Introduction
Organisational workplace interventions in occupational health can be defined as changes in the working conditions, work processes, organisational policies and procedures, work tasks or the work environment in order to reduce health hazards or improve workers’ well-being [
1,
2]. In contrast, individual-oriented interventions aim to change individual behaviour, attitudes, skills or involve the use of personal equipment [
3]. Participatory interventions are generally based on the idea that participants of the intervention should contribute their expertise to the determination of the intervention contents, design and implementation [
4,
5]. Therefore, it is assumed that participatory organisational interventions are more appropriate for reducing occupational risks at the source and integrating the specific intervention contents into the work routines of the organisation [
6].
However, organisational-level workplace interventions are complex social interventions, i.e., they involve several interacting components at both the individual and the aggregate social level which affect the intervention delivery and contents in an unpredictable manner [
7]. The challenges associated with complex social interventions are not simply related to whether a particular set of intervention activities are effective or not, but to the circumstances under which those activities are planned, carried out and received by the target groups. Therefore, organisational interventions require not only an evidence-based approach which ensures that the planned measures are effective and the attainment of goals for the targeted groups are feasible [
2,
8], but also an appropriate strategy which serves as a guide for the actual implementation of the intervention [
9,
10]. At the same time, the complexity inherent to organisational interventions demands an evaluation of the expected outcomes in terms of effect size estimates and the intervention process itself as well. Whereas the outcome evaluation is usually performed by statistical analysis suitable for a specific study design, the evaluation of the intervention process is much more heterogeneous. Several process variables have been used in previous research to assess the implementation quality of interventions in occupational health including contextual factors, barriers and facilitators, initiation of the intervention, ownership, appropriateness, participation, protocol adherence, communication, management support and readiness for change, among many others [
11,
12].
Health services providers in several European countries are currently confronted with some form of personnel shortage due to several factors including an ageing workforce and the relatively high levels of psychosocial risk and ensuing mental health symptoms including long working hours, job insecurity, burnout and stress in the health care sector [
13]. Work in healthcare is concomitant with increased emotional demands [
14], high cognitive and time demands [
15] and low job rewards [
16]. As a response to this situation in the health-care labour market, some health services providers have been implementing in recent years some form of age management practices to reduce the impact of this personnel shortage [
17]. In a previous study with health care organisations in Germany, the UK and Finland, it was found that the most frequent age management measures concerned reductions of the working time or re-arrangement of work scheduling [
17]. Nonetheless, the findings suggested that health care organisations do not usually attempt to decrease work demands, modify the work environment or adopt a life-course approach with special emphasis on age and career phases or healthcare workers [
17].
From a more general perspective, these type of age management programmes can be interpreted as occupational health interventions focusing on working time arrangements and shifting schedules of healthcare workers. However, there are several research gaps regarding the expected primary outcomes resulting from such programmes. For instance, the mechanisms are not specified whereby the rescheduling of working time arrangements should reduce personnel shortages in healthcare settings. Even though the redesign of shift work schedules may have some beneficial effects on outcomes such as work-life balance and work stress [
18,
19], it is unclear whether previous age management programmes actually targeted such outcomes as antecedents of staff turnover or early retirement intentions. In addition, to the knowledge of the authors, previous age management programmes in healthcare settings have not been evaluated in randomised controlled trials and, therefore, there is a high risk of bias in the corresponding literature.
Hence, the present study contributes to previous research in occupational health interventions in two ways: First, the study presents the results of a participatory organisational intervention which explicitly addressed the age and career phases of healthcare workers to improve their perceived work ability and, hence, increase the chances of longer employment careers of workers. Second, the present study reports not only the effect size estimates of the intervention, it also focuses on the psychosocial load in terms of the Effort-Reward Imbalance (ERI) model and provides a detailed process evaluation which addressed the context and actual delivery of the intervention. The ERI model of work stress assumes that the perception of lack of reciprocity in terms of high efforts and low rewards at work elicit stress reactions [
20]. The ERI model postulates also that failure to withdraw from work obligations, i.e., to be overcommitted to one’s own work duties, represents a health-adverse coping pattern [
20]. Hence, the ERI model is based on three dimensions, namely, efforts, low rewards and overcommitment. Against this background the present study provides additional information which may help to understand the impact of the intervention on work stress and how and why the observed outcomes may have come about [
2,
8,
10]. In addition, considering that the present intervention started before the declaration of the COVID-19 pandemic, it was possible to assess the potential impact of the pandemic on the primary and secondary intervention outcomes. This is important given the fact that healthcare workers may have been more exposed to a stressful work environment during the early stages of the Sars-CoV-2 outbreak [
21].
Discussion
The statistical analysis did not provide support to the hypothesis that the intervention would improve the self-assessed physical and mental work ability of healthcare workers in the intervention wards. There was also no evidence that the intervention was beneficial for reducing the psychosocial load at work in terms of the efforts, rewards and overcommitment perceived by the workers receiving the intervention. Furthermore, the results obtained from the process evaluation did not suggest that the implemented measures were specifically addressing the self-assessed physical and mental work ability. Since the intervention lacked an evidence-based approach to organisational change, the actual delivery of the measures was not embedded in a systematic plan to monitor and ensure the implementation of measures. For the majority of healthcare workers (about 70%), the intervention study did not have any impact whatsoever on their working conditions. This finding indicates that the main pillar of the intervention, i.e., the organisation and moderation of single workshops with selected workers and supervisors in the wards, was not conducive to the pursued improvements of the self-assessed physical and mental work ability. Even though the disruptions related to COVID-19 resulted in few workers being transferred to the intensive care units, the findings suggested that that ward transfers as such had a negative impact on the self-assessed work ability, independently of the cause behind the transfer (Table
2). Although few wards reported that the COVID-19 related disruptions hindered or delayed the implementation of some measures, these was no indication that these disruptions prevented the delivery of the proposed measures in all wards and institutions.
The failure of the intervention to attain the proposed goals can be explained by taking into account some theoretical considerations pertaining to the conduction of complex social interventions. It seems that the main deficiency of the present intervention was the lack of an evidence-based set of statements providing the rationale of why certain intervention measures may be causally related to the expected outcomes. The intervention was rather vague concerning the specific causal mechanisms which were thought to lead to the expected outcomes. The recurrence to the work ability concept and the definition of specific primary endpoints were indeed explicit, but there was no decision guide for the relevant actors (workers, supervisors, management) as to the type of measures which would result in improvements of the self-assessed work ability of healthcare workers. In particular, the analysis of the single intervention measures revealed that the healthcare workers participating in the workshops did not have a clear idea of what an intervention implies, namely, to change some aspects of the working environment [
32,
33]. Most “measures” could not be related to specific changes of the working conditions, but were rather complaints, individual requests, anecdotes or issues being debated during the workshops (about 46% of all “measures”, Table
5). Even though one of the strengths of the intervention was the participatory approach, there was not systematic approach of how to select the most effective measures leading to improvements of the self-assessed work ability. As a matter of fact, the analysis suggested that some wards may have even experienced a worsening of the working conditions (Table
4).
Table 5
Intervention level of measures and their implementation status in the intervention wards
1-Individual | 8 (26%) | 23 (74%) | 0 (0%) | 31 (6%) |
2-PEC | 17 (19%) | 67 (76%) | 4 (5%) | 88 (17%) |
3-SAP | 4 (14%) | 19 (66%) | 6 (21%) | 29 (6%) |
4-TRA | 7 (14%) | 40 (82%) | 2 (4%) | 49 (9%) |
5-WPT | 11 (19%) | 41 (72%) | 5 (9%) | 57 (11%) |
6-Other organisational | 12 (43%) | 15 (54%) | 1 (4%) | 28 (5%) |
7-Unclassifiable | 59 (24%) | 175 (72%) | 10 (4%) | 244 (46%) |
Total | 118 (22%) | 380 (72%) | 28 (5%) | 526 (100%) |
Moreover, despite the fact that the contents discussed during the workshops did focus on key issues including the adequacy of work processes, extent of job duties, leadership or health issues of healthcare workers in the intervention wards, the final decision on the implementation of measures was not always made by the workers themselves. In particular, measures which affected more structural aspects of the work environment, e.g., definition of work tasks, work load, coordination of work within and between wards, were competency of the initiative circles. From this perspective, the intervention approach did not explicitly defined feedback or consultation mechanisms between the wards and the initiative circles. Hence, it seems that the key limitation concerning the efficacy of the workshops as the centrepiece of the intervention was not primarily due to the specific contents and themes discussed in the intervention, but rather to different factors associated with the identification and selection of effective measures, the lack of a theoretical rationale for defining and prioritising the measures and the partial detachment of decision-making power from the intervention wards.
It has to be acknowledged that the receptivity and engagement of the healthcare workers themselves [
34], as measured by the concepts of workshop-related efficacy expectation and prospective outcome expectations, was rather high, as additional analyses focusing on the collective self-efficacy beliefs of the workshop participants indicated [
35]. There was some evidence suggesting that the higher the workshop-related efficacy expectation, the larger the number of proposed measures was. Hence, it seems that the intervention activities were strongly supported by workers and supervisors. Since collective self-efficacy refers to people’s shared beliefs in their collective power to produce desired results by collective action [
36], there seemed to have been sufficient receptivity and engagement among workers to bring about changes in the working conditions at the ward level. However, as stated above, the participation of workers was not embedded in a general framework of causal mechanisms relating the proposed measures and the intended goals and, therefore, there was indeterminacy regarding the results to be expected from the collective action efforts. In addition, the intervention approach consisted of a single workshop and, therefore, it did not provide continuous support throughout the intervention period to enable workers revise the adequacy of measures and ensure their delivery in the intervention wards.
The analysis of the associations between the psychosocial workload and the physical and mental work ability did not provide evidence that the intervention resulted in a reduction of job efforts and overcommitment, or in the improvement of the rewards obtained at work in the intervention wards. Nonetheless, these results emphasised once more that these psychosocial risk factors had a large impact on the perceived work ability and, consequently, confirmed previous research findings obtained with larger samples of the employed population in Germany and Finland which reported substantial associations between the effort-reward-imbalance ratio and the general work ability [
37,
38]. Moreover, the observation that high efforts and low rewards are related to an increased likelihood of sick leave [
39] and the intention to claim disability pension [
37] underlines the importance of maintaining lower levels of psychosocial workload. At the same time, these findings point to potential causal mechanisms which may inform the design of future organisational interventions which explicitly address the work characteristics specifically associated with job efforts, rewards and overcommitment. On the basis of previous interventions it can be expected that such an evidence-based approach may be more effective in attaining beneficial outcomes at the individual and organisational level [
40,
41].
Implications
By taking into consideration the results of the present participatory organisational intervention, it can be concluded that at least three major criteria may help organisations and researchers with the design and planning of more effective participatory organisational interventions.
1.
Definition of the intervention goals. The present intervention pursued to improve the self-assessed physical and mental ability. However, the concept of work ability is rather an attitude, i.e., a cognitive appraisal process of one’s prospects of coping with the physical and mental job demands [
42]. The intervention approach implicitly assumed that this cognitive appraisal could be changed by delivering the modifications proposed by the healthcare workers. However, previous research in social psychology has indicated that the modification of attitudes is a challenging task which requires strong stimuli and effective environmental modifications [
43]. Even though it appeared plausible to assume that the self-assessed work ability could be changed by modifying the working conditions as proposed by the workers, the actual causal pathways of how those proposed measures would serve the intervention goals were not identified. In addition, it was assumed that the measures proposed in the workshops would be beneficial for all workers in the wards, independently of whether they participated in the workshops or not. However, since the workers participating in the survey were in rare instances also participants of the workshops, the lack of treatment effects indicates that this assumption was not tenable, i.e., no spill-over of benefits were observed for all workers in the intervention wards. Accordingly, workplace interventions should carefully take into consideration the feasibility of achieving the goals and, accordingly, identify the set of modifications which may be causally related to the intended goals.
2.
Intervention approach. The process evaluation indicated that the intervention limited itself to the preparation of some activities including the establishment of initiatives circles, the conduction of interviews and the organisation and moderation of single workshops. However, the intervention approach did not take into consideration that workplace interventions require continuous monitoring, evaluating and adjusting of the intervention processes and contents [
5,
9]. As the process evaluation revealed, the delivery of the intervention was unsuccessful since no mechanisms were installed to manage the organisational change activities, i.e., to check and revise contradictory or ineffective measures, bypass unforeseeable events (e.g., COVID-19 pandemic) or facilitate the coordination of the intervention measures within and across wards [
8]. Over time, the intervention efforts waned and the majority of proposed measures were not implemented. Consequently, any approach to conduct organisational interventions should be conceived as an ongoing organisational change process based on a feedback system including at the very least the phases of preparation of the intervention, selection of methods, action planning, monitoring of the implementation and evaluation [
9].
3.
The action plan. The intervention did not provide explicit guidance regarding timelines for the implementation of measures or the allocation of resources necessary to deliver the measures. The vagueness of the goals was accompanied thus by the vagueness of action plans to implement the proposed measures. The intervention failed to identify the specific activities which were needed in order to bring about specific organisational changes such as procurement of resources, budgeting, the clear delegation and assignment of tasks and responsibilities or the mechanisms used to maintain and enforce the measures [
8]. Form this perspective, workplace interventions would greatly benefit by adopting a project-based approach in which tasks, responsibilities, resource allocation and timelines are specifically determined in order to bring about actual changes in the organisational structure and processes.
Strengths and limitations
The major strengths of the present intervention are the study design and the extensive process evaluation. In contrast to other organisational interventions which are prone to confounding due to several factors including lack of control groups, randomisation or treatment contamination [
1], the results of the present study are robust given the successful implementation of a cluster-randomised controlled design [
22]. In addition, the process evaluation was based on previous literature and included group and individual levels of variation [
35] which allowed an in-depth analysis of the most important factors which may have contributed to the observed results. On the other hand, there are two major limitations. First, the information on the intervention measures proposed in the workshops was collected by the consultants and not by independent observers. Even though the information on the workshop contents and implementation status was systematically collected, there were ambiguities in the description of single measures. However, since the measures were classified independently by each author of the present study, the impact of those ambiguities can be considered low. Second, due to organisational constraints it was not feasible to collect detailed information of the work in the initiatives circles. Hence, the analysis of the decision-making process leading to the acceptance or rejection of measures could not be performed.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.