Individual and organisational factors in the psychosocial work environment are associated with home care staffs’ job strain: a Swedish cross-sectional study

Background Home care staff (HCS) provide essential service to enable older adults to age in place. However, unreasonable demands in the work environment to deliver a safe, effective service with high quality has a negative impact on the individual employee’s well-being and the care provided to the older adults. The psychosocial work environment is associated with employees´ well-being, although, knowledge regarding which individual and organisational factors that contribute to job strain for HCS is limited. These factors need to be identified to develop targeted interventions and create sustainable work situations for HCS. This study aimed to explore how HCS´s perceived job strain is associated with, and to what extent can be explained by, individual and organisational factors of the psychosocial work environment and psychosomatic health. Method An explorative cross-sectional questionnaire survey design was used in a large Swedish county. Five home care agencies with a total of 481 HCS were asked to respond to a questionnaire regarding their perceived level of job strain (Strain in Dementia Care Scale), psychosocial work environment (QPSNordic34+), and psychosomatic health (Satisfaction with Work Questionnaire). Multiple linear regression (MLR) analyses were conducted to explore the association between job strain and individual and organisational factors. Results In total, 226 (46%) HCS responded to the questionnaire. Both individual and organisational factors were significant predictors of job strain and explained a variance ranging between 39 to 51% (p = 0.001). The organisational factor job demand and the individual factor feeling worried and restless was most frequently represented in these MRL models. A higher job strain was also associated with adverse outcomes regarding leadership, organisational culture and climate, and control at work. Conclusion This study indicates that there is an intertwined complexity of individual and organisational factors that are associated with the HCS´s perception of job strain. Implementation of new multidimensional work strategies, such as a reablement approach, could support the development of efficient strategies for HCS and reduce the level of job strain. Policy changes for the provision of home care are also needed to support the development of a sustainable and healthy psychosocial work environment. Supplementary Information The online version contains supplementary material available at 10.1186/s12913-022-08699-4.


Background
To enable ageing in place for older adults, the provision of home care can be crucial. However, with an increasing ageing population in combination with an estimated shortage of nurse assistants and nurse aides in the home care workforce [1,2], the health and social care welfare system will require financial resources as well as an increase of employees [1,3,4]. To retain and attract new employees within home care, in particular nurse assistants and nurse aides, it is of essence to develop a sustainable work environment that also is attractive. However, before a new sustainable work environment can be developed, information and understanding about the home care staffs' (HCS) current work situation is required. Therefore, we want to explore how HCS perceives their psychosocial work environment and factors in the work environment that are associated with job strain.
In Sweden, home care for older adults is subsidised and provided by both public and for-profit organisations. The purpose with home care is to provide support in activities of daily living (ADL), social aspects, and medical distribution [1]. Nurse assistants and nurse aides are commonly employed within home care in Sweden, and women in this cohort are overrepresented (< 80%). The HCS´s work is independent, where the staff has limited collaboration with colleagues and managers as they perform their work in the homes of their clients´ [1,5]. This creates demands on the individual worker to manage time, assess necessary skills and resources to support each client, and having to rely on one's own competence and capacity [6]. Even so, to be employed as HCS in Sweden today, limited or no formal education within health and care is required, and formal education is usually equal to secondary school level [7,8].
Throughout the last decades, organisational and regulatory changes have occurred within the home care service system in Sweden [1,9]. Some of these changes have had negative consequences and increased the strain on HCS to uphold a sustainable service, which impacts both the individual and organisational psychosocial work environment [9]. One example is the expanded amount of tasks assigned to the HCS in combination with organisational restrictions, which have led to an increased workload that requires greater competencies to provide expected and proficient care [1,5,10,11].
The psychosocial work environment and the organisational structure are multifaceted phenomena that are significantly associated with the health and well-being of care staff in different settings [7,[12][13][14][15]. There are wellestablished relationships regarding demand, resource, and support [16,17] and how this impacts the work situation and health for the staff [18][19][20]. The composition and balance of the psychological work environment and the organisational structures, empowerment, and good management, are essential parts to prevent physical and psychological ill-health and can increase the job satisfaction for HCS [1,5,[21][22][23][24]. In contrast to job satisfaction, job strain is a concept that deals with the perception of stress and strain related to work and reflects the perceived burden of the work environment [25]. Research targeting a diversity of occupations has identified that job strain is related to several factors in the work environment, such as leadership, conflicts at work, education and work experience [25][26][27][28]. In addition, experiencing high job demands, working independently without appropriate support, feeling out of control, or being unable to manage issues during work can further influence the perceived levels of job strain [16,26,27]. Combined, these issues increase the risk to negatively impact a person's well-being and lead to health problems such as insomnia, increased incidences of stress and burnout, psychological disruptions, or longterm sick-leave [7,10,14,23,26,27,29].
Research targeting job strain for nurse assistants and nurse aides has mainly focused on those working in the context of residential nursing homes rather than those working within home care [6,13,25,28]. In residential nursing homes, personal factors regarding level of education and feeling worried and stressed, in combination with work-related factors of working day time, and insufficient organisational and environmental support, caring climate, and leadership, have been associated with high job strain [14,29]. The consequences of high job strain does not only negatively affect the individual but it also impacts the organisation in terms of high turnover rates [1,29]. Subsequently, the older adult who is receiving the service can also be affected in terms of unattainable relationships with the staff and reduced quality of the received care, which could lead to a decreased quality of life [22,29,30].
Within a changing welfare organisation that demands stricter requirements for the provision of home care, and with tasks that command a higher level of competencies for the HCS, it is essential to investigate which individual and organisational factors of the psychosocial work environment that may contribute to a higher or lower job strain. There is limited research regarding job strain within the setting of home care as well as which factors that are associated with job strain for HCS. Hence, by gaining knowledge regarding which individual and organisational factors that impact the HCSs' perceived job strain, more efficient and appropriate interventions could be developed to create a sustainable work situation and thereby facilitate the possibility to have a healthy and satisfied workforce.
The aim of this exploratory study is twofold; 1) to explore how HCS perceive their level of job strain, and 2) to examine how job strain is associated with, and to what extent job strain can be explained by, individual and organisational factors of the psychosocial work environment and psychosomatic health factors.

Methods
This is an explorative study using a cross-sectional questionnaire survey design. Descriptive statistics were used to describe characteristics of the participating HCS, as well as levels of HCSs' perceived job strain, their psychosocial work environment, and psychosomatic health factors. Inferential statistics were used for pre-analysis and to explore individual and organisational factors that were associated with job strain by conducting multiple linear regression (MLR) analysis.

Setting and sample
Via the research & development units in Stockholm county, 26 municipalities were invited to participate in the study. Five of the invited municipalities agreed to participate, and within these municipalities, five home care agencies (public and for-profit) consisting of 17 home care units agreed to partake. Agencies that were included met the following criteria 1) having at least 30 employees (home care staff) who had been employed at least three months and had a contract equalling ≥ 50% employment (20 h/week), and 2) provide home care service to at least 30 older adults.
In total, 481 HCS employed at these units were requested to respond to the questionnaire.

Instruments
The last author (A-M.B) compiled a survey that consisted of three standardised instruments (described below) to assess the dependent variable job strain and the independent individual and organisational variables.
The instruments were chosen as they focus on job strain, work environment, and perceived health. They are all considered to be concise and easy to administrate and have previously been used with diverse professions, including HCS, in a Swedish context. None of the instruments require a license.

Dependent variables Job Strain
Strain in dementia care scale (SDCS) The SDCS is a self-reported instrument and addresses the perceived level of job strain among care staff. The SDCS was developed by a multi-country research group to explore nurses' experience of strain when working in dementia care [10,31]. SDCS has later been used for care staff working with older adults with and without dementia in residential nursing homes and home care settings [6,13,31,32]. The aim of the SDCS is to identify the level of job strain within different aspects of the work environment. The instrument's developer refers to Knapp´s definition of strain "as the effects of stress 'the wear and tear itself '", but also explains that the causality relationship between stress and strain are inter-related with a complex system of variables associated to stress [32].
The SDCS consists of 27 statements concerning the care staffs' work situation. For each statement (item), two aspects are investigated; 1) how frequently the situation occurs and 2) how much stress each situation generates. Both aspects are measured on a four-point Likert scale; 1 = never/no stress to 4 = very often/high stress. The 27 statements are allocated into one of the following five factors: Frustrated empathy-F1 (7 items), Difficulty understanding and interpreting-F2 (7 items), Balancing competing needs-F3 (5 items), Balancing emotional involvement-F4 (4 items), and Lack of recognition-F5 (4 items) [31].
The level of perceived job strain is calculated by multiplying the response of frequency and stress for each statement which creates an output between 1-16. A higher number indicates a higher perception of job strain [6,13,31,32]. The score of the total job strain, and for each factor, is calculated by adding the scores of each included statement and then dividing the sum with the number of statements included in the whole SDCS, or within each factor [31].
Psychometric properties for SDCS have been deemed acceptable. Internal consistency is valid at α 0.91 to α 0.94 for the whole SDCS and between α 0.53 to α 0.90 for the five factors [6,13,31,32]. No cut-off scores exist for the SDCS; however, previous research presents mean scores between 2.7 and 6.85 [6,8,13,32]. In the present study, a small adaption of the instrument was made to cohere with the HCS´s work environment, and the word resident/client was changed to older adult, which has been done previously [6].
In this study, the total score of SDCS (referred to as the total job strain) and the sum scores of each SDCS factor (F1-F5) is respectively represented as a dependent variable (DV), giving a total of six DV's. There is one MLR model for each one of the six DV's.

Independent variables
Independent variables are divided into individual and organisational factors. Individual factors consist of characteristic data of the participating HCS, and the HCS's self-related health, while the organisational factors consist of psychosocial factors at work. Every variable from the individual and organisational factors are considered potential independent variables in the forthcoming MLR models.

Individual factors
Characteristic data Characteristic data of the participants included information such as gender, age, Swedish as a first language, formal education, education in care/ caring, work time, permanent position, and years of work experience within home care services.

Satisfaction with work questionnaire (SWQ)
The SWQ is an instrument that assesses the work environment for staff who are working with care for older adults [33]. Four questions from the SWQ subscale "psychosomatic health aspect" (PH) were extracted and included in the survey. These four PH questions focus on the staffs' experiences during the last 90 days regarding; 1. feeling unhappy and depressed, 2. having sleep problems, 3. feeling worried and restless, and 4. feeling physically fatigued after work. The responses are rated on a five-point Likert-scale with the options 1 = very often to 5 = never [33].

Organisational factors
General Nordic questionnaire for psychological and social factors at work (QPSNordic) The QPS Nordic measures the psychosocial factors of the work environment and is developed from organisational theories to investigate the relationship between work, health and productivity [34]. It is a general self-assessed instrument that evaluates an employee's perception of their psychological and social work-life in combination with the organisational work relationships within a person's work-life [35].
The QPS Nordic questionnaire consist of 118 work-related items [35,36]. In the present study, the shorter version of QPS Nordic 34+ , consisting of 37 questions, was used. The first 34 questions are assembled into eight subscales: Job demands (4 items), Role expectations (3 items), Control at work (6 items), Predictability at work (1 item), Mastery of work (1 item), Social interactions (4 items), Leadership (2 items), Organisational culture & climate (8 items), and Perception of group work (2 items). Responses to questions 1-34 are made on a fivepoint Likert-scale, ranging from 1 = very seldom/never to 5 = quite to very often/always. The last three questions (35 to 37) are not included in any subscale and treated as single items and have the response options 1 = do not agree/not at all to 5 = fully agree/very much. The response frequency is presented in percentages for each item or as a reduced three-scale, where responses 1 and 2, as well as 4 and 5, are combined, or as mean scores for each subscale of QPS Nordic 34+ [34][35][36]. Limited evidence exists for the validity of the QPS Nordic 34+ subscales [37,38].
In this study, we chose to treat the single item number 37 as an individual factor instead as an organisational factor since the question is not formulated as a stress situation at work but rather stress on a more general basis.

Data collection
Data were collected with one web-based and one paper questionnaire with the following structure: characteristic information, the SDCS, the four PH questions, and the QPS Nordic Included home care agencies received oral and written information about the project. A consent letter and a link to the web-based survey was sent out to all HCS at these agencies. The web-survey was accessible between April to June 2018 and consent was given by starting the survey. Due to a low response rate during this period, a paper questionnaire was utilized during September to November 2018. A researcher from the team visited all units and informed participants about the questionnaire. Staff who had participated in the web-based survey were asked not to partake a second time. The staff deposited the completed paper questionnaire in a secure response box at the unit, which was collected after two weeks.
Data from non-responders were not collected.

Data analysis
All data were checked by the first and third author and transferred to SPSS version 26.0 [39] where all analyses were performed. If participants had changed an answer by crossing over or erasing their response, it was treated as missing if consensus was not reached between the two researchers.

Pre-analysis
Missing data Missing data was investigated for the factor groups (F1-F5) in SDCS, the subscales in QPS Nordic

34+
, and the PH responses. All five SDCS factors had 20 to 25% missing data; the subscales in QPS Nordic 34+ had 1.8 to 8.4% missing data. Some data was systematically missing; however, most responses were missing at random. A multiple imputation (MI) method was used to replace missing data. MI substitutes values and constricts ambiguity about missing values [40] and improves the validity, increases precision, and enables robust statistics [41]. The outcome of the MI is a mean value that is based on five imputation cases, which is presented in a pooled data set [40]. Missing data from characteristic data was not imputed.
Internal consistency of SDCS & QPSNordic34 + Internal consistency for the total score and factor/subscale scores for SDCS and QPS Nordic 34+ were measured with Cronbach´s Alpha (α), with a cut-of at α 0.70, but with minimal acceptance at α 0.65 to 0.70 [42]. The QPS Nordic 34+ had inconsistent results where only the total QPS Nordic 34+ and four out of nine subscales had valid outcomes (α 0.68 to 0.89): Job demand, Control at work, Leadership and Perception of group work. Three subscales were re-built to reach a valid α by testing the sets of items that corresponded to the subscale of the full version of QPS Nordic [38], these were: Role expectation (α 0.69), Social interactions (α 0.73), and Organisational culture and climate (α 0.84). The two subscales Prediction at work and Mastery of work only consisted of one or two items initially, hence, they were not possible to re-build. All items that in the end did not belong to a subscale were treated as single items.

Responses of SWQ
The responses of the SWQ were inverted so that a higher score indicated poorer health.

Statistics
Descriptive statistics were used for the participants' characteristics and for all items in each instrument, using frequencies and percentage for categorical variables, and means and standard deviations for continuous variables.
Inferential statistics such as chi-square and ANOVA test were applied for analysing differences of job strain between the categorical variables, multi-choice responses and interval characteristics. Normal distribution and homoscedasticity were examined to determine if a MLR could be conducted. Multicollinearity was applied to ensure basic assumptions for MLR and the included variables, with the cut-off set to r < 0.80, VIF < 5, and Tolerance < 0.40 [43].

Multiple linear regression analyses
In total, six MLR models were assessed, where total job strain and each of the five SDCS factors (F1 to F5) were used as a DV in each of the six models. Individual factors in terms of characteristics of participants, the four PH questions, and the QPS Nordic 34+ single item 37, as well as organisational factors in terms of subscales and single items in the QPS Nordic 34+ , were all considered possible independent variables in all six MLR models. Initial analysis identified eligible independent variables for each forthcoming MLR model. Each independent variable had to meet the criteria for the unstandardized regression coefficient (B) p-value at < 0.05, as well as having a significant correlation (p = 0.05) with the DV. An Adjusted R 2 above 0.20 was considered good based on previous research within the field where subjectively perceived data has been used [44].

Results
The outcomes are presented below in the following order: a description of the sample, descriptive outcomes of the DV's regarding mean level of perceived job strain from the total SDCS score and the five SDCS factors (F1-F5), as well as from specific items in SDCS. This is followed by outcomes of the individual and organisational independent variables, and finally, the results of the six MLR analyses are presented.

Description of the sample
Out of a possible 481 staff members, 226 (47%) from 17 units responded to the survey. A majority of the participants were women (80%), had education in care/caring (86%), and had on average worked 13 years within home care (Table 1).

Descriptive outcomes of the dependent variables
Our first aim was to explore the perceived level of job strain for HCS.
The mean job strain level for the HCS in this study was 4.43 (SD 1.8). Regarding the sub-factors of SDCS, the factor Lack of recognition (F5) was rated with the highest mean of job strain, 5.34 (SD 2.67). F5 also includes the item I want to do much more for older persons than my employers will allow, which received the highest mean job strain score, 6.55, in the whole SDCS. The factor Difficulties understanding and interpreting (F2) received the lowest mean of job strain 3.08 (SD 1.66) and also included the item that received the lowest mean in the whole SDCS, which was I have difficulties understanding older persons' needs (2.69) ( Table 2).

Descriptive outcomes of independent variables Individual factors
Descriptive information of individual factors regarding the participants' characteristics, outcomes of the PH questions, and the QPS Nordic 34+ single item 37 are presented in Table 1.
More than 70% of the HCS perceived that they sometimes or very often felt unhappy and depressed Table 1 Characteristics of participants and mean score of total job strain (SDCS), outcomes of the four psychosomatic health aspects, and the QPS single item 37 Note: Characteristic data presented in n (%) and y = years. Strain of Dementia Care Scale (SDCS) presented with mean and standard deviation (SD) for the total job strain in the whole sample and each characteristic domain except for variables with years. Psychosomatic Health aspects (PH) presented in % with a reduced response scale 1 & 2 = never/rather seldom, 3 = sometimes, 4 & 5 = rather/very often. QPS Nordic 34+ item presented in % with a reduced response scale 1 = not at all to 5 = very much. Missing value for each variable: a 0,4 to 2,6%, b 4,4 to 7%, c 19% (PH1), had problems sleeping (PH2), and felt worried and restless (PH3). In addition, 40% perceived that they often felt rather/very stressed (item 37) ( Table 1). No significant differences were found within the different characteristic groups regarding total mean SDCS score.

Organisational factors
Four out of seven subscales in the QPS Nordic 34+ were rated with low mean scores, indicating a negative outcome. These four subscales were Job demand (2.55), Leadership (2.72), Organisational culture and climate (2.78) and Control at work (2.87). More than 50% of the HCS perceived they had too much to do (item 2), were not able to influence the amount of work (item 10), the work pace (item 11), or decide when to take a break (item 12) (Table 3). In addition, nearly half of HCS perceived not receiving sufficient support from their managers to develop their skills (item 20) or being encouraged to participate in important decisions (item 21), and 55% of the HCS perceived that the managements' interest in the staff 's health and well-being (item 34) was absent (Table 3).  ) was found for the subgroups Perception of group work (3.88), as well as Role expectation (3.47). Within these two subscales, the majority of the HCS (70%) rated that they quite often or very often appreciated belonging to their group at work (item 26) and they considered the group to be good at solving problems (item 27). Furthermore, 67% of the HCS reported having well-defined goals for their work (item 7), and 90% knew what was required of them at work (item 8) (Table 3).

Individual and organisational factors associated with job strain
The second aim of this study was to explore the associations of individual and organisational factors with job strain. The results from the six MLR models are presented in Table 4, showing that both individual and organisational factors are associated with job strain in all six models.
The MLR model with Job strain (total SDCS scoring) as the DV, identified five variables that explained 51% of the variance. The organisational factors were Job demand (subscale) and Incompatible requests (item 9), while the individual factors were Feeling stressed (item 37) and the two PH questions related to the perception of Feeling unhappy and depressed (PH1) and Worried and restless (PH3).
In the MLR model with Lack of recognition (F5) as the DV, six variables explained 46% of the variance. F5 addresses HCSs' desire to do more for their clients and being recognised for their work by the client's family, their manager, or by others. A higher level of lack of recognition was associated with the organisational factors perceiving incompatible requests from others (item 9) and the two subscales Job demand and Organisational culture and climate. Organisational culture and climate includes aspects such as if one is in a supportive and comfortable work environment, being encouraged to improve things at the workplace, or being rewarded for a Note: The scale ranges from 1 = very seldom/never to 5 = very often/always for items 1-34, 1 = do not agree to 5 = fully agree for item 35 and 36, and 1 = not at all to 5 = very much for item 37. Mean and SD for each subscale. Mean, SD for each item, and percentage frequency on the reduced response scale for each item. Cronbach´s Alpha (α) for all subscales. ↑ Indicates positively formulated questions, where a higher score relates to a positive impact. ↓ indicates negatively formulated questions, where higher scores relate to a negative impact  well-done job. Additionally, the individual factors of feeling stressed (item 37), feeling worried and restless (PH3), and the level of education, were also significantly associated with a high lack of recognition (Table 4). The MLR model with Frustrated empathy (F1), Balancing competing needs (F3) and Balancing emotional involvement (F4) as DVs, identified individual and organisational factors that explained 39, 40 and 43% of the variance (Table 4). Finally, the model with Difficulty understanding and interpreting (F2) as the DV, had the lowest explanation rating (20%), but included the two most common variables in all MLR models: the organisational factor Job demand and the individual factor Feeling worried and restless (PH3).
The organisational factor Job Demand (subscale) was associated with the DV´s in all six MLR models and the individual factor Feeling worried and restless (PH 3) was associated with job strain in five out of six MLR models. In three out of six MLR models the individual factor Feeling stressed (item 37) and the organisational factor Perception of receiving incompatible requests (item 9) were associated with job strain.

Discussion
The aim of this study was to explore how HCS perceive their level of job strain, and how job strain was associated with, and to what extent job strain can be explained by, individual and organisational factors of the psychosocial work environment and psychosomatic health factors. As follows, we will focus the discussion on three areas. First, we will discuss the complexity of job strain, secondly, the most frequent factors in the MLR models that explained job strain, and lastly, strategies that can be implemented to reduce job strain among HCS.

Complexity of job strain
The results of this explorative study show that HCSs' perceived job strain is explained by a combination of both individual and organisational factors within the psychosocial work environment. The variety of factors that explain job strain for HCS in this study ranges from perceiving high job demands, organisational culture and climate, and role expectations, to education level and feeling stressed, unhappy, and worried. Other studies have also pointed out that leadership factors, lower competence level, low control of decision-making, and unmet needs are associated with higher job strain and a negative psychosocial environment [6,7]. Hence, job strain is a multifaceted and a complex concern. Even so, job strain is more complex than just considering the above stated variables. Therefore, we will continue the discussion to acknowledge factors in the context that would be possible triggers to the outcomes of job strain.

Contextual factors influencing the complexity of job strain
The contextual factors that potentially influences the HCS job strain lies within the paradigm of today´s eldercare, where aging-in-place with home care is prioritised to living in nursing homes [1,4,8]. This has not only created new demands and challenges for the older adult, but also for the HCS [1]. Since older adults who now are living at home are frailer with multiple health conditions, leading to complex health issues, more advanced care and support is required [1,5,45]. Hence, this leads to implications for the HCS in regards of an increase of different tasks imposed on the HCS, increased workload, as well as demands of increased competencies [1,4,9,10,47]. In addition, the HCS have strict time regulations for their support, deficiencies in collaboration with health professions, a staffing shortage, and are not provided with skill development opportunities [11,46,47]. Hence, the combination of all these factors reflects the complexity of job strain, as multitude aspects have to be considered in relation to each other.
Because of the present situation, there is a need for changes to ensure a sustainable work environment for HCS. When updating national guidelines and policies for home care systems and developing organizational changes to improve the HCS´s psychosocial work environment, a span of approaches must be taken into consideration, to include both individual and organisational aspects.

Frequent factors explaining job strain
Since job strain is complex and influenced by a combination of individual and organisational factors, we will address the variables that most frequently explained job strain in this study. These were the two individual factors: Frequently feeling worried and restless and Feeling stressed and the organisational factors: Job demands, Role expectations, Organisational culture and climate, and Perception of receiving incompatible requests.

Job demand and individual factors
The organisational factor of the QPS Nordic 34+ subcategory Job Demand was associated with high job strain in all six MLR models (Table 4) and the two individual factors feeling worried and restless and feeling stressed were found to be associated with job strain in five out of six MLR models. The subfactor Job Demand includes questions about irregularity and amount of work as well as one's competencies to manage the work tasks.
In this study, 70 to 86% of the HCS´s perceived a high job demand, as well as feeling worried, restless, and stressed. These results align with previous Swedish research involving staff in both home care and residential care, who also emphasized that high job demands and feelings of worrying were associated with higher job strain [6,13,16,27]. When not being able to manage or influence your work and not having time for unexpected situations, the combined situation leads to higher job strain for HCS [48] In addition, it has been concluded that the workload for HCS's many times are too high, being one reason why HCS´s consider quitting their work [49,50].

Job strain and ill-health
A majority of the participants in this study reported that they felt depressed and had sleeping problems, which were strongly associated with perceiving a higher level of job strain. These results are troublesome, as research has shown that a higher job strain increases the probability of depressive symptoms and major depression [18,51]. The combination of psychosomatic issues, high job strain and demands, low control of work, and low support, are factors leading to a risk for ill-health [16,51]. Ill-health for HCS can lead to consequences that not only affect the individual but also the recipients of the service as well as the organisation. A newly published report from the Swedish National Board of Health and Well-fare [11] concludes that HCS in Sweden are on sick leave twice as often compared to other professions in the health and well-fare sector. Sick leave influences the stability of staffing, which could lead to decreased quality of care and safety for the older adults who are receiving the service [11]. In addition, attention should be drawn to how the staffs' health could be viewed as a resource [29,52]; where a healthy staff can contribute with positive outcomes for the organisation in terms of ensuring staff continuity and thereby providing a safe, efficient service with higher quality to the older adults. Research targeting how working conditions for HCS could influence the delivered care perceived by older adults is limited. However, a newly published Swedish study within this field found a positive association between low job strain for the HCS and older adults' perception of overall satisfaction, staff treatment and sense of security [53].

Organisational support and the managers' role
In addition to the organisational factor Job demand, the organisational factor Organisational culture and climate was also significantly associated with the two MLR models with Frustrated empathy (F1) and Lack of recognition (F5) as DVs (Table 4), where a negative perception of support and recognition contributed to a higher job strain. Managers who overlook their staff, are not concerned about the staffs' health, do not enable communication, or have staff that experience an unencouraging and unsupportive environment, can create additional problems to an already strained situation [7,28,54]. The independent work and the limited contact with colleagues and managers can further contribute to a more strained situation [55]. In this study, the HCS perceive a lack of continuous and sufficient communication within the work setting and perceived that the management was not interested in their health and wellbeing. Hence, increased communication and support from managers, as well as from the organisation overall, is crucial for dealing with negative job strain and to promote positive health and well-being among HCS [28]. In the setting of home care, a more active communication is required by managers compared to other workplaces, such as residential care. In home care, there is a lack of spontaneous meetings, which negatively influence the HCSs' possibility to address issues that occur throughout the day and also the possibilities to receive support from the manager.
The support and leaderships behaviours of the HCSs' managers are associated with the level of job strain and a negative psychosocial work environment can result in psychological disruptions and long-term sick-leave [7,13,14,28,56]. By enabling a supportive leadership, as well as a supportive organisation, there may be possibilities to reduce job strain and develop a positive atmosphere that protects against adverse health effects among staff [28].

Strategies to reduce job strain among HCS Communication and reflection
Efficient strategies to decrease job strain regarding Frustrated empathy and Lack of support could be to enable time and resources for group sessions with a focus on self-reflection and communication. Previous studies have shown that HCS perceive limited time for self-reflection, developing necessary communication skills, and meeting colleagues before or after their shift, all of which they considered important [5,48,57]. Efficient strategies to improve the HCS situation could be to provide and enable time and resources to develop sessions for selfreflection and communication. A supportive relationship among colleagues in combination with sharing and learning together, could build health-promoting relationships and considered to be strategies to improve the psychosocial work environment [29]. Previous studies indicate that intervention programs that focus on knowledge translation regarding evidence-based and person-centred care can be a support to reduce job strain [8].

Reablement programs to address job strain in home care
Another strategy to reduce job strain would be to apply a reablement approach, where teamwork, structured planning, and communication is central [58]. Reablement has been tested and implemented in home care organisations in several countries [59][60][61][62]. Research has identified that HCS find reablement beneficial for them as their work becomes more efficient, older adults have a more positive attitude towards them, and their roles shift from delivering home care to providing a more person-centred support, reinforcing their roles within health care [57,63].
Reablement facilitates a structure to the HCS work as the base entails a provision of person-centred care where older adults are empowered to do needed or desired everyday activities by themselves or in collaboration with the HCS, rather than having HCS doing activities for them [59,64,65]. In addition, reablement is also a useful approach to incorporate opportunities to share experiences and reflect upon these experiences with colleagues as a part of the work strategy [57,66]. Reablement has so far only been implemented on a small scale in Sweden, although research is ongoing [66,67]. However, to learn more about how reablement can support the improvement of, and contribution to, a sustainable work environment for HCS, more research is needed to explore if and how reablement could contribute to reduce job strain for HCS.

Limitations and strengths
A limitation was the low response rate in this study, which was 47%. Although only one large county was targeted, the 266 respondents were distributed in 5 municipalities, among 5 home care agencies with a total of 17 home care units. However, the participants characteristics: gender, age, years of experience working within home care, having Swedish as a first language, and education level, reflects the population in other studies that has included home care staff in Sweden [6,54,56].
With regards to limitations, there was missing SDCS data in this study. Similar issues have occurred in previous studies [6,13,31], although it is unclear if data in those studies were missing at random and to what degree. In this study, 20-25% of the SDCS data were missing. This problem concerned the second aspect of the SDCS response option about stress in the specific situation, where the data was mostly missing at random. This could be a design error in the questionnaire, as the two aspects were responded to on the same row as the question, resulting in the possibility that the participant did not observe the need to respond to two different aspects (Supplement 1). Orrung-Wallin et al. [13] had similar problems but considered the questionnaire to provide valuable information although results could be systematically biased. Even if SDCS provides valuable information, systematic bias due to a design fault in a questionnaire is considered problematic for its validity. To deal with the missing data in this study, a MI method was used instead of single imputation as in previous studies [6,13,31].
The strength of using MI is that it is considered a more valid and robust procedure where values are randomly replaced by plausible values [68] and account for uncertainties associated with the imputed value [69]. Although, with such a high number of missing data in one aspect, other methods could be considered and evaluated to better manage the problem.
In regards of instruments that assesses the psychosocial work environment for home care staff, other tools such as the COPSOQ [70] are available However, we chose to use the QPS Nordic + 34 because of the familiarity of the instrument and the possibility to compare the results of the present study with previous studies in this area.

Conclusion
This study indicates that there is an intertwined complexity of individual and organisational factors that affect HCSs' perception of job strain. This complexity requires implementation of new work strategies that are multidimensional, aiming to reduce the level of job strain and thereby create a positive psychosocial work environment for HCS. Future research should focus on implementing reablement in Sweden and explore if reablement can be a method to reduce job strain for HCS.
In addition, the strained situation for HCS implies a pressing need for policy changes within the home care system. Policymakers should focus on developing more appropriate and sustainable strategies to improve the psychosocial work environment for HCS.