Workplace health promotion interventions for nurses in Germany: a systematic review based on the RE-AIM framework

The electronic databases PubMed and PubPsych were searched on January 11th, 2021 for potential articles. Search terms used for relevant studies were (nurs* OR "professional care" OR "professional caregiver") AND ("workplace health promotion" OR "work health promotion" OR WHP OR WHPP OR prevention OR “preventive health program" OR "preventive health care" OR "intervention program") AND (health* OR violence* OR "work ability" OR disease OR morbidity OR "risk factor" OR burden OR stress) AND (german*). Original studies in German or English language, published between January 01^st, 2010 and January 11^th, 2021 were taken into account. Results were completed by a manual search upon included studies’ references.

In our review we defined workplace health promotion as behavioural measures offered at the workplace, addressing individual coping skills in the field of physical-activity-promoting work and physically active employees, stress-management and -strengthening resources, healthy diet in everyday work, and addiction prevention [40]. Studies which met the following inclusion criteria were examined: (1) target group or subgroup analysis: professional nurses in Germany, (2) setting: acute medical care hospital, inpatient LTC facilities and/or home-based LTC, (3) intervention study (RCTs, non-RCTs, non-controlled intervention studies and pilot studies), (4) primary outcome: physical health, mental health and/or violence experience. Articles that showed at least one of the following exclusion criteria were not considered for further analysis: (1) no original study (e.g. review or editorial), (2) interventions that were prilimilary addressing health and safety protection at the workplace (according to social code (SGB VII)), as another recognized pillar of a holistic workplace health management in Germany, (3) studies outside of Germany. Comparators were not defined in advance.

The study selection process after the elimination of duplicates was conducted with the software tool for systematic reviews “Rayyan” [41]. Two authors (MG, TK) independently performed the title and abstract screening as well as the subsequent full-text screening including the record of reasoned exclusion. Any discrepancies were resolved by discussion and consensus with a third researcher (AS). The selection process was illustrated in a PRISMA Flow Chart [39]. Data extraction of the included articles was separately performed by two authors (MG, TK) and crosschecked in each case.

In order to answer research question 1 on the setting specific availability of WHP for nurses, extracted data of the studies included were author and publication year, the setting in which the study was conducted (acute medical care hospital, inpatient LTC facilities, home-based LTC, cross-setting) and the health issue addressed in the study (physical health, mental health and/or violence experience), In addition, the interventions were presented against the background of the quality criteria for prevention measures of the statutory health insurers [40]. The quality criteria include planning and concept quality (target group; content; participants material; theoretical framework/evidence of the intervention), process quality (group size, contraindications, number, duration and frequency of units, location) and structural quality (provider qualification).

To answer question 2 on the appraisal of the respective studies, the study design and the comparators (ususal care, non-intervention, comparison intervention, no control group) were extracted and the results of the studies were presented according to the RE-AIM framework [42, 43]. Table 1 shows the chosen indicators for each RE-AIM dimension that were extracted in the present review. Missing information in the original studies on one dimension was described as "not reported".

The Delphi List [44] was applied in order to evaluate the selected articles and to identify the risk of bias of the included studies. The Delphi List consists of nine items, which are answered with "yes", "no" or “don’t know”. Two authors (MG, TK) independently applied the checklist. In case of disagreements in the ratings of the nine items, they were resolved after reconsideration and, if necessary, discussed with a third author (AS). Finally, the percentage of checklist items answered with “yes” was calculated for each study. If the study scored ≥ 50% by fulfilling at least five quality requirements, a “low risk of bias” was considered.

During the initial search 444 publications were identified. After duplicates’ removal 426 publications were included in further screening. After screening titles and abstracts, 16 full-texts were again considered, of which seven were included in the analysis. In addition, four studies were identified by cross-referencing, what resulted in a total of eleven studies (see Fig. 1).

The eleven included studies in the review refer to eight different research projects on WHP among nurses (see Table 2). From three different projects, two publications each were included in the review [45‐50]. Based on the eleven publications, seven studies were only conducted in acute medical care hospitals [45, 46, 49‐53] and three only in inpatient LTC facilities [47, 48, 54]. One study was designated as a cross-setting study (inpatient LTC facilities and home-based LTC) but due to an institutional drop out results were only available for the setting of inpatient LTC facilities [55]. In terms of outcomes, most studies solely aimed to improve mental health [47‐53]. Three studies reported on interventions aiming at improving only physical health [45, 46, 54] and one study targeted both mental and physical health [55]. Violence experiences were not the content of any of the included studies.

Regarding the quality criteria assessed, information on the intervention provider’s qualification was given in about half of the studies [45, 46, 50‐53]. Even though all interventions targeted nurses, some studies still addressed specific subgroups, such as nurses with physical complaints [45, 46], managerial roles [47, 48], working in a specific setting [51, 54] or being at a defined age [49, 50]. Manuals included (psychosocial) stress training programs [45‐52], dealing with difficult residents, communication, and leadership [47, 48], physiotherapy exercises [45, 46], collegial counselling [53], ergonomics training [54], as well as a multi-component program [55]. Nine studies reported an underlying theoretical framework of their intervention [45‐53]. Regarding the process quality, the intervention groups were designed for six [45, 46, 52] to 19 [53] participants. Contraindications for the participation were reported depending on the content of the intervention in five studies [45, 46, 50, 51, 54]. The number of intervention units varied from four [53, 54] to twelve [47, 48, 51], with a duration of the units between 45 min [45, 46, 53] and eight hours [52]. The frequency of units was mostly weekly, except in one study [52]. Seven studies were conducted as inhouse training programs [47, 48, 50, 52‐55].

With regard to study design, five studies were conducted as randomized controlled trials (RCT) [45, 46, 50‐52], three studies had a quasi-experimental design [47, 48, 53] and three studies as a single-group longitudinal study [49, 54, 55]. Five studies were designated as pilot studies [46, 49, 51, 53, 54]. According to the Delphi List study quality regarding the fulfilment of quality requirements for intervention studies varied from 0 to 78% (Table 3).

Regarding the Reach-dimension, three studies did not report or incompletely reported participants’ characteristics [47‐49] (see Table 3). The sample sizes at baseline varied from 9 [49] to 202 subjects [55]. The participants’ age ranged from 31.3 [51] to 52.6 [50] years. Overall, the proportion of female nurses in the studies was between 70.6% [53] and 100% [54]. The primary outcomes of the studies in regard to the impact of the intervention assessed (Effectiveness), were the functional status of the locomotor system and pain severity [45, 46], perceived job stress [51], mental health-related quality of life [50], well-being [50, 52], different competences [48], as well as ressources, irritation and burnout [53]. Four studies did not report on the effectiveness of the respective intervention [46, 47, 49, 55]. The vast majority of the outcome variables examined in the studies showed no statistically significant time x group interaction effects. In some cases, significant differences were found, e.g. with regard to impairment by pain on everyday movement [45], perceived job stress [45], mental health-related quality of life [44] relationship to residents [41] or sense of community [53]. Regarding Adoption, seven studies [47, 48, 50, 52‐55] reported the amount and type of the participating institutions. The number of institutions varied between one [52, 53] and eleven, with the targeted implementation in about 150 facilities [47]. Nine studies did not report on Implementation. In two studies [47, 49] it was stated that the intervention was modified (e.g., shortening of intervention period). Seven studies did not report follow-up results in order to evaluate interventions’ Maintenance [47‐50, 52, 53, 55]. In the remaining studies, long-term changes on targeted outcomes were assessed after three [45], six [51, 54], twelve [51], and 24 months [46]. The available results indicate for example a perceived reduction of job stress after a stress management training compared to the waiting control group [51]. Regarding physical health, results on Maintenance indicate a sustained reduction of time in stressful trunk postures [54] as well as a reduced pain severity on everyday movement [45] and a reduced restriction of maximum degree movement [46] by the respective intervention.

To our knowledge, our study is the first review on workplace health promotion for nurses in Germany. The national focus of the review is due to the specific social law basis for workplace health promotion in Germany. However, this focus is associated with the limitation that no conclusions can be drawn regarding international comparisons. The strength of the study lies in the consideration of the nursing settings, the quality criteria for prevention measures of the statutory health insurers, different health burdens and also the RE-AIM criteria. Nevertheless, our review has some limitations. Thus, very few studies could be included in the review, which are hardly comparable due to different approaches and also different reporting quality of the results. A further limitation arises from the challenge of conceptually differentiating the nursing settings. Knowing that very different patient groups are cared for within home-based LTC and LTC facilities care as well as in hospitals (e.g. people with disabilities, children, sick and healthy elderly people, etc.) and there are also “mixed forms or hybrid forms” (e.g. geropsychiatry) we have decided not to make any further distinctions. We have decided to take the perspective of WHP providers according to which there are usually programmes for employees in home-based LTC, LTC facilities as well as acute medical hospitals. Due to the small number of studies, a further target group-specific differentiation of the results (e.g. according to the qualification of the nurses or according to trainees) was not possible. Beyond, our research was limited to scientific publications. Project reports that were not published as scientific publication in one of the databases used were not taken into account.