved. A return plan without an escalation history may miss the work factors that contributed to the absence or made the return fragile.

Step 8: train managers with realistic cases

Manager training should not be a slide deck about empathy alone. Use short scenarios that force decisions: an operator crying before a confined-space entry, a supervisor after a fatality exposure, a worker making vague self-harm comments, a team member reporting bullying, a driver showing exhaustion, or an engineer returning after mental-health absence.

For each scenario, ask managers to identify the level, immediate task control, words they would use, information they would avoid collecting, owner they would contact, and the work factor they would review. The training should include active listening, but it should end in accountable routing.

In more than 250 cultural transformation projects supported by Andreza Araujo, a recurring weakness is the leader who wants to help but does not know where the boundary sits. Scenario training gives that leader a practiced route, which reduces both avoidance and overreach.

Step 9: test the protocol with one site before rollout

Run a 30-day pilot at one site, business unit, or function. Track how many concerns were routed, how quickly owners responded, whether privacy rules were followed, whether safety-critical tasks were paused when needed, and whether recurring work factors were identified.

The pilot should not celebrate volume by itself. A good protocol may increase early reporting because people finally know where to go, yet the real measure is whether leaders act earlier and whether the organization removes the work conditions that keep producing distress.

Compare the pilot with workplace mental health campaign traps. Awareness without escalation creates emotional visibility without operational control. The pilot must prove that concern moves from conversation to responsible action.

Mental health escalation protocol checklist

• Escalation levels are clear enough for non-clinical leaders to apply.
• Urgent risk triggers emergency or crisis procedures immediately.
• Managers describe observable work impact and avoid diagnosis.
• HR, EHS, occupational health, security and operations ownership is defined.
• Forms exclude unnecessary clinical or personal details.
• Each case asks whether a psychosocial work factor needs control.
• Level 2 and Level 3 cases receive a documented 24-hour review.
• Managers practice realistic scenarios before the protocol goes live.
• The pilot measures response quality, not only the number of cases reported.

Conclusion

A workplace mental health escalation protocol is effective when it protects three boundaries at the same time: the worker's dignity, the manager's non-clinical role, and the company's duty to control work-related risk. If any one of those boundaries is missing, the process becomes either too soft to act or too intrusive to trust.

The practical goal is not to medicalize management. The goal is to make the next right action visible when a concern appears, especially when safety-critical work, violence risk, post-incident distress, bullying, overload, fatigue, or return-to-work fragility is present.