Continuous Psychosocial Risk Monitoring: A Framework for Anonymised, Evidence-Based Workplace Risk Oversight

Research Paper — Sandora

A structured description of the Sandora approach to continuous monitoring, anonymous data aggregation and compliance-oriented risk evidence generation in organisational environments.

Published: 22 May 2026
Author: Meline Lopes, Sandora
Version: 1.0
Format: Conceptual framework — defensive publication

Abstract

Psychosocial risks in the workplace — including chronic workload overload, interpersonal conflict, social isolation and perceived lack of organisational safety — represent a growing public health and governance challenge. Existing monitoring approaches rely predominantly on periodic surveys, reactive complaint mechanisms and point-in-time assessments, resulting in significant detection gaps between the emergence of risk and organisational response. This paper describes a continuous, anonymised monitoring framework designed to generate structured, real-time evidence of psychosocial risk at the organisational and team levels. The framework combines short-cycle anonymous data collection, multi-layer aggregation with re-identification protection, composite risk scoring across established psychosocial dimensions, and compliance-oriented reporting outputs. The approach is intended to serve both employer compliance obligations and occupational health collaboration requirements, and is designed to be applicable across organisational contexts regardless of sector or size.

Contents

1. Introduction and Rationale
2. The Problem: Structural Gaps in Current Approaches
3. Conceptual Foundations
4. Framework Architecture
5. Anonymisation and Privacy Design
6. Risk Scoring Approach
7. Compliance Integration Layer
8. Occupational Health Collaboration Model
9. Practical and Operational Considerations
10. Scope, Limitations and Ethical Boundaries
11. Conclusion
12. References

Section 1

Introduction and Rationale

The relationship between working conditions and mental health outcomes has been extensively documented across occupational health research for several decades. What has changed significantly in recent years is the regulatory and governance context within which employers are now expected to respond to this evidence. Across multiple jurisdictions, legislation has moved from general obligations around physical safety to explicit requirements for the identification, assessment and management of psychosocial risk factors at the workplace level.

In parallel, the direct economic costs of unmanaged psychosocial risk have become increasingly quantifiable. Mental health conditions now account for the largest share of work-related sickness absence and disability pension expenditure in many European countries. The indirect costs — reduced organisational productivity, increased staff turnover, reputational damage and escalating occupational health intervention needs — compound the direct burden significantly.

Despite this dual pressure from regulatory obligation and economic necessity, the operational tools available to most organisations for managing psychosocial risk remain structurally inadequate. The dominant approaches — annual employee surveys, reactive HR case management, clinical referral only after symptom presentation — are characterised by significant temporal gaps between the emergence of risk and organisational detection. By the time a risk becomes visible through these mechanisms, its consequences have typically already begun to materialise.

This paper describes a framework that addresses this structural gap directly: a continuous, anonymised monitoring system designed to generate structured, real-time evidence of psychosocial risk states at the organisational and team levels. The framework is conceptualised as infrastructure — not an intervention in itself, but the data layer that enables timely, evidence-based intervention by those with the authority and expertise to act on it.

Publication purpose

This document constitutes a public description of the Sandora framework as implemented and developed by Sandora. It is published to establish a clear date of prior disclosure and to contribute to the shared understanding of continuous psychosocial risk monitoring as an organisational practice. Specific implementation details, scoring algorithms, proprietary configurations and system architecture remain the intellectual property of Sandora and are not disclosed herein.

Section 2

The Problem: Structural Gaps in Current Approaches

To understand the design rationale for a continuous monitoring framework, it is necessary to examine in detail why existing approaches are insufficient. The limitations are not incidental — they reflect fundamental structural choices in how psychosocial risk has historically been conceptualised as a monitoring problem.

2.1 Temporal insufficiency of periodic assessment

The predominant model for organisational psychosocial risk assessment involves periodic data collection, typically annual or bi-annual in frequency. This approach is derived from clinical and epidemiological research methodologies and is well-suited to population-level trend analysis. It is poorly suited, however, to organisational risk management, where the relevant unit of analysis is not a population average but a specific team or work unit at a specific point in time.

Psychosocial risk states are dynamic. They emerge and escalate in response to organisational events — changes in workload, leadership transitions, team restructuring, conflict escalation — that can occur and evolve substantially within a period of weeks. Annual assessment intervals mean that risk states may develop, cause harm, and partially resolve before any structured data collection occurs. The result is a systematic detection failure at the precise moment when intervention would be most effective.

2.2 Reactive architecture of complaint-based systems

A second dominant approach relies on complaint and escalation mechanisms: formal HR processes, grievance procedures, whistleblowing channels and occupational health referrals initiated by employees. These systems share a common structural feature: they are activated by symptom presentation rather than risk identification. By definition, they only register a psychosocial risk event after an employee has reached a threshold of distress sufficient to initiate a formal process.

This architecture has well-documented limitations beyond its reactivity. Psychological barriers to formal reporting — stigma, fear of professional consequences, perceived futility and cultural norms around disclosure — mean that formal complaint rates significantly understate the true prevalence of workplace psychosocial risk. The absence of complaints is systematically misread as evidence of absence of risk.

2.3 Fragmentation across systems and data sources

Where organisations do collect psychosocial risk-relevant data, it is typically distributed across multiple disconnected systems: HR information systems, absence management platforms, performance management tools, occupational health provider records and informal management observation. These data sources do not communicate with one another, generate data in incompatible formats and are rarely synthesised into a coherent risk picture at the team or organisational level.

The consequence is that risk signals that would be meaningful if observed in combination remain invisible as isolated data points. An increase in short-term absence, a pattern of deadline extensions, elevated turnover intent in engagement data and a rise in informal management escalations may each individually fall below any threshold for action, while their combination represents a coherent early warning of significant psychosocial risk accumulation.

2.4 The evidence gap for compliance purposes

Regulatory frameworks in multiple jurisdictions have moved to require not merely that employers manage psychosocial risk, but that they are able to demonstrate having done so. This documentation requirement creates a specific challenge for organisations relying on informal or reactive mechanisms: the absence of structured, timestamped evidence of monitoring and assessment activities constitutes non-compliance even where actual organisational practice may have been reasonable.

Continuous monitoring frameworks address this directly by generating structured, auditable records of ongoing risk assessment activities as an inherent output of normal operations — rather than requiring retrospective documentation of activities that were never designed to produce records.

Section 3

Conceptual Foundations

The Sandora framework draws on several established bodies of knowledge in occupational health psychology, privacy-preserving data systems and organisational risk management. The conceptual integration of these domains constitutes the foundational contribution of the framework.

3.1 Psychosocial risk dimensions

The identification of psychosocial risk factors relevant to the framework is grounded in the established literature on work-related mental health determinants. The primary dimensions monitored by the framework correspond to risk factors identified across major occupational health research traditions, including the demand-control-support model, the effort-reward imbalance model and the Copenhagen Psychosocial Questionnaire framework.

These include, but are not limited to, the following dimensions:

• Workload and pace: quantitative demands, time pressure and the ability of employees to influence the pace and volume of their own work.
• Work organisation and content: clarity of role expectations, degree of autonomy and the meaningful variation of tasks.
• Social environment: quality of interpersonal relationships at the team level, perceived social support and the presence of interpersonal conflict or exclusion dynamics.
• Psychological safety: the degree to which the work environment is perceived as safe for expressing concerns, making errors or raising issues without fear of negative consequences.
• Leadership and recognition: perceived fairness of treatment, clarity of feedback and the degree to which contributions are recognised.
• Work-life interface: the degree to which work demands intrude on or are compatible with non-work life responsibilities.

These dimensions are not assessed in isolation but are understood as interacting elements of a systemic risk state at the team level. The framework is designed to capture their combined and time-varying configuration rather than their individual point-in-time values.

3.2 Continuous monitoring as a risk management paradigm

The concept of continuous monitoring as applied to psychosocial risk represents an adaptation of principles established in physical risk management, financial risk oversight and information security. In these domains, the value of continuous monitoring over periodic assessment is well established: risks that develop and escalate dynamically require monitoring systems whose detection granularity is at least as fine as the timescale of risk evolution.

Applied to psychosocial risk, this principle implies monitoring cycles that are short enough to detect meaningful changes in risk state within the timeframe in which intervention would alter outcomes — typically measured in weeks rather than months. The framework is designed around this temporal constraint.

3.3 Privacy-preserving aggregate analytics

A foundational design constraint for any psychosocial monitoring system is the protection of individual respondent privacy. This constraint operates on two levels: as an ethical requirement grounded in respect for individual autonomy and data rights, and as a functional requirement for system validity — a monitoring system whose outputs employees do not trust to be genuinely anonymous will systematically produce biased data, as individuals modify their responses in anticipation of identification.

The framework addresses this through a layered anonymisation architecture that ensures individual responses are never directly accessible to any organisational user, and that aggregation occurs under conditions that prevent statistical re-identification of individual respondents. The specific mechanisms of this architecture are proprietary and not described in this document; the design principles they implement are drawn from established practice in privacy-preserving analytics and differential privacy research.

Section 4

Framework Architecture

The Sandora framework is organised around four functional layers that operate in a continuous, iterative cycle. Each layer performs a distinct function and produces outputs that serve as inputs to the subsequent layer.

Data collection → Anonymisation layer → Risk scoring → Compliance output → Monitoring loop

The Sandora continuous monitoring cycle. The loop operates at short intervals, generating an ongoing time-series of risk states rather than point-in-time assessments.

Layer 1

Continuous data collection

Short, frequent anonymous check-ins capture key psychosocial indicators across established risk dimensions. Collection cadence is designed to match the temporal dynamics of psychosocial risk evolution rather than organisational reporting cycles.

Layer 2

Anonymisation and aggregation

Individual responses are processed through a multi-layer anonymisation architecture before any aggregation occurs. Minimum group thresholds and additional privacy-preserving mechanisms prevent re-identification at every stage of processing.

Layer 3

Dynamic risk scoring

Aggregated, anonymised data is processed through a composite scoring model that generates a continuous risk state representation across defined psychosocial dimensions. Scoring incorporates both current state and temporal trend components.

Layer 4

Compliance-oriented output

Risk scores and associated evidence are translated into structured outputs aligned with regulatory documentation requirements. Outputs are designed to support both internal organisational decision-making and external compliance demonstration.

The four layers are not sequential stages in a batch process but components of a continuously operating system. At any given point, different parts of the system are simultaneously collecting new data, processing recent responses, updating risk scores and generating updated outputs. This architecture is what enables the framework to provide a genuinely continuous — rather than periodically refreshed — view of organisational risk state.

4.1 The monitoring loop

A defining characteristic of the framework is its iterative, time-series nature. The output of each monitoring cycle is not a static report but an updated observation in an ongoing time series. This means the framework produces not only a current risk state reading but a history of risk state evolution, from which trend dynamics, escalation trajectories and the effects of interventions can be observed and interpreted.

This time-series property is significant for both operational and compliance purposes. Operationally, it enables the detection of emerging risk trajectories before they reach acute levels — the framework identifies not only where risk is high, but where risk is increasing and at what rate. For compliance purposes, the time series constitutes a documented record of continuous monitoring activity, demonstrating ongoing risk oversight rather than episodic assessment.

Section 5

Anonymisation and Privacy Design

The anonymisation architecture is the most operationally critical component of the framework. Its adequacy determines both the validity of the data collected — through its effect on respondent trust and behaviour — and the ethical permissibility of the system’s operation.

5.1 Design principles

The anonymisation layer is designed around the following principles, which are implemented through mechanisms that are proprietary to the Sandora system and not described in this document:

Individual non-exposure

No individual employee response is ever directly accessible to any organisational user, including HR administrators, line managers or senior leadership. This constraint is absolute and is not configurable at the organisational level.

Group threshold protection

Aggregated outputs are only generated for groups that meet minimum size thresholds, below which the risk of statistical re-identification from aggregate data is considered unacceptable. Threshold values are set conservatively and are not disclosed publicly.

Temporal aggregation constraints

The system applies constraints on the temporal granularity of data available in outputs to prevent re-identification through pattern analysis over time, particularly in small or stable team configurations.

Response aggregation before scoring

The sequence of operations in the system is designed such that individual-level data is never present at the point where scoring and output generation occur. Scoring is performed on pre-aggregated data, not on individual response records.

No behavioural inference

The system does not generate any outputs that attribute individual-level characteristics, behaviours or risk states to specific employees. All outputs describe collective states of defined organisational units.

5.2 The trust-validity relationship

A monitoring system’s data quality is fundamentally dependent on the degree to which respondents trust that their anonymity is genuinely protected. This relationship creates a design constraint that is distinct from, and in some respects more demanding than, the purely technical privacy requirement. A system that is technically anonymous but is not perceived as such by respondents will produce systematically biased data, as employees modify responses to avoid perceived identification risk.

The framework addresses this through a combination of architectural design (the principles above), transparent communication of those principles to employees, and organisational governance requirements that prevent configurations that could undermine perceived anonymity — for example, by creating organisational units small enough that even aggregate outputs would identify individuals.

Section 6

Risk Scoring Approach

The risk scoring component of the framework translates aggregated, anonymised response data into structured risk state representations suitable for organisational decision-making and compliance documentation. The scoring approach is described here at the conceptual level; specific model parameters, weighting schemes, threshold values and algorithmic implementations are proprietary and are not disclosed in this document.

6.1 Composite nature of the score

The framework generates risk scores that are composite across multiple psychosocial dimensions rather than single-metric or unidimensional. This compositional approach reflects the empirical understanding that psychosocial risk at the workplace level is not adequately captured by any single indicator. High workload combined with strong social support, for example, produces a different risk profile than high workload combined with social isolation — a unidimensional workload score would obscure this distinction.

The composite score is designed to reflect both the current state across dimensions and the relationship between them. Combinations of dimension states that are jointly associated with elevated risk receive scores that reflect this combination effect rather than being reducible to the sum of their components.

6.2 Temporal component

A distinguishing feature of the scoring approach is the explicit incorporation of temporal trajectory as a component of risk assessment, not merely as context. The system does not only assess the current level of a risk indicator but also the direction and rate of change of that indicator over recent monitoring cycles.

This temporal component reflects the clinical and organisational observation that the trajectory of a risk state is independently meaningful from its current level. A team with a moderately elevated risk score that has been stable for several months represents a different risk situation from one with the same current score that has risen sharply over the past four weeks. The framework is designed to make this distinction operationally visible to users.

6.3 Organisational comparability

The scoring approach is designed to support meaningful comparison of risk states across different teams and organisational units over time. This comparability property is a prerequisite for the use of monitoring data in organisational risk prioritisation — without it, scores can only be interpreted in isolation rather than as inputs to a portfolio view of organisational risk.

Scope limitation: The risk scores generated by the framework are designed to support organisational risk management and compliance documentation. They are not clinical diagnostic instruments and should not be interpreted as such. The framework does not generate individual-level clinical assessments, does not constitute a replacement for occupational health clinical evaluation, and its outputs should not be used to make decisions about individual employees.

Section 7

Compliance Integration Layer

A core design requirement of the framework is that its outputs be directly useful for compliance purposes — not merely providing data that could, with additional processing, support compliance activities, but generating structured records that directly address the documentation obligations employers face under relevant regulatory frameworks.

7.1 Regulatory alignment

The framework’s compliance outputs are designed to address obligations arising under multiple regulatory frameworks applicable to workplace psychosocial risk. These include, but are not limited to:

• Obligations under occupational safety and health legislation requiring employers to systematically identify, assess and document psychosocial risk factors (including applicable national implementations of EU OSH framework directives and equivalent national legislation).
• Obligations under occupational health legislation requiring structured monitoring and assessment activities to be conducted collaboratively with certified occupational health providers.
• Obligations arising under equality and anti-discrimination legislation to maintain work environments free from psychosocial harm, including harassment, exclusion and psychological safety failures.
• Emerging obligations under corporate governance and ESG reporting frameworks requiring organisations to demonstrate structured oversight of workforce wellbeing and psychosocial risk.

7.2 Documentation outputs

The compliance layer generates structured documentation outputs as a continuous by-product of normal system operation. These outputs include timestamped records of monitoring activities, risk assessment summaries aligned to regulatory risk identification requirements, trend reports documenting the evolution of risk states over time, and intervention documentation linking monitoring observations to organisational responses.

A critical design principle of the documentation outputs is that they represent a genuine record of ongoing monitoring activity rather than post-hoc reconstruction. The timestamps, data provenance and audit trail embedded in the outputs are generated at the time of the corresponding monitoring event, creating a documentary record whose integrity is verifiable independently of any subsequent organisational account of what monitoring was conducted.

7.3 Upstream positioning relative to reactive compliance tools

It is important to locate the framework correctly within the broader ecosystem of organisational compliance tools. Whistleblowing systems, formal complaint channels, HR case management tools and occupational health referral pathways address psychosocial risk at the point at which it has already escalated to the level of individual disclosure or clinical presentation.

The framework operates at an earlier stage in the risk lifecycle — providing structured evidence of risk states before they escalate to the threshold that activates reactive tools. It is not a replacement for those tools but a complementary layer that provides the early-warning function their architecture does not support. Organisations that implement the framework alongside existing compliance infrastructure gain structured evidence of both the risk state that preceded any formal complaint and the monitoring activities that were in place at the time.

Section 8

Occupational Health Collaboration Model

The framework is designed not as a standalone employer tool but as infrastructure that enhances the effectiveness of the collaboration between employers and occupational health (OHC) providers that is required — and in many jurisdictions legally mandated — as part of systematic workplace health management.

8.1 The OHC collaboration gap

Certified OHC providers are typically engaged to deliver two distinct functions in relation to psychosocial risk: population-level risk monitoring across their client workforce, and individual-level clinical assessment and intervention for employees presenting with symptoms. Both functions are compromised by the same data gap that affects employer risk management: the absence of structured, continuous psychosocial risk data at the team level.

Without this data, OHC providers are forced to rely on periodic survey instruments, clinical presentations and informal management reports to form a picture of client workforce psychosocial risk state. The result is that OHC engagement tends to be reactive — concentrated at the point of individual clinical presentation — rather than proactive and prevention-oriented.

8.2 The framework as a shared data layer

The framework creates a structured, continuous psychosocial risk data layer that can serve as a shared information resource for both employer risk management and OHC provider service delivery. In this model:

• The employer uses the framework’s outputs for operational risk management, compliance documentation and internal intervention planning.
• The OHC provider uses the same data layer to understand the organisational and team-level psychosocial risk context for employees they see clinically, to prioritise proactive outreach in higher-risk environments, and to document the monitoring activities they are obligated to deliver.
• The combination of continuous monitoring data and OHC clinical expertise enables a referral pathway that connects individuals to clinical support earlier — when intervention is more effective — and with richer contextual information about the organisational risk environment.

B2B2C architecture

The collaboration model described here reflects a B2B2C architecture in which the framework serves as infrastructure at the employer-OHC interface, enabling both parties to deliver their respective obligations more effectively and enabling employees to be supported earlier in the development of psychosocial risk than would be possible through either employer-only or OHC-only mechanisms.

Section 9

Practical and Operational Considerations

9.1 Implementation prerequisites

Effective deployment of the framework requires several organisational prerequisites that are not themselves part of the technical system but that determine whether the technical system can generate valid and useful outputs. These include:

• Minimum viable team size: The anonymisation architecture requires that monitoring units meet minimum group size thresholds. Organisations with very small team structures require adaptation to ensure that monitoring units are constituted in ways that satisfy anonymity requirements while remaining organisationally meaningful.
• Communication and transparency: Employees must be clearly informed about what data is collected, how it is processed, what protections apply, and how outputs are used by the organisation. Inadequate communication compromises both the ethical basis of the system and the data quality it produces.
• Governance structure: Organisations require a defined governance structure for the use of monitoring outputs — including clarity about who has access to what level of aggregation, how outputs are used in decision-making, and what processes govern any interventions triggered by monitoring signals.

9.2 Integration with existing organisational systems

The framework is designed to be deployable as a standalone monitoring layer without requiring integration with existing HR, OHC or case management systems. Where integration is possible and appropriate, it enables enriched contextual interpretation of monitoring outputs — for example, correlating risk state changes with identifiable organisational events such as restructuring, leadership changes or workload peaks.

9.3 Scalability across organisational contexts

The framework’s architecture is designed to be applicable across a wide range of organisational contexts — varying by size, sector, geographic location and regulatory environment. The core monitoring cycle, anonymisation architecture and scoring approach remain consistent across contexts; what varies is the calibration of specific parameters to reflect context-appropriate risk dimensions, relevant regulatory frameworks and organisational governance norms.

Section 10

Scope, Limitations and Ethical Boundaries

The framework described in this document represents a significant operational advance over existing approaches to psychosocial risk monitoring. It does not, however, address all aspects of the psychosocial risk management challenge, and its application has boundaries that are important to understand clearly.

10.1 What the framework does not do

Individual assessment

The framework does not and cannot identify which individual employees are experiencing psychosocial distress. Its outputs describe collective risk states of organisational units. Individual-level assessment remains the domain of occupational health clinical evaluation.

Causal attribution

Risk state observations generated by the framework identify the presence and intensity of risk states but do not independently establish their causal origins. Interpretation of monitoring outputs requires organisational context and expertise that the system itself does not provide.

Intervention design

The framework generates evidence that informs intervention decisions but does not prescribe specific interventions. Intervention design remains a human judgment function requiring organisational, HR and clinical expertise as appropriate to the situation.

Clinical diagnosis

Monitoring outputs should not be used as the basis for clinical diagnosis or to support decisions about individual employment status. The framework is an organisational risk management tool, not a clinical instrument.

10.2 Ethical framework

The framework’s design is governed by the following ethical principles, which are embedded in its architecture and cannot be overridden by organisational configuration:

• Individual anonymity is an absolute protection, not a configurable preference.
• Data collected through the system is used exclusively for the purposes disclosed to employees at the time of collection.
• The system is not designed to monitor, assess or generate outputs about the performance, behaviour or suitability of individual employees.
• The system is not designed to surveil employees or generate data for purposes unrelated to psychosocial risk management and occupational health support.

10.3 Limitations of this document

This document describes the Sandora framework at a conceptual and architectural level. It does not disclose specific implementation details, algorithmic parameters, scoring model configurations, proprietary threshold values or system architecture details. These elements remain the intellectual property of Sandora and are described only internally, subject to appropriate confidentiality protections.

Section 11

Conclusion

The challenge of psychosocial risk in the workplace is not a new one. What is new is the convergence of three developments that together create both the imperative and the operational possibility for a fundamentally different approach to managing it: explicit and enforceable regulatory obligations requiring structured, documented monitoring; increasing quantification of the economic costs of unmanaged psychosocial risk; and the technical capability to implement genuinely continuous, privacy-preserving monitoring at organisational scale.

The framework described in this paper responds to this convergence with an architecture that addresses the structural gaps of existing approaches: the temporal insufficiency of periodic assessment, the reactive architecture of complaint-based systems, the fragmentation of risk-relevant data across organisational systems, and the absence of structured compliance evidence that continuous monitoring generates as an inherent output.

The Sandora framework positions psychosocial risk monitoring as infrastructure — a continuous, anonymised data layer that enables timely, evidence-based action by the employers, managers, HR professionals and occupational health providers who are responsible for workforce wellbeing. It does not replace the human judgment, clinical expertise and organisational accountability that effective psychosocial risk management requires. It provides the structured, continuous evidence base on which those human functions can operate effectively.

As regulatory expectations in this domain continue to develop across European and other jurisdictions, organisations that have established continuous monitoring infrastructure will find themselves better positioned — practically, evidentially and reputationally — to meet those expectations. The cost of waiting for the regulatory floor to rise before investing in monitoring capability is paid not in compliance penalties alone, but in the human cost of risks that continuous monitoring would have made visible in time to act.

References

References

This paper draws on the established literature in occupational health psychology, privacy-preserving analytics and regulatory frameworks for workplace risk management. Selected key references are listed below.

Bakker, A.B. & Demerouti, E. (2007). The Job Demands-Resources model: State of the art. Journal of Managerial Psychology, 22(3), 309–328.

Cox, T., Griffiths, A. & Rial-González, E. (2000). Research on Work-Related Stress. European Agency for Safety and Health at Work.

Dwork, C. & Roth, A. (2014). The Algorithmic Foundations of Differential Privacy. Foundations and Trends in Theoretical Computer Science, 9(3–4), 211–407.

Edmondson, A.C. (1999). Psychological Safety and Learning Behavior in Work Teams. Administrative Science Quarterly, 44(2), 350–383.

European Agency for Safety and Health at Work (EU-OSHA). (2022). Psychosocial Risks and Stress at Work. European Union.

Karasek, R. & Theorell, T. (1990). Healthy Work: Stress, Productivity and the Reconstruction of Working Life. Basic Books.

Ministry of Social Affairs and Health, Finland. (2020). National Mental Health Strategy and Programme for Suicide Prevention 2020–2030. Publications 2020:15.

Ministry of Social Affairs and Health, Finland. (2023). Healthy Work: Framework Plan for Occupational Safety and Health Divisions 2024–2027. Publications 2023:23.

OECD/EU. (2018). Health at a Glance: Europe 2018 — State of Health in the EU Cycle. OECD Publishing.

Siegrist, J. (1996). Adverse health effects of high-effort/low-reward conditions. Journal of Occupational Health Psychology, 1(1), 27–41.

Toppinen-Tanner, S. (Director of Programme, FIOH). (2020). Contribution to the preparation of the Finnish National Mental Health Strategy 2020–2030. Finnish Institute of Occupational Health.

Published by Sandora · sandora.me/research/sandora-framework · © 2026 Sandora. All rights reserved.

This document is published for the purpose of public disclosure and does not constitute a patent application or a waiver of intellectual property rights with respect to proprietary implementation details not described herein.

For enquiries: meline@sandora.me

Published: 2026-05-22 | Version 1.0 | sandora.me/research