The building blocks of occupational health and safety management: profiles of preconditions and their links to practices among Swedish welfare organizations Open Access

Many employers still struggle to implement occupational health and safety management (OHSM) effectively, despite decades of legislation and organizational policies aimed at ensuring healthy workplaces. Understanding the preconditions that enable successful OHSM is more critical than ever as work environments grow increasingly complex, due to digitalization, psychosocial stressors and organizational restructuring. Across the globe, governments have implemented regulations to ensure workplace safety and health (ILO, 2025). For instance, in Sweden, the Work Environment Act mandates employers to conduct OHSM activities (Lundqvist et al., 2024). Similar regulatory frameworks exist in other countries (e.g. the Work Health and Safety Act, 2011 in Australia, Health and Safety at Work Act, 1974 in the United Kingdom, or the Workplace Safety and Health Act in Singapore) reflecting a growing recognition of the importance of active management of the work environment to ensure workplace safety and health. Swedish law mandates that all employers conduct OHSM consisting of systematic evaluations, improvements and monitoring of employee occupational health issues. The regulatory framework from 2015, updated in 2025 (SWEA, 2023), requires preventive OHSM to address the organizational and social aspects of the work environment, i.e. how work is organized, designed and managed.

These legal requirements are backed up by research demonstrating that successful OHSM is generally associated with positive outcomes for both employees and organizations (Robson et al., 2007; Paulsson et al., 2023; Ståhl et al., 2025; Stöllman et al., 2025). Dahler-Larsen et al. (2020) have shown that implementing systematic OHSM enhances the likelihood of addressing psychosocial risk factors. Successful OHSM activities lead to improved workplace perceptions (e.g. better safety climate) (Robson et al., 2007), increased productivity (Nkrumah et al., 2021) and a reduction in injuries (Bunn et al., 2001) and disability-related costs (Robson et al., 2007). Furthermore, there is evidence linking OHSM to enhanced organizational commitment, decreased work alienation and improved job performance (Kaynak et al., 2016). It has nevertheless been argued (Frick, 2014) that there is insufficient integration of the Swedish OHSM regulations into practice, making it relevant to study which preconditions are linked to high-quality OHSM.

Various preconditions within organizations can significantly influence the management of occupational health and safety (OHS), playing a crucial role in determining the extent and quality of OHSM activities. One key organizational precondition is the availability of resources (EASHW, 2010; Ståhl et al., 2025). Managers often cite lack of time as a barrier to conducting OHSM activities effectively (Hellman et al., 2019; Larsson et al., 2015). Financial resources also play a critical role, as budget allocations can directly impact OHSM initiatives (Hasle et al., 2006; Lundqvist et al., 2024). Additionally, the level of support and commitment from upper managementand employee participation has been identified as a significant factor in the success of OHSM initiatives (Gallagher et al., 2003; Fadhel and Alqurs, 2025; Nielsen et al., 2025; von Thiele Schwarz et al., 2021). In a recent study, managerial support was found to be a difference-making factor for the overall success of OHSM initiatives, while employee participation determined the level of success (Akerstrom et al., 2024a). Moreover, the level of knowledge and expertise within the organization regarding OHS matters can significantly impact OHSM initiatives and their effectiveness (EASHW, 2010; Walker and Tait, 2004; Ståhl et al., 2025). This involves training as well as access to working methods or established routines.

While regulatory frameworks underscore the need for OHSM, less is known about the specific organizational preconditions that determine whether OHSM efforts succeed or fail. Previous research has identified individual factors such as leadership support and resources (EASHW, 2010; Hellman et al., 2019), yet these studies have not systematically examined how different combinations of such preconditions coexist within organizations. Nor have they explored how these profiles of organizational preconditions relate to actual OHSM quality. This study addresses these limitations by using latent profile analysis (LPA) to empirically identify distinct profiles of organizational preconditions for OHSM.

Therefore, the aim of this study was to identify profiles of organizational preconditions for OHSM and investigate their association to the success and quality of their OHSM, defined as work involving a high percentage of promotive/preventive initiatives conducted mainly at an organizational rather than individual level, and including some form of direct or indirect employee participation (SWEA, 2001; Nielsen et al., 2025). In doing so, this study contributes to the literature by offering a typology of organizational preconditions that may guide tailored interventions and policies.

This was done by answer the following research questions (RQs):

Unlike prior studies that examine variables in isolation, this approach allows for nuanced understanding of how multiple preconditions cluster together and influence OHSM. It thereby offers a classification framework for organizational readiness or capability to implement effective practices, as well as inform which preconditions should be prioritized for best outcomes. Understanding these preconditions is crucial for developing more effective OHSM strategies and ultimately improving workplace OHS outcomes.

In this study, a digital questionnaire was distributed to all municipalities, regions and privately owned businesses that operate within the public sector. The survey was investigating the employers’ work routines, resources, use of work environment economics within and the perceived success and quality of the employers’ preventive OHSM. The questionnaire was distributed via e-mail (retrieved from public records and registers from the Swedish Association of Local Authorities and Regions, The Health and Social Care Inspectorate and The Swedish National Agency for Education) and advertised on social media, the project webpage and through occupational networks. After having given informed consent for participation, the organizations nominated one or more employer representatives with the most thorough knowledge of preventive OHSM and best insight on its performance to answer the survey. The survey was open for two months (April–June 2023) and used a quota sampling strategy with the goal of reaching about 150 responses evenly distributed within the different sectors (private, municipality and region). The development of the questionnaire, the data collection and its validity and reliability has previously been described in detail (Akerstrom et al., 2024b). The questionnaire was developed and tested for the purpose of this project and showed a satisfactory response distribution, as well as high validity and reliability. For details, see Akerstrom et al. (2024b).

The survey was distributed to organizations within the welfare sector (healthcare, social services, education, public administration and workplaces providing services connected to welfare). The public sector is the main provider of welfare services in Sweden and includes 290 municipalities and 21 regions. Municipalities handle services such as education, social services and elderly care, while regions manage services such as healthcare and regional development (SALAR, 2018). In addition to these public actors, privately owned businesses may also operate within the publicly funded welfare sector.

Our study captured 113 responses from employer representatives, with no missing data in the indicator variables (see below). In most cases (72%), HR was appointed to be the employer representative answering the survey. The majority (71%) of the participating organizations were large, with 250 or more employees. Organizations represented both the private (21%) and public sector (79%).

In this study, five items representing different organizational preconditions for OHSM were selected from the larger questionnaire described above, to be used as indicators in the latent profile analysis. Another five items were selected to represent outcomes connected to the perceived success and quality of the OHSM. The choice of variables to include in the analysis was guided by the resources and recommendations highlighted in the Swedish regulatory provision on preventive OHSM (SWEA, 2001) and in the literature (von Thiele Schwarz et al., 2021). Table 1 lists the selected variables, their operationalization in the survey and applied response categories.

To investigate the latent profiles of organizations in terms of preconditions for OHSM (described above), we employed latent profile analysis (LPA) using R package tidyLPA (Rosenberg et al., 2018). Given the relatively low sample size, we decided to apply a less computationally intensive model, with equal variances of indicators between clusters and covariances constrained to 0 (i.e. the variables were not allowed to covary over and above their association as part of the same profile). This specification corresponds to Model 1 in tidyLPA. Prior to modeling, the indicators building the profiles were standardized, i.e. we used z scores with a mean of 0 and standard deviation of 1.

In line with established LPA practices (Spurk et al., 2020), we used a combination of statistical fit indices and theoretical interpretability to determine the optimal number of profiles. Due to the exploratory nature of this analysis, we generated a series of LPA models with an increasing number of latent profiles (between one and seven) and iteratively compared each successive model k with the previous one (k − 1). As recommended (Spurk et al., 2020), multiple criteria were used to determine the optimal number of profiles: Akaike’s information criterion (AIC), Bayesian information criterion (BIC), the sample-size adjusted BIC (SABIC) and the bootstrapped likelihood ratio test (BLRT). A better model fit is indicated by lower AIC, BIC and SABIC values. The BLRT tested whether adding an additional profile significantly (p < 0.05) improved model fit. Profile clarity was examined via the entropy criterion, which indicates how accurately the organizations were categorized into their respective profiles, with values closer to 1 indicating a better profile separation (Spurk et al., 2020). Finally, when making decisions about the optimal number of profiles, we considered meaningfulness (e.g. being consistent with theory and common sense) and parsimony of the latent profiles, as well as the size of the smallest profile (Collins and Lanza, 2010).

The profiles were interpreted and described descriptively using the standardized scores of the indicators. Additionally, each profile was characterized by descriptively comparing sector prevalence (public, private), organization size (i.e. number of employed persons) and industry distribution.

Finally, we compared the profiles in regard to the outcomes of interest. First, we looked at the participants’ perceptions of the role of OHSM in their organization, i.e. to what extent they believe it promotes health. Second, we investigated the practices in terms of the proportion of implemented promotive/preventive initiatives and the proportion of organizational level initiatives within the OHSM. Finally, the profiles were compared regarding the extent to which the content of preventive/promotive work environment initiatives was based on the dialogue with (1) occupational health service, (2) internal expert resource (e.g. HR, work environment coordinator, etc.), (3) safety delegates or union representatives and (4) employees. The first two represent external and internal support, while the latter two correspond with participatory approach targeting the employees.

For the purpose of contrasting profiles in the above-listed outcomes, we used non-parametric Kruskal–Wallis test due to uneven distribution of organizations among the groups, as well as the ranked nature of the outcomes. Post-hoc pairwise comparisons using Dunn’s test were conducted, with adjustments for multiple comparisons using Holm methods. Additionally, due to low numbers of organizations in several profiles, inference was supported by inspecting effect sizes rather than purely statistical significance. To compute effect sizes, Cohen’s r were calculated due to multiple comparisons, with values of 0.10 indicating small effect, 0.30 − medium effect, while 0.50 and higher − large effect.

Table 2 summarizes the model fit statistics for the 1–7 profile solutions. Solutions with six or more profiles were discarded because BLRT suggested that adding additional profiles over a five-profile solution did not improve model fit. We closely inspected solutions with 3–5 profiles. While the five-profile solution obtained the best fit statistics, as indicated by lowest AIC, BIC and SABIC, it was ultimately rejected, as the size of the smallest profile was only three organizations. In addition, the fifth profile led to less interpretability without a significant improvement in model quality. The four-profile solution was selected as optimal, based on a combination of the low AIC and SABIC, high entropy (0.838) and significant improvement in fit over the three-profile solution (BLRT, p = 0.0099).

The selected four-profile solution represents distinct organizational preconditions available for OHSM (see Figure 1). Table 3 presents mean scores for each variable within each profile (both raw and standardized scores), demonstrating the distinct characteristics of each profile.

1. Profile 1: Prosperous foundations (n = 16)

Organizations in this profile, compared to other profiles, exhibit the highest scores across all variables (z-scores ranging between 0.53 and 1.34, Table 3), indicating superior provision of resources, presence of leadership support to invest in improving OHS, required budget allocations, as well as access to knowledge and appropriate working methods. Among these preconditions, leadership support is relatively lower but still higher than in other profiles, Table 3.

1. Profile 2: Deficient in know-how (n = 16)

Organizations in this profile have low scores across all preconditions (z-scores ranging from −1.87 to −0.29, Table 3), indicating challenges in implementing and sustaining an effective OHSM. The scores are the lowest in terms of accessing relevant knowledge and appropriate working methods, implying overall insufficient foundations for OHSM (Table 3).

1. Profile 3: Balanced but limited (n = 67)

This largest profile includes organizations with average and balanced foundations across all five variables (z-scores ranging from −0.28 to 0.13, Table 3). However, the preconditions of organizations in this profile are poorer than those in the “Prosperous foundations” profile when it comes to budget, resources, knowledge and work practices for OHS, as well as leadership support for the investment in OHSM (z-scores −0.16, 0.11, 0.13, 0.07 and −0.28 compared to 1.02, 1.12, 1.24, 1.34 and 0.53, Table 3), indicating that there is room for growth in this area.

1. Profile 4: Lacking investments (n = 14)

Organizations in this profile have the lowest scores when it comes to leadership support to invest in OHS (a z-score of −1.57), as well as sufficient budget for it (a z-score of −1.14), Table 3. However, they do have access to appropriate working methods (a z-score of −0.19), which is their only present precondition for OHSM. Nevertheless, the level is moderate, i.e. it is less than in the “Prosperous foundations” profile but significantly better than in the “Deficient in know-how” profile. Compared to the “Deficient in know-how” profile, they also have more knowledge, but do not differ when it comes to available resources, Table 3.

Table 4 provides descriptive statistics regarding characteristics of organizations in each profile solution. In terms of sector, both private and public organizations were most represented in the “Balanced but limited” profile, due to its size. However, as Table 4 indicates, public sector organizations were more likely to be found in profiles with poorer preconditions (“Deficient in know-how” and “Lacking investments’) than in the one with the strongest preconditions (“Prosperous foundations”). Conversely, private organizations were represented more in the “Prosperous foundations” profile than in the ones with poorer preconditions (only two organizations in the “Deficient in know-how” profile and none in the “Lacking investments” profile). When it comes to organizational size, only large organizations appear in the “Lacking investments” profile. In contrast, small organizations appear only in “Prosperous foundations” and “Balanced but limited” with better OHSM preconditions. In regard to industry, while the “Balanced but limited” profile dominates across all industries due to its overall prevalence in the sample, some variations emerge upon closer examination. For example, the fact that healthcare organizations appear to have higher representation in the “Prosperous foundations” profile, a presence in the “Balanced but limited” profile, and are absent from the “Lacking investments” profile indicates that the majority of the organizations in this sector possess relatively good preconditions for OHSM.

A Kruskal–Wallis test revealed differences in participants’ perceptions of whether OHSM promotes health across the four organizational profiles, χ² (3) = 25.26, p < 0.001 (see Figure 2). Post-hoc pairwise comparisons revealed that organizations in the “Prosperous foundations” profile had perceptions of significantly higher OHSM performance in promoting health compared to organizations in the “Deficient in know-how” profile (Z = 3.97, p < 0.001, Holm-adjusted p < 0.001, r = 0.38) and the “Lacking investments” profile (Z = 3.39, p = 0.004, Holm-adjusted p = 0.003, r = 0.32). Compared to organizations in the “Balanced but limited” profile, there was a small, but not statistically significant effect (z = 1.34, p = 0.182, r = 0.13). However, organizations in the “Balanced but limited” profile perceived their OHSM performance to be better than those in the “Deficient in know-how” profile (z = 3.71, p < 0.001, Holm-adjusted p = 0.001, r = 0.35) and the “Lacking investments” profile (z = 2.95, p = 0.003, Holm-adjusted p = 0.009, r = 0.28). No significant difference was observed between organizations in the “Deficient in know-how” profile and the “Lacking investments” profile (z = −0.451, p = 0.652, r = −0.04).

In the next step, we examined if there were differences in the reported percentage of promotive and preventive work conducted as part of OHSM across the four profiles. The analysis revealed a statistically significant difference among the profiles in the distribution of the categories (percentage of promotive and preventive work below 25%, 26–50%, 51–75%, or 76–100%), χ² (3) = 12.40, p = 0.006 (see Figure 3). The post-hoc pairwise comparisons indicate that organizations in the “Deficient in know-how” profile reported significantly lower percentage of promotive and preventive work than organizations in the “Prosperous foundations” profile (z = −3.11, p = 0.002, Holm-adjusted p = 0.009, r = 0.30) and “Balanced but limited” profile (z = −3.26, p = 0.001, Holm-adjusted p = 0.007, r = −0.31), but only marginally differed from organizations in the “Lacking investments” profile (z = −2.20, p = 0.028, Holm-adjusted p = 0.112, r = −0.21). No other comparisons were statistically significant, and effect sizes were negligible.

When comparing the profiles in regard to the extent of organizational-level initiatives, we observed that the Kruskal–Wallis test was marginally significant, χ² (3) = 6.42, p = 0.093. However, none of the post-hoc pairwise comparisons were statistically significant. Looking at effect sizes, organizations in the “Deficient in know-how” profile reported a lower percentage of organizational-level OHS initiatives than those in “Prosperous foundations” (r = −0.20) and “Balanced but limited” (r = −0.22) profiles.

Figure 4 demonstrates the differences among the profiles regarding the extent to which the content of preventive/promotive work environment initiatives is based on the dialogue with (1) occupational health service, (2) internal expert resource (e.g. HR, work environment coordinator, etc.), (3) safety delegates or union representatives and (4) employees. Kruskal–Wallis tests were performed separately for each of these outcomes.

There were no differences among the profiles with regards to the external occupational health services’ role for the content of work environment initiatives, χ²(3) = 0.74, p = 0.864. Regarding the extent to which the content of promotive/preventive OHS work is based on dialogue with internal expert resources, the results indicated a statistically significant difference among the organizations clustered in the four profiles, χ²(3) = 10.98, p = 0.012. Post-hoc pairwise comparisons revealed that organizations in the “Deficient in know-how” profile reported a statistically lower extent of such dialogue than those in the “Prosperous foundations” profile (z = −2.69, p = 0.007, Holm-adjusted p = 0.036, r = −0.26) and “Balanced but limited” profile (z = −3.08, p = 0.002, Holm-adjusted p = 0.012, r = −0.29). The comparison between the “Deficient in know-how” profile and the “Lacking investments” profile (z = −1.37, p = 0.172, Holm-adjusted p = 0.686, r = −0.130) did not reveal any statistically significant difference, the small effect size indicating only a minor difference. No other differences were significant, and the effect sizes indicated negligible differences.

In the next step, we evaluated differences in terms of the extent to which the content of OHSM is based on dialogue with safety delegates or union representatives. The test statistic was significant, χ²(3) = 22.03, p < 0.001. Post-hoc pairwise comparisons between the “Prosperous foundations” and the “Deficient in know-how” profiles yielded a statistically significant result (z = 3.96, p < 0.001, Holm-adjusted p < 0.001, r = 0.37), suggesting that organizations in the “Prosperous foundations” profile reported significantly higher levels of dialogue with safety delegates compared to those in the “Deficient in know-how” profile. The comparison between the “Prosperous foundations” profile and the “Lacking investments” profile showed a similar direction and was marginally significant (z = 2.44, p = 0.015, Holm-adjusted p = 0.058, r = 0.23). Organizations in the “Deficient in know-how” profile have significantly less dialogue than those in the “Balanced but limited” profile (z = −3.98, p < 0.001, Holm-adjusted p < 0.001, r = −0.37). The remaining comparisons did not reach statistical significance after adjustments. However, there was a small difference (r = 0.19) between the “Balanced but limited” profile and the “Lacking investments” profile, indicating more such dialogue in the former profile.

The final analysis revealed a statistically significant difference among the profiles in terms of employee involvement in the content of promotive/preventive OHS, χ²(3) = 16.37, p < 0.001. Post-hoc pairwise comparisons indicated that the “Prosperous foundations” profile had significantly higher scores of employee participation than the “Deficient in know-how” profile (z = 3.79, p < 0.001, Holm-adjusted p < 0.001, r = 0.36) and the “Lacking investments” profile (z = 2.54, p = 0.011, Holm-adjusted p = 0.045, r = 0.24). Employee involvement was significantly lower for organizations in the “Deficient in know-how” profile than in the “Balanced but limited” profile (z = −2.91, p = 0.004, Holm-adjusted p = 0.018, r = −0.28) but these differences are moderate. The remaining comparisons were not statistically significant after applying the adjustments for multiple comparisons. However, a small effect occurred (r = 0.18) suggesting higher employee involvement in organizations belonging to the “Prosperous foundations” profile than those in the “Balanced but limited” profile.

This study aimed to identify configurations of preconditions for OHSM among Swedish organizations and explore their associations with OHS practices. The results revealed four distinct profiles of organizational preconditions for OHSM, each with unique characteristics and preconditions for OHSM. The identified pattern of findings underscores the complexity and variability in how organizations approach OHSM, moving beyond simplistic categorizations of “good” versus “poor” practices. The “Prosperous foundations” profile, characterized by strong preconditions, demonstrates optimal access to resources supporting OHSM. This profile aligns with what Uhrenholdt Madsen et al. (2020) identified in their realist review as organizations with “integrated systems” where OHS management becomes embedded within organizational processes. Similarly, our findings parallel the “blooming profile” identified in recent latent profile analyses of occupational health and well-being strategies (Beauchamp Legault and Chênevert, 2024), where organizational commitment to health and wellbeing is positively associated with employee outcomes.

In contrast, the “Deficient in know-how” profile constitutes a set of poor preconditions with especially limited access to OHSM knowledge and methods, highlighting the multifaceted challenges that some organizations face. This profile shares characteristics with the “wasteland profile” identified in the review by Beauchamp Legault and Chênevert (2024), where employees report higher levels of absenteeism, emotional exhaustion and lower job satisfaction. In their systematic review of safety management practices, Tawfeeq et al. (2024), mapped such workplaces as organizations overlooking “human safety elements”.

The largest group, the “Balanced but limited” profile, exhibits moderate preconditions, suggesting that many organizations possess some foundations for OHSM but may have room for improvement. This middle-ground position corresponds to what has been previously identified among Swedish organizations, where basic compliance exists but strategic integration of safety practices remains underdeveloped (Andersen et al., 2019; Frick, 2014). This profile shares some characteristics with the “Beginner” level described in recent maturity models of OHS (Kusma et al., 2024), where organizations have initiated implementation but have not yet achieved full integration.

In the “Lacking investments” profile, we discovered organizations with acceptable working practices but simultaneously other preconditions at poor levels, especially budget and leadership support for investment in OHSM with an acceptable OHSM performance. This profile illuminates a pattern not extensively documented in previous research: the disconnect between operational practices and strategic commitment. Recent research by Uhrenholdt Madsen et al. (2020) provides a possible explanation for this phenomenon, suggesting that formal systems may exist without substantive implementation when leadership commitment is lacking, similar to our “Lacking investments” profile according to above.

As seen above, similar profile has also been identified in other contexts and countries with other legal frameworks indicating that these findings may also be relevant outside the Swedish context. Also, the need for sufficient preconditions for OHSM have been stressed in a resent systematic review in the European context (Ståhl et al., 2025).

The association of the above-outlined profiles with different sectors, industries and organizational sizes, particularly the overrepresentation of public sector organizations in profiles with poorer preconditions, points to systemic differences that may require tailored interventions. Furthermore, the observed variations across profiles in perceptions of OHSM’s role, the balance of preventive and promotive work and the extent of internal participation in setting OHSM activities indicate that these preconditions may have implications for an organization’s overall approach to and implementation of OHSM. Below, we elaborate on the contributions derived from these findings.

This study makes several important implications. First, by identifying distinct profiles of organizational preconditions for OHSM, our research provides a more nuanced understanding of how the levels of various factors coincide to create different opportunities for workplaces to manage the work environment. This approach, thus, goes beyond the traditional binary classification of organizations as having either “good” or “poor” OHSM and aligns with recent theoretical developments that emphasize the multi-dimensional nature of OHSM (Tremblay and Badri, 2018). Consequently, although the levels of all five preconditions correlated positively with each other, LPA allowed us to detect uniqueness across two groups of organizations with poor preconditions. In organizations in the “Deficient in know-how” profile, all preconditions were low and especially lacking appropriate working methods and knowledge, whereas in the “Lacking investments” profile the working methods were available while the other preconditions were poor, especially leadership support for investment in OHSM. This means that organizations with seemingly similar overall poor OHSM precondition levels may actually face distinct challenges and require tailored interventions to improve their OHSM, as a one-size-fits-all approach may be ineffective. Specifically, in the “Deficient in know-how” profile, interventions should prioritize developing evidence-based OHSM methods, as well as investing in training and education to build the necessary knowledge base for those responsible for OHSM. For organizations in the “Lacking investments” profile, it seems especially valuable to focus on engaging management and educating them about the importance of investing in OHSM. This differentiated approach reflects recent literature suggesting that effective OHS implementation requires context-specific strategies (von Thiele Schwarz et al., 2021).

Second, we indicate potential associations between organizational preconditions and the performance of the OHSM, such as the proportion of preventive and promotive work or the degree of participation. Our findings join earlier research demonstrating how organizational contexts shape the scope and quality of OHSM implementation (Lundqvist et al., 2024; Ståhl et al., 2025). We extend this body of literature by demonstrating differences (albeit small ones) in OHSM between the two profiles with poorer preconditions: “Deficient in know-how” versus “Lacking investments”. We revealed that organizations in the “Deficient in know-how” profile had the lowest percentage of promotive and preventive work and lowest participation of internal experts in creating the content of such initiatives. These findings imply that for this profile, even relatively better leadership support for investment in OHSM initiatives (as compared to the “Lacking investments” profile) cannot substitute for the lack of relevant knowledge and appropriate working methods for OHSM. This pattern suggests the importance of investing in organizational capabilities (Rydell et al., 2019), which could be the difference-making factor in implementing OHSM initiatives.

Third, to the best of our knowledge, this research is among the first to indicate an association between more favorable OHSM preconditions and employee participation in developing the content of OHSM initiatives. In this study, we not only investigated the extent of such participation, but also examined the involvement of different roles. While there was no difference across the four profiles regarding the role of external support from occupational health services, higher involvement in the content of OHS initiatives was more generally seen from internal expert resources, safety delegates or union representatives and employees in organizations in the two profiles with favorable preconditions (“Prosperous foundations” and “Balanced but limited”). Additionally, organizations belonging to the “Prosperous foundations” profile were found to have greater dialogue with employees about the content of OHS initiatives. This pattern may suggest a culture where OHS is prioritized, necessary resources are provided, and the importance of frontline perspectives is recognized, thereby encouraging open dialogue with employees and strengthening the safety climate. Alternatively, it could indicate a positive reinforcement cycle: direct participation leads to better mapping of required OHS initiatives and an increased organizational learning, which may require improvements in the organizational preconditions.

Finally, for policymakers, our findings regarding the overrepresentation of public sector organizations in profiles with poorer preconditions suggest the need for sector-specific policy frameworks, aligning with recent International Labor Organization guidelines emphasizing tailored approaches based on organizational characteristics (ILO, 2025). The observed pattern most probably contributes to a higher level of perceived stress and psychosocial problems among public sector employees than those employed in the private sector (Dahler-Larsen et al., 2020). The differences in preconditions observed between public and private sectors suggest a need for systemic changes in how OHSM is perceived and prioritized within public institutions. Additionally, given the resource-constraints in the public sector, our findings can be used to understand which preconditions should be prioritized. For instance, focusing on strengthening knowledge and methods by forming partnerships with academia (O’Toole, 2015) may help public sector entities overcome barriers to accessing evidence-based recommendations and practices. Ultimately, these findings underscore the need for tailored approaches/profile-specific training to OHSM that account for the distinct characteristics and challenges of the public and private sectors.

The strength of this paper is the use of LPA, which introduced a case-centered approach to understanding organizational preconditions for OHSM, contributing to more comprehensive knowledge regarding factors linked to OHSM perceptions and practice. This approach allowed us to move beyond examining the impact of individual factors in isolation and offered a more comprehensive view of organizational preconditions for OHSM by looking at clusters of preconditions. Applying a quantitative questionnaire to investigate the employers’ preventive OHSM also enabled us to include more cases than would have been possible with a qualitative case study. Another strength is the use of a specifically developed, and tested questionnaire for this project. Although we have used self-reports from one person about an organization to understand the preconditions and the outcomes, additional analyses showed that whenever both employer- and employee representatives answered the survey, their reports were highly correlated (Akerstrom et al., 2024b), which testifies to the validity and quality of the collected data.

This research is not without limitations. The relatively small sample size affected the size of some of the profiles, and this may limit the generalizability of the findings. Also, the vast majority of the participating workplaces were found within the public sector, reflecting the true distribution within Sweden, which also affects the findings when comparing between sectors. Additional profiles could have been revealed with a larger and more diverse sample. Furthermore, the cross-sectional nature of the study precludes causal inferences about the relationships between the studied preconditions and the designated outcomes. To further develop knowledge in this areas, future research could employ longitudinal designs to examine how organizational preconditions for OHSM evolve over time (e.g. as a result of investments or interventions) and how these changes impact OHSM. Finally, the outcomes related to the percentage of organizational-level initiatives or preventive/promotive work across all OHSM initiatives come from self-report. Future investigations should strive to incorporate organizational register data to be able to objectively assess the impact of the preconditions on actual OHSM work.

In conclusion, this study provides insights into the complex interplay between organizational preconditions and their influence on the quality of OHSM. By identifying distinct profiles of preconditions, our research highlights the importance of profile-specific interventions to increase the preconditions for OHSM and especially investing in organizational OHSM knowledge and appropriate working methods. The absence of these elements is associated with lower quality OHS initiatives.

Participants gave their informed consent before answering the survey.

We would like to thank Jonathan Severin, Annemarie Hultberg and Henrik Eklund for their contributions to the data collection. We would also like to thank AFA Insurance for funding the research.