Workplace interventions focusing on how to plan, organize and design the work environment in hospital settings: A systematic review

1.1Aim

The aim of the study was to summarize the evidence for workplace interventions with regards to outcomes, workplace interventions and intervention strategies in hospital settings.

Research question 1: How do workplace interventions affect employee and work-related outcomes in hospital settings?

Research question 2: Which workplace interventions and intervention strategies are used in workplace interventions in hospital settings?

2.1Study design

A systematic literature review with a narrative summary [18]. This study is presented according to the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) 2020 statement [19].

2.2Eligibility criteria

Studies published in English between January 2010 and September 2021 were considered. Research indicates a long-term gradual increase of complexity in hospital settings [20, 21]. The start date of the searches was a pragmatic choice used to increase the relevance of the included studies. The search strategy was structured according to PICO (population, intervention, comparators, and outcome).

The population in the included studies were employees of all professions working within hospitals (e.g., physicians, nurses, nursing assistants, medical administrators, occupational therapists, physiotherapists, technicians). The intervention were workplace interventions (randomized controlled trials (RCTs), non-RCTs and mixed methods studies) focusing on planning, organizing and designing the workplace context to benefit work conditions and employee health. Intervention studies with any comparator/no comparator were included. To be included, studies where work conditions and work-related outcomes (e.g., stress, burnout, physical pain, or sick leave) were evaluated.

Inclusion criteria: Workplace interventions (RCTs, non-RCTs and mixed methods) focused on planning, organizing and designing the workplace context aiming to improve work conditions and employee health in hospital settings were included. Exclusion criteria: Individual focused work environment interventions intended to support the hospital employee, such as coping strategies to manage work demands, were excluded. Study designs other than intervention studies were excluded.

2.3Information sources and search strategy

Literature search strategies were developed and conducted by a research librarian (EN) to reflect the concepts outlined. A combination of title and abstract keywords were used as well as controlled vocabulary whenever possible. A set of key articles were identified before the search process and were used to generate search terms and to test the effectiveness of the strategies in each database. The MEDLINE strategy was developed with input from the project team, then peer reviewed by a second librarian, not otherwise associated with the project. After the MEDLINE strategy was finalized, it was then adapted to the syntax and subject headings of the other databases/platforms.

The following databases/platforms were searched in September 2021: CINAHL with Full Text (EBSCOHost), MEDLINE (EBSCOHost), PsycInfo (ProQuest), Scopus (Elsevier), and Web of Science Core Collection (SCI-EXPANDED, SSCI, A&HCI, CPCI-S, CPCI-SSH, ESCI). See appendix for complete documentation of the search strategies. In addition, the CENTRAL trials registry of the Cochrane Collaboration (Wiley) was searched for ongoing or recently completed trials. Before undertaking this review and to avoid research waste, the PROSPERO database and Google Scholar were also searched for ongoing or recently completed systematic reviews on the same topic. The searches were limited to intervention studies published in English from 2010 onwards. As relevant studies were identified, the reviewers checked for additional relevant cited and citing articles.

2.4Selection process

Records found during the search phase were exported to a reference management software (EndNote) to enable the identification and removal of duplicates [22]. Prior to the formal screening process, a calibration exercise was undertaken to pilot and refine the screening questions. Records were then screened using Rayyan, a web-based application for systematic reviews [23] based on the previously described inclusion/exclusion criteria. The primary investigator (PH) performed an initial screening on the title/abstract level. The last author (KAJ) screened random samples, as well as titles/abstracts with uncertainties. Disagreements were resolved by discussion until consensus was reached. Two team members (PH and KAJ) independently conducted full-text assessment of included records, and any disagreements were resolved by discussion until consensus was reached.

2.5Quality assessment

Full-text studies were analyzed independently by two authors (PH and KAJ) with the Mixed Methods Appraisal Tool (MMAT). MMAT is a critical analysis tool for studies, allowing studies with different study designs to be included in a systematic literature review. It permits to appraise the methodological quality of five categories of studies: qualitative research, randomized controlled trials, non-randomized studies, quantitative descriptive studies, and mixed methods studies [24].

The quality assessment with MMAT was conducted in two steps. The first step was conducted by responding to two screening questions similar for all study designs (S1. Are there clear research questions? S2. Do the collected data allow to address the research questions?). The response options were yes, no or can't tell. Responding no or can't tell to one or both screening questions might indicate that the paper is not an empirical study, and thus cannot be appraised using the MMAT. The second step was conducted by choosing the correct study category (RCT, non-RCT or Mixed methods in our study). Regardless of study design each study was evaluated by five quality criteria questions, with the response options yes, no or can't tell. The can’t tell response category meant that the paper did not report appropriate information to answer yes or no [24]. Any disagreements during the analysis process were resolved by discussion until consensus was reached.

In the MMAT it is discouraged to calculate an overall score from the ratings of each criterion. Instead, it is advised to provide a more detailed presentation of the ratings of each criterion to better inform the quality of the included studies [24].

2.6Data extraction

Data on outcome and interventions were extracted from the intervention studies. Data on intervention strategies were extracted with a matrix of intervention strategies based on principles from Gustafson and von Thiele Schwarz [25, 26]. Start-up criteria, evaluation methods, interventions, and outcome characteristics were evaluated in each study. The matrix contained twenty-six items answered with yes, no/unknown or not applicable. Data extraction was conducted independently by two authors (PH and DJ). Any disagreements during data extraction were resolved by discussion until consensus was reached.

2.7Data synthesis

Since a large heterogeneity in outcomes and interventions were expected, the data synthesis was conducted with a narrative summary [18]. Employee outcomes were presented in three categories: physical outcomes, psychosocial outcomes and other outcomes, together with related workplace interventions. Intervention strategies were presented in three categories: start-ups, evaluation methods and intervention characteristics.

3.1Study selection

The database searches identified 4,793 records. After the removal of duplicates, 3,234 records were screened on the title/abstract level. Of these, 40 full-text documents were reviewed, and finally 26 papers [27–52] were included for analysis. See Fig. 1 for a flow diagram of the review process.

Fig. 1

PRISMA 2020 flow diagram of review process.

3.2Characteristics of the included studies

The study design of the included studies were RCTs (n = 12), non-RCTs (n = 13), and mixed methods (n = 1), published between 2010 and 2021 with a study duration between 4 weeks and 13 years. The studies were conducted in Europe (Italy, France, The Netherlands, Denmark, and Sweden), Asia (Thailand, China, Korea, Vietnam, and Japan), Canada and the USA. The sample sizes ranged from n = 8 to n = 11545 (Table 4).

3.3Quality assessment

The quality assessment showed that all the included RCTs, except Bagaipour-Divshali [28], had an appropriately performed randomization. The RCT by West [51] did not have comparable groups at baseline and the study by Sohn [47] did not present data on baseline characteristics. Complete outcome data were presented in all the included RCTs. The quality assessment identified a general lack of blinding in the included RCTs. Only five of twelve RCTs clearly stated that the assessors were blinded to the intervention. The study by van der Meer [49] did not have blinded assessors and no information about assessor blinding was presented by Bagaipour-Divshali [28], Chanchai [31], MacIntyre [40], Uchiyama [48], van Der Molen [50] and West [51]. All included RCTs, but Bagaipour-Divshali [28] and Chanchai [31], presented data showing that the participants adhered to the assigned intervention (Table 1).

The quality assessment showed that all non-RCTs had participants that were representative of the target population and that the measurements were appropriate in relation to outcome and intervention. All the included non-RCTs, except d’Ettore [32], presented complete outcome data. Confounders were accounted for in all studies, except Pierce [45]. In Mehrdad [42] the intervention was not administered as intended and Abdollahi [27] and d’Ettore [32] did not present information about this. All other non-RCTs administered the interventions as intended (Table 2).

One study with mixed methods design, by Michélsen [43] was included. The quality assessment showed that the study had an adequate rational for the use of mixed methods design to address the research question. The quality assessment also showed that the different components of the study were effectively integrated to answer the research question. The study did not present data to evaluate if the integration of qualitative and quantitative components were adequately interpreted. The divergences and inconsistencies between qualitative and quantitative data were not adequately addressed. The study did adhere to quality criteria in the qualitative and quantitative traditions (Table 3).

3.4Workplace interventions and outcomes

Improvements were identified in a wide variety of outcomes. Statistically significant improvements were found in physical outcomes like decreases in pain, discomfort, musculoskeletal disorders, and injuries [27, 29, 31, 39, 46, 47, 52], as well as increases in the use of patient handling equipment [37, 52]. The interventions were heterogeneous and included ergonomic education [29], ergonomic education/practical ergonomic changes [27, 31], work stress reduction [27], workstation re-design [31], changes in administrative work [31], minimal lift policy and lift equipment [46, 52] and minimal lift support (peer-coaches) [52]. The Jacobsen study was an exception concerning positive outcomes, as they did not find any changes in low back pain or injuries despite improvement in the use of assistive devices [37]. Interventions related to physical outcomes consisted of different combinations of these interventions (Table 4).

Two studies of ergonomic interventions evaluating operating table height for anesthetists, were exceptions as they did not consist of combination interventions [39, 47]. Higher operating tables reduced discomfort [47], and decreased intubation discomfort [39] and lower operating tables decreased ventilation discomfort [39] (Table 4).

Statistically significant improvements were found in psychosocial outcomes like decreases in work-related burnout and improvements in psychosocial work factors like demands, support, and job satisfaction [28, 30, 33, 34, 43–45, 48, 51]. The interventions were heterogeneous and included organizational interventions like job crafting [34, 44], task rotations [30], bed reductions [30], lean improvement [44], work breaks [44], improvement of workflow processes [45], and divisions of duties [43]. Psychosocial interventions like leadership [28, 45], cooperation/communication [30, 43], team culture [34, 45], rewards [48], support [48], role play [33], debriefing [33], reflection groups [51], and clarifications of responsibilities and goals [48] were also used. All interventions related to psychosocial outcomes consisted of different combinations of these interventions (Table 4).

Statistically significant improvements were also found in other outcomes. Education to reduce needle stick injuries, as a single intervention, increased the number of reported needle stick injuries [42]. A combination of education and needle stick devices decreased the number of needle stick injuries more than only education [50]. Interventions to reduce work stress decreased the number of needle stick injuries further [32] (Table 4).

Exposure to work violence decreased through education in an intensive and emergency department [36] and through a restricted access policy in a tertiary hospital [38]. Education and participative working groups to prevent hand eczema led to a more preventive behavior [49]. A policy update in combination with employee participation in improving drug handling decreased the number of contaminated surfaces [35] (Table 4).

Work-related viral/bacterial infections decreased with the use of N95 respirators and medical masks [40, 41]. The interventions show that N95 respirators gives better protection against viral and bacterial infections, compared to medical masks [40] and that medical masks are better than cloth masks and control [41] (Table 4).

3.5Intervention strategies

4.1Result discussion

The results of this systematic literature review support the use of workplace interventions, with regards to planning, organizing and designing the workplace context, to improve work conditions, health, and well-being in hospital settings.

Previous research has shown workplace interventions reduce stress, distress and burnout in physicians and nurses in hospital settings [12–16]. The result of this systematic literature review adds new and broader knowledge with regards to outcomes from workplace interventions in hospital settings. The results support improvements in physical outcomes like pain, discomfort, musculoskeletal disorders and injuries. The results also support improvements in other psychosocial outcomes like work demands, support and job satisfaction. There were also improvements in other outcomes like needle stick injuries, exposure to violence and viral/bacterial infections.

The results are based on intervention studies that were judged as having moderate to good quality overall. The quality of the included RCTs were considered as moderate to good based on their fulfillment of the quality criteria in MMAT (Table 1). The quality assessment of the included RCTs showed a somewhat poorer result for the assessor blinding criteria, with only five of the twelve RCTs clearly stating that the assessors were blinded to the intervention [33, 37, 39, 41, 47] (Table 1). Four of these five RCTs showed positive outcomes [33, 37, 39, 47]. The included non-RCTs were judged as having good quality, based on the high fulfillment of the quality criteria in MMAT (Table 2). The included mixed methods study was judged as having moderate quality, since it only fulfilled three out of five criteria in MMAT (Table 3). The moderate to good quality of the included studies strengthens the value of the results. The studies were conducted in many different countries, with different study designs, different sample sizes, different types of interventions, and different types of outcomes (Table 4). Improvements were shown in almost all the included studies despite a large study heterogeneity, which further strengthens the results supporting workplace interventions as a means to improve work conditions.

The results of this systematic literature review add knowledge with regards to workplace interventions in hospital settings. The workplace interventions were heterogeneous and almost all included different combinations of ergonomic, environmental, psychosocial and organizational interventions. This finding corresponds with van der Beek et al. supporting that considerations must be given to physical, psychological, and psychosocial factors etc. to understand risk exposure [2]. Generalizability of workplace interventions has been questioned in previous research, since interventions like these usually are tailored to a specific organizational context [53]. The results of this systematic literature review support this criticism. The result supports that simple problems can, in some cases, be solved with simple interventions. An example of a simple intervention is changing operating table heights to decrease work discomfort in anesthetists [39, 47]. Most of the included studies contained complex problems that were solved with combinations of interventions. This illustrates the importance of adapting intervention strategies and combinations of interventions to the specific contextual problems at each workplace, supporting a pragmatist complexity approach in workplace interventions.

The results of this study also add knowledge with regards to intervention strategies. In the design of workplace interventions, a clear work structure was created, and time was set aside to enable the intervention to be conducted. The type of problem and level of complexity were addressed when choosing an evaluation method to ensure that relevant perspectives were included in the evaluation. The interventions were based on some form of evidence, had a multifactorial approach (several interventions at the same time) and a participative approach (employer and employee participating in identifying and solving work environment problems).

4.2Strengths and limitations

When interpreting the results of this review, the following aspects should be considered. To ensure the rigor of our study, the literature search strategies were developed and conducted by a research librarian (EN), and how we conducted the various stages of the review were reported in accordance with the PRISMA 2020 Guidelines [19]. The search term “health personnel” was used to include professions other than physicians and nurses, which was a strength in this study. A potential problem with focusing interventions towards one profession is that an intervention that leads to positive outcomes in one profession might lead to negative outcomes in another. Such a conflicting result was shown in the study by Michélsen, et al, 2014 [43]. Their intervention led to decreases in mental strain in physicians and midwives but worsened the social climate among nurses (Table 4). Drop-outs of the included studies were not included in the quality assessment with MMAT, which can be seen as a limitation in the review. The included workplace intervention studies in this systematic review had a complex study structure. The complexity of the included studies was a challenge in the data extraction process. Meta-analyses of the results would have been ideal but was impossible due to a large heterogeneity with regards to intervention characteristics, measurement methods and outcomes. The result is therefore presented with a narrative summary [18], which can be seen as a limitation.

4.3Practical implications

There are practical implications for OHS from this systematic review with regards to workplace interventions. The results support OHS to work in compliance with the ILO definition. In the start-up and design of workplace interventions, basic conditions must be established. A clear work structure should be created, and time should be set aside to enable the intervention to be conducted. The type of problem and level of complexity should be addressed when choosing an evaluation method to ensure that the context specific problems are identified. Simple work environment problems can, in some cases, be solved with simple interventions. Combinations of interventions tailored to the context specific problems should be used to solve complex work environment problems. They should be based on some form of evidence, with a multifactorial and participatory approach and iterative processes.

The results of this study show that if these criteria are met in practical OHS support in workplace interventions, positive outcomes can be expected. With adequate intervention strategies and workplace interventions tailored to the workplace, OHS will be able to support employers and employees complying with the UN-SDG promoting safe and secure working environments in hospital settings.

4.4Future research

Future research should focus on the intervention strategies and evaluating health outcomes in relation to the intervention strategies used in the study. Broad evaluations, multiple professions and the whole department (or several departments) should be used instead of, for example, a single nursing ward in order to capture the complexity. Studies with single outcomes, single professions or single nursing wards should be avoided to minimize the risk of non-evaluated negative outcomes.

It is of great importance that future research should be designed to evaluate interventions in relation to the work environment complexity in hospital settings.