Association between workload and support utilisation - A longitudinal study on emergency medical service personnel

doi:10.21203/rs.3.rs-8279546/v1

Download PDF

Research Article

Association between workload and support utilisation - A longitudinal study on emergency medical service personnel

https://doi.org/10.21203/rs.3.rs-8279546/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background

Social support is a key protective factor against mental illness. Yet, qualitative studies have suggested that emergency medical service personnel working in ambulance contexts underutilise workplace-provided social support. Although a possible barrier to seeking support, no previous quantitative study has examined workload’s association with support utilisation in high-strain emergency settings. This study assesses the longitudinal association between emergency medical service personnel’s workload and utilisation of workplace-provided formal support in an ambulance context.

Methods

341 emergency medical service personnel responded to quarterly surveys on support utilisation. The survey data was combined with organisational records of emergency responses, providing a measure of workload at a station level. Logistic mixed models were performed to assess associations between workload, measured by number of emergency responses, and formal support utilisation.

Results

Emergency medical service personnel at higher-workload stations were more likely to use support than those at lower-workload stations (OR = 2.03, CI = 1.26; 3.35). This effect persisted after adjusting for confounders (OR = 1.93, CI = 1.12; 3.35).

Conclusion

This is the first study examining emergency medical service personnel’s workload and social support utilisation in a longitudinal design. Contrasting qualitative findings, this study presents evidence of greater support utilisation under higher workload. This could reflect a need to conserve resources under pressure, and underlines the importance of accessible support in high-risk occupations.

Social support

workload

post-traumatic stress

emergency medical service personnel

help seeking

first responders

occupational stress prevention

Operational stressors in emergency medical service work

First responder work is associated with a high level of psychological distress due to exposure to critical incidents at work (1). In particular, emergency medical service (EMS) personnel have been found to respond to more emergency calls than police officers and firefighters combined (2). Furthermore, they may also be subject to greater risk of mental health problems compared to other first responders (3, 4). EMS work includes several risk factors of psychosocial stress. The daily operational work includes risk of encountering critical incidents, defined as overwhelming operational tasks that interferes with individual functioning, including both high and low intensity situations (5). However, EMS personnel’s mental health risk is not only constituted by the exposure to critical incidents. Workload in itself is similarly regarded as an isolated stressor for emergency services, contributing to impaired mental health (6–8), and may even be a greater stressor than the nature of the work in itself (7). Responding to more calls has been suggested to potentially increase the level of distress for the worker (8). This has been elaborated by findings of a Swedish cross-sectional study of quantitative design. Here it was suggested that EMS personnel potentially experience persistent stress symptoms due to frequent work exposure to potentially traumatic incidents (9).

Managing the negative psychological effects of the daily work of EMS personnel has been addressed from multiple angles. The issue of ongoing exposure to operational work has been proposed to be potentially detrimental to coping with the exposure to critical incidents, highlighting the need for downtime (7, 10, 11). However, it may be difficult to secure downtime in modern ambulance organisations, due to the challenges in regards to recruitment of a sufficient workforce and lack of resources (7, 12). Not only do these challenges reduce the possibility of downtime, but they can also result in workload intensification and a worsening of mental health in the EMS workforce.

Social support as a preventive measure against stressors

Social support has long been regarded as an adaptive strategy of managing work stressors (13–15). The phenomenon is defined both as the exchange of resources between at least two people with a perception of an intention to enhance the recipient's well-being (16) and as a signal of caring for the recipient, hereunder signalling being a member of a beneficial network (17). Social support may therefore foster social cohesion and act as a protective factor against mental health problems. Further, social support has been presented in central theories as facilitating psychological and social functioning by fulfilling the need for autonomy, competence and relatedness (18, 19), buffering the association between stressor and stress (20, 21), and as an action performed in order to preserve one’s resources, i.e. mental well-being or work capacity (13, 22). The protective effect of social support has been investigated in numerous studies of different theoretical framework, though with mixed results (23). Empirical studies of social support have however produced mixed findings, with limited evidence for its buffering role between stressors and strain. Explanations have pointed to conceptual mismatches, for instance assessing support from individuals who lack influence over the stressor such as the type of demand not aligning with the source of support, or to measurement issues (23). Beyond stress, social support has also been linked to post-traumatic stress disorder (PTSD). A recent meta-analysis indicated a reciprocal relationship between social support and PTSD symptoms (24). Yet, a re-analysis questioned this conclusion, suggesting that much of the observed association may be spurious (25).

This study’s theoretical framework is primarily based on COR, as it nuances the motivation and intention behind engaging in social support - an issue central to the present investigation. In professional groups, PTSD are often characterized by more prolonged developmental trajectories rather than sudden onset of full-syndrome disorders (26). This implies that individuals may attempt to manage emerging symptomatology through coping strategies, which further underscores the relevance of the Conservation of Resources framework for this study. In COR, resources are defined broadly as objects, personal characteristics, conditions, or energies that are valued in themselves or serve to protect other valued resources (27). In the workplace, such resources might include emotional energy or resilience needed to sustain satisfactory work performance. The theory proposes that resource loss or the threat of loss elicits stress, whereas access to resources buffers against further loss (27). Thus, COR stresses help-seeking behaviour as a self-preserving action when facing stressors (13, 22).

In sum, social support is conceptualized as both an individual coping resource and an interpersonal act that strengthens social bonds at work. Nevertheless, its protective effects have proven inconsistent across theoretical frameworks and outcomes, underscoring the need for further examination in occupational contexts.

Social support in ambulance settings

Quantitative studies of EMS personnel have primarily examined social support using measures of perceived support. Such studies show that higher levels of perceived social support are associated with lower work stress, PTSD, and depression (8, 28, 29). However, perceived support measures do not capture whether workers actually use social support, nor from which sources. This distinction matters because the protective effect of social support is not uniform. According to the “matching hypothesis”, support is most effective when the type of support matches the specific demands or stressors faced (30, 31). Understanding which types of support are used in practice is thus essential to evaluate the role of support.

The actual utilisation of social support among EMS personnel have been investigated further through qualitative studies (7, 10, 11). Here, it has been suggested that EMS personnel generally prefer informal, ad hoc sources of support at work (i.e. talking with a colleague or manager at the station), over more formal sources (i.e. peer-support programmes, sessions with a crisis psychologist, or psychological debriefing) (7, 10). Workplace-provided formal support has further been highlighted as important, but less accessible and visible to the workers (11, 32).

Formal support services can be provided by the workplace as a strategic measure to ensure the prevention of mental illness due to work factors, e.g. by offering debriefing after particularly traumatic events. Such initiatives usually focus directly on enhancing the individual’s resources, provided by a person with certain qualifications (33, 34), whereas informal support also can be used due to personal relations and micro-cultures as well as a personal need to manage stressors. From a practical perspective, it is relevant to consider how formal and informal support are accessed and experienced in everyday work life. Formal support sources could be more time-consuming than, e.g. talking informally with a colleague, and require setting a meeting. Further, a preliminary conversation with a manager, referring the employee to the support source, could also be required to access these services. In order to seek formal support, and thus more specialised support than a brief conversation with a colleague, it might require that the employee or a co-worker realises that the current issues or mental state appear to require more time or qualification to be treated. Identifying barriers to the utilisation of formal social support initiatives is therefore important in order to enhance alleviation of symptoms of mental illness due to work exposure.

Workload as a barrier to using formal support

Although the benefits of social support appear well-established, qualitative studies of EMS personnel indicate organisational, cultural, and individual barriers to using social support at work, particularly formalised support initiatives (7, 10, 11). Lack of time and uncertainty about where and how to access help have been identified as central barriers to support utilisation (11, 35), especially regarding formalised types (7, 10). Consequently, EMS personnel may rely more frequently on ad hoc support from trusted informal relations, partly due to accessibility (10 2025). Importantly, such barriers may be particularly relevant in contexts with high workload, where time prioritisation is critical. Theoretically, it has been suggested that stressful working conditions may undermine the availability of resources, including social support in the work environment (36). This dynamic has been observed empirically in the context of ambulance work, as Rikkers and Lawrence (35) found that more than one third of EMS personnel reported difficulty getting time off work as a barrier to help-seeking, a rate higher than in other emergency professions. In addition, 27.5% indicated that seeking help would disrupt their operational work, further underscoring workload and time pressure as barriers. Thus, utilisation of social support is not only a matter of willingness but also of opportunity: high job demands may limit workers’ ability to engage with support. This raises the question of whether working under higher workload in ambulance services predicts the actual utilisation of social support.

To our knowledge, no quantitative studies have examined the longitudinal effects of workload on workers’ utilisation of formal support at work. Establishing such evidence is pivotal to improve understanding of key organisational factors that may prevent EMS personnel from accessing support. Therefore, the present study investigates the effects of workload on the usage of formal social support at work over the course of one year. We hypothesise that workload will be negatively associated with the utilisation of formal support, with EMS personnel at low-workload stations reporting higher levels of support utilisation than those at high-workload stations.

Participants and procedures

This longitudinal survey study is part of the cohort study You Don’t Stand Alone (37), investigating critical incidents and development of PTSD in a large, public Danish ambulance organisation over a three-year period. We have previously published a cross-sectional study based on the baseline data from the DUSA cohort, which examined individual-level associations between social support utilisation and PTSD symptoms (38). The present study enhances our knowledge on support utilisation in ambulance workers by linking longitudinal organisational data of workload with five waves of longitudinal measures of support utilisation to investigate how workload affects help-seeking over time.

The organisation represents 21.5 percent of all 3271 Danish EMS personnel employed in ambulance services (39), including ambulance rescuer students.

All operational EMSpersonnel (N = 703) in the ambulance organisation were invited to the baseline survey. The sample consists of EMS personnel in operational ambulance duty, including emergency service transport personnel, ambulance rescuer students, ambulance rescuer assistants, ambulance rescuers, and paramedics. The respondents are employed at a station in one of the seven areas in the organisation, apart from a minority of workers employed across stations. Managers were not included as operational EMS personnel, as their engagement in operational duty is lower than that of the employees, and as their utilisation of support was expected to differ from that of the employees.

While all employees were invited to participate at baseline, the survey remained open to new respondents at later waves in order to capture employees who had not responded initially or who had joined the organisation after baseline. Employees who responded to the baseline survey were invited to the 3-, 6-, 9-, and 12-month follow-up surveys, as well as new responders were invited to subsequent follow-up surveys. Inclusion required participation in a minimum of two survey waves.

A total of 462 employees participated at baseline. Of these, 11 were registered as managers and were excluded from the study, yielding a baseline sample of 451 respondents. The analytic sample included 341 participants who provided repeated responses on the outcome measure. This group consisted of 334 individuals enrolled at baseline and 7 who joined the study after baseline but met the same inclusion criteria. At the time of the 12-month follow-up, 642 were employed in operational duty in the organisation, yielding a response rate of 53.2%. Study data were collected and managed using REDCap electronic data capture tools (40, 41) hosted at Region of Southern Denmark.

(Insert Fig. 1)

(Figure titel) Fig. 1: Overview of data collection

Ethics

Participants were informed of the purpose and nature of the survey through an online information sheet, and participation was based on written consent. The project complies with GDPR requirements (the Danish Data Protection Authority, # 20/47381). The study was presented to the Scientific Ethics Committee, which received the formal response that, according to Danish law, the study was not subject to approval by the committee (# 20222000-78).

Measurements

Outcome variable: Social support utilisation

Social support utilisation was the main outcome measured with a modified version of the General Help Seeking Questionnaire (42) targeted towards the context of ambulance work. For this study, we focused on the formalised support types provided by the workplace (formal collegial support by a colleague trained in providing support, formal support by a manager trained in providing support, crisis psychologist through work, debriefing/defusing). Debriefing and defusing is initialised automatically in the ambulance organisation when pre-defined categories of critical incidents are encountered. The management informed that it was customary for all invited workers to participate. The utilisation of debriefing and defusing thus differed from the other formal support types and did not reflect help seeking behaviour in the same sense. We therefore excluded measure of debriefing and defusing from our analyses. Social support utilisation was measured quarterly as depicted in Fig. 1. Answers were given as yes/no for each type of support.

Explanatory variables

Measure of exposure: Workload

Workload was the main explanatory variable of our analyses. The measure of workload was constructed as an index based on organisational data of all emergency responses from three months before the baseline survey and up to the date of the 12-month follow-up survey. The data reflected every single emergency response at the level of station. Data was categorised by date and time. Emergency response data was also coded in relation to response type: acute car, acute medical car, day shift ambulance, 24-hour shift ambulance, effective ambulance, and medical transportation. In Denmark, emergency responses are categorised into A-, B-, or C-responses, indicating the urgency and fatality of the incident, the ambulance is called out to. Because our hypothesis was focused on the overall workload, we chose to include all responses into a sum score for each time point, divided by the number of days for each time period as well as divided by number of EMS workers at the station. This operationalisation reflects an index approximating the workload associated with being employed at the particular station, ranging from 0.10 to 0.38 at the baseline level.

Adjustment variables

Based on extant research, we included several adjustment variables to assess the robustness of the associations between workload and support utilisation. We included age at baseline (measured as whole years) and gender (measured as male, female, or other gender orientation), because both age and gender have been shown to predict support utilisation (43, 44). Post-traumatic stress symptoms (PTSS) and perceived social capital were also included in the analyses. PTSS was included because it have been proposed to affect individual social behaviour patterns including utilisation of support both among EMS personnel and trauma-exposed individuals in general (45, 46). PTSS was measured using the validated Danish version of the International Trauma Questionnaire (ITQ), 6-item version, rated on a five-point Likert scale from “Not at all” (0) to “Extremely” (4) (47, 48). The scale measures PTSS during the past month with two items for each of the three symptom clusters of PTSD: re-experiencing, avoidance, and hypervigilance (47). The scale has shown both good construct validity (49) and criterion validity in different trauma populations (50). Compared to the DSM-5, it has been found to produce statistically significantly lower diagnostic rates (51), which is considered relevant to reduce the risk of over-reporting in a non-clinical population. ITQ has been recommended specifically for assessing PTSS among EMS personnel due to the construct and phrasing of symptoms that resonates with the population’s work exposure (52). The overall symptom level of PTSS was assessed by summing the six items into a sum scale from 0–24, and thus coded as a continuous variable. PTSS was measured as a time variant variable at all five survey waves. The scale showed acceptable to good internal consistency with Cronbach’s Alpha ranging from 0.76 to 0.83 across time points.

Social capital was included in the analyses because the phenomenon can be expected to affect the accessibility of social support sources at work by facilitating co-operation and relational support (53). Individual differences in perceived social capital could thus affect the inclination to rely on help from one’s workplace when distressed. Social capital was measured with the Copenhagen Questionnaire of Workplace Social Capital (54). The full scale measures the experience of respect, justice and the ability to collaborate across different subdimensions of the organisation, i.e. leadership and workers (three items) and groups of workers (four items). As EMS personnel usually work closely in pairs, we added a subscale focusing on the social capital of this partnership. This was done by adapting the wording from the co-worker social capital scale to target the partnership (four items). For all three subscales, the items were answered on a Likert scale from 0 = “never” to 4 = “always”. The scales were summed into a continuous variable, according to manual, as a measure of the overall social capital of the daily workplace. The overall social capital scale was measured at baseline and treated as a time-constant variable. Cronbach’s alpha of the scale showed good internal consistency (α = .88).

Statistical methods

Preliminary analyses and assumption checks

Missing analyses were performed with Little’s MCAR test. Visual inspection of QQ-plots and residual plots was used to assess assumptions of normality and homoscedasticity in the linear models applied for further analyses. Kolmogorov-Smirnov test was used to test distribution of samples. Residual diagnostics, including influential cases of outliers and uniformity of residuals, was conducted using the DHARMa (55) and the Performance package (56). Risk of multicollinearity was assessed by calculation of variance inflation factor in multivariable regression.

Latent class analysis was conducted on three binary indicators of formal support utilisation at baseline (peer support, leadership support, and psychological crisis support). The purpose was to explore whether the indicators formed distinct latent classes of utilisation, thereby identifying natural patterns in the data, or whether they could be treated as a single outcome in subsequent analyses. Models with one to six classes were estimated.

Preliminary analyses of the predictor variable examined the number of emergency responses, both overall and within different categories (A-, B-, and C) at each time point. This was done to investigate potential variations in emergency response data across stations and over time, as well as the data's suitability for primary analyses.

Estimation of whether participation was influenced by workload the year before baseline was performed with simple linear regression. This was done using linear regression with workload as the explanatory variable of the participation percentage at each station.

We assessed attrition on the outcome variable (formal social support utilisation) by testing whether key baseline characteristics predicted missingness using univariable logistic regressions. This was defined as having responded to outcome measures at baseline, but not at any of the follow-up surveys. Workload, age, gender, social capital, and PTSS at baseline were entered as predictors of dropout.

Main analyses

Descriptive statistics and percentage analyses were performed to generate frequencies, means, and standard deviations. These statistics were calculated for the entire sample and for subsamples of EMS personnel depending on their status on workload (low/high). To compare differences between stations experiencing higher versus lower levels of workload, the workload variable was dichotomised using a median split (median = 0.19) for each of the measurement periods. Stations with values above the median were coded as 1 (high workload), and those below or equal to the median as 0 (low workload), with the latter serving as the reference group. This approach allowed station-level workload status to vary over time, reflecting changes in response activity, while maintaining a fixed threshold for classification based on the overall distribution of workload across the study period. This ensured consistent interpretation of what constituted "high pressure", while capturing temporal variation in exposure.

Main analyses were performed with generalised linear mixed models with a logit link function and binomial distribution, estimated via Laplace approximation, using the lme4 package (57). This approach allows for modelling of a binary outcome while accounting for random effects, capturing the hierarchical structure of the data, based on maximum likelihood estimation. Support utilisation at each time point was entered as the binary outcome. Workload for each quarterly period leading up to each survey wave was set as explanatory variable to assess the longitudinal associations between periods of workload and probability of using formal support at work. We did not include lagged effects (e.g. workload predicting support utilisation at later survey waves), as this was beyond the scope of the present study. A random effect term with the respondent-ID as subject was included to allow different intercepts for each respondent (58).

The analysis was conducted in two steps. At step one, only workload was entered as explanatory variable. At step two, the model was adjusted for effects of age, gender, and workplace social capital at baseline as time-constant variables, as well as PTSS as a time variant adjustment factor. Time was also entered at step two to assess the effect of time. Apart from the latent class analyses, which were performed using M-plus (version 8.11), all analyses were performed in R-studio (version 4.4.2). ChatGPT-5 was used in order to assist with initial coding of data.

Preliminary analyses and assumption checks

The rate of missing data was estimated at 8.1%. Although it is difficult to establish a true indication of Missing Completely at Random data, the statistically non-significant results of Little’s MCAR test indicated data missing at random (Chi2 = 1185.3 (df = 1236), p = .846). This supports our use of mixed models and thereby decreasing the risk of possible bias inflation due to missing data.

Data visualisation with QQ- and residual plots, as well as results of the KS-test, indicated normal distribution of data, and apparent fitness for the chosen analyses. The outlier test did not identify statistically significantly potentially influential outliers. Manual assessment did not indicate response errors. VIFs ranged from 1.00–1.07, suggesting no multicollinearity.

The results of the LCA showed that fit indices improved from the 1-class to the 2-class model (BIC decreased from 741.1 to 731.8; entropy = 0.894), and both the Lo-Mendell-Rubin Adjusted Likelihood Ratio Test (LRT) and Bootstrapped LRT were statistically significant (p < .001), supporting the 2-class solution (see Additional files 1–3). However, the 3-, 4-, 5-, and 6-class models did not yield meaningful improvements in log-likelihood or fit and were not supported by the LRTs (see Additional files 2, 4–7). Critically, the 2-class model showed strong class imbalance (approximately 6% vs. 94%), and one of the indicators exhibited a boundary threshold (probability = 0 or 1), indicating potential overfitting. The classes primarily distinguished between general utilisation and non-utilisation of support, rather than distinct patterns of support-seeking. As such, the latent classes lacked conceptual richness, indicating that the support types could be treated as a single variable without risking overlooking substantially different patterns of support utilisation of the sample. As the distributions of the single support types were heavily uneven, and as our main interest was in formal support utilisation in general, we opted to construct a composite binary outcome. The three items were first summed and then dichotomised into a variable indicating 0 = no utilisation of formal support and 1 = utilisation of at least one form of formal support at each of the five time points. This allowed us to investigate potential differences in using versus not using formal support.

Preliminary analyses of patterns of the number of emergency responses showed generally stable trends across stations, with a slight increase in the most critical responses and a decrease in the less severe responses. There was a minor decline across stations in responses from 2022 to 2023, but otherwise, the number of responses per station remained relatively stable, indicating that the index was a suitable proxy measure for workload within the time period of the study.

Analysis of whether the explanatory variable influenced participation in the survey showed that workload the year before baseline had a statistically significant effect on participation in the survey (B = 0.26, std. B = 0.42, p < .05). This indicates a higher tendency to participate in the survey among the respondents working at stations with a higher workload.

Attrition analysis showed that none of the examined predictors (age, gender, social capital, pre-baseline workload, baseline PTSS) were statistically significantly associated with non-response on the outcome. Odds ratios ranged from 0.98 to 1.29 (all p > .12).

Descriptive results

[insert Table 1]

Table 1 shows descriptive results of the sample. The sample was mainly consisting of men (82%), the mean age was 42.1 years, and the mean seniority was 16.6 years. Generally, formal social support was more commonly used by workers at busier stations. The mean index of workload was 0.18-19 (range at baseline = 0.10–0.38) for all respondents across all time points. The average of PTSS at baseline was relatively low with a mean of 2.99 (SD = 3.15). This was slightly higher for respondents belonging to the group of higher workload (Baseline mean = 3.07 (SD = 3.16)) than for those with lower workload (Baseline mean = 2.89 (SD = 3.15)).

Main results

[insert Table 2]

Table 2
Results from main analyses Presenting results from generalised mixed models of binomial distribution with formal collegial support as outcome. The simple model includes only the predictor of interest. The multiple model adjusts for covariates. Random intercepts account for repeated observations within individuals. Presented as Odds Ratios (OR), coefficients, standard errors (Std. error), 95% confidence intervals (CI), and significance levels (p). Statistically significant results (p < .05) are flagged with asterisks.
Simple model (observations = 1225, 341 individuals)						Multiple model (observations = 1117, 312 individuals)
Predictor variable	OR	Coefficient	Std. error	95% CI	p	OR	Coefficient	Std. error	95% CI	p
Workload	2.03	0.71	0.26	1.26 ; 3.35	.006 **	1.93	0.66	0.28	1.12 ; 3.35	.019 *
Age						1.00	-0.00	0.02	0.97 ; 1.03	.912
Gender						1.88	0.64	0.40	0.86 ; 4.16	.114
Workplace social capital						1.01	0.01	0.02	0.96 ; 1.05	.790
PTSS						1.14	0.13	0.04	1.06 ; 1.23	.001 ***
Time effect						0.76	-0.27	0.07	0.67 ; 0.88	.000 ***

Working at a high-workload station statistically significantly increased the likelihood of using formal support (OR = 2.03, p < .01, AIC = 1068.0, BIC = 1083.3; ICC = 0.50). The effect remained statistically significant but attenuated after adjustment (OR = 1.93, p < .05). The association was small-to-moderate by Chen et al.’s (59) definition. Model fit indices indicated better fit of the adjusted model (AIC = 963.2; BIC = 1003.3; ICC = 0.51).

The findings did not support our hypothesis that support was less utilised by the busiest workers. Age, gender, and social capital showed no statistically significant effects. PTSS had a statistically significant but small positive effect. Time indicated a trend towards decreasing support utilisation over the study period.

This study found that EMS personnel operating under the highest levels of workload were more likely to utilise formalised support initiatives provided at work. Contrary to previous research portraying workload as a barrier to support-seeking (7, 10, 11), our results suggest that high job demands may act as an incentive rather than an impediment under certain conditions. This finding challenges the prevailing view on support utilisation among EMS personnel, and underscores the need to nuance our theoretical and practical understanding of how workload shapes help-seeking behaviour in high-risk occupations.

Adaptive responses and resource mobilisation under higher workload

Drawing on the COR theory, our findings can be interpreted as evidence of adaptive coping under strain. COR emphasises that individuals strive to protect and replenish their personal resources when confronted with high demands. Within this framework, help-seeking may be seen as a resource mobilisation strategy that is activated under pressure in order to prevent further depletion. Rather than workload simply depleting resources, it may also serve as a trigger for mobilising external ones. In this sense, our study illustrates a particular aspect of COR, namely that high demands can act not only as a barrier, but also as an incentive, to engage external resources. The utilisation of formalised support may, in addition to reflect distress, serve as a strategy to conserve capacity and maintain function. Hence, individuals are more likely to mobilise external resources under pressure in order to avoid further depletion (13). Thus, our findings suggest that workers may turn to available support systems not in spite of a high workload, but because of it. To our knowledge, no previous studies have examined workload as a predictor of employees’ utilisation of support, making this study the first to address this question in an ambulance service context.

The practical and relational conditions of working under high workload may influence how and from whom support is sought. One explanation is that workers at busy stations encounter a greater volume of potentially distressing events (7), and insufficient time to reflect or decompress after such events has been shown to increase psychological load (10, 60). This heightened strain may create both the need and the motivation to seek support. Qualitative findings further suggest that support is prioritised when stressors become increasingly difficult to manage alone (61). At the same time, structured support may be particularly valuable in high-workload contexts because it helps maintain professional functioning. In this sense, support can relieve distress while simultaneously reinforcing professional identity and perceived efficacy (62), aligning with COR’s notion of resource conservation. Thus, seeking support may not merely reflect vulnerability, but also agency - a proactive attempt to sustain functional capacity under strain. The increased use of support observed in our study may therefore reflect how workers in emergency settings engage with support during adversity to protect their personal resources.

Beyond this, our findings may also reflect strategic substitution of support sources. Unlike informal peer interactions, formalised support is typically framed within a structured time and a defined relational role of a provider with a certain level of training in crisis support. In line with COR, this may be understood as a substitution of resources: when a source of support is less accessible, other strategies are activated to preserve functioning and contain stress (36). This mechanism could be founded in an intention to spare stressed colleagues, as well as a potential realisation that one’s need for support surpasses what can be carried out in the time between emergency calls. Our findings can be seen through this framework, where workers at busier stations are more likely to seek support with a more pre-defined purpose and setting, in this case the formalised types, whereas this mechanism seem less prevalent at stations with workload. Targeted help-seeking may serve to limit further loss of resources in the immediate environment by redirecting the coping burden from interpersonal relations to institutional structures with the explicit purpose of support. This pattern may reflect strategic substitution of support sources, yet the underlying mechanisms cannot be firmly established from the present data. Future research, including qualitative studies, is needed to examine how emergency personnel make sense of and navigate different forms of support under varying workload conditions.

The use of organisational data in predicting workplace behaviour

The use of objective measures in epidemiological studies is often motivated (63). To our knowledge, no previous study has used organisational data of emergency responses as a predictor of formal support utilisation over time. By linking emergency response volume to self-reported support utilisation, we provide a non-reactive indicator of operational strain and behavioural response.

This study offers a foundation for integrating organisational data into psychosocial research, and suggests that such approaches can complement self-reported data by uncovering patterns otherwise inaccessible. Whereas earlier studies have predominantly relied on retrospective self-reports and qualitative interviews, our study provides prospective, quantitative insights into the relationship between an objective measure of workload and help-seeking behaviour, thereby offering novel evidence to a largely qualitative field. This may have added to the contrast in findings between this study and prior research. The observed pattern suggests that even organisational-level indicators can yield meaningful insights into the conditions that drive help-seeking behaviour.

Practical implications

Our results indicate that utilisation of support increased under higher workload, suggesting that traditional barriers such as stigma or fear of judgment (10, 32, 64) may be less inhibiting in contexts of intensified demand. One possible explanation is that when workload intensifies, the urgency of sustaining competence and effective functioning outweighs concerns about stigma, leading employees to engage with available support despite potential reservations. Another factor that may have contributed to this pattern is the organisational context. The ambulance service in this study has implemented preventive initiatives and invested in mental health literacy, which may have lowered thresholds for seeking help and improved awareness of how and where to access support (65).

Taken together, our findings support the notion that structural availability of support is not sufficient in itself. For support to be effectively utilised, it must also be visible, legitimate, and embedded in organisational culture. This highlights the importance of organisational efforts not only to provide support systems but also to normalise and integrate help-seeking into everyday practice. In practical terms, this may include leadership endorsement of support use, training initiatives that build mental health literacy, and proactive communication about available services. In high-demand contexts, such initiatives could be particularly important, since workload and time pressure can limit the practical possibility of engaging with support.

Limitations

This study has several limitations and alternative interpretations that should be considered.

First, the utilisation of formal support was measured at the individual level via survey data, while workload was measured at the station level, though generated into a proxy of emergency responses per individual. This mismatch in levels may obscure individual variation in emergency load, as individual paramedics at the same station may have experienced different degrees of workload. Although this operationalisation reflects the organisational context and collective workload, it restricts inference of subjective or task-specific workload at the individual level. Linking single employees’ emergency response data to survey data in future studies would enable more granular insights into how objective exposure predicts support utilisation. However, as we found direct effects of workload at a station level on the individual’s likelihood of utilising formal support sources, we do believe it is possible to operationalise workload at the general station level and still finding relevant information of its impact on workplace mental health and preventive initiatives. As the utilisation of formal support is to some extent relying on the workplace’ readiness to provide support, the measure of support utilisation might reflect the capacity of both the individual and its surroundings at work. This is grounded in the fact that engaging in e.g. formal managerial- or collegial support requires both knowledge of the support initiative as well as capacity to prioritise it from both parts.

Moreover, our use of emergency response per employee was based on a fixed employee count per station, not accounting for staff fluctuations over the study period, potentially introducing measurement imprecision. This is expected to affect the results to a lesser extent, as it is customary to cover understaffed shifts across stations.

The intraclass correlation coefficients of our models suggest that a substantial portion of the variance in formal support utilisation is explained by factors beyond workload. Prior research points to organisational culture, perceived stigma, and individual characteristics such as self-efficacy or attachment style as possible determinants of help-seeking (7, 66, 67). These factors were omitted from our analyses in order to avoid overcorrection. Given that support to some extent is granted based on event intensity, it is essential to examine how subjective appraisal, coping self-efficacy, and perceived support effectiveness shape outcomes. Future studies should consider applying more direct and differentiated measures of workplace culture, exposure to critical incidents, perceived stigma, and individual-level dispositions.

As the outcome and adjustment factors of the study relied on self-reported support utilisation, we sought to minimise recall bias by distributing the surveys over shorter timeframes. The longitudinal design with repeated measures every third month and inclusion of objective exposure data help mitigate common method bias and recall bias (63), though residual risk remains.

Our initial analysis showed that workload the year before baseline had a statistically significant effect on participation in the survey. This indicated a higher inclination to participate among the busiest stations, contrary to our expectations. Analysis of attrition did not find indication that dropout of the study was associated with demographics, workload, or mental health factors. Further, the sample represents over 50% of the target organisation at baseline and over 20% of the Danish EMS workforce overall, and was overall comparable to the reference population, thus enhancing generalisability and practical relevance of the study. We therefore believe the results to not reflect the dynamics of employees at stations with lower workload, but rather that they present insights to employee dynamics, even among the busiest workers.

This study challenges the prevailing view that workload deters help-seeking and instead point to a more dynamic coping process in high-demand contexts. These findings empirically support the Conservation of Resources model, highlighting the importance of situational context in interpreting formal support utilisation. In settings with higher workload, help-seeking might not only be understood as a sign of personal vulnerability, but also as a potential functional adaptation to sustain competence and effective functioning; an insight that carries direct implications for workforce mental health strategies.

The study demonstrates the potential value of integrating organisational-level indicators in the study of help-seeking behaviour in emergency services. Future research should further explore the effectiveness, timing, and context of support utilisation, and examine how these factors interact with long-term mental health trajectories in high-risk occupations.

COR = Conservation of Resources

EMS = Emergency medical service

ITQ = International Trauma Questionnaire

LRT = Likelihood Ratio Test

PTSD = Post-traumatic stress disorder

PTSS = Post-traumatic stress symptoms

Ethics approval and consent to participate

The project complies with GDPR requirements (the Danish Data Protection Authority, # 20/47381). The study was presented to the Scientific Ethics Committee, which received the formal response that, according to Danish law, the study was not subject to approval by the committee (# 20222000-78).

Participants were informed of the purpose and nature of the survey through an online information sheet, and participation was based on written consent.

Consent for publication

Not applicable.

Avaibility of data and materials

The datasets used in this paper are not publicly available. The dataset contains sensitive clinical and personal information that might identify individual participants. We do not have the ethics committee’s or our participants’ consent to grant access to the collected data to third parties outside the research project.

Competing Interests

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Funding

This project has been generously funded by the Danish Working Environment Research Fund (#20-2022-03 20225100185). Jesper Pihl-Thingvad is funded by the research fund of Region of Southern Denmark, (#22/8605 efond1398).

Author Contributions

PM-N, MLV, and JP-T contributed to the conceptualisation and design of the study. JP-T collected the data, PM-N analysed the data, and PM-N interpreted the data. PM-N made the first article draft and J-PT, MLV, LPSA, NL and AE made critical revisions to the manuscript. All authors approved the final manuscript, and all authors agree on full accountability of the content of the manuscript.

Acknowledgements

The authors would like to thank Ambulance Syd for their assistance with data collection for the current study.

Additional files

Additional file 1

File format: .docx

Title: Supplementary table 1

Description: Overview of formal support utilisation (measured at baseline)

Additional file 2

File format: .docx

Title: Supplementary table 2

Description: Fit statistics for a 1 through 6 class model for latent class analysis of formal support utilisation at work

Additional file 3

File format: .docx

Title: Supplementary figure 1

Description: Graphic representation of the 2-class structure of formal support utilisation

Additional file 4

File format: .docx

Title: Supplementary figure 2

Description: Graphic representation of the 3-class structure of formal support utilisation

Additional file 5

File format: .docx

Title: Supplementary figure 3

Description: Graphic representation of the 4-class structure of formal support utilisation

Additional file 6

File format: .docx

Title: Supplementary figure 4

Description: Graphic representation of the 5-class structure of formal support utilisation

Additional file 7

File format: .docx

Title: Supplementary figure 5

Description: Graphic representation of the 6-class structure of formal support utilisation

Lewis-Schroeder NF, Kieran K, Murphy BL, Wolff JD, Robinson MA, Kaufman ML. Conceptualization, Assessment, and Treatment of Traumatic Stress in First Responders: A Review of Critical Issues. Harv Rev Psychiatry. 2018;26(4):216–27.
James AE, Wright PL. Occupational stress in the ambulance service. Health Manpow Manage. 1991;17(4):4–11.
Berger W, Coutinho ES, Figueira I, Marques-Portella C, Luz MP, Neylan TC, et al. Rescuers at risk: a systematic review and meta-regression analysis of the worldwide current prevalence and correlates of PTSD in rescue workers. Soc Psychiatry Psychiatr Epidemiol. 2012;47(6):1001–11.
Marmar CR, Weiss DS, Metzler TJ, Ronfeldt HM, Foreman C. Stress Responses of Emergency Services Personnel to the Lorna Prieta Earthquake Interstate 880 Freeway Collapse and Control Traumatic Incidents. J Trauma Stress. 1996;9(1):63–84.
Alexander DA, Klein S. Ambulance personnel and critical incidents: impact of accident and emergency work on mental health and emotional well-being. Br J Psychiatry. 2001;178(1):76–81.
Hart PM, Wearing AJ, Headey B. Assessing police work experiences: Development of the police daily hassles and uplifts scales. J Criminal Justice. 1993;6:553–72.
Lawn S, Roberts L, Willis E, Couzner L, Mohammadi L, Goble E. The effects of emergency medical service work on the psychological, physical, and social well-being of ambulance personnel: a systematic review of qualitative research. BMC Psychiatry. 2020;20(1):348.
Regehr C, Millar D. Situation critical: High demand, low control, and low support in paramedic organizations. Traumatology. 2007;13(1):49–58.
Jonsson A, Segesten K, Mattsson B. Post-traumatic stress among Swedish ambulance personnel. Emerg Med J. 2003;20:79–84.
Auth NM, Booker MJ, Wild J, Riley R. Mental health and help seeking among trauma-exposed emergency service staff: a qualitative evidence synthesis. BMJ Open. 2022;12(2):e047814.
Johnston S, Waite P, Laing J, Rashid L, Wilkins A, Hooper C et al. Why Do Emergency Medical Service Employees (Not) Seek Organizational Help for Mental Health Support? A Systematic Review. Int J Environ Res Public Health. 2025;22(4).
Brooks SK, Dunn R, Amlot R, Greenberg N, Rubin GJ. Social and occupational factors associated with psychological distress and disorder among disaster responders: a systematic review. BMC Psychol. 2016;4:18.
Halbesleben JRB, Neveu J-P, Paustian-Underdahl SC, Westman M. Getting to the COR. J Manag. 2014;40(5):1334–64.
Karasek R, Theorell T. Healthy Work: Stress, Productivity, and the Reconstruction of Working Life. New York: Basic Books Inc; 1990.
Lazarus RS, Folkman S, Stress. Appraisal, and Coping: Springer Publishing Company; 1984.
Schumaker SA, Brownell A. Toward a Theory of Social Support: Closing Conceptual Gaps. J Soc Issues. 1984;40(4):11–36.
Cobb S. Social Support as a Moderator of Life Stress. Psychosom Med. 1976;38(5):300–14.
Ryan RM, Deci EL. Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. Am Psychol. 2000;55:68–78.
Van den Broeck A, Vansteenkiste M, De Witte H, Lens W. Explaining the relationships between job characteristics, burnout, and engagement: The role of basic psychological need satisfaction. Work Stress. 2008;22(3):277–94.
Bakker AB, Demerouti E. The Job Demands-Resources model: State of the art. J Managerial Psychol. 2007;22:309–28.
Demerouti E, Bakker AB. The Job Demands–Resources model: Challenges for future research. SA J Industrial Psychol. 2011;37(2).
Hobfoll SE. Conservation of Resources - A New Attempt at Conceptualizing Stress. Am Psychol. 1989;44(3):513–24.
Jolly PM, Kong DT, Kim KY. Social support at work: An integrative review. J Organizational Behav. 2020;42(2):229–51.
Wang Y, Chung MC, Wang N, Yu X, Kenardy J. Social support and posttraumatic stress disorder: A meta-analysis of longitudinal studies. Clin Psychol Rev. 2021;85:101998.
Sorjonen K, Melin B. Prospective associations between social support and posttraumatic stress disorder may be spurious: A re-meta-analysis of longitudinal studies and a simulation. J Affect Disord. 2023;340:174–80.
Bonde JPE, Jensen JH, Smid GE, Flachs EM, Elklit A, Mors O, et al. Time course of symptoms in posttraumatic stress disorder with delayed expression: A systematic review. Acta Psychiatr Scand. 2022;145(2):116–31.
Hobfoll SE. The Influence of Culture, Community, and the Nested-Self in the Stress Process: Advancing Conservation of Resources Theory. Appl Psychol. 2001;50(3):337–421.
Skogstad M, Skorstad M, Lie A, Conradi HS, Heir T, Weisaeth L. Work-related post-traumatic stress disorder. Occup Med (Lond). 2013;63(3):175–82.
van der Ploug E, Kleber RJ. Acute and chronic job stressors among ambulance personnel: predictors of health symptoms. Occup Environ Med. 2003;60:40–6.
Cohen S, Willis TA, Stress. Social Support, and the Buffering Hypothesis. Psychol Bull. 1985;98(2):310–57.
Cutrona CE, Russell DW. Type of social support and specific stress: Toward a theory of optimal matching. In: Sarason BR, Sarason IG, Pierce GR, editors. Social support: An interactional view. Wiley; 1990. pp. 319–66.
Regehr C, Goldberg G, Hughes J. Exposure to human tragedy, empathy, and trauma in ambulance paramedics. Am J Orthopsychiatry. 2002;72(4):505–13.
Harvey SB, Devilly GJ, Forbes D, Glozier N, McFarlane A, Phillips J, et al. Expert guidelines: Diagnosis and treatment of post-traumatic stress disorder in emergency service workers. the Royal Australian and New Zealand College of Psychiatrists; 2015.
NICE. Mental wellbeing at work. NICE - National Institute for Health and Care Excellence; 2022.
Rikkers W, Lawrence D. Barriers to seeking help for an emotional or mental health condition among Australian emergency services workers. Australasian J Disaster Trauma Stud. 2022;26(1):23–40.
Hobfoll SE. Social and Psychological Resources and Adaptation. Rev Gen Psychol. 2002;6(4):307–24.
Pihl-Thingvad J, Vang ML, Moller SR, Hansen NB. Critical Incidents Scale for Ambulance Work - Denmark (CISAW-D): the development of a screening tool for work exposure to critical events in operative ambulance personnel. Br Paramed J. 2022;7(3):26–33.
Melander P, Vang ML, Lindekilde N, Andersen LPS, Elklit A, Pihl-Thingvad J. Social support utilization's effect on post-traumatic stress symptoms: a Danish cross-sectional study of 383 ambulance personnel. Front Psychiatry. 2024;15:1425254.
Østerlund I. Ambulancekaos landet over: Så mange reddere mangler: FOA; 2023 [Available from: https://www.fagbladetfoa.dk/Artikler/2023/09/22/Ambulancekaos-landet-over-Saa-mange-reddere-mangler
Harris PA, Taylor R, Minor BL, Elliott V, Fernandez M, O'Neal L, et al. The REDCap consortium: Building an international community of software platform partners. J Biomed Inf. 2019;95:103208.
Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inf. 2009;42(2):377–81.
Wilson CJ, Deane FP, Ciarrochi J. Measuring Help-Seeking Intentions: Properties of the General Help-Seeking Questionnaire < Wilson, Deane, Ciarrochi & Rickwood, 2005.pdf>. Can J Counselling. 2005;39:15–28.
Addis ME, Mahalik JR. Men, masculinity, and the contexts of help seeking. Am Psychol. 2003;58(1):5–14.
Liddon L, Kingerlee R, Barry JA. Gender differences in preferences for psychological treatment, coping strategies, and triggers to help-seeking. Br J Clin Psychol. 2018;57(1):42–58.
Rikkers W, Lawrence D. Mental health help-seeking experiences and service use among Australian first responders. Australian J Psychol. 2021;73(2):125–33.
Sheerin C, Berenz EC, Knudsen GP, Reichborn-Kjennerud T, Kendler KS, Aggen SH, et al. A population-based study of help seeking and self-medication among trauma-exposed individuals. Psychol Addict Behav. 2016;30(7):771–7.
Cloitre M, Shevlin M, Brewin CR, Bisson JI, Roberts NP, Maercker A, et al. The International Trauma Questionnaire: development of a self-report measure of ICD-11 PTSD and complex PTSD. Acta Psychiatr Scand. 2018;138(6):536–46.
Vang ML, Dokkedahl SB, Lokkegaard SS, Jakobsen AV, Moller L, Auning-Hansen MA, et al. Validation of ICD-11 PTSD and DSO using the International Trauma Questionnaire in five clinical samples recruited in Denmark. Eur J Psychotraumatol. 2021;12(1):1894806.
Karatzias T, Shevlin M, Fyvie C, Hyland P, Efthymiadou E, Wilson D, et al. An initial psychometric assessment of an ICD-11 based measure of PTSD and complex PTSD (ICD-TQ): Evidence of construct validity. J Anxiety Disord. 2016;44:73–9.
Hansen M, Hyland P, Armour C, Shevlin M, Elklit A. Less is more? Assessing the validity of the ICD-11 model of PTSD across multiple trauma samples. Eur J Psychotraumatol. 2015;6:28766.
Hyland P, Shevlin M, Brewin CR, Cloitre M, Downes AJ, Jumbe S, et al. Validation of post-traumatic stress disorder (PTSD) and complex PTSD using the International Trauma Questionnaire. Acta Psychiatr Scand. 2017;136(3):313–22.
Baqai K. Resilience over recovery: A feasibility study on a selftaught resilience programme for paramedics. University of Central Lancashire; 2020.
Putnam RD. Bowling Alone: America’s Declining Social Capital. In: Crothers L, Lockhart C, editors. Culture and Politics. New York: Palgrave Macmillan; 2000.
Borg V, Mateu N, Clausen T. Udvikling af en ny metode til undersøgelse af social kapital på arbejdspladsen [Developing a new method for assessment of workplace social capital]. Copenhagen, Denmark: The National Research Center for the Working Environment; 2014.
Hartig F. DHARMa: residual diagnostics for hierarchical (multi-level/mixed) regression models https://cran.r-project.org/web/packages/DHARMa/vignettes/DHARMa.html: cran.r-project; 2024 [.
Lüdecke D, Ben-Shachar M, Patil I, Waggoner P, Makowski D. performance: An R Package for Assessment, Comparison and Testing of Statistical Models. J Open Source Softw. 2021;6(60).
Bates D, Mächler M, Bolker B, Walker S. Fitting Linear Mixed-Effects Models Usinglme4. J Stat Softw. 2015;67(1).
Zuur AF, Ieno EN, Walker NJ, Saveliev AA, Smith GM. In: Gail M, Krickeberg K, Samet JM, Wong AT W, editors. Mixed Effects Models and Extensions in Ecology with R. Springer Science + Business Media; 2009.
Chen H, Cohen P, Chen S. How Big is a Big Odds Ratio? Interpreting the Magnitudes of Odds Ratios in Epidemiological Studies. Commun Stat - Simul Comput. 2010;39(4):860–4.
Halpern J, Maunder RG, Schwartz B, Gurevich M. Downtime after critical incidents in emergency medical technicians/paramedics. Biomed Res Int. 2014;2014:483140.
Jones S, Agud K, McSweeney J. Barriers and Facilitators to Seeking Mental Health Care Among First Responders: Removing the Darkness. J Am Psychiatr Nurses Assoc. 2020;26(1):43–54.
Viswesvaran C, Sanchez JI, Fisher J. The role of social support in the process of work stress: A meta-analysis. J Vocat Behav. 1999;54:314–34.
Podsakoff PM, MacKenzie SB, Lee JY, Podsakoff NP. Common method biases in behavioral research: a critical review of the literature and recommended remedies. J Appl Psychol. 2003;88(5):879–903.
Clompus SR, Albarran JW. Exploring the nature of resilience in paramedic practice: A psycho-social study. Int Emerg Nurs. 2016;28:1–7.
Jorm AF. Mental health literacy: empowering the community to take action for better mental health. Am Psychol. 2012;67(3):231–43.
Mikulincer M, Shaver PR. Attachment orientations and emotion regulation. Curr Opin Psychol. 2019;25:6–10.
Vollrath M. Personality and stress. Scand J Psychol. 2001;42(4):335–47.

Table 1 is available in the Supplementary Files section.

No competing interests reported.

Download PDF

Reviewers agreed at journal
05 Jan, 2026
Reviewers invited by journal
10 Dec, 2025
Editor invited by journal
09 Dec, 2025
Editor assigned by journal
05 Dec, 2025
Submission checks completed at journal
05 Dec, 2025
First submitted to journal
04 Dec, 2025

You are reading this latest preprint version

Association between workload and support utilisation - A longitudinal study on emergency medical service personnel

Status:

Version 1

Abstract

Background

Methods

Results

Conclusion

Figures

Background

Methods

Results

Discussion

Conclusion

Abbreviations

Declarations

References

Table 1

Additional Declarations

Supplementary Files

Status:

Version 1