Taming the flood of findings: What makes for a really useful literature review in occupational health psychology?

Work & Stress was established in 1987 as one of the first journals specifically dedicated to publishing research in occupational health psychology (OHP), that is, psychological research aiming to improve the quality of work life, and to protect and promote the safety, health and well-being of workers (NIOSH, 2022). Other journals have not held back in publishing research in this area. To date a wealth of findings is available on a wide range of OHP-relevant topics like work stress, organisational and individual-level interventions, safety at work, recovery, work characteristics, and interpersonal relations at work. Moreover, much of this research can be retrieved in a split second using search engines such as PsycInfo and Google Scholar; a situation that could hardly be foreseen back in the days when Work & Stress was established!

The downside of the availability of this treasure trove of findings is that these often diverge and sometimes contradict each other. Creating an overview can be confusing and the precise state of affairs in a particular area may be unclear. What do we know about a particular phenomenon? What is still under discussion? What are the research gaps that still need to be addressed? Responding to the growing need for summarising, reviewing, interpreting and integrating the findings that are available, over the years Work & Stress has paid much attention to publishing review studies. For example, over the past three years about one in five papers published in the journal was a review study. Moreover, these reviews figure prominently among the most highly-sought articles in Work & Stress. Clearly, there is a great need for such studies. Unfortunately, whereas to date we receive many submissions that present reviews, not all of these are equally interesting and useful. This raises questions like (a) What makes for a really valuable review study in the area of OHP?, and (b) Which recommendations can be formulated for new review studies to bring the research field of occupational health further?

What sort of review studies are most useful?

A review study can be defined as any study attempting to synthesise the current state of affairs on a particular topic. As Grant and Booth (2009) note, under this label comes a bewildering number of different approaches to reviewing the literature. These vary on several dimensions, with the so-called scoping, rapid and mapping reviews on the broad, not-so-comprehensive and statistically unrefined end (cf. Tricco et al., 2015), and state-of-the art meta-analyses on the focused, systematic and statistically sophisticated end. At Work & Stress we do not care much about the label attached to a particular review study, and while we admire a high level of statistical sophistication, we are not intimidated by it. Rather, it is important that the approach that is used meshes well with the state of affairs in a particular area or with the topic(s) under study. For example, a subject on which a large number of results is available (e.g. the associations between leadership style and work engagement, Li et al., 2021) lends itself well to a comprehensive, systematic and quantitative approach (i.e. a meta-analysis). Conversely, in relatively young and emerging fields where only few primary studies are available, when a body of literature has not yet comprehensively been reviewed, but where there is an urgent need for results to be disseminated quickly (consider the early days of the Covid-19 pandemic), a rapid or scoping review will be appropriate (Peters et al., 2015).

After some forty years of research in OHP, many topics have generated a sufficiently large number of primary findings to warrant a quantitative approach. Our community is building steadily on a body of knowledge on the associations between characteristics of the psychosocial work environment and worker well-being (in a broad sense), taking into account characteristics of the worker, their family environment, and the organisational context. Heated debates concerning particular pressing issues that need to be decided upon quickly, are largely absent. This may suggest that reviews in OHP should almost by definition take a quantitative/meta-analytic approach. However, more narrative/scoping/integrative review studies will often be appropriate as well, e.g. to explore the state of affairs in relatively young and emerging areas (such as the relations between aging and employability, e.g. De Lange et al., 2021), among particular occupational groups, or when only few primary studies are available. Similarly, in fields where the body of literature concerns a mixture of qualitative, quantitative or mixed-method studies, a systematic, qualitative approach may be suitable as well. For example, studies in the field of organisational interventions use a plethora of methods to understand what makes the intervention work, and quantitative approaches may be less suitable to synthesise insights from a complex field (Fox et al., 2022).

Quality assessment. According to Grant and Booth (2009), the essence of doing a systematic review is “Gathering research, getting rid of rubbish and summarising the best of what remains” (p. 92). As the shortcomings of primary studies carry over to review studies, it is important to assess the quality of the studies included in any review. If many primary studies are available, there is no point in including the not-so-good studies; even if the findings of the good differ from those obtained for the not-so-good studies, conclusions will usually be based on the first only. If only few primary studies are available, inclusion of the not-so-good studies may be justifiable (Grant & Booth, 2009). However, even then it is important that the limitations of the underlying material are acknowledged, or better still, to translate these limitations into specific recommendations for future research in this area.

What sort of research questions are of most interest?

Review studies can be descriptive and objective (basically providing a summary and integration of previous research, focusing mainly on the “what is the case”-question), but also more integrative (e.g. focusing on integrating definitions, cf. Verschuren et al., 2021) or interpretative/critical (e.g. aiming at theoretical innovation, addressing the “given what we know, what does this mean for phenomenon x?”). All these types of review studies are imperative for progress in our field, although even primarily descriptive studies should always address the “so what/what’s next?” question; whereas it is fine to know where things stand, it is at least as important to know where we should go from there. A crucial part of any review is thus to develop a research agenda and provide directions for future research.

How temporally stable are findings? Apart from being objective versus interpretative, the topic to be addressed in a review is obviously essential. Just like other studies, reviews should provide new insights. There is no point in providing updates of existing, high-quality review studies, unless there is reason to believe that a new review will extend our knowledge, for example, by taking only high-quality studies into account (e.g. Lesener et al., 2020).

Although such reviews can extend the empirical basis for particular conclusions, it is probably at least as important – also from a practical point of view – to see whether findings are contingent upon other factors, such as the type of population, cultural features or quality of included methodology and research designs (cf. Li et al., 2021). Importantly, OHP aims to improve the quality of work life and to promote the safety, health and well-being of workers (NIOSH, 2022; italics ours); i.e. we consciously want to bring about change. This implies that the findings on the topics we study are not set in stone. This applies to descriptive findings (means, standard deviations, either for groups as a whole or for its constituent parts), the associations among concepts and, indeed, the processes generating these associations and descriptive information.

Indeed, meta-analytic findings that applied two decades ago may not be relevant today, i.e. every once in a while there is a need to update matters or to at least compare older to more recent findings. For example, in the pre-covid days working from home often involved a voluntary decision, e.g. when workers needed high levels of concentration to work on tasks like writing a paper. Colleagues and superiors were reluctant to interfere, and children were at school. However, at present many workers must work from home, children may be quarantined at home (and may even need to be home-schooled), and colleagues and supervisors have become used to routinely entering our home office through our webcams. No longer are our homes the quiet refuges they were in the pre-covid days, which could imply that the beneficial effects of working from home become weaker (cf. Murphy, 2021; Shifrin & Michel, 2022).

Note that a comparison of findings across time may be interesting in itself. E.g. work intensification refers to “the process of continuously increasing job demands that have to be attended to in shorter time” (Bunner et al., 2018). Both scientifically and practically it would be interesting to see whether work intensification has increased over the years, whether there is an upper bound regarding this intensification, whether differences in intensification among organisations and departments change, and whether this affects its associations with outcomes such as performance, stress and safety behaviour of employees. This sort of questions could adequately be addressed using a quantitative meta-analysis that explicitly focuses on the role of time. Since many concepts in OHP have been around for quite a while, it should increasingly be possible to address this sort of questions involving a comparison of findings pertaining to different time points.

Another issue related to the OHP aim of making changes to improve the quality of work life and promote the safety, health and well-being of workers (NIOSH, 2022) is how we generate knowledge about how to implement changes. Regardless of the level of intervention, this knowledge generation requires reviews to consider the intervention processes and the context within which interventions are implemented to ensure sustainable effects, be it processes for integrating organisational interventions into daily work practices (Nielsen & Abildgaard, 2013) or workers translating learned cognitions, emotions and behaviours to the workplace setting post-training (Nielsen & Shepherd, 2022).

The present issue

What makes for a really valuable review study in the area of OHP? It’s complicated. As indicated above, ideally reviews should meet many criteria. Obviously, a review must have been conducted well, cover the appropriate literature and address an OHP-relevant topic. We feel that the value of such studies would be enhanced even more if (a) there is a good match between the type of review that is conducted and the degree of maturity of – i.e. the number of primary studies available in – a particular area; (b) there is good match between the type of review and the methods and study designs in a particular field, encompassing the breath of studies in a given field; (c) the review provides insight in both the “what is the case” and the “so what/whats next?” question; and (d) they cover a longer time period (e.g. more than a decade), provide a comparison of older versus more recent findings, or provide relevant new integrative perspectives that can be used for further theory building. Clearly, review studies meeting all these requirements are rare and not meeting all these requirements certainly does not invalidate a particular review. The present issue includes five review studies, each of which is really valuable in its own way.

In the first contribution to this special issue, Fikretoglu et al. (2022) present an example of a narrative review, exploring the impact of fidelity, i.e. whether interventions are implemented according to plan in individual-level interventions. Understanding whether interventions are implemented according to plan makes a significant contribution to moving beyond answering the simple question of whether an intervention worked or not, to understanding how an intervention needs to be implemented in order for it to achieve its intended outcomes.

Fox et al. (2022) conducted a systematic literature review of group- and organisational level interventions, identifying that different targets of interventions, be it flexible work and scheduling changes; job and task modifications; relational and team dynamic initiatives; or participatory process interventions, all have the potential to improve worker well-being. In line with Fikretoglu et al. (2022) the intervention process, in this review identified as increased control and opportunities for workers’ voice and participation, were important determinants for improving worker well-being.

In the third contribution to this issue, Shifrin and Michel (2022) meta-analytically reviewed the effects of flexible work arrangements on a variety of positive outcomes in different domains. They found that such arrangements are associated with better physical health, lower absenteeism, and fewer somatic symptoms, suggesting that organisations that offer their employees flexible work arrangements can promote the health of the latter.

The fourth contribution focuses on the cognitive functioning of workers having had a clinical burnout. In their meta-analysis, Gavelin et al. (2022) shown that clinical burnout is associated with small to moderate impairments in episodic, short-term and working memory, executive functioning, attention and processing speed, and fluency. From a practical point of view this implies that cognitive dysfunction needs to be considered in the treatment and management of burnout to support return-to-work.

In the final paper included in this issue, Lam et al. (2022) present a systematic, narrative review on the impact of informal eldercare on workers in paid employment. Although juggling the responsibilities in the work and eldercare domains can be burdensome, overall this impact was positive. The impact of demands and resources on employee work, health and well-being outcomes varied across different levels of analysis (i.e. the individual versus the group versus the organisation). Moreover, the effects of resources tend to have received more attention than those of demands – findings that can only be obtained in a review study.

In conjunction, these five reviews provide an excellent illustration of the sort of questions that review studies can deal with, and the sort of conclusions only they can provide. Moreover, they also show that no single methodological approach is best; depending on the maturity of a field, the specific research question and the number of primary studies that are available, a variety of approaches can be used. We are looking forward to receiving many more such submissions in the future!