This paradox highlights a frequent disconnect: official figures appear excellent, yet serious accidents continue to occur in both industrial and service organisations.
Why Are Safety Indicators Good but Accidents Still Occur?
In many organisations, everything seems to be under control. The accident frequency rate remains below 27 per 1,000 employees, as reported by the CNAM in 2024. Incident severity is under control, ISO 45001 audits are validated, quality of working life programmes are in place, and procedures are regularly updated.
Yet reality is harsher. In France, just over 500,000 workplace accidents occur each year, around 750 of them fatal. Recurrent incidents, non compliance with rules in the field, silence following events, and a significant increase in psychosocial issues such as stress and fatigue are also observed.
So what do these indicators really tell us? And more importantly, what do they leave unsaid?
The Limits of Traditional Safety Indicators
These indicators, often referred to as lagging or retrospective, are based on what has already happened: reported accidents, measurable outcomes, and compliance with formal rules. They help us look in the rear view mirror, but they do not allow us to anticipate future problems. Studies such as those conducted by FONCSI show that their correlation with serious accidents is weak. In short, a good past result does not predict a safe future.
The Seven Blind Spots of Safety Indicators in Organisations
Recent research by INRS, FONCSI and Eurogip converges on one key point: a large share of risks escape traditional indicators, particularly those related to human and organisational dimensions.
1. Underreporting of Incidents and Near Misses
Underreporting is the first major pitfall. When rates become performance targets, teams hesitate to report minor incidents or near misses. Eurogip’s work shows that in some countries and sectors, only a fraction of accidents are actually recorded, meaning many events remain off the radar.
2. Mental Factors Invisible in Statistics
Mental factors such as fatigue and stress affect a large proportion of employees. Recent surveys indicate that around 64 percent of employees report feeling stressed at work at least once a week. These factors reduce vigilance without leaving any trace in accident statistics, while significantly increasing risk exposure.
3. Lack of Psychological Safety
Safety culture is also an issue. When employees fear blame or do not feel able to say “stop”, they remain silent. This lack of psychological safety, a key concept popularised by Amy Edmondson, hinders reporting and prevents essential information from being shared.
4. The Gap Between Prescribed Work and Real Work
Actual work often differs from what is prescribed on paper. Teams improvise in response to unforeseen events or pressure, but these adaptations are invisible in indicators. Ergonomic studies and accident analyses by INRS highlight the role these gaps play in the genesis of incidents.
5. Weak Signals and the Gradual Decline in Vigilance
Weak signals are more subtle. Everything appears stable until a sudden breakdown occurs. After good results, vigilance naturally declines, as noted in FHOS analyses, delaying the detection of drifts before an accident happens.
6. Managerial Biases
Managerial biases also come into play. A form of complacency can set in, or production may implicitly take precedence over safety, influencing daily decisions without being visible in indicators.
7. Organisational Gaps and Governance Failures
Finally, organisational gaps represent a major blind spot. Conflicting objectives, implicit trade offs, or insufficiently monitored subcontractors create risk zones that traditional indicators fail to capture.
When Good Safety Indicators Increase Real Risk
Ironically, positive indicators can create a false sense of security. Several studies show that good indicators can significantly reduce field audits and vigilance. Conversely, when management encourages listening and feedback, the reporting of safety related information increases markedly. Behavioural analyses confirm that the vast majority of accidents, often more than 80 to 90 percent, could be avoided through better practices and anticipation.
How to Complement Safety Indicators to Better Prevent Accidents
High performing organisations complement these indicators with leading or forward looking measures: field observations, anonymous safety culture surveys, open discussions, and analyses of human and organisational gaps. Behavioural analyses show that most accidents could be avoided through better behaviours. When managers provide meaning, lead by example, and create a climate of trust, employees report risk situations more readily because they feel safe to speak without fear of sanction.
Why Does Mental Health Directly Influence Workplace Safety?
In 2025, the link can no longer be ignored. Fear of speaking up is a risk in itself. Mental overload triggers errors, and chronic stress undermines vigilance. Physical safety therefore depends on mental well being and overall organisational design, now integrated into the Single Occupational Risk Assessment Document.
Reassuring indicators do not guarantee a safe organisation. They can sometimes reveal organisational silence.
Our Contribution
At C2D Prévention, we support industrial and service organisations in overcoming this paradox. We deploy cultural diagnostics through validated questionnaires, management focused training programmes, and hybrid dashboards integrating psychosocial risks and weak signals. Our interventions help reduce blind spots by creating secure spaces for dialogue. Contact us !
Key Takeaways
- Traditional indicators measure declared past events, but overlook more than 50 percent of risks through seven main blind spots: underreporting, mental factors, safety culture, real work, weak signals, managerial biases, and organisational gaps.
- Good figures can create false reassurance, reducing vigilance and field audits.
- Complement them with leading indicators: field observations, psychological safety surveys, and open dialogue.
- Physical and mental safety are closely linked. Integrate psychosocial risks into risk assessment to anticipate incidents.
FAQ
Why can organisations have good safety indicators and still experience serious accidents?
Because traditional safety indicators mainly measure past and reported events. They fail to capture the human, organisational, and mental factors that precede serious accidents, such as fatigue, fear of speaking up, or on site trade offs.
What do traditional safety indicators actually measure?
They primarily measure regulatory compliance, reported accidents, final outcomes, and adherence to procedures. They provide a retrospective but very limited view of day to day safety.
Are safety indicators useless?
No. They are necessary but insufficient. Problems arise when they are used alone, without complementary indicators that help anticipate future risks and understand real work conditions.
Why do employees not report dangerous situations?
Underreporting is often linked to fear of reprimand, pressure on results, a blame oriented culture, or the belief that reporting will not change anything. This creates misleading organisational silence.
What is a “leading” safety indicator?
A leading indicator is forward looking. It focuses on behaviours, perceptions, weak signals, and working conditions before an accident occurs, unlike traditional lagging indicators.
What are the main blind spots of safety indicators?
The most frequent are underreporting, fatigue and stress, safety culture, the gap between prescribed and real work, weak signals, managerial biases, and organisational vulnerabilities.
Sources :
C2D Prevention. (2025, 17 décembre). Comment comprendre et mettre en œuvre la sécurité psychologique au travail en 2025. https://www.c2dprevention.com/blog/comment-comprendre-et-mettre-en-oeuvre-la-securite-psychologique-au-travail-en-2025/
CNAM. (2023). Tableaux de synthèse des statistiques nationales de sinistralité AT/MP 2023. https://www.formaprev21.fr/media/original/2023-synthese-sinistralite-pour-chaque-ctn-297171.pdf
DARES. (2025, 22 octobre). Les accidents du travail. https://dares.travail-emploi.gouv.fr/donnees/les-accidents-du-travail
Drieets Île-de-France. (2025). Prévention des accidents graves et mortels au travail 2025. https://idf.drieets.gouv.fr/sites/idf.drieets.gouv.fr/IMG/pdf/06_risquespro-dp_2025.pdf
Eurogip. (2023). Estimations actualisées du phénomène de sous-déclaration des accidents du travail en Europe. https://eurogip.fr/wp-content/uploads/2023/12/EUROGIP-2023-Sous-declaration-des-AT-en-Europe.pdf
FONCSI. (2022). Facteurs humains et organisationnels : État de l’art. https://www.foncsi.org/fr/publications/cahiers-securite-industrielle/facteurs-humains-et-organisationnels/CSI-FHOS-etat-art.pdf
INRS. (2025). Hygiène et sécurité du travail n°279. https://www.inrs.fr/dam/jcr:387a799c-5161-4bf3-98ed-d7a7a40c1f5f/HST279_COMPLET.pdf
INRS. (2025). Analyse des accidents du travail. https://www.inrs.fr/demarche/analyse-accidents-travail/ce-qu-il-faut-retenir.html
Le Point. (2025, 26 novembre). Le nombre d’accidents du travail en net repli en France. https://www.lepoint.fr/economie/le-nombre-d-accidents-du-travail-en-net-repli-en-france-26-11-2025-2604070_28.php
myRHline. (2025, 29 juin). Stress au travail en 2025 : 64% des salariés restent sous pression. https://myrhline.com/type-article/stress-travail-2025/
Syndex. (2025, 12 octobre). Enquête sur les risques graves en entreprise. https://www.syndex.fr/espace-presse/syndex-devoile-les-resultats-dune-enquete-sur-les-risques-graves-en-entreprise
UNSA. (2025, 24 novembre). En France, encore trop de morts au travail. https://www.unsa.org/En-France-encore-trop-de-morts-au-travail.html
Vie-publique. (2025, 1er décembre). Accidents du travail mortels : 764 décès en 2024. https://www.vie-publique.fr/en-bref/301137-accidents-du-travail-mortels-764-deces-en-2024
Aegide International. (2025, 13 mai). Culture de sécurité : Faire évoluer les comportements pour éviter 90% des accidents. https://www.aegide-international.com/culture-securite-faire-evoluer-les-comportements-eviter-90-des-accidents/
Edmondson, A. (1999). Psychological safety and learning behavior in work teams. Administrative Science Quarterly, 44(2), 350-383. [Cité via ]
PMC-NIH. (2025). Safety-specific transformational leadership and safety performance. https://pmc.ncbi.nlm.nih.gov/articles/PMC12337517/
Carcept-Prev. (2025, 27 janvier). AT-MP : La fréquence des accidents du travail a encore diminué en 2023. https://www.carcept-prev.fr/actualites/mp-la-frequence-des-accidents-du-travail-encore-diminue-en-2023
INRS. (s.d.). Accidents du travail : Mieux les connaître pour les prévenir (DO 48). https://www.inrs.fr/dam/jcr:c9d40ec9-a588-45e1-8825-95f34fd47c12/do48.pdf
Info.gouv.fr. (s.d.). Stop aux accidents du travail graves et mortels. https://www.info.gouv.fr/upload/media/content/0001/07/7dd59d11ed5eafd87b816663396d4f8015c73d27.pdf