How to Audit Psychosocial Risks in a Plant

ISO 45003:2021 made psychosocial risk part of occupational health and safety management, which means a plant cannot treat overload, harassment, role conflict, and fatigue as soft HR concerns. This guide shows EHS and HR teams how to audit psychosocial risks in an industrial site without reducing the work to a survey, a wellness talk, or a compliance folder.

Why a psychosocial risk audit must inspect work design

A psychosocial risk audit is a structured review of how work is organized, led, measured, and experienced, with the purpose of finding exposures that can harm psychological health and safety. ISO 45003, HSE Management Standards, and ESENER name the topic directly, while the HSE Management Standards group work-related stress around six work-design areas: demands, control, support, relationships, role, and change.

The mistake is to audit feelings without auditing the system that produces them. A plant may have a respectful poster campaign and still expose people to twelve-hour shifts, rotating priorities, unclear handovers, weak supervision, and production targets that punish any attempt to slow down.

As Andreza Araujo argues in Safety Culture Diagnosis, perception data matters because workers know which rules survive pressure. The audit should capture that perception, but it must connect answers to staffing, workload, maintenance planning, contractor interfaces, shift routines, and leadership behavior.

Step 1: Define the audit boundary and sponsor

The first step is to define which plant, department, shift, contractor group, and leadership layer the audit will cover. A psychosocial audit without a boundary becomes a general climate conversation, and general conversations rarely produce controls.

The sponsor should be operational, not only HR. HR can coordinate confidentiality and employee relations, but the plant manager or operations director must own decisions that affect headcount, deadlines, shift patterns, supervisor span of control, and production pressure.

Write a one-page audit charter before collecting evidence. It should state the scope, business reason, confidentiality rules, data sources, decision authority, timeline, and the minimum control plan expected after the findings are validated.

Step 2: Build a risk-factor map before asking questions

A risk-factor map turns broad psychosocial language into observable work conditions. Use ISO 45003:2021, the HSE Management Standards, and local incident evidence to list exposures such as excessive demands, low control, poor support, role ambiguity, interpersonal conflict, harassment, fatigue, and unmanaged change.

Across 25+ years leading EHS at multinationals, Andreza Araujo has observed that organizations often ask employees whether they feel stressed before asking whether the job is realistically designed. That order is backwards because the audit should begin with the source of exposure, not with the worker's capacity to endure it.

For each risk factor, define the evidence you will inspect. Demands may require overtime records, backlog, absenteeism, maintenance delays, and peak-shift staffing. Relationships may require grievance data, interview themes, supervisor behavior, and evidence from workplace bullying investigations.

Step 3: Select indicators that reveal pressure early

Psychosocial indicators should reveal pressure before absence, turnover, or a serious event becomes the first visible signal. Useful indicators include overtime concentration, shift swaps, absenteeism spikes, grievance frequency, medical restrictions, rework, quality escapes, near misses linked to fatigue, and recurring deadline changes.

Do not let the dashboard become a hiding place. A monthly green score can conceal one supervisor with 40 percent turnover, one maintenance crew with chronic callouts, or one production line where temporary workers absorb the worst shifts.

Connect each indicator to a decision. If overtime exceeds the agreed trigger for two cycles, who can approve staffing, maintenance reprioritization, or production rescheduling? The same logic that applies to workload risk indicators should apply here, because psychosocial risk becomes manageable only when evidence changes work.

Step 4: Design interviews that protect candor

Interviews should test how work actually happens under pressure, not whether employees can recite the policy. The best questions ask what changed during peak demand, who can say no, what happens after conflict, and which risks people stopped reporting because nothing changed.

In more than 250 cultural-transformation projects supported by Andreza Araujo's team, candor usually depends less on the questionnaire and more on whether workers believe retaliation is possible. Confidentiality, neutral facilitation, separated supervisor and worker groups, and careful reporting language protect the evidence.

Use interviews to explain the numbers, not to replace them. If overtime is high, ask what work is being absorbed, which maintenance tasks are deferred, and where employees feel they must choose between quality, production, and recovery.

Step 5: Inspect the work system at the workplace

The field review should look at the physical and organizational conditions that shape psychosocial exposure. Walk the line, observe shift handover, review break access, inspect control-room interruptions, listen to radio traffic, and compare the written procedure with the actual pace of work.

The trap is believing that psychosocial risk cannot be seen. It can be inferred from queues, alarms, skipped breaks, conflict at handover, supervisors carrying too many direct reports, operators waiting for decisions, and contractors receiving incomplete instructions.

Pair the field review with adjacent safety evidence. Repeated line-of-fire events, LOTO mistakes, dropped objects, or poor permit quality may point to workload, distraction, unclear roles, or weak supervision rather than to isolated technical failure.

Step 6: Separate hazards from consequences

A good audit separates the psychosocial hazard from its consequence. Burnout, anxiety complaints, turnover, and absence may be consequences, while excessive demand, poor control, bullying, role ambiguity, and unmanaged change are closer to the source.

This distinction protects the audit from becoming a medicalized report that sends everyone to counseling while the plant keeps the same exposure. Care resources matter, but they cannot substitute for work redesign.

During the PepsiCo South America tenure, where the accident ratio fell 50 percent in six months, Andreza Araujo learned that prevention improves when leaders move upstream from injury data to weak signals. The same principle applies here: treat absence and complaints as late signals, then trace them back to work design.

Step 7: Rate risk with severity, exposure, and control strength

Psychosocial risk rating should combine severity, exposure frequency, number of people exposed, vulnerability of the group, and strength of existing controls. A small team under chronic harassment may deserve a higher priority than a large department with moderate survey discomfort.

A simple color matrix is not enough if it treats every complaint as equal or averages away a high-risk subgroup. The audit team should document the rationale, including which evidence supports the rating and which uncertainty remains.

250+ companies across 30+ countries form part of Andreza Araujo's practical experience with culture transformation, and one repeated lesson is that weak controls often hide behind acceptable averages. Segment by shift, supervisor, contractor status, gender, tenure, and work area before approving the risk level.

Step 8: Convert findings into controls with named owners

Controls must address the source of psychosocial exposure. Examples include staffing correction, deadline governance, supervisor training, harassment investigation protocol, escalation rules, protected breaks, change-impact review, workload triggers, and clearer role definitions.

As Andreza Araujo argues in Safety Culture: From Theory to Practice, culture is shaped by repeated leadership choices. A finding with no owner, budget, or decision date tells workers that the organization wanted a diagnosis without the burden of change.

Use the hierarchy of controls as a discipline. Eliminate or reduce the demand where possible, redesign the work, strengthen supervision and support, and only then add individual support resources such as coaching or EAP referral.

Step 9: Verify whether the control changed exposure

Verification should test whether exposure changed, not whether the action was closed. A completed training session does not prove workload fell, conflict reduced, or employees gained real control over priorities.

Set a 30, 60, and 90-day verification rhythm. Review indicators, repeat targeted interviews, inspect the workplace again, and ask supervisors what decisions changed because of the audit.

Each month without verification allows the plant to mistake activity for risk reduction, while psychosocial exposures keep shaping fatigue, silence, conflict, and operational error.

Comparison: Survey-only review versus risk-control audit

The survey-only model can be useful for listening, although it becomes weak when it stops there. A risk-control audit listens, verifies, segments, and changes the operating conditions that produced the exposure.

What EHS and HR should do next

A plant-level psychosocial risk audit works when it treats overload, poor control, harassment, role ambiguity, fatigue, and change as occupational exposures that deserve evidence and controls. The goal is not to prove that people are stressed, but to prove which parts of the work system are creating preventable harm.

Start with one high-pressure area, build the risk-factor map, collect indicators and interviews, inspect the work, then approve a control plan that operations can execute. If your organization needs a deeper method, Andreza Araujo's work through ACS Global Ventures and Andreza Araujo's Safety School can support the move from awareness to disciplined psychosocial risk management.