nursing staff competence, psychosocial work environment ...169399/fulltext01.pdfnursing staff...

ACTAUNIVERSITATISUPSALIENSISUPPSALA2006

Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Medicine 210

Nursing Staff Competence,Psychosocial Work Environmentand Quality of Elderly Care: Impactof an Educational Intervention

HENNA HASSON

ISSN 1651-6206ISBN 91-554-6750-4urn:nbn:se:uu:diva-7383

List of Papers

This thesis is based on the following papers, which will be referred to in the text by their Roman numerals.

I. Verho H, Arnetz J.E. Validation and application of an instrument for measuring patient relatives’ perception of quality of geriatric care. International Journal for Quality in Health Care 2003;15(3):197-206.

II. Hasson H, Arnetz J.E. Nursing staff competence, work strain, stress and satisfaction in elderly care: a comparison of home-based care and nursing homes. Journal of Clinical Nursing2006; in press.

III. Arnetz J.E., Hasson H, Evaluation of an educational "toolbox" for improving nursing staff competence and psychosocial work environment in elderly care: results of a prospective, non-randomized controlled intervention. International Journal of Nursing Studies 2006; in press.

IV. Hasson H, Arnetz J.E. The impact of an educational intervention for elderly care nurses on care recipients' and family relatives' ratings of quality of care: a prospective, controlled intervention study. International Journal of Nursing Studies 2006; in press.

Reprints were made with permission from the publishers.

Contents

Introduction.....................................................................................................9Aims of the thesis .....................................................................................10The Swedish system of elderly care .........................................................10

A brief historical description ...............................................................10Laws regulating elderly care................................................................12Division of responsibilities ..................................................................13Costs of elderly care ............................................................................13Home-based care .................................................................................14Nursing homes .....................................................................................15

The main actors in elderly care ................................................................15Care recipients .....................................................................................15Family members of care recipients......................................................16Staff .....................................................................................................17

Work environment....................................................................................17Work environment regulations ............................................................18Stress and emotional strain ..................................................................18Physical strain......................................................................................19Competence .........................................................................................19Work satisfaction .................................................................................21

Quality of care..........................................................................................21Theoretical quality definitions.............................................................21Quality of elderly care .........................................................................22Quality regulations...............................................................................23Quality measurements .........................................................................23

Educational interventions in elderly care .................................................25The theoretical framework .......................................................................26

Hypothesized impact of the intervention according to the theoretical models..................................................................................................30

Methods ........................................................................................................32Designs .....................................................................................................33Settings .....................................................................................................34

The Örebro University Hospital ..........................................................34The elderly care organization of Bengtsfors........................................34The elderly care organization of Uddevalla.........................................35

Participants ...............................................................................................36Nursing staff ........................................................................................36

Care recipients .....................................................................................36Family members of care recipients......................................................37

A brief description of the intervention .....................................................37Data collection procedures .......................................................................38

Interviews and focus group discussions...............................................38Questionnaire studies...........................................................................39Nursing home costs and administrative data .......................................40

Outcome measures ...................................................................................40Staff competence and the psychosocial work environment.................40Quality of care .....................................................................................41Nursing home costs..............................................................................42Administrative data..............................................................................43

Statistical analysis ....................................................................................43Papers I-IV...........................................................................................43Additional analyses..............................................................................44

Power calculations....................................................................................45Ethical approval........................................................................................45

Results...........................................................................................................46Paper I ......................................................................................................46Paper II .....................................................................................................46Paper III....................................................................................................47Paper IV ...................................................................................................48Additional findings...................................................................................48

General discussion ........................................................................................50The impact of the educational intervention ..............................................50

Psychosocial work environment ..........................................................50Quality of care .....................................................................................53

Perceptions of competence and work environment..................................55Perceptions of quality of care...................................................................57Costs and administrative data of nursing homes ......................................58Methodological considerations.................................................................60

Considerations about the questionnaire studies ...................................60Considerations about the toolbox intervention design.........................61Statistical considerations .....................................................................62

Ethical considerations ..............................................................................63Future research .........................................................................................63Conclusions ..............................................................................................64

Acknowledgements.......................................................................................66

Sammanfattning (in Swedish).......................................................................70

References.....................................................................................................73

Abbreviations and definitions

ANOVA Analysis of variance, CI Confidence Interval (usually 95% CI) DCM Demand-Control Model HWO Healthy Work Organization NS Non-Significant OECD Organisation for Economic Co-operation and Development QI Quality Improvement QWC Quality-Work-Competence questionnaire RAI Resident Assessment Instrument SD Standard Deviation SEK Swedish kronor SPSS Statistical Package for the Social Sciences TOOL Tools for Optimal Organizational Load (name of the re-

search project in this thesis) TQM Total Quality Management UK The United Kingdom US The United States USD American dollars USÖ Örebro University Hospital VAS Visual Analogue Scale VIF Violent Incident Form

Term used in this thesis Definition Home help services Services offered to people living in their own homes or in

sheltered housing. Home nursing care Nursing care for people living in their private homes or in

sheltered housing. Home-based care All care and services offered in private homes or in sheltered

housing, i.e. home help services and home nursing care. Sheltered housing A group accommodation with services and care offered by

personnel in home help services or home nursing care. Al-ternative terms: ordinary home, service accommodation andhome for the elderly.

Nursing home A group accommodation with 24-hour nursing surveillance.

Nurses’ aides An employee whose level of education varies from little orno formal nursing training to a three-year upper secondary school Healthcare program. Alternative terms: nursing aide, nursing auxiliary and auxiliary nurse.

Practical nurses An employee whose education includes a three-year uppersecondary school Healthcare program. Alternative terms:nursing assistant, assistant nurse and enrolled nurse.

Registered nurses An employee whose education includes a three-year univer-sity program.

Nursing staff Nurses’ aides, practical nurses, and registered nurses.

9

Introduction

The need for competence development for elderly care nursing staff has increased during recent years (Socialstyrelsen 2004). Nursing care tasks, both in care recipients’ private homes as well as in nursing homes, have be-come more medically demanding as increasing numbers of elderly suffer from neurodegenerative disorders (Bergmark et al. 2000, Socialstyrelsen 2006e). Recent reports from Swedish elderly care have revealed low levels of formal staff competence (Socialstyrelsen 2004), as well as limited oppor-tunities for skills development at work (Socialstyrelsen 2005d). Insufficient competence regarding work tasks might lead to less control over the work situation and increased stress levels (Theorell 2002). Other reports have showed that the psychosocial work environment is poor in many elderly care organizations in Sweden (Bäckman 2001, Gustafsson & Szebehely 2005).

There have also been several recent alarming stories regarding poor qual-ity and negative care recipient outcomes in Swedish elderly care (Parker 2000). Prior research has reported an association between quality of care and staff competence (McAiney 1998), stress (Cohen-Mansfield & Rosenthal 1989) and work satisfaction (Chou et al. 2003, Redfern et al. 2002), i.e. lower quality of care has been related to lower competence, less satisfaction, and higher stress levels among nursing staff.

Few educational interventions in elderly care have been evaluated on out-come measures such as staff work satisfaction, psychosocial work environ-ment or quality of care (Aylward et al. 2003). In addition, earlier interven-tion projects aimed at improving elderly care nursing staff competence have also been criticized for methodological limitations such as small sample sizes, lack of a control group and lack of long-term evaluation (Aylward et al. 2003). Difficulties in transferring the knowledge gained from educational interventions to work practices have also been reported (Broad 1997). In this thesis an educational intervention with a “toolbox” design for nursing staff in home-based care and in nursing homes was developed and implemented. The impact of the intervention on nursing staff competence, the psychosocial work environment, and the quality of care was evaluated.

10

Aims of the thesis The overall aim of this thesis was to develop, implement, and evaluate an educational intervention for nursing staff in elderly care. In a prospective, controlled study, evaluation of the intervention was based on nursing staff ratings of their competence and psychosocial work environment, as well as on care recipients’ and their relatives’ ratings of the quality of elderly care. The aims of the papers included in the thesis were:

to validate a questionnaire concerning care recipient relatives’ per-ceptions of quality of care and to investigate their perceptions of the quality of elderly care (Paper I)

to compare self-rated competence, work strain, stress, and work satisfaction between nursing staff working in home-based care and those working in nursing homes (Paper II)

to evaluate the possible impact of an educational workplace inter-vention on nursing staff perceptions of their competence, work sat-isfaction, and psychosocial work environment (Paper III)

to evaluate the possible impact of the educational intervention on care recipients’ and relatives’ perceptions of the quality of elderly care (Paper IV)

Additional analyses that were not included in any of the papers were con-ducted for this thesis. The aims of the additional analyses were to investi-gate:

the possible association between nursing home costs and the work environment and between nursing home costs and the quality of care.

The Swedish system of elderly care In this chapter, the Swedish system of elderly care provision regarding the aspects relevant to this thesis will be briefly described.

A brief historical descriptionAround 1910 “old people’s homes” were becoming common to provide sub-sistence, lodging, and care for older people (Edebalk & Lindgren 1994), and in 1918 every local authority was advised to have a facility that functioned

11

as an old people’s home (Möller 1996). After the Second World War, there were attempts to reform and develop these homes (Möller 1996), but due to an economic crisis, the program was not fully implemented (Edebalk & Lindgren 1994). Instead, in the 1950s, a new ideology for elderly people living in their own homes and receiving home help services became popular (Socialstyrelsen 2006c). The home help service expanded throughout the 1950s (Edebalk & Lindgren 1994, Möller 1996), and continued to expand until the end of 1970 (Socialstyrelsen 2006c).

Political decisions in the 1960s dictated that the local municipalities would allocate resources to home help services and county councils for insti-tutional care (Bergmark 1997, Edebalk & Lindgren 1994). The ideology of having elderly people stay in their homes was powerful, and the county council institutions were only for those who could not be cared for at home (Edebalk & Lindgren 1994). Thus, county councils have traditionally pro-vided healthcare for the elderly, and municipalities have provided services (Riksrevisionsverket 2003).

In the 1970s elderly care, with respect to both municipalities and county councils, underwent rapid expansion (Edebalk & Lindgren 1994, Riksrevi-sionsverket 2003). Municipalities offered the elderly, including individuals not severely disabled, home help services such as shopping, preparing food, and cleaning (Bergmark et al. 2000), thus emphasizing the social needs of elderly citizens (Parker 2000). Since very old people continued to live in their own homes, the work of care assistants started to include tasks helping with personal needs, such as toilet assistance. This imposed requirements for further development of staff skills (Edebalk & Lindgren 1994).

In the 1980s, the expansion of elderly care began to slow down. The in-creased cost of home help services led to decreasing amounts of people re-ceiving home help. Home help began to focus on the most frail elderly peo-ple (Bergmark 1997, Edebalk & Lindgren 1994), with services also offered during the evenings and nights (Sveriges Kommuner och Landsting 2005b). During the 1980s the proportion of elderly people living in institutions de-creased dramatically, while the proportion of those living in service apart-ments with social service increased (Edebalk & Lindgren 1994).

Developments during the 1990s In 1992, the Elderly Reform Bill (ÄDEL) transferred the responsibility for nursing homes and other institutions for long-term medical care from the county councils to the municipalities (Thorslund et al. 1997). Municipalities also received the responsibility of providing healthcare and rehabilitation in nursing homes. Municipalities could also, in agreement with the county council, take on the responsibility for home nursing or primary healthcare for elderly individuals in their own homes (Riksrevisionsverket 2003, Sveriges Kommuner och Landsting 2005a). Approximately 55,000 staff members and 400 care units with 30,000 residents were transferred from

12

county councils to municipalities (Riksrevisionsverket 2003, Sveriges Kommuner och Landsting 2005b).

The first half of the 1990s was characterized by an economic recession in Sweden. National and local authorities made cutbacks in many areas of the welfare sector (Bergmark et al. 2000). County councils reduced the number of hospital beds, and the average length of stay in hospital emergency care units was shortened (Riksrevisionsverket 2003). The shortened hospital stays implied that a substantial part of advanced care was transferred to the mu-nicipalities. Elderly individuals who had previously been cared for in hospi-tals were transferred to municipal nursing homes (Socialstyrelsen 2006c). At the same time, municipalities substantially decreased the number of beds in nursing homes (Socialstyrelsen 2005e). The ambition was to allow elderly people to live in their own homes as long as possible (Thorslund et al. 1997). Municipal elderly care services were meant for the individuals with greatest needs. Younger and healthier elderly persons that were in less need of care and services were given lower priority when allocating municipal resources (Socialstyrelsen 2005e). This created a greater need for private resources, such as help from families, for those not receiving enough, or any, municipal care (OECD 2005).

Several other European countries (Kerkstra & Hutten 1996, Meijer et al.2000, Oberski et al. 1999, Paasivaara et al. 2003), as well as countries in the OECD (OECD 2005) have also had the ambition to build up community-based services to replace care in hospitals and long-term care institutions. For instance, in the United Kingdom the number of hospital beds available for long-term care has decreased, and communities have received the re-sponsibilities for management of elderly care services (Higgs et al. 1998). The main reason for the shift to home-based care is considered to be its lower cost in comparison to institutional care (Kerkstra & Hutten 1996, Mei-jer et al. 2000).

In conclusion, elderly care in Sweden, in line with international develop-ments, is increasingly being provided outside the traditional elderly care institutions. Elderly care is often provided in private homes. At the same time, elderly care staff has to deal with increasingly complex age-related disorders, both social and medical. Together, these changes have put a great emphasis on the need to enhance skills and knowledge among staff. The role of relatives of elderly care recipients has also changed, and they are viewed today as an important component in ensuring proper care of the elderly.

Laws regulating elderly careThere are two main laws regulating care and services for elderly people in Sweden. These laws are of a framework nature, and are thus designed to designate the state’s and municipalities’ responsibilities to provide for citi-zens’ needs, rather than emphasizing citizens’ rights (Parker 2000).

13

The Social Services Act (2001) grants all individuals assistance with dif-ferent aspects of living if their needs cannot be provided in any other way. The act states that municipalities are ultimately responsible for ensuring that people receive the support and assistance they need. The social services pro-vided should enable elderly people to continue living in their home environ-ments, or, if this is no longer possible, offer them sheltered housing, group accommodation, or nursing homes with services and care.

The Health and Medical Services Act (1982) maintains that medical care shall be available to all members of society. According to this act, healthcare and medical services aim to maintain a good standard of health among the entire population.

The Swedish legislation emphasizes universality and aims to serve all in-dividuals if needs arise, regardless of age, income or place of residence (Parker 2000). However, the laws include few elements of detailed regula-tions and allow local authorities autonomy in decision-making (Thorslund et al. 1997).

Division of responsibilities Responsibility for elderly care is divided between three levels of govern-ment. At the national level, the parliament and government set out aims and directives in the form of legislation and economic policy documents. At the regional level, the county councils are responsible for healthcare, if an agreement with municipalities taking over the responsibility has not been made, and at the local level the municipalities are legally obligated to meet the social service and housing needs of all elderly people living in the mu-nicipality (Socialstyrelsen 2005e). However, as stated in the previous sec-tions, the laws regulating elderly care do not provide details regarding mu-nicipalities’ obligations. All services and care for elderly people are based on a needs’ assessment, but the municipalities have the authority to decide crite-ria for the amount and type of help offered to elderly individuals (Thorslundet al. 1997). Thus, available elderly care services may vary greatly between municipalities.

The National Board of Health and Welfare and County Administrative Boards have a joint duty to supervise the provision of elderly care and the Swedish work environment authority is responsible for supervision of the work environment factors (Ansvarskommittén 2003).

Costs of elderly careSeveral approaches, such as an increase in private care providers and a de-crease in nursing home places in comparison to care in private homes, have been tried in order to decrease costs and increase efficiency in Swedish eld-erly care (Hörstedt et al. 1996, Socialstyrelsen 2006e).

14

Elderly care accounts for approximately 20% of the total expenditures of municipalities (Ansvarskommittén 2003), even though the total costs for care of and services for elderly have decreased somewhat during recent years (Socialstyrelsen 2006e). In 2001, the municipalities’ total costs were 82.8 billion SEK and in 2004 they were 79.6 billion SEK (in constant prices, ref-erence year 2004) (Socialstyrelsen 2006e)1. The decrease in total costs is mainly explained by a decrease in nursing home places (Socialstyrelsen 2006e). Nursing home care is the segment of elderly care that accounts for the majority, approximately 66%, of all municipal costs for elderly care (Socialstyrelsen 2006e). The median cost per resident in a nursing home was 483.700 SEK2 per year in 2004. However, large differences have been re-ported between municipalities’ costs for elderly care (Hörstedt et al. 1996, Socialstyrelsen 2006e).

Nursing home costs have been associated with different elements of care provision. An economic model associates greater costs with higher quality of care (Fleming 1990). In this model, quality of care is measured with struc-tural (input) measures of care provision such as staffing levels (Fleming 1990). Thus, for instance, higher staffing levels are expected to result in higher costs. Earlier empirical research on nursing home cost and quality has focused on the structural measures of quality (Nyman 1988, Weech-Maldonado et al. 2006). A quality management model expects higher quality to be associated with lower costs (Daigh 1991). Quality is conceptualized as freedom from deficiencies and is measured as outcome variables, such as the number of care recipients’ pressure ulcers (Weech-Maldonado et al. 2006). Extra time and resources are required to produce products and services with high quality. However, the extra resources serve as prevention of defects, which leads to lower cost (Daigh 1991). Few prior studies have evaluated nursing home cost in relation to the outcome measures (Nyman 1988, Weech-Maldonado et al. 2006).

In conclusion, attempts to reduce the cost of elderly care have been made, but the possible implications for the outcome measures of quality of care have not been empirically investigated.

Home-based care In general, home help services for elderly living in their own homes include practical assistance such as cleaning, laundry, shopping, bank and post office errands, food distribution, transportation service and security alarms. Home help services also include personal care such as assistance with eating, get-

1 Approximately corresponding to 11.3 and 10.8 billion USD: mean exchange rate SEK/USD in 2004: 7,35. Riksbanken (2006) http://www.riksbank.com/templates/stat.aspx?id=17212. Accessed 1 October 2006.

2 Approximately 65.813 USD.

15

ting dressed, and personal hygiene (Westlund & Edvardsson 1998). Com-plementary services also include daytime activities that can be offered to elderly individuals living in their own homes (Sveriges Kommuner och Landsting 2005a).

Home nursing care for elderly people living in their own homes includes care and rehabilitation by registered nurses (Sveriges Kommuner och Landsting 2005a). The Elderly Reform Bill made it possible for municipali-ties to take the responsibility for home nursing care. In 2004, half of the mu-nicipalities had the responsibility for home nursing care (Sveriges Kommu-ner och Landsting 2005a). As a result, home nursing has become a part of the social services and often belongs to the same organization as the home help services (Kerkstra & Hutten 1996). The needs’ assessment procedure precedes a decision of receiving home nursing care (Thorslund et al. 1997), but no referral from a doctor is needed (Kerkstra & Hutten 1996).

After the developments of the 1990s, the nature of home-based care has changed (Blomberg et al. 2000). It has been reported that it became more difficult for elderly to obtain home services. In addition, increased nursing elements in home help services, as well as more active help from families and voluntary organizations, have been identified (Blomberg et al. 2000).

Nursing homes For elderly people with more extensive care requirements, nursing home placement can be offered by the municipalities (Sveriges Kommuner och Landsting 2005a). Between the years 2000 and 2003, cutbacks in the mu-nicipal budgets reduced the nursing home places by approximately 7400 (Socialstyrelsen 2005e). On a national level, there are indications that nurs-ing homes have been replaced by home help services, which have increased by approximately 7200 care recipients during the same time period (Socialstyrelsen 2005e). However, at the municipal level, only half of the municipalities that have cut back on nursing home care have expanded home help services. A comparison of ten western countries at the beginning of the 1990s showed that Sweden had the highest percentage of elderly people in the population, but a relatively low ratio of nursing home beds to elderly individuals (Ribbe et al. 1997).

The main actors in elderly care

Care recipientsThe tendency towards concentrating municipal resources on elderly people with the most extreme needs implies that a greater proportion of both home

16

care recipients and residents in nursing homes have physical and cognitive disabilities than previously (Bergmark et al. 2000). The majority of elderly people living in nursing homes have some form of dementia, often combined with other disabilities (Socialstyrelsen 2006e, Sveriges Kommuner och Landsting 2005a). Elderly people receiving home-based services or care often constitute a more heterogeneous group, where some need advanced medical care and others need help with personal care or general housework (Socialstyrelsen 2005e).

A total of 237,100 persons received elderly care in Sweden in 2004, i.e. 15% of the population aged 65 and above (Socialstyrelsen 2005f, 2005g). The total amount of care recipients has been unchanged during the 2000s. Meanwhile, the amount of persons receiving home help services has in-creased by 10% and those receiving home nursing care have increased by 5%. In contrast, elderly people living in nursing homes have decreased by approximately 11% (Socialstyrelsen 2005g). Approximately 75% of all care recipients were older than 80 years (Socialstyrelsen 2005g) and a majority were women (Socialstyrelsen 2006e). Approximately 40% of those receiving home help services received help between one to nine hours per month and 23% received between 10 and 25 hours per month (Socialstyrelsen 2005g).

Family members of care recipients Family members of elderly care recipients are playing increasingly important roles in care provision, since many of the elderly receive services and care both from the municipalities and from their relatives (Socialstyrelsen 2006e). Family members also act as monitors of the quality of care and report possi-ble inaccuracies (Bowers 1988, Socialstyrelsen 2006e). In addition, the Health and Medical Services Act states that if information cannot be sup-plied to the patient, the relatives should be informed instead, which further indicates relatives’ important role in care provision.

In Swedish home care, the official help from municipalities in 1994 was estimated to be 40% of the total help to persons over 75 years, and in 2000 the corresponding proportion was estimated to be 30% (Socialstyrelsen 2005e, Sveriges Kommuner och Landsting 2005a). Thus, the informal help seems to have increased during the last ten years. Informal care is most often provided by women between 45-64 years of age, and 10-20% of the Swed-ish population were estimated to be informal caregivers to a relative (Socialstyrelsen 2006b). International studies have reported that family members retain emotional ties and often visit their relative even when the relative has been transferred to a nursing home. Residents with little or no contact at all with their families are a minority (Bowers 1988, Ejaz et al. 2003).

17

StaffSeveral staff categories are engaged in elderly care. Swedish municipalities usually employ practical nurses, nurses’ aides, needs’ assessments officers, and supervisory staff. Since the Elderly Reform Bill came into effect in 1992, occupational therapists and registered nurses have also been employed by the municipalities. However, physicians are employed by county councils (Sveriges Kommuner och Landsting 2005b).

In many municipalities practical nurses and nurses’ aides comprise the largest professional groups (Fahlstrom & Kamwendo 2003). There is a high percentage of women working in elderly care, and a low percentage (less than 10%) of men (Socialstyrelsen 2005e). The proportion of men working in elderly care is even lower in small municipalities (Socialstyrelsen 2005e). The mean staff age is relatively high and approximately 22% of staff mem-bers are 55 years or older (Socialstyrelsen 2005e). In 2004, approximately 239,500 persons, including part-time staff and staff on sickness or parental leave, worked in municipal elderly care (Socialstyrelsen 2006e). Between 1995 and 2004, the total amount of staff has increased by 15% (Sveriges Kommuner och Landsting 2005a).

The National Board of Health and Welfare (2006e) recently reported that the greatest challenge in elderly care was the problem of ensuring sufficient and competent staff. Care work in general, and elderly care in particular, seem to have lost their attraction, and a constant need to recruit large num-bers of care staff for elderly care has been reported (Socialstyrelsen 2005e). A shortage of staff with adequate knowledge and skills in several other member states in the European Union (Kerkstra & Hutten 1996) and in OECD countries (OECD 2005) has also been reported, both in nursing homes and in home care.

Work environmentThis thesis deals mostly with the psychosocial work environment. However, some aspects of the physical work environment and physical strain are also considered.

The psychosocial work environment has been defined as psychological work demands, influence and control over work, good contact with and sup-port from supervisor and fellow workers, stimulation from work and oppor-tunities for development (Källestål 2004, Theorell 2003). In elderly care settings, a lack of sufficient leadership and support (Arbetarskyddsstyrelsen 2000, Meyer & Muntaner 1999, Oberski et al. 1999), influence and control over the work (Bäckman 2001, Gustafsson & Szebehely 2005, Johansson 1995), as well as stimulation from the work (Johansson 1995) have been reported. In addition, an increased degree of dependency among elderly care

18

clients has increased the staff’s physical workload (Oberski et al. 1999). In fact, high physical and emotional demands at work were considered to be the main reason that many staff members were not able to work full-time (Gustafsson & Szebehely 2005).

Work environment regulations The Work Environment Act (1977) requires that work organization and work content should be designed so that staff is not exposed to physical strain or mental stress that may lead to illness or accidents. The employer’s obligations, such as investigation of work-related injuries and ensuring that staff has sufficient knowledge of work conditions and possible hazards, are also stated in the act.

The Swedish Work Environment Authority’s regulation regarding sys-tematic work environment efforts (Arbetsmiljöverket 2001) gives more spe-cific rules for employers to take their responsibility for work environment. Another regulation (Arbetsmiljöverket 1990) on home-based care states that staff working in private homes should have sufficient knowledge regarding the work tasks in the home environment. The Social Service Act also em-phasizes that staff should have suitable training and experience to perform their work tasks in social services.

As mentioned earlier, the Swedish Work Environment Authority has the responsibility for supervision of the work environment in Swedish work-places and they may carry out workplace inspections for that purpose.

Stress and emotional strain There does not seem to be a single, accepted definition of stress (Hasson 2005). However, most of the stress theories agree that “stress is adaptive, that it is associated with threatening or harmful events, and that it is typically characterized by aversive or unpleasant feelings and mood” (Dougall & Baum 2001 p. 321). Stress has been shown to affect emotions, health, pro-ductivity, and performance (Dougall & Baum 2001). There are several theo-retical frameworks for work stress (Nelson et al. 2001) but it is not within the scope of this thesis to describe these models. However, the Demand-Control Model (DCM) will be briefly presented, since the concept of skills utilization is included in the model. The Demand-Control Model suggests that the impact of psychosocial work demands on staff health is moderated by the level of decision latitude (work control). The authors proposed that jobs with high stress levels are characterized by high job demands and low employee control (Karasek 1979, Karasek et al. 1988). Skills utilization and development are a part of the control (decision latitude), i.e. if staff has op-portunities to develop their competence they also have greater possibilities to control the different situations occurring at work (Theorell 2002).

19

Prior research has found high stress levels among staff in nursing homes (Morgan et al. 2002, Proctor et al. 1999, van den Berg et al. 2006) and home care (Denton et al. 2002a, Laamanen et al. 1999). It has also been proposed that work stress and stress-related symptoms such as fatigue and exhaustion have increased during the past years among staff in Swedish elderly care (Arbetarskyddsstyrelsen 2000, Bäckman 2001). In addition, nursing staff have reported higher stress levels than the working population in general (Wall et al. 1997). However, lower levels of stress have also been identified among staff in psychiatric services for elderly people (Baillon et al. 1999), in units for severely demented people (Hansson et al. 1995), and among home care workers (Bartoldus et al. 1989, Johansson 1995).

In the occupational context, heightened stress or prolonged stress have been proposed to be related to decreased work satisfaction (Boswell 1992), increased absenteeism (Vagg & Spielberger 1998), higher staff turnover (Sjogren et al. 2005, Yin & Yang 2002), a greater number of injuries (Clarkeet al. 2002), and poorer quality of care (Cohen-Mansfield & Rosenthal 1989).

Physical strain A general increase in physical workload in Swedish elderly care has been identified both in home care and in nursing homes since residents’ care dependency has increased (Gustafsson & Szebehely 2005, Socialstyrelsen 2005e). Increased physical workload also seems to be a trend in other Euro-pean countries (Meijer et al. 2000). A questionnaire study of elderly care staff in eight Swedish municipalities revealed that between 20-30% of home care and nursing home staff reported that they were often physically tired after work (Gustafsson & Szebehely 2005).

The most common physically strenuous aspects of elderly care work have been reported to be heavy lifts and inappropriate working positions (Bäckman 2001). One reason for the heavy physical work strain might be the use of poor or inadequate equipment when transferring care recipients (Arbetarskyddsstyrelsen 2000). In addition, toilet areas that are too small, both in private homes and in nursing homes, caused unsuitable working po-sitions. The limited knowledge of staff regarding ergonomics has also been reported (Arbetarskyddsstyrelsen 2000).

CompetenceAs described earlier, several of the laws regulating elderly care require that staff have the appropriate education and experience for executing their work tasks. However, the formal level of education among elderly care nursing staff is reported to be low, with only 56% of nursing assistants and practical nurses having a formal upper secondary (high school) education suitable for

20

their profession (Socialstyrelsen 2004). The proportion of staff with formal education is higher in nursing homes compared to home help services (Socialstyrelsen 2004). Nursing assistants’ and practical nurses’ opportuni-ties for skills development at work are also limited (Socialstyrelsen 2005d). In addition, the guidance given to these occupational groups has been re-ported to be limited, since there are few doctors and registered nurses working in elderly care settings (Riksdagens revisorer 2002, Socialstyrelsen 2005a).

In a recent review of Swedish literature regarding the work of elderly care nursing staff, several different competence areas necessary to nursing work were identified (Socialstyrelsen 2006d). Nursing staff needs to be able to carry out tasks related to nursing care, social relations, housework, medical tasks, and administrative work (Socialstyrelsen 2006d). Work competence in elderly care has also been defined as knowledge that is needed to take care of a home and the physical body of another individual as well as skills to listen to and identify needs and wishes of a care recipient (Runesson & Eli-asson-Lappalainen 2000). In addition, the ability to individualize the care for each care recipient is needed, together with theoretical knowledge of medi-cal, psychiatric, and social aspects of work (Runesson & Eliasson-Lappalainen 2000). Thus, the definition of work competence for elderly care nursing staff is broad and the fact that care recipients today have more com-plex medical and psychiatric needs than previously has placed new demands on nursing staff competence (Proctor et al. 1999, Robertson & Cummings 1996, Socialstyrelsen 2004, Sung et al. 2005). Areas such as dementia and dementia treatment (Brodaty et al. 2003, Sung et al. 2005), mental health problems (Hsu et al. 2005), medical work tasks such as injections, physical diseases with special treatment (Brulin et al. 1998), aggressive behavior (Skovdahl et al. 2003), palliative care (Ersek et al. 1999, Raudonis et al. 2002), and employees’ rights (Wills et al. 1998) have been reported to be in need of competence development.

As described earlier, sufficient competence regarding work tasks im-proves staff opportunities to control different work situations (Theorell 2002). An earlier study showed that nursing aides’ performance of work tasks that were not included in their education contributed to uncertainty (Brulin et al. 2000). In addition, risk for burnout and negative attitudes has been shown to be higher for nurses with low education levels, since the less qualified nurses have a greater exposure to residents’ behavioral distur-bances, and they are less skilled in managing these behaviors than nurses with higher levels of education (Astrom et al. 1991). Other studies have as-sociated lack of skills and training opportunities with low work satisfaction (Brodaty et al. 2003), high staff turnover (D'Eramo A et al. 2001, Sung et al.2005), and low quality of care (McAiney 1998, Rantz et al. 1999).

21

Work satisfaction Earlier studies have found a number of organizational, professional, and personal variables to be related to work satisfaction (Lu et al. 2005). For instance, work stress has been proposed to be closely related to work satis-faction (Blegen 1993, Lu et al. 2005). In addition, social support (Chou et al. 2002, Navaie-Waliser et al. 2004), feedback (Jansen et al. 1996), interaction with care recipients (Denton et al. 2002a, Moyle et al. 2003, Navaie-Waliseret al. 2004), development opportunities (Jansen et al. 1996, Moyle et al.2003), leadership (Grieshaber et al. 1995), and influence on work (Friedmanet al. 1999, Hansson et al. 1995) have been reported to be positively related to work satisfaction. Personal variables, such as age and education level have also been proposed to be related to work satisfaction (Grieshaber et al.1995). However, work environment variables have been reported to be more related to work satisfaction than personal characteristics, implying that adap-tation of work environment is crucial to improving work satisfaction (Friedman et al. 1999).

Prior research has reported high and moderate work satisfaction in home care (Denton et al. 2002a, Jansen et al. 1996) and in nursing homes (Brodatyet al. 2003, Grieshaber et al. 1995, Moyle et al. 2003). However, low satis-faction (Haggstrom et al. 2004) has also been reported. For instance, almost half of the registered nurses in 12 municipalities in southern Sweden re-ported being dissatisfied with their working situations (Weman et al. 2004).

Quality of care Theoretical quality definitions There are several theoretical concepts for defining quality and a great many empirical evaluations. One of the most frequently used quality models in service organizational research defines quality as having three aspects: cus-tomer outcome, customer process, and prerequisites for the service(Edvardsson 1996, 1998). Edvardsson (1998) argues that the customer judges the quality of a service based on these outcomes and he/she is a co-producer of a service in the customer process. Outcome and process quality are dependent on the organization’s prerequisites for the service, i.e. the resources of the service provider, in terms of technical and physical re-sources as well as administrative routines and procedures (Edvardsson 1998).

In the field of medical research, Donabedian’s definition of quality is probably the most well-established model. Donabedian (1966) proposed a quality of care framework including structural, process, and outcome meas-ures of quality. Structural factors are organizational characteristics such as staffing levels, management, facility size, and consumer characteristics such

22

as functional level of clients (Dellefield 2000, Schirm et al. 1999). Process refers to what is done for care recipients, including aspects such as quality assurance work, cleaning, infection controls, and restraint use (Dellefield 2000). Outcome refers to the results of the care and services, such as preva-lence of pressure ulcers, malnutrition, and falls as well as client and family satisfaction (Dellefield 2000, Schirm et al. 1999). Donabedian’s framework proposed that the three quality aspects are causally linked; structure leads to process, and process leads to outcomes. Later on, it was also proposed that structure directly influences outcomes (Sainfort et al. 1994). Thus, the basic assumption is that good organizational structure and care processes are linked to desirable outcomes (Davis 1991).

Regardless of the different research traditions in service organizational re-search and medical research, the models proposed above have similar char-acteristics. The models include outcomes as one measure of quality. It is proposed that outcomes are influenced by service delivery processes and organizational factors. The main difference between the models seems to be the view of the customer/patient. Service organizational research treats the customer as co-producer in the service delivery processes. The traditional medical view of the patient has included a more passive role. However, in recent years the medical and nursing research has also reinforced the patient position in Sweden and other countries (Nordgren 2003, Sitzia & Wood 1997). This is in line with a general trend towards individualism in the soci-ety, and in fact the word patient has been replaced in many organizations by terms such as consumer, customer, client, or service user (Nordgren 2003, Sitzia & Wood 1997).

Quality of elderly care A great variation in quality of elderly care has been reported between the OECD countries and between care providers within a country (OECD 2005). Swedish elderly care does not have a long tradition of quality assurance work or standardized quality evaluation. However, recent scandals regarding unacceptable conditions in elderly care have increased the interest in quality controls (Parker 2000). In general, quality measurement and improvement strategies in elderly care have been reported to lag behind the methods used in emergency care (OECD 2005).

The Swedish County Administrative Boards conducted inspections of elderly care organizations in almost all of the Swedish municipalities in 2004 (Socialstyrelsen 2005b). Deficiencies were identified in half of the inspec-tions, with the most common situations regarding staff behavior and poor documentation processes (Socialstyrelsen 2005b). Reports of inaccurate municipality healthcare, according to the Lex Maria (see below), consisted most often of improper medication, late or incorrect diagnoses, slip and fall accidents and poor cooperation between different caregivers resulting in

23

unmet care needs (Socialstyrelsen 2006e). Inadequate care in institutional settings in other OECD countries has included inadequate treatment of chronic pain, pressure ulcers, inappropriate use of chemical or physical re-straints, inadequate housing, and poor social relationships (OECD 2005).

Quality regulations The Social Services Act and the Health and Medical Services Act require “good quality” for the services and care provided. The acts also require sys-tematic and continuous quality development work. Quality development work is further defined in The National Board of Health and Welfare’s rec-ommendations for social services (Socialstyrelsen 1998) and for healthcare (Socialstyrelsen 2005c). It is stated that an elderly care organization should have a system for establishing quality goals, as well as for planning, per-forming, evaluating, and developing quality (Socialstyrelsen 1998). The goal of the quality system is to ensure that the individual’s care and service needs are met, and the system should include methods to secure aspects such as needs-assessment procedures, responsibilities of staff members and organi-zations, adequate staff competence, as well as reports of adverse events and complaints (Socialstyrelsen 1998).

Lex Sarah, a chapter in The Social Services Act (2001), also states that persons who are active in elderly care shall verify that good care and secure living conditions are offered. The act requires that inappropriate care must be reported, which is also required of staff working in healthcare according to the chapter called Lex Maria in The Health and Medical Services Act (1982).

Quality measurements Quality evaluations often utilize structure, process, and outcome measures of quality. These aspects of quality can be measured according to an expert-oriented approach containing objective and often technical aspects of quality (Gummesson 1992). In the healthcare context expert-oriented objective measures have commonly been used (Socialstyrelsen 2003). Many elderly care organizations in several of the OECD countries have used quality re-quirements such as room size, staff ratios, prevalence of pressure ulcers, infections, slip and fall accidents, restraints and anti-psychotic drug use (OECD 2005).

Another approach, a user-based quality definition, implies that quality is what gives satisfaction to a customer and that quality is measurable in the sense of asking a customer about the service experience (Gummesson 1992). The users’ answers can be used as input for quality development (Gummesson 1992). The user-based view on quality measurement is tradi-tionally included in service organizational research (Seth et al. 2005) and

24

also increasingly in medical and nursing research (Axelsson 2000). In fact, half of the Swedish municipalities reported using client satisfaction as a tool to evaluate quality of elderly care (Socialstyrelsen 2001).

Elderly care recipients’ perceptions of quality It has been recognized that the elderly care recipients’ viewpoints on quality and content of care can offer important information for care organizations in identifying potential areas for improvement (Aharony & Strasser 1993, Duffy et al. 2001). Several Swedish elderly care organizations started with measurements of clients’ perceptions of quality in the 1990s. However, no nationwide studies have been conducted (Socialdepartementet 2002, Social-styrelsen 2006e). In the USA a requirement that nursing home quality meas-urements should incorporate assessment of resident satisfaction has been codified in legislation (Bowers et al. 2001a).

Prior research has in general found positive care recipients’ perceptions of quality of care (Curry & Stark 2000, Curtis et al. 2005, Pearson et al. 1993, Socialdepartementet 2002). However, critical ratings have also been reported concerning some aspects of care, such as information received from staff (McCartan-Quinn et al. 1996), continuity of staff (Socialdepartementet 2002), social activities available (McCartan-Quinn et al. 1996), opportunities to influence care (Socialstyrelsen 2006a), and opportunities to participate in decision-making regarding the care provision (Wressle et al. 2006).

Several characteristics of elderly care recipients have been found chal-lenging when including them in quality measurement activities (Bowers et al. 2001a). For a great proportion of elderly care recipients, dementia may limit their ability to report perceptions of quality (Bowers et al. 2001a, Wilde et al. 1995). However, a majority of long-term care residents has been reported to be able to produce logical responses to a quality of care survey on incontinence care (Simmons & Schnelle 1999). Another common prob-lem regarding quality measurements with elderly people has been a tendency toward high quality ratings (Chong 2003, Owens & Batchelor 1996). It is reported that elderly care recipients give answers consistent with prevailing social norms rather than answers that are their real personal responses (Forbes & Neufeld 1997); they may not be willing to express critical views due to vulnerability to staff (Grau et al. 1995, Owens & Batchelor 1996). Elderly people have also been reported to have low expectations regarding care and therefore to be more easily satisfied with services and care than younger people with higher expectations (Grau et al. 1995).

Family members’ perceptions of quality It has been recognized that family members can also be a source of informa-tion regarding the quality of care, especially since they have been shown to be very interactive in the care processes (Ejaz et al. 2003). Relatives’ views on quality of care can be an alternative source of information when patients

25

have difficulties expressing their views (Ejaz et al. 2003, Lubart et al.2004a). Relatives’ perceptions of quality may also be a complement to care recipients’ perceptions, since the views of these groups have been shown to differ from each other (Ejaz et al. 2003). It is also recognized that family members are likely to influence or make decisions for elderly care recipients, which further points to the importance of family perspectives (Ejaz et al.2003). However, like care recipients, the family members may also be un-willing to report dissatisfaction because of fear of staff retaliation against their relative (Grau et al. 1995, Hertzberg & Ekman 1996).

Earlier research on family members’ perceptions of elderly care quality has found that relatives were highly satisfied with nursing home care (Lubartet al. 2004b, Maas et al. 1991) and with home care (Chiu 1997). However, relatives have also reported dissatisfaction with specific aspects of care, such as their own possibilities to participate in care provision (Maas et al. 1991), information they receive from staff members regarding care recipients’ ill-nesses and treatments (Wellwood et al. 1995), activities provided to care recipients (Maas et al. 1991), and ignorance of care recipients’ personal preferences when planning care and services (Curry & Stark 2000).

Educational interventions in elderly care It has been proposed that interventions to strengthen staff care provision in terms of improving the working conditions and increasing staff competence are possible ways to improve care quality (Wiener 2003). However, few evaluations of educational initiatives in elderly care settings have been con-ducted (Aylward et al. 2003, Beck et al. 1999, Ross et al. 2001). Aylward etal. (2003) reviewed literature regarding continuing educational programs in long-term care facilities and found 48 educational studies. The Beck et al.(1999) literature review of dementia training identified six interventions evaluating dementia training programs for nursing assistants in nursing homes.

Earlier educational interventions have been reported to suffer from seri-ous methodological weaknesses, such as small sample sizes, lack of a control group, or low response rates (Aylward et al. 2003, Beck et al. 1999). In addi-tion, a lack of long-term measurements has been reported to limit the evalua-tion of the effects of the educational interventions (Aylward et al. 2003). For instance, two thirds of the reviewed educational studies by Aylward et al.(2003) included only immediate post-intervention evaluations and lacked long-term follow-up evaluations.

It has thus been suggested that more rigorous research on the implementa-tion and effectiveness of continuing education is needed (Aylward et al.2003). However, there are several factors that have been reported to chal-lenge the provision of staff education in elderly care organizations. For in-

26

stance, lack of fiscal and human resources, lack of time for staff to attend lectures, and poor staff motivation regarding education have been reported to make it difficult to conduct educational interventions in elderly care (Fitzpatrick & Roberts 2004, Ross et al. 2001).

The focus of previous educational interventions has often been on resi-dent mental health. A total of 19 of the 48 interventions identified in the Aylward et al. (2003) literature review focused on resident mental health issues. The reviewed interventions also focused on chemical or physical restraint reduction, incontinence, and oral care. However, none of the re-viewed interventions focused on staff work environment issues, except for a few studies aiming to change staff attitudes towards elderly people. Adding staff work environment perspectives to interventions aiming to improve resi-dents’ wellbeing might be important, since work environment is one crucial factor for staff ability in providing care (Wiener 2003).

The educational interventions have often been conducted in a staff train-ing program format (Aylward et al. 2003). However, a need for innovative education formats has been identified (Ross et al. 2001, Stolee et al. 2005), primarily since difficulties transferring and implementing new knowledge into practice after a staff education program have been identified (Broad 1997).

Most of the reviewed educational interventions (Aylward et al. 2003) were evaluated on staff knowledge, attitudes, or behavior, and only two stud-ies on staff work satisfaction. A total of 14 interventions were evaluated on resident outcomes, and the majority of the interventions only used one out-come variable for evaluation (Aylward et al. 2003). A literature review re-garding staff educational programs in dementia care suggested that future intervention evaluations should include outcome measures of improvements in staff knowledge (Beck et al. 1999). They recommended measures of fac-tors such as staff, resident, and family satisfaction in order to evaluate the effect of increased staff knowledge.

The theoretical framework The theoretical framework of this thesis is based on the organizational re-search regarding healthcare organizations and the quality management theo-ries presented earlier. The basic assumption of the quality theories was that good work organization, care processes, and adequate staff competence are linked to desirable outcomes (Davis 1991, Donabedian 1966, Edvardsson 1998). This approach will be further developed with the theoretical models presented below.

Sainfort et al. (2001) developed a theoretical model that integrates the to-tal quality management (TQM), quality improvement (QI) and healthy work organization (HWO) theories, all of which emphasize the importance of

27

management practices, employee empowerment, and supportive organiza-tional culture. Sainfort et al. (2001) proposed that adopting quality im-provement methods, such as total quality management, not only improve the quality of care, but also contribute to the transformation of an organization into a healthy work organization. By “healthy work organization” the au-thors mean an organization with both a healthy workforce and financial suc-cess. The “Organizational Health and the Total Quality Management (TQM) Framework” is theoretical, but according to the authors it is based on prior empirical research investigating relationships between work organizations, quality of care, and staff outcomes in hospital settings.

Work Organization• Staffing• Job design• Reward system• Etc

Environment• Setting of care• Direct environment• External environment• Etc

Organizational Culture• Management style• Organizational values• Organizational climate• Etc

Technology• Technological systems• Quality/reliability• Training• Etc

Patient• Health• Safety• Satisfaction• Quality of life

Organization• Performance• Quality

Staff• Health• Safety• Performance• Satisfaction• Quality of working life

Care Processes• Performance• Quality

Work Organization• Staffing• Job design• Reward system• Etc

Environment• Setting of care• Direct environment• External environment• Etc

Organizational Culture• Management style• Organizational values• Organizational climate• Etc

Technology• Technological systems• Quality/reliability• Training• Etc

Patient• Health• Safety• Satisfaction• Quality of life

Organization• Performance• Quality

Staff• Health• Safety• Performance• Satisfaction• Quality of working life

Care Processes• Performance• Quality

Figure 1. Organizational Health and the Total Quality Management (TQM) Frame-work by Sainfort et al. (2001).

At the left of the “Organizational Health and the Total Quality Management (TQM) Framework” are the organizational variables – work organization, organizational culture, technology, and the environmental factors affecting the organization. The organizational variables interact with the intermediate outcomes that are in the middle of the model. The intermediate outcomes are staff outcomes and care processes. According to the model, staff health and care processes will be directly affected by the organizational variables. At

28

the right of the model are the final outcomes – patient health, safety, satisfac-tion, quality of life, and organizational performance and quality. Both the organizational variables and the intermediate variables affect the final out-comes directly. The organizational factors may also affect the final outcomes through their impact on the intermediate variables. In addition, the authors proposed that the intermediate and the final outcomes may also influence the organizational variables. However they did not, for some reason, draw that relationship to the “Organizational Health and the Total Quality Manage-ment (TQM) Framework”.

Since this model was developed for healthcare organizations in general, some aspects of the model, such as technological factors, may not be as rele-vant in elderly care settings as in other healthcare environments. For this reason, another theoretical model will be presented. The “Model of hypothe-sized relationships between work satisfaction, work stress, quality of care and resident/patient wellbeing” was developed especially for elderly care settings by Hannan et al. (2001). The model has basic features that are simi-lar to those of the framework proposed by Sainfort et al. (2001), but it is based on the findings in the authors’ literature review regarding work and quality of care in elderly care settings (Hannan et al. 2001).

Institutional factors• Management• Training & education• Supervision• Group cohesiveness• Staffing levels• Workload

Staff outcomes• Work satisfaction• Work stress• Role conflict & ambiguity• Work commitment• Control over work• Sickness absence

Home factors• Job & home stress• Social support

Staff moderators• Age• Gender• Ethnic group• Staff grade• Tenure• Part/full-time• Marital status• No/ages of children• Other dependants

Quality of care

Resident/patient wellbeing

Resident/patient moderators• Age• Gender• Ethnic group• Dependency• Social support• Marital status• No/ages of children &grandchildren• Other dependants

Institutional factors• Management• Training & education• Supervision• Group cohesiveness• Staffing levels• Workload

Staff outcomes• Work satisfaction• Work stress• Role conflict & ambiguity• Work commitment• Control over work• Sickness absence

Home factors• Job & home stress• Social support

Staff moderators• Age• Gender• Ethnic group• Staff grade• Tenure• Part/full-time• Marital status• No/ages of children• Other dependants

Quality of care

Resident/patient wellbeing

Resident/patient moderators• Age• Gender• Ethnic group• Dependency• Social support• Marital status• No/ages of children &grandchildren• Other dependants

Figure 2. Model of hypothesized relationships between work satisfaction, work stress, quality of care, and resident/patient wellbeing by Hannan et al. (2001).

The model consists of input factors in the left side. These are institutional factors, such as management, training, staffing levels as well as home factors such as stress regarding balance between home and job and support received from home environment. The input factors are suggested to be directly re-lated to staff outcomes, such as work satisfaction, work stress, and sickness absence. The authors proposed that the effect of the institutional factors is expected to be stronger than that of the home factors since more support was

29

found for this in the literature review. Staff outcomes are suggested to be directly related to quality of care and resident wellbeing. However, this rela-tionship may be influenced by staff and resident characteristics, which act as moderators. The solid arrows in the model indicate relationships that are relatively well-established in the literature and the broken arrows indicate relations that are suggested, but not clearly established.

The model does not consider that the outcome factors may, in turn, affect the organizational variables, which was proposed by Sainfort et al. The rela-tionship between staff and resident outcome factors and the organizational variables has also been proposed in an empirical study (Robertson et al.1995). Robertson (1995) considered the influence of these factors to be a “mutual feedback system”. The author suggested that factors like low work satisfaction lead to a decline in quality of care, which further diminishes the staff work satisfaction. Thus, work satisfaction and quality of care reinforce each other in a workplace (Robertson et al. 1995).

Neither of the theoretical models presented above emphasizes the influ-ence of family members of care recipients on the outcome measures. The model by Hannan et al. (2001) recognizes marital status of care recipient and number of children as patient moderators, but does not develop the role of relatives and family members as factors influencing the care quality. There-fore a modified version of the “Model of hypothesized relationships between work satisfaction, work stress, quality of care and resident/patient wellbeing” by Hannan et al. (2001) is proposed.

Institutional factors

Staffoutcomes

Quality of care

Care recipient wellbeing

Family members of care recipient

Institutional factors

Staffoutcomes

Quality of care

Care recipient wellbeing

Family members of care recipient

Figure 3. Modified version of “Model of hypothesized relationships between work satisfaction, work stress, quality of care, and resident/patient wellbeing” by Hannan et al. (2001).

The modified model proposes that family members of the elderly care re-cipients have a possible influence on staff outcome, quality of care, and care recipient outcomes. Edvardsson (1998) proposed that customers are co-producers of the services provided. This thesis considers both care recipients and their family members as co-producers of elderly care services. The modified model places staff competence and skills as staff outcomes. Staff competence is related to staff moderators such as staff grade and can be in-fluenced by institutional factors such as training and education. Staff compe-

30

tence in turn affects work satisfaction, stress, and work performance, in addi-tion to quality of care and care recipient wellbeing.

The model also proposes that care recipient wellbeing affects staff out-comes and the institutional factors, as was proposed by Sainfort et al. (2001) and Robertson (1995). The modified model functions as a theoretical frame-work for this thesis. Some variables from the original model of Hannan et al.(2001), such as home factors, staff, and patient moderators, are not empiri-cally measured in this thesis and thus not included in the modified version of the model. These variables are still considered to be possible confounding variables. The modified model is mainly used to give structure to the evalua-tion of the intervention (Papers III and IV).

In conclusion, the empirical research behind the theoretical models focus-ing on the outcome variables of this thesis reported that staff competence correlates negatively with staff work stress levels, i.e., the lower one’s per-ceptions of competence and skills, the higher the feelings of work stress (Dunn et al. 1994, Morgan et al. 2002). In addition, staff have reported be-coming more comfortable and confident at work after having an opportunity to improve their skills (Morgan et al. 2002). Work stress has been revealed to be closely related to work satisfaction, i.e. increased levels of work-related stress have been associated with lower levels of work satisfaction (Boswell 1992, Lu et al. 2005). A strong negative association has also been reported between work stress and quality of care (Cohen-Mansfield & Rosenthal 1989), and work satisfaction and quality of care (Chou et al.2003).

Hypothesized impact of the intervention according to the theoretical models The model presented above was a foundation for the hypothesis constructed for the evaluation of the educational intervention in this thesis. The toolbox intervention was hypothesized to have a direct positive impact on staff com-petence and several aspects of the psychosocial work environment, such as social climate and staff possibilities to influence and participate. The inter-vention was also hypothesized to have an indirect positive impact on staff stress and work satisfaction through improved competence and psychosocial work environment.

It was also hypothesized that the intervention would also have a direct positive impact on certain aspects of quality of care, since these areas of care were the focus of some of the toolbox instruments. This could be the case, for instance, for activities provided for care recipients, since an instrument for recording care recipients’ preferences for leisure activities was included in the toolbox. Using the instrument, staff could gather information regard-ing care recipients’ preferences for activities and adapt the activities to better

31

suit their preferences. This could improve care recipient and their relatives’ quality ratings regarding the activities.

The intervention was also hypothesized to have an indirect positive im-pact on some other aspects of quality of care such as care recipients’ oral health and pressure ulcers, since staff knowledge and skills in these areas was expected to improve when using the corresponding instruments included in the toolbox.

Finally, the general improvement of staff psychosocial work environment and work satisfaction was hypothesized to improve the quality of care as proposed by the theoretical models. The improvements in quality of care would in turn improve care recipients’ wellbeing, which would eventually have a positive impact on both institutional factors and staff outcomes.

32

Methods

The four papers included in this thesis are based on a research project enti-tled “Tools for Optimal Organizational Load” (TOOL), which was con-ducted in two municipal elderly care organizations. The project aimed to assess an educational intervention on staff perceptions of the psychosocial work environment and care recipient and family members’ perceptions of quality of care. All four papers included in the thesis present results from questionnaire studies conducted in these municipalities between 2001 and 2004. Paper I also includes questionnaire data from a university hospital study that was aimed at investigating patient relatives’ perceptions of quality of care. Figure 4 illustrates the study design, settings, and participants in each of the papers.

Figure 4. Study setting, participants and design in Papers I-IV.

In addition, this thesis includes analyses of nursing home costs’ possible correlation with administrative data concerning resident and staff outcomes, staff ratings of the psychosocial work environment, and care recipients’ and

PAPER IPatient relativesÖrebro University hospital n=38

PAPER II

PAPER III

2001Care recipient relativesBengtsforsPilot study, n=184

Care recipient relativesBengtsforsMain study, n=134

2003

2001

Home care staffBengtsforsn=75

Home care staffUddevallan=223

Nursing home staffBengtsforsn=192

Nursing home staffUddevallan=373

2003

2004

Nursing staffBengtsforsn=270



Nursing staffUddevallan=606


PAPER IV

2001

2003

2004

Care recipientBengtsforsn=84



Care recipientUddevallan=378


Care recipient relativesBengtsforsn=149



Care recipient relativesUddevallan=582


PAPER IPatient relativesÖrebro University hospital n=38

PAPER II

PAPER III

2001Care recipient relativesBengtsforsPilot study, n=184

Care recipient relativesBengtsforsMain study, n=134

2003

2001

Home care staffBengtsforsn=75

Home care staffUddevallan=223

Nursing home staffBengtsforsn=192

Nursing home staffUddevallan=373

2003

2004






PAPER IV

2001

2003

2004











33

their relatives’ ratings of quality of care. The data concerning nursing home costs and administrative data about staff and resident outcomes were col-lected from the two municipalities involved in the TOOL project. The results of these analyses were not presented in any of the papers, but are presented as additional findings in this thesis.

DesignsPapers I and II were cross-sectional questionnaire studies. Papers III and IV were based on a prospective, non-randomized, controlled intervention. The intervention study used a quasi-experimental design, since the individuals and workplaces included in the intervention and control groups were not randomly assigned. The timeline of the intervention project and the study design are presented in Figure 5. The additional analyses regarding nursing home cost had cross-sectional designs.

Baselinemeasurement (I)

Toolboxintroduced (I)

6-monthevaluation (I)

Baselineevaluation (R)


12-monthevaluation (R)

Order from the Swedish Work Environment Authority (I)

Pilot questionnairestudy (I)

Focus groupdiscussions (I)

Order from the Swedish Work Environment Authority (R)

1998 Fall 2000 April 2001 Sept. 2001 2002 Feb. 2003 Sept. 2003 Sept. 2004

Baselinemeasurement (I)

Toolboxintroduced (I)


Baselineevaluation (R)


12-monthevaluation (R)

Order from the Swedish Work Environment Authority (I)

Pilot questionnairestudy (I)

Focus groupdiscussions (I)

Order from the Swedish Work Environment Authority (R)

1998 Fall 2000 April 2001 Sept. 2001 2002 Feb. 2003 Sept. 2003 Sept. 2004

Figure 5. Timeline for the intervention project, (I) = intervention municipality, (R) = reference municipality.

The TOOL project started in the elderly care organization of the intervention municipality (Bengtsfors). This organization had received orders from the Swedish Work Environment Authority to evaluate the working situation of the elderly care staff. High levels of sickness absence and work strain among nursing staff were also reported in the municipality’s internal reports. In response to the authority’s demand, the municipality’s occupational health service contacted the research team in 1998 asking for assistance in carrying out an investigation of the psychosocial work environment in the elderly care organization. The results of the investigation became a starting point for an educational intervention, “Toolbox”, which was implemented in the elderly care organization in Bengtsfors in 2003.

34

Settings The Örebro University Hospital The university hospital study (Paper I) was conducted in five departments at the Örebro University Hospital (USÖ). The departments that participated in the study were geriatrics, medicine, neurology, orthopedics, and the inten-sive care unit. USÖ had conducted hospital-wide patient questionnaire stud-ies to measure patient perceptions of quality of care. There was a need for a patient relatives’ questionnaire, since a large proportion of patients on these departments were not capable of replying to a questionnaire.

The elderly care organization of Bengtsfors Bengtsfors is a municipality in western Sweden with 10,335 inhabitants in 2004 (Bengtsfors kommun 2004). The amount of inhabitants has decreased in the municipality during the past ten years by approximately 10%. Those who have moved from the municipality have mainly been young people, with the result that approximately 25% of the population is over 65 years (Wigren et al. 2002). This can be compared to 17% in the surrounding county and in Sweden in general (Socialstyrelsen 2006c, Wigren et al.2002). During recent years many industrial companies have also closed down in Bengtsfors and the economical situation of the municipality has been strained (Kommunstyrelsen 2001, Wigren et al. 2002). The unemploy-ment rate was approximately 10% in both 2001 and 2004 (Bengtsfors kom-mun 2004, Kommunstyrelsen 2001).

The municipal elderly care organization in Bengtsfors made several changes in the provision of home care and nursing home care during the research project. In 2002 and 2003 the number of nursing home places was reduced and home care was reorganized into one unit instead of several geo-graphical units (Bengtsfors kommun 2003). All staff having permanent tenure kept their jobs, but the amount of part-time staff and substitutes was reduced (Bengtsfors kommun 2003). In addition, several new managers were employed for administration of social services and for the elderly care or-ganization, since turnover in these positions was high (Bengtsfors kommun 2004). The municipality also took measures for improving the process of elderly individuals’ need assessments and offered education in dementia care for some nursing staff (Bengtsfors kommun 2004). A systematic work envi-ronment plan was introduced in 2003 in all workplaces (Bengtsfors kommun 2003).

The elderly care organization of Bengtsfors includes one service apart-ment complex that is taken care of by a non-profit association, but home care services are offered by the municipal elderly care (Private communication Nils Andrée 6 October 2006). There are eight service apartments in total, and the care recipients in these apartments and their relatives were included

35

in the present study. The numbers of care recipients and staff members in the elderly care organization of Bengtsfors are presented in Table 1.

Table 1. Numbers of elderly care recipients and staff members in the elderly care organizations in Bengtsfors and Uddevalla.

Bengtsfors Uddevalla

2001 2004 2001 2004

Amount of care recipients ^ in: home help services 174 189 687 811 home nursing care * 79 (25) 112 (41) 420 (71) 316 (49) nursing homes 280 216 774 756 Total amount of care recipients † 479 446 1532 1616 Amount of staff members (part-time and full-time) in: home help services 95 78 home nursing care 8 18

406 458

nursing homes 279 271 593 626 Total amount of staff members 382 367 999 1 084

^ = References: (Socialstyrelsen 2002, 2005g) * = care recipients receiving both home nursing care and home help services, within parentheses the amount of care recipients receiving only home nursing care. † = the sum of care recipients receiving home help services, home nursing care (excl. those receiving both home nursing care and home help services) and individuals in nursing homes. = References: Bengtsfors: (Private communication Sara Molander 9 October 2006), Uddevalla: (Private communication Pär Levander 25

Oktober 2006). =the municipality’s statistics calculated only for home help services and home nursing care together

The elderly care organization of Bengtsfors was the intervention organiza-tion in the present project. The workplace toolbox was introduced at all workplaces in the elderly care organization in the intervention municipality.

The elderly care organization of Uddevalla The Uddevalla elderly care organization was included in the research project as a reference municipality in 2003. The organization had also received or-ders from the Swedish Work Environment Authority to evaluate the working situation of the elderly care staff. Reference organization elderly care staff had no knowledge of, or exposure to, the intervention.

Uddevalla is a municipality that had 50,068 inhabitants in 2004 (Uddevalla kommun 2004). The population trend has been upward; there has been an increase in the amount of inhabitants during recent years. For pur-poses of comparison there were 49,255 inhabitants in the municipality in 2001 (Uddevalla kommun 2001). The proportion of people over 65 years has been approximately 21% of the population during recent years (Kvalitetsnätverk Västkust 2003). The unemployment rate increased from 6.4% in 2001 to 6.8% in 2004 (Uddevalla kommun 2001, 2004).

36

In 2003, new nursing home places were opened for elderly with dementia, and attempts to divide the staff ratios more equally between nursing homes were conducted (some nursing homes received higher staff ratios while oth-ers were reduced) (Uddevalla kommun 2003). During 2004 some of the workplaces within the elderly care organization tested new work scheduling models such as a 3-3 system, (in which people work three days and then have three days off), some measures to improve care recipients’ meals and nutrition, palliative care, and care of individuals with dementia were con-ducted (Uddevalla kommun 2004).

There were no private care providers in elderly care in 2004 (Private communication Pär Levander 25 Oktober 2006). Numbers of care recipients and staff members in the elderly care organization of Uddevalla are pre-sented in Table 1.

Participants

Nursing staff In the analyses used in this thesis nursing staff working in the two munici-palities’ elderly care organizations were included. Nursing staff was defined as registered nurses, practical nurses and nurses’ aides. The other staff cate-gories, such as occupational therapists and needs assessment officers, were excluded from the analyses. All types of care settings, i.e. home help ser-vices, home nursing care, and nursing homes were included. However, staff that had been away from the workplace for longer than three months due to maternity or sick leave were excluded from the questionnaire studies. These individuals were excluded since we aimed to measure more recent percep-tions. Response rates and background characteristics of the staff participants for each questionnaire study are presented in Table 3 in Paper III.

Care recipients Elderly care recipients who were capable of responding to a questionnaire (on their own or with assistance) were considered eligible for the studies. Nursing staff and their supervisors together made the decisions regarding capability of care recipients to reply to the questionnaire. Care recipients were permitted to receive assistance reading the questions and/or writing responses. Response rates and background characteristics of the care recipi-ent respondents for each questionnaire study are presented in Table 2 in Pa-per IV.

37

Family members of care recipients Participants in the Örebro University Hospital study were relatives of pa-tients whose illness prevented them from responding to a questionnaire con-cerning their views on quality of care in the five units of the hospital. Re-sponse rate for the USÖ study is presented in Table 1 in Paper I, and back-ground characteristics of the participating relatives in USÖ study are pre-sented in Table 2 in Paper I.

For the municipality studies, relatives’ questionnaires were delivered to individuals who were registered as primary family members of care recipi-ents. One relative per care recipient was registered as a primary family member and therefore one relative per care recipient was invited to partici-pate in the studies. Response rates and background characteristics of the participating relatives for each questionnaire study in Bengtsfors and Udde-valla are presented in Table 3 in Paper IV.

A brief description of the intervention The content of the educational intervention, the toolbox, was based on the results from the staff, care recipient, and relatives’ baseline (2001) question-naire studies in Bengtsfors. The results revealed areas in need of improve-ment regarding staff competence, psychosocial work environment, and qual-ity of care. For instance, staff reported a need to improve their competence in areas such as care recipients’ medication, oral hygiene and strategies for handling aggressive care recipients as well as care recipients with dementia. A work group comprised of researchers and elderly care management repre-sentatives assembled a toolbox of practical instruments for use at interven-tion workplaces. For a more detailed description of the toolbox development, see Paper III.

A total of 16 instruments for nursing staff were included in the toolbox. The instruments were practical in nature and the aim was that staff could use these to improve their regular work practices. A description of the instru-ments is available in Table 2 in Paper III. The instruments were organized into four categories: (1) instruments for improving the working situation for nursing staff; (2) educational materials regarding, for example dementia, and nutrition, (3) instruments for meeting care recipients’ social and physical needs; and (4) staff, care recipients’, and relatives’ baseline questionnaires accompanied by guidelines for interpreting the questionnaire results.

The intervention toolbox was a binder containing detailed descriptions of each of the toolbox instruments. Each description stated the instrument’s purpose, use, cost (when applicable), and the name of a contact person for further information. Whenever possible, the actual instrument or educational

38

material was included in the binder; otherwise, the description was accom-panied by appropriate references.

In February 2003 the toolboxes were distributed to workplaces in the in-tervention organization. Each workplace was free to decide to what extent they would work with the toolbox instruments, and no guidelines were given by the researchers or the management of the elderly care organization re-garding work with the instruments. However, the intervention workplaces were asked to fill in a short checklist every other month regarding their work with the toolbox instruments. Staff members were asked to document which instruments the workplace had used and to record the time allocated for this work. The reference organization did not receive any information regarding the toolbox intervention.

Data collection procedures The results obtained in the articles included in the present thesis are based on data gathered with self-administered questionnaires. Prior to the question-naire, structured interviews and focus group discussions were conducted in order to identify possible relevant items to be added to the previously vali-dated questionnaires. The staff questionnaire was based on the Quality-Work-Competence questionnaire (QWC), which has been used extensively and validated among healthcare staff (Arnetz 1999, Arnetz & Wiholm 1997). A previously validated Pyramid patient questionnaire (Arnetz & Arnetz 1996) was a foundation for the elderly care recipient questionnaire and for the relative questionnaire. In addition, a Pyramid questionnaire for parents of children receiving hospital treatment (Ygge & Arnetz 2001) was used as a base for the relatives’ questionnaire.

The results obtained in the additional analyses included in the thesis are based on nursing home costs and administrative data regarding staff and resident outcomes. The data collection of these measures will be described at the end of this chapter.

Interviews and focus group discussions When developing a questionnaire for patient relatives at the Örebro Univer-sity Hospital, a total of 22 structured interviews with patients’ relatives were carried out by a nurse in each of five departments at USÖ. The purpose of the interviews was to develop relevant questionnaire items for the relatives’ questionnaire.

In further developing the relatives’ questionnaire to community elderly care, one focus group discussion with family members of elderly care recipi-ents was conducted in Bengtsfors. Another focus group discussion was con-ducted with elderly care recipients in an effort to adapt the Pyramid patient

39

questionnaire to municipality elderly care recipients. In addition, a total of six focus group discussions with different staff groups were conducted in order to add questions to the staff questionnaire. All focus group discussions were conducted in the last quarter of 2000 in Bengtsfors.

The participants in the focus groups discussions were recruited by the lo-cal administrators. All focus group discussions included six-eight partici-pants in each group. All focus group discussions were tape-recorded and lasted approximately 90 minutes. Two members from the research team were present at each of the focus groups. One person moderated the discussions and the other person made notes on a notice board concerning the themes discussed. These notes were visible to all participants. All focus groups fol-lowed a structured plan in which the following main subject areas were dis-cussed: organizational stress, work environment, quality of care/services, and organizational management.

A team consisting of researchers read all the researchers’ notes made in the focus group discussions. The tape recordings from each of the focus group discussions were listened to by a research assistant. The assistant made notes regarding the topics discussed, and these were compared with the researchers’ notes. Additional questions were developed and included in the staff’s, care recipients’, and family members’ questionnaires, while some questions that were not found to be relevant in community settings were omitted.

Questionnaire studies All questionnaire responses were anonymous and no reminders were sent. Each respondent received an addressed, postage-paid envelope in which to return the questionnaire to the research unit. The staff questionnaires were distributed at workplaces to all currently employed elderly care staff. Work-place managers at each respective workplace distributed questionnaires di-rectly to each employee. Questionnaires were mailed to the homes of staff that had been on maternity or sick leave for less than three months.

Care recipient questionnaires were distributed by nursing staff after dis-cussions with their supervisors regarding the capability of care recipients to reply to the questionnaire. Care recipients were permitted to receive assis-tance in reading the questions and/or writing responses from a friend or a relative but not from staff. Less than half of eligible care recipients re-sponded to the baseline questionnaire in 2001. For this reason, individuals were recruited to help the care recipients with the questionnaire in both of the follow-up measurements in the intervention municipality and in the sec-ond follow-up in the reference municipality. These persons were directed to read the questions and response alternatives and to fill in the care recipient’s reply, but not to interpret or answer the questions themselves.

40

The relatives’ questionnaire at USÖ was delivered to family members by staff members at each of the units. One family member per patient received a questionnaire. Relatives of elderly care recipients in the municipalities re-ceived the questionnaire by mail. The researchers received a list of relatives’ addresses in Bengtsfors and posted the questionnaires. In Uddevalla the local administrators mailed the questionnaires to relatives. In both municipalities, university envelopes were used for mailing the relatives’ questionnaires.

Nursing home costs and administrative data Nursing home costs and administrative data regarding staff and resident out-comes were collected for all nursing homes in Bengtsfors and Uddevalla for 2003. The administrative data were collected from the municipalities’ ad-ministrative registers. Measures that would be comparable between the mu-nicipalities were chosen. Costs were collected from the municipality finan-cial administrators. The administrators in both municipalities were asked to provide a data file with nursing homes’ total costs, staff costs, and operating costs for all nursing homes in the municipality. They were also asked to pro-vide any information that would be necessary when comparing costs be-tween nursing homes, for instance costs not included in the data files or pos-sible differences between nursing homes on cost calculations.

Outcome measures Nursing staff competence and psychosocial work environment as well as care recipients’ and their family members’ perceptions of quality of care were the main outcome measures of this thesis. In addition, nursing home costs and administrative data regarding staff and resident outcomes were measured. The outcome measures will be described in the following chap-ters.

Staff competence and the psychosocial work environment Staff respondents were asked to rate their knowledge regarding eighteen aspects of their work (analyzed in Papers II and III). In addition, staff were asked to rate the emotional strain for eight aspects of their work and the physical strain for seven aspects (analyzed in Paper II).

The staff questionnaire included 11 measurement indices for different as-pects of their psychosocial work environment. In addition, a visual analogue scale for overall work satisfaction was included in the staff questionnaire. An overview of the staff work environment indices is presented in Table 2. The items included in each of the measurement indices are presented in Pa-per III, Table 1.

41

Table 2. Psychosocial work environment indices included in the staff questionnaire.

Not all of the indices were included in all of the four papers. In Paper II four indices – work-related exhaustion, work stress, skills development, and men-tal energy – were analyzed. In Paper III all of the 11 indices were analyzed for 2003 and 2004, but for 2001 only six indices were included in the ques-tionnaire. These were work-related exhaustion, work stress, participation, efficiency, leadership, and work climate.

Quality of care The care recipients’ questionnaire consisted of four quality of care meas-urement indices (see Table 3). The component items included in the indices are presented in Table 1, Paper IV. Care recipients were also asked to give an overall rating for the quality of care on a visual analogue scale.

The family members’ questionnaire consisted of seven quality of care in-dices (Table 3). An overview of the items included in each of the indices is presented in Table 1, Paper IV. Relatives were also asked to give an overall rating to the quality of care on a visual analogue scale.

The names of two measurement indices in the family members’ question-naire are not consistent in Paper I when compared to those in Paper IV. The “nursing staff” index name was changed to “staff professional skills” and “caring processes” was changed to “care”. In addition, an index measuring family members’ perceptions of the staff work environment was presented in Paper I. However, this index was not analyzed in Paper IV, since the focus of that article was only on quality of care.

Work-related exhaustion Work stress Staff participationEfficiency LeadershipWork climate Mental energy * Goal clarity * Employeeship * Performance feedback * Skills development * Overall work satisfaction (VAS) * not included in the questionnaire in 2001

42

Table 3. Quality of care measurement indices included in care recipient and family members’ questionnaires.

Quality of care indices rated by:

CARE RECIPIENTS Information Staff behavior CareActivityOverall quality grade (VAS) FAMILY RELATIVES Information Staff professional skills (in Paper I: nursing staff) Care (in Paper I: caring processes) ActivityContact Social support Relative participation Overall quality grade (VAS)

Nursing home costsNursing homes’ total costs, staff costs, and operating costs were collected. However, the information received from the municipalities’ financial admin-istrators revealed that costs for cleaning were either calculated as staff costs or operating costs depending on whether the cleaning was conducted by regular nursing staff or an external cleaning firm. Thus, it was decided that the costs would only be analyzed as the facility’s total costs so that the dif-ferences in calculation of cost for cleaning would not affect the results.

The total cost was supposed to consist of staff and operating costs and ex-clude costs for facility rental. Rent was not included, since great differences were reported in rent levels due to differences in internally charged transfer rents. In Bengtsfors the cost reports did not include costs for nursing home supervisors and administrative staff. These costs could not be provided by the municipality, but were included in the total cost reports received from Uddevalla. Thus, it was decided that the total costs of nursing homes were not comparable between the municipalities. Analyses of nursing home costs were only conducted for the municipality with the greater number of nursing homes (Uddevalla).

There were some differences between the nursing homes in Uddevalla re-garding how the costs of care recipients’ meals were reported. This affects the reports of the total facility costs. Most of the nursing homes prepared the meals by themselves or bought the meals from another producer, and the costs were calculated for the facility. However, some nursing homes re-ceived the meals or part of the meals from another nursing home without actually paying for the service. This meant that some nursing homes did not calculate any costs for care recipients’ meals, while other nursing homes

43

included these costs in their reports. The financial administrators could not provide any approximate cost for these meals. Thus, the costs of the different nursing homes in Uddevalla may not be totally comparable.

The total costs of a facility were divided by the number of its resident places in the facility in 2003 in order to increase the comparability of the costs between the nursing homes. The local administrators were asked to report possible changes in the amount of resident places during the year 2003, which was taken into consideration when calculating the costs per resident day. All resident places were estimated to be used, since the mu-nicipality reported a waiting list for nursing home places (Uddevalla kom-mun 2003).

Administrative dataAdministrative data regarding staff and resident outcomes were: staff sick-ness absence, prevalence of adverse medication events, staff work injuries, and care recipients’ falls. Sickness absence was divided into four categories in the municipalities’ registers: 3 days or less absence, 4-14 days, 15-90 days, and 91 days or more absence. The total amount of absence days at the nursing home level in each of the four categories was divided by the total number of staff members at that nursing home in order to get a measure that would be more comparable between the nursing homes.

In both municipalities some nursing homes lacked data for one or more of the measures. Several attempts by the author were made to obtain the miss-ing data. In some cases the data were provided later on by a nursing home and in other cases nursing homes reported no data for the measure. However, when the analyses were conducted, one nursing home still lacked a measure for prevalence of falls. This nursing home was located in Bengtsfors and thus was not included in the analyses, since a decision had been previously made to include only nursing homes in Uddevalla.

Statistical analysis The SPSS statistical package (versions 11.0, 12.0, 14.0) was used for all statistical analyses. Statistical significance for all analyses was set at p<.05 (two-tailed).

Papers I-IV In all of the papers, Chi square statistics were used to compare discrete vari-ables, such as respondent background characteristics, between groups. In addition, Chi square statistics were used to compare the questionnaire re-

44

spondents to the total population of staff, care recipients, and relatives in each municipality, respectively.

All continuous variables (measurement indices, overall work satisfaction and overall quality ratings) were examined for normality using the Kolmo-gorov-Smirnov test in Papers II-IV. Since not all of the continuous variables were normally distributed, both non-parametric and parametric tests were performed simultaneously in Papers II-IV, yielding identical results. Possible differences in mean values of the continuous variables between two groups were assessed using the non-parametric Mann-Whitney rank sum test and the parametric independent samples t-tests. When mean values between three or more groups were compared one-way analysis of variance, ANOVA, with Tukey and Bonferroni post-hoc tests, and Kruskall Wallis rank sum tests were used. Changes over time and possible interaction effects between mu-nicipalities were assessed using multivariate ANOVA (Papers III and IV).

Pearson product-moment correlations (r), were used to examine correla-tions between measurement indices (Papers I and II).

Forward stepping multiple regression analysis was used to determine predictors of care recipient relatives’ overall quality ratings (Paper I) and of staff work satisfaction ratings (Paper II). The fit of the regression model was studied by examining scatterplots of residuals for each independent variable, plotted against the values of the respective predictor (Paper II).

In Paper I, exploratory and confirmatory factor analysis was conducted in order to study the questionnaire item’s suitability for building indices. In all of the papers, internal reliability of the indices was measured using Cron-bach’s alpha. In Paper I, second order factor analysis using principal com-ponents analysis with scree plots and varimax rotation was carried out to further investigate inter-scale correlations.

Additional analysesNursing home costs’ possible correlation to the administrative data, staff ratings of the work environment, and care recipients and their relatives’ rat-ings of the quality of care were analyzed in an aggregated data file. The data file was aggregated at the nursing home level and comprised of total facility cost per care recipient per day, administrative data at the nursing home level, nursing home mean values of the staff work environment indices, and care recipients’ and relatives’ ratings of the quality of care indices.

Non-parametric correlation tests, Spearman rho, were conducted in order to investigate whether the costs were correlated to the administrative data or to the work environment or quality of care indices.

45

Power calculations The powers of the studies in Paper II-IV were calculated with the sample power program included in the statistical package SPSS (version 12.0).

The statistical power is a function of statistical significance, sample size and effect size, i.e. existing differences between the groups (Kazdin 2003). It has been proposed that the power should be at least .80 when statistical sig-nificance is .05 (Cohen 1965, Kazdin 2003).

To calculate effect size, the difference between study groups in mean val-ues for a dependent variable was estimated to be approximately 10 percent-age points on a scale from 0% to 100% (standard deviation of approximately 20.0). This calculation gives an effect size of .50. The statistical significance was set at .05 level (2-tailed). The power was calculated for the minimum sample sizes in Papers II and IV and a power over .80 was estimated. Thus, the power may be even higher since the actual sample sizes were larger. In Paper III, it was estimated that at least 180 respondents per measurement would be needed to obtain a power of 100% in order to detect an increase of 9-percentage points in the scale measuring skills’ development.

Ethical approval No ethical approval was requested for the university hospital study, since the study was a part of an internal quality development project at USÖ. The TOOL project obtained ethical approval from the research ethics committees of Uppsala University and Gothenburg University (dossier number 00-206).

46

Results

Paper I Seven quality of care indices and one index for relatives’ ratings of staff work environment were constructed. The content validity of the question-naire was considered good, since some of the indices were found to be simi-lar to prior studies with patients’ relatives. In addition, the majority of rela-tives considered the questionnaire to be relevant and of reasonable length. Internal reliability of the indices exceeded 0.70 in all but one index (Table 3, Paper I). However, correlations between some of the indices were high, indi-cating some overlap between these indices (Table 4, Paper I). A second or-der factor analysis resulted in three distinct index groupings: personnel, rela-tive’s role and care content (Figure 1, Paper I). These three dimensions seem to summarize relatives’ perceptions of the quality of elderly care.

Correlation analysis confirmed a higher correlation between relatives’ overall quality rating and work environment variables, compared to variables concerning relatives’ participation. Furthermore, relatives’ ratings of the staff work environment was the most important predictor of a positive over-all quality rating in all three surveys. For relatives with more frequent con-tact with elderly care staff, the index nursing staff was the strongest predic-tor, followed by work environment, information, and contact. When the three-quality dimensions were entered into a regression model, the personnel dimension was the only significant predictor of a positive overall quality rating.

Relatives of care recipients in the two questionnaire studies in Bengtsfors gave highest ratings to the caring processes index (Figure 2, Paper I). In the university hospital study patients’ relatives gave highest ratings to social support. The indices information and participation received lower ratings from relatives in all three studies. In addition, relatives in the Bengtsfors studies gave lowest ratings to the activity index.

Paper II In general, a larger percentage of home care staff in both municipalities rated their knowledge as insufficient, compared to staff in nursing homes (Table 3, Paper II). Staff in both care settings rated their opportunities for skills devel-

47

opment as limited (Table 5, Paper II). Staff expressed similar competence development needs in both care settings. There were several areas, such as dementia, psychiatric illnesses, threats and violence, leadership and laws regarding healthcare and social welfare, in which the majority of nursing staff ( 50%) rated their knowledge as insufficient in both care settings.

In general, staff in nursing homes rated their work as significantly more physically and emotionally strenuous than staff in home care did (Table 4, Paper II). Transfer of care recipients to and from bed, transfers to and from a chair, caring for individuals requiring different levels of assistance and as-sisting at the toilet were rated as the most physically strenuous aspects of work in both care settings. Lack of time for work tasks, caring for individu-als with dementia, caring for individuals requiring different levels of assis-tance were rated as the most emotionally strenuous aspects of work in both care settings.

There were no significant differences in mean values for staff ratings of work-related exhaustion, mental energy, and overall work satisfaction be-tween home care and nursing homes or between municipalities (Table 5, Paper II). Nursing home staff ratings of work stress were within the target level range, but below the target level minimum for staff in home care, indi-cating low work stress in home care. Mean values for work-related exhaus-tion were above the target level in both care settings, indicating high work-related exhaustion. Mean values for staff ratings of their mental energy were above the target value in both settings, which indicates high mental energy. Staff ratings of work satisfaction were moderate both in home care and in nursing homes.

Work-related exhaustion, skills development, work stress, and mental en-ergy were significant predictors of work satisfaction ratings in both care settings (Table 6, Paper II). The strongest predictor of work satisfaction in both care settings was work-related exhaustion. The relationship was in-verse, indicating that higher ratings of work-related exhaustion predicted lower ratings of work satisfaction.

Paper III In the intervention municipality (Bengtsfors), nursing staff ratings of their competence and the psychosocial work environment improved significantly over time. Staff ratings of work-related exhaustion and work stress decreased significantly over time (Table 5, Paper III). Mean values for participation, efficiency, leadership, work climate, performance feedback, skills develop-ment, and work satisfaction increased significantly over time (Table 5, Paper III). Staff ratings of their knowledge showed significant improvements in 10 out of 18 areas in the intervention municipality in 2004 as compared to 2003 (Table 4, Paper III).

48

In the reference municipality, staff ratings of knowledge showed im-provements in only one area (Table 4, Paper III). No significant changes were found over time in the reference municipality regarding staff ratings of their psychosocial work environment (Table 5, Paper III).

A significant interaction effect between municipalities over time (2003–2004) was found for staff ratings of participation, leadership, performance feedback, and skills development, which indicates a significant difference over time between the municipalities in staff ratings of these four scales (Ta-ble 5, Paper III).

A significant increase in staff reports of utilization of the toolbox instru-ments between 2003 and 2004 was found in the intervention municipality. In 2003, 42% and in 2004, 53% of the respondents reported that their work-place had used the toolbox instruments.

Paper IV There were no significant changes over time in care recipients’ (Table 4, Paper IV) or relatives’ (Table 5, Paper IV) ratings of the quality of care in the intervention municipality. In the reference municipality, some significant changes over time were found regarding care recipients’ (Table 4, Paper IV) and relatives’ (Table 5, Paper IV) ratings. However, no significant interac-tion effects between municipalities over time (2003-2004) were found for care recipients’ (Table 4, Paper IV) or relatives’ (Table 5, Paper IV) ratings of the quality of care indices.

Some ceiling effects on the index care and the overall quality grade were found in care recipients’ and relatives’ baseline surveys.

Additional findings The total facility costs, amount of resident places, and total cost per resident places per day for the 17 nursing homes in Uddevalla are presented in Table 4.

49

Table 4. The total facility costs, number of resident places and total cost per resident per day for nursing homes in Uddevalla 2003

Total cost SEK 1000’s

(USD 1000’s3)

Number of residentplaces

Number of resident places

per year

Total cost per resident places per day SEK (USD)

Mean 13,168 (1,627) 37 13,351 1,030 (127) Minimum 3,754 (464) 8 2,920 735 (91) Maximum 23,370 (2889) 60 21,900 1,427 (176)

The total costs per resident place per day were significantly positively corre-lated to staff ratings of goal quality (rho .512 = p<0.05), and relatives’ rat-ings of the quality of care indices care (rho .576 = p<0.05), contact (rho .655 = p<0.01), social support (rho .553 = p<0.05), and participation (rho .592 = p<0.05). The total costs per resident places per day were not significantly correlated to any of the administrative measures or care recipients’ ratings of quality of care.

3 Mean exchange rate SEK/USD in 2003: 8,09. Riksbanken (2006) http://www.riksbank.com/templates/stat.aspx?id=17212. Accessed 1 October 2006.

50

General discussion

An overall interpretation of the findings is provided in this chapter. A more detailed discussion of the findings is included in each of the four papers in-cluded in the thesis.

The impact of the educational intervention Psychosocial work environment It was hypothesized that the toolbox intervention would have a positive im-pact on staff ratings of their competence, the psychosocial work environ-ment, as well as on work stress and work satisfaction. The hypotheses were supported by our findings, which showed that nursing staff ratings of their competence and most of the psychosocial work environment indices im-proved significantly over time in the intervention municipality. In addition, staff ratings of work-related exhaustion and work stress decreased and their ratings of work satisfaction increased significantly. In the reference munici-pality, staff rated one of 18 competence questions significantly higher in the follow-up measurement, while their work environment ratings were un-changed. Significant interaction effects between the municipalities over time for four psychosocial work environment indices offered further evidence for the positive work environment changes in the intervention municipality.

The introduction of the educational toolbox is suggested to be related to the improvements in the psychosocial work environment in the intervention municipality. However, it is not possible to rule out the possible effects of other changes occurring in the organizations at the same time. In elderly care in the intervention municipality, several organizational changes, including new supervisors, and decreases in the number of staff members and nursing home places, occurred during the intervention period. Prior research has found that this type of change can cause stress and lower work satisfaction among staff working in the organization (Brown et al. 2003, Petterson et al.2005). Thus, the other changes occurring at the same period of time as the intervention would probably have a negative impact on staff. However, some staff in the elderly care intervention received education in dementia care during the intervention period (Bengtsfors kommun 2004). Based on regula-tions implemented in 2001 (Arbetsmiljöverket 2001), systematic work envi-

51

ronment improvement efforts were also introduced (Bengtsfors kommun 2003). It is possible that the dementia care education or the systematic work environment work improved the staff work situation. However, it seems unlikely that this effect alone would be reflected in all aspects of staff rat-ings. On the contrary, almost all of the competence areas that showed im-provements in the intervention staff ratings could be associated with the toolbox instruments, which indicates that the intervention was related to the improvements. The organizational changes occurring in the reference mu-nicipality during the study period were in general more positive than in the intervention municipality (Uddevalla kommun 2004), but no changes in staff work environment ratings were found in the reference municipality. This offers further support that the toolbox was related to the positive changes in the intervention municipality.

It is possible that there were changes in Swedish elderly care in general during the intervention period. The negative work environment trend that was noticed during the 1990s was reported to have stabilized at the begin-ning of the 2000s (Gustafsson & Szebehely 2005). Thus, it is possible that the toolbox intervention was implemented and evaluated during a period of time when a general positive trend regarding staff work situation in munici-pality elderly care was noticed. The positive changes in staff ratings of their work environment in the intervention municipality could be a result of that trend. However, no corresponding improvements were found in the reference municipality. Thus, it is unlikely that the changes in the intervention municipality would be a result of a general improvements in elderly care work environment in Sweden.

There are some features of the toolbox intervention that could have ac-counted for the changes in staff ratings. It has been proposed that interven-tions have best results when the real needs of staff are investigated prior to designing the intervention (Cowan et al. 2004, Donabedian 1996, Gustafsonet al. 2003). The contents of the toolbox intervention were based on reported staff needs in the baseline measurement. Staff may have felt that the instru-ments included in the toolbox met their competence development needs, thus motivating them to use the toolbox. In addition, different development needs were reported between the intervention workplaces, and in order to further meet staff needs, the project design allowed each workplace to decide which instruments they would work with, based on the specific needs experienced by each work group. A standardized educational program was not considered as effective in meeting the different workplace development needs, since difficulties in tailoring educational program for staff with different needs and backgrounds have been reported (Maslin-Prothero 1997, Ross et al. 2001). In addition, staff participation in the intervention processes has also been reported to be crucial for successful intervention (Dahl-Jorgensen & Saksvik 2005). The toolbox design allowed nursing staff themselves to make deci-sions regarding the development work at their workplace. This was an im-

52

portant aspect for improving staff participation, but it was also necessary since supervisor turnover was high in the intervention organization. It is possible that staff felt empowered by the possibility to influence and make decisions regarding their workplace development work.

The toolbox intervention also lacked some characteristics that have been reported in prior studies to be crucial for successful intervention. Leadershipcommitment has been reported to be one of the most important aspects for intervention projects (Donabedian 1996, Gustafson et al. 2003, Mentes & Tripp-Reimer 2002). Leadership commitment in the present project was probably low in some workplaces, due to high turnover rates among manag-ers. It is possible that the present intervention would have had an even stronger impact if all of the workplaces had had stable leadership. It is also possible that those few workplaces in intervention organization that had the same supervisor during the whole study period had greater improvements than the other workplaces. However, this thesis did not compare workplaces within the intervention municipality. Sufficient resources in terms of staff members, money, and time should also be provided in order to achieve the best results with an intervention (Gustafson et al. 2003). Lack of resources could have been one of the main reasons for workplaces not participating in the work with the toolbox intervention. No additional resources or time for planning and conducting the toolbox work was provided at the workplaces. However, since the instruments included in the toolbox intervention were in general free of charge, the intervention was estimated to be economically possible for the workplaces. The intervention was a workplace intervention, i.e. all skills development would be carried out right at the workplace. This was intentional in order to encourage individual workplace choice as well as to limit costs.

Staff reported that use of the toolbox instruments increased significantly over time. However, this thesis did not include analyses at the workplace level regarding the use of the toolbox intervention. Such analyses could identify characteristics of workplaces using the toolbox to a greater extent or investigate whether improvements in work environment were greater at workplaces that used the toolbox to a greater extent. Such analyses are needed in order to investigate the causality of the intervention, i.e. whether workplaces that worked more with the toolbox achieved greater improve-ments in work environment ratings. It would also be interesting to study whether workplaces with higher initial work environment ratings were more likely to use the toolbox instruments. These analyses would also highlight possible differences between home-based care and nursing homes regarding the extent of toolbox use.

This is among the first of any controlled studies investigating the impact of an educational intervention on self-rated staff competence, the psychoso-cial work environment, and work satisfaction in elderly care. It is suggested that an educational toolbox, consisting of practical instruments adapted to

53

local needs, might be an effective intervention for improving staff compe-tence and the psychosocial work environment in elderly care.

Quality of care It was also hypothesized that the intervention would have a positive impact on quality of care through improved staff knowledge, psychosocial work environment, and work satisfaction. The theoretical basis of the thesis pro-posed that staff outcomes such as health, satisfaction, and quality of working life directly affect the quality of care and care recipient wellbeing (Hannan et al. 2001, Sainfort et al. 2001). However, our findings did not support that hypothesis, since care recipients’ and their relatives’ ratings of quality of care did not change significantly in the intervention organization after the introduction of the intervention.

There are several possible explanations for these findings. Perhaps staff competence, work satisfaction, and work environment are not related to quality of care as proposed by the theoretical frameworks. The theoretical models were based on empirical research regarding work organization, staff work environment and quality of care. Several literature reviews from hospi-tal settings have concluded that factors such as total nursing hours per pa-tient, skill mix among staff (Lang et al. 2004), quality of interaction among staff (Lundstrom et al. 2002) and quality of work environment (Mitchell & Shortell 1997) seem to be related to quality of patient care. In addition, re-search from so called magnet organisations have further supported the rela-tion between work organization and quality of care and patient outcomes. Magnet hospitals are characterized by low staff turnover, strong nurse autonomy, decentralised decision making, adequate staffing levels and in-vestments in nurse education, and have reported lower patient mortality (Aiken et al. 1994) and higher patient satisfaction (Rondeau & Wagar 2006) than other hospitals. However, one recent review (Hoff et al. 2004) con-cluded that hospital organizational variables like leadership or teamwork do not seem to have an influence on medical errors or patient safety. In conclu-sion, the majority of the prior empirical research from hospital settings give support for an association between staff work-related factors and quality of patient care. However, several different measures have been included in the studies which make a comparison of results difficult. In addition, several of the measures used are not relevant in elderly care settings.

In elderly care, a smaller amount of research regarding the links between staff perceptions of their jobs and the well-being of care recipients has been conducted. Hannan et als (2001) concluded in their literature review that increased work satisfaction and decreased stress among nurses probably improve quality of care, but added that the association seems to be complex and there might be several factors, such as management style and group co-hesiveness, that influence the relationship between these factors. Some re-

54

cent empirical studies have found that greater resident satisfaction was asso-ciated with higher staff job satisfaction (Redfern et al. 2002, Sikorska-Simmons 2006). However, Sikorska-Simmons concluded that the causal nature of this relationship needs to be further investigated (Sikorska-Simmons 2006). In addition, many of the prior studies investigating the rela-tionship of work organization and quality of care have been criticized for not specifying the mechanism linking these factors (Mitchell & Shortell 1997) and for not showing that organizational factors come before quality of care (Flood 1994). It seems that the empirical evidence for the relationship be-tween nursing work and quality of care, especially in elderly care settings, is mainly based on cross-sectional research, which limits the understanding of the causal relationship between these factors. In addition, there are some studies that did not find an association between work environment and qual-ity of elderly care (Goodell & Van Ess Coeling 1994, Jenkins & Allen 1998). This could imply that staff competence and quality of working life are not related to care recipients’ or relatives’ ratings of quality of care. How-ever, future, longitudinal studies are needed to highlight the causal links and mechanism between staff work variables and quality of care.

It is also possible that nursing staff did not change their working behavior after gaining new knowledge, in ways that were appreciated by care recipi-ents and their relatives. The actual behavior of staff was not evaluated in the present study. Prior studies have found that staff had difficulties putting knowledge gained in educational interventions into practice (Broad 1997). Clearly, more research is needed to investigate whether knowledge gained by staff in educational interventions changes their work behavior and im-proves the quality of care. Elderly care providers should be aware of the possibility that staff education does not automatically improve quality of care or satisfaction among care recipients or their relatives. In addition, the actual patient outcomes, such as occurrence of pressure ulcers or amount of activities provided to care recipients, were not evaluated. This could have provided different results, since it has been suggested that care recipients’ and family members’ perceptions of quality of care are not sensitive to changes (Boumans et al. 2005). Several prior educational interventions did not result in any changes in care recipients’ (Boumans et al. 2005, Laitinenet al. 1996, Molloy et al. 2000) or family members’ (Finnema et al. 2001, Molloy et al. 2000) perceptions. Future studies could combine more objec-tive measures of quality such as administrative data with care recipients’ and relatives’ perceptions, as has been proposed by some earlier studies (Changet al. 2006, Socialdepartementet 2002).

The fact that no significant effects were found on care recipients’ and their relatives’ ratings of the quality of care does not necessarily mean that the quality of care was unchanged during the study period. There were sev-eral methodological aspects that could have limited the evaluation of the

55

intervention’s possible impact on quality of care. These are discussed under the methodological considerations.

Perceptions of competence and work environment The findings of the present study showed that the majority of nursing staff rated their knowledge as insufficient in areas such as dementia, psychiatric illnesses, threats and violence and laws regarding healthcare and social wel-fare. Similar competence development needs were identified in home-based care and in nursing homes. However, a greater proportion of staff in home-based care reported less sufficient knowledge compared to staff in nursing homes. These findings indicate that in these municipalities, staff working in home-based care had greater needs for competence development than staff in nursing homes. In addition, a larger proportion of staff in Bengtsfors rated their knowledge as insufficient when compared to staff in Uddevalla. Clearly, the organization (municipality), not just the care setting, may affect nursing staff ratings and competence levels may vary greatly between mu-nicipalities. Staff ratings of their opportunities for skills development were low in both care settings and both municipalities. These findings are in line with prior studies that have found, for instance, that less than half of home help service staff felt that they had sufficient training and work introduction that they needed to cope with the demands at work (Laamanen et al. 1999). The National Board of Health and Welfare has also reported poor compe-tence development opportunities for nursing staff in elderly care (Socialstyrelsen 2005d). In addition, it has been reported to be difficult to recruit staff with adequate competence and knowledge to elderly care (Kerkstra & Hutten 1996, OECD 2005, Socialstyrelsen 2006e). Swedish regulations emphasize that staff should have sufficient competence, but do not consist of detailed regulation (Thorslund et al. 1997).

The most emotionally and physically strenuous aspects of work in both care settings were: lack of time for work tasks, caring for individuals with dementia, caring for individuals requiring different levels of assistance, and physical transfers. In general, home care staff ratings of work strain were significantly lower than those of nursing home staff. This could be due to the fact that a larger proportion of care recipients in nursing homes are more dependent and cognitively impaired (Socialstyrelsen 2005e). This gives us very specific information about causes of work strain that should be consid-ered in all efforts to improve the work situation of elderly care staff. Prior studies have also reported that elderly care staff experience lack of time at work (Bowers et al. 2001b, Brulin et al. 2000, Laamanen et al. 1999). It is interesting that staff in our study reported lack of time to be one of the most strenuous aspects of their work, but at the same time they rated the work stress index in general as low. The work stress index measures staff percep-

56

tions of time for planning and executing work tasks and the findings indicate that staff experienced that they had sufficient time at work, i.e. the stress levels were low. However, time at work was measured on different scales in these items (VAS vs. four Likert scale questions), which could explain the differences in staff ratings. However, since time pressure seems to be a common issue in elderly care, further research should focus on the possible reasons, such as workload and work routines.

Staff ratings of work-related exhaustion were high in both municipalities and in both care settings. Thus, nursing staff experienced high exhaustion levels after work, but responded that they had sufficient time for work tasks. These findings indicate that other factors than lack of time caused high ex-haustion levels for staff. Staff ratings of their opportunities for participation, the leadership in the organization, goal clarity at the workplace, directives and feedback received from the leaders, and the possibilities for skills devel-opment were rated low in both municipalities. Thus, there were several as-pects of the psychosocial work environment that were rated as limited in these elderly care settings, which could have caused the high exhaustion levels. Prior studies in nursing homes and home care have associated high exhaustion and stress with time pressure (Morgan et al. 2002, Rafnsdottir et al. 2004), but also with factors such as insufficient skills (Dunn et al. 1994, Morgan et al. 2002), unclear work directives (Brulin et al. 2000), hierarchic organization (Brulin et al. 2000), limited control and impact on work tasks (Denton et al. 2002b), heavy workload (Morgan et al. 2002), uncooperative family members (Walcott-McQuigg & Ervin 1992), and cognitively im-paired residents (Novak & Chappell 1996, van den Berg et al. 2006).

Staff ratings of work satisfaction were moderate both in home care and in nursing homes. The predictors of work satisfaction were the same in both care settings. The strongest predictor of work satisfaction in both care set-tings was work-related exhaustion. The relationship was inverse, indicating that higher ratings of work-related exhaustion predicted lower ratings of work satisfaction. Skills development, work stress, and mental energy also predicted staff ratings of work satisfaction. Work satisfaction has in previous studies also been influenced by stress and work pressure (Chou et al. 2002, Karsh et al. 2005). It seems that work-related exhaustion levels should be reduced in order to improve work satisfaction. In addition, stress manage-ment and increased opportunities for competence development should result in higher satisfaction levels.

Our findings showed that home care staff reported lower strain than nurs-ing home staff. Otherwise few differences regarding staff ratings of the psy-chosocial work environment were found between the care settings. The comparison of our findings to prior research is limited, since few studies have compared the work environments between these care settings. One recent study found that home care staff in Swedish elderly care rated the work environment more positive in terms of workload, the variety of work

57

tasks and stimulation from the work, when compared to staff in nursing homes (Gustafsson & Szebehely 2005). However, home care staff rated their influence on working conditions lower than nursing home staff did (Gustafsson & Szebehely 2005). More research could be conducted in order to investigate possible work environment issues that should be considered in future interventions in each care environment.

Perceptions of quality of care Care recipients and their relatives gave highest ratings to the basic nursing elements of elderly care, such as personal hygiene, meals, and physical transfers. It seems that these actors perceived that the elderly care recipients were well taken care of regarding the elementary aspects of care. Prior stud-ies of Swedish elderly care have found that relatives and care recipients in general give positive ratings to nursing care (Socialstyrelsen 2006a). In fact it has been reported that staff prioritize the nursing tasks of their work in-stead of, for instance, the social aspects (Socialstyrelsen 2006d), which could be the reason for high ratings regarding nursing care.

The results of the present study also showed that both care recipients and their family members gave lower ratings to information they received from staff and for the activities that were offered to care recipients. Information and activities have also been given low ratings by care recipients and fami-lies in prior studies (Finnema et al. 2001, Maas et al. 1991, Ryan & Scullion 2000, Wellwood et al. 1995). The findings could indicate that these aspects of care were not prioritized by nursing staff as the basic nursing care was. Relatives of Swedish nursing home residents reported communication diffi-culties with nursing staff, and felt a responsibility to initiate and establish interaction with nursing staff (Hertzberg et al. 2001). Relatives also felt that staff had no time to talk to them (Hertzberg et al. 2001).

Our findings also revealed that family members of care recipients gave low ratings for their own possibilities to participate in the care processes, which is also supported by earlier Swedish studies (Socialstyrelsen 2006a) and studies from the USA (Maas et al. 1991). In this thesis relatives of care recipients were considered to be some of the main actors in elderly care, together with staff and care recipients. Relatives have been reported to give a large amount of informal care to their elderly relatives (Socialstyrelsen 2006e). In addition they are actively involved in care decisions for the eld-erly care recipients (Ejaz et al. 2003). The role of family members is further emphasized in the Swedish Health and Medical Services Act (1982), which states that if information regarding care cannot be supplied to the patient, the relatives should be informed instead. However, it seems that the family members in our studies were not satisfied with the information they received or with the opportunities they had to influence care provision. There are

58

several prior studies investigating the relationship between nursing staff and relatives of residents (Bowers 1988, Hertzberg et al. 2001, 2003, Ryan & Scullion 2000). Staff members seem to recognize that family members might be important in patient care (Astedt-Kurki et al. 2001, Weman et al. 2004). However, it has been reported that staff give low priority to working with family members, which was considered to be time-consuming and demand-ing (Hertzberg et al. 2003). It has been proposed that cooperation with rela-tives and staff in elderly care is not fully utilized (Hertzberg & Ekman 2000), and that relatives are undervalued as a resource (Ryan & Scullion 2000). Lack of support and appreciation from management for the work with relatives has been reported (Hertzberg et al. 2003). It is suggested that more structured ways to inform and involve the relatives together with strong su-pervisory support are needed in order to increase the involvement and poten-tial help of relatives of elderly care recipients.

This thesis did not focus on comparing quality of care between home-based care and nursing homes. However, such a comparison could be bene-ficial, since the structure of care provision in nursing homes versus private homes has changed during recent years (Socialstyrelsen 2005e). The ambi-tion of allowing elderly people to live in their own homes has had a powerful impact (Thorslund et al. 1997). It is suggested that studies should be con-ducted comparing the perceptions of care recipients and their families in these care settings in order to investigate how these actors experience the increased home care.

Costs and administrative data of nursing homesAnalyses were conducted to investigate possible correlations between nurs-ing home costs and administrative data regarding staff and resident out-comes, such as prevalence of staff sickness absence and care recipients’ fall incidence, staff ratings of the psychosocial work environment, and care re-cipients’ and their relatives’ ratings of quality of care. The findings revealed few significant correlations between costs and the other variables. The total costs per resident places per day were significantly positively correlated to staff ratings of goal quality at the workplace and relatives’ ratings of four quality of care indices. No significant correlations were found between nurs-ing home costs and administrative data or care recipients’ ratings of quality of care. These results must be interpreted with caution for several reasons. First of all, only nursing homes in one of the municipalities, Uddevalla, were included in the analyses due to incomparable measures of cost between the municipalities. Thus, the results are based on 17 nursing homes. In addition, the measure of nursing home total cost was not completely comparable between the nursing homes in Uddevalla, since some differences were re-ported in the system for cost accounting for care recipient meals. The finan-

59

cial administrators could not provide costs for these meals and no attempts to estimate the cost were included in the present analyses. These analyses must be viewed as a first attempt in identifying associations between costs and other organizational measures in the Swedish elderly care context.

There are few studies from other countries that have investigated nursing home costs in relation to outcome measures of quality of care, and a need for such studies has been reported (Weech-Maldonado et al. 2006). Two earlier studies from the USA found that nursing home costs were significantly asso-ciated with care recipients’ outcome measures such as worsening of pressure ulcers and mood declines (Mukamel & Spector 2000, Weech-Maldonado et al. 2006). The relationships were complex, but were in general non-monotonic with an inverted U-shape, implying that the nursing homes were able to achieve both lower costs and better care recipient outcomes (Mukamel & Spector 2000, Weech-Maldonado et al. 2006). Another study found that availability of individualized activities at the nursing home was associated with lower costs (Nyman 1988). However, that study did not find any of the care recipient outcome measures, such as quality of life, to be significantly related to higher costs (Nyman 1988). The author concluded that costs were not higher in nursing homes with higher quality of care (Nyman 1988). There is one Swedish study comparing nursing home costs, which found that labor costs per resident were lower in care units with a greater number of residents (Svensson et al. 1996). However, no measures of quality of care were included in the study. There is a need for future studies investigating whether nursing home costs are related to quality of care or care recipient outcomes. In addition, research regarding nursing home costs and staff work environment, for instance in terms of competence develop-ment opportunities for nursing staff, could increase the knowledge about how cost is related to staff outcomes. However, in order to conduct such studies it is important to have measures that can be compared between the care units and between care providers. Having these measures correct and comparable is also crucial for the care providers when investigating the effi-ciency of elderly care.

In this thesis an attempt was made to collect administrative data of staff and resident outcome measures. However, in both municipalities some nurs-ing homes lacked data for one or more of the measures and several attempts by the author were necessary in order to receive the missing data. In some cases the data was not collected by a nursing home, despite municipality policy and requirements by the laws regulating work environment and eld-erly care. Difficulties in obtaining outcome measures make any evaluation of elderly care quality challenging. In fact, only half of the Swedish municipali-ties were reported to have recorded adverse events as required by the Health and Medical Services Act (Socialstyrelsen 2001). In addition, quality com-parisons between municipalities have been difficult to conduct, since differ-ent measures and methods have been used (Socialstyrelsen 2001). OECD

60

(2005) concludes that measurement of quality of care is still in its infancy in elderly care. A need for systematic objective quality data collection in eld-erly care has been recognized by The National Board of Health and Welfare (Socialstyrelsen 2006e).

Methodological considerations First of all, this study encompassed only two municipal elderly care organi-zations, which limits generalization of the results. These municipalities and the elderly care organizations differed regarding several characteristics. However, similar results were found in both municipalities regarding staff ratings of their competence and the psychosocial work environment, and care recipients’ and their relatives’ ratings of quality of care, which lends support to the findings of the thesis.

It is also important to remember that almost all data in this thesis were self-rated, i.e. staff, care recipients’, and family members’ ratings were evaluated in this way. Thus, for instance staff competence was not measured with objective measures. It is possible that different results might be ob-tained if more objective measures had been used.

Considerations about the questionnaire studies The QWC questionnaires that were used to measure staff perceptions were previously validated. In addition, the Pyramid questionnaire that was the base for the care recipient and family members’ questionnaires used in the present studies had been validated. The family members’ questionnaire de-veloped in Paper I demonstrated good reliability and validity. The reliability and validity of a questionnaire is important for the ability to draw reliable conclusions, for instance in terms of the instruments’ ability to detect differ-ences between groups and changes occurring during the intervention.

Some methodological limitations need to be taken into consideration when interpreting the results of the questionnaire studies. First of all, the proportion of care recipients and relatives that responded to the surveys was low. It could be concluded that the care recipient respondents did not differ significantly from the total population of care recipients regarding age and gender, but no other information about those who did not respond was avail-able. Thus, it was difficult to determine whether the study samples were representative of the total populations of care recipients and relatives. The evaluation of the intervention was based on a limited number of care recipi-ents and family members and it is possible that different results could have been obtained if a larger proportion of care recipients and relatives had re-sponded to the surveys.

61

It is possible that the quality of care indices were not sensitive enough to detect changes in care recipients’ and relatives’ views. Ceiling effects were found regarding care recipients’ and relatives’ ratings for some of the quality of care indices. It is therefore possible that the ratings at baseline were so high on these scales that additional improvements were difficult to detect in the follow-up measurements. However, ceiling effects were not found on all of the scales used and still no improvements were found. Perhaps it takes a longer time than 18 months for improved staff knowledge to affect the qual-ity of care, as perceived by care recipients and relatives.

The same questionnaires were used in home care and in nursing home set-tings. Attempts to modify the quality of care questionnaires to be suitable for both settings were made, for instance care recipients and relatives from both environments participated in the focus group discussion prior to the ques-tionnaire studies. However, it is possible that some of the questions were not relevant in both care settings. For instance, an evaluation of possibilities for going outdoors might be irrelevant for care recipients capable of going out by themselves. There is also a possibility that elderly individuals could not distinguish between services and care offered by municipalities and those offered by family members. For instance, their evaluation of social activities available may include an evaluation of activities provided by other actors than the municipality.

Finally, the validated target levels for the QWC questionnaire indicating the acceptable staff mean percentage scores for the measurement indices were based on healthcare staff in general and not on elderly care nursing staff. It is possible that the target levels are not exactly the same in these settings, since the work environment itself differs.

Considerations about the toolbox intervention design The toolbox intervention seems to be in accordance with the description of a complex multifactor intervention as described by Rowlands et al. (2005) and Blackwood (2006). Several components, such as staff education, improve-ments of psychosocial work environment, and quality of care were included, which may cause difficulties in the evaluation process (Blackwood 2006, Rowlands et al. 2005). For instance, the mechanism of these components influencing each other and the outcome variables might be difficult to iden-tify (Blackwood 2006, Rowlands et al. 2005). One of the goals of the tool-box intervention was that it should be applicable in staff work practices, which may have negatively influenced the study design and the intervention evaluation, as will be briefly described below.

The fact that the study groups were not randomized made it more difficult to rule out the possible effects of other changes occurring in the organiza-tions during the study period. A randomization of the workplaces within a single municipality to intervention and reference groups would have been an

62

alternative. However, that would probably have led to leakage of informa-tion from intervention workplaces to reference sites, which perhaps was not the case in the present design, since the two municipalities were not geo-graphically close to one other. Clearly, the presence of a reference organiza-tion was a strength, and made a comparison of the results from the interven-tion municipality possible, even though the baseline measurements for the two municipalities did not coincide.

The municipalities and the elderly care organizations involved in the pro-ject differed in several aspects. An attempt was made to compare these or-ganizations. However, there are several characteristics that could have influ-enced the results of the study. For instance, we had no measures for staff-care recipient ratios, care recipient case mix, i.e., care requirements or mu-nicipalities’ criteria for offering services for elderly individuals in these mu-nicipalities. It is possible, for instance, that the care recipient case mix dif-fered between the measurement points or between the municipalities, which could have influenced staff ratings of work strain or stress.

In the present intervention the reference municipality also received a part of the intervention, since the staff’s, care recipients,’ and relatives’ question-naire studies were carried out in both municipalities. The respective results were presented to both organizations and it is possible that the reference organization used the results for development work. However, the ratings of work environment did not improve in the reference municipality. This strengthens the differences found in the post measurements and multivariate analysis and the possibility that the differences were due to the toolbox in-tervention.

The fact that several outcome variables were evaluated was a strength in the present intervention. Most of the earlier educational interventions in eld-erly care have been evaluated on only one outcome variable (Aylward et al.2003). However, the fact that several concepts were included may lead to superficial conceptualization and interpretation of the results. It is not possi-ble to completely define these terms and fully analyze all the possible as-pects of these concepts in one thesis.

A strength of the intervention evaluation was also the two perspectives on quality of care (care recipients’ and their family members’), since a large proportion of care recipients were not able to answer the questionnaire.

Statistical considerations The powers of Studies II-IV were estimated to be sufficient to detect differ-ences between the study groups when differences existed within the popula-tions. No power calculation was carried out for Study I. It is possible that the power was low regarding the hospital study, since we only had 38 respon-dents.

63

The large number of statistical tests conducted in the present thesis might explain the number of significant findings, since multiple tests increase the chance of finding differences between studied groups even though no true differences exist in population (Kazdin 2003). However, the number of sig-nificant differences was not that high in the present studies, and in general the findings were consistently in the hypothesized direction. This indicates that our findings were systematic and not based on chance.

Ethical considerations The TOOL project was approved by research ethics committees, but no ap-proval was sought for the relatives study at USÖ, since that was a hospital internal quality project. There are some ethical issues that need to be dis-cussed. First of all, the questionnaire surveys were anonymous, but the work units in the elderly care organizations were quite small. The staff question-naire asked the respondents to answer some background questions such as occupation, age, and years of work experience. It is possible that staff mem-bers who were a minority of some kind at their workplaces were suspicious of the anonymity. The same could have been the case for care recipients and their relatives. However, the questionnaires were accompanied by an intro-ductory letter explaining that results would only be analyzed at group level.

Participation in the questionnaire studies was voluntary. However, super-visors distributed the questionnaires to staff members, which may have made staff feel obligated to respond. In a similar way care recipients might have been afraid that staff would find out whether they answered the question-naire or not. However, each respondent received a postage-paid return enve-lope. Each completed questionnaire was enclosed in the envelope and sent directly to a data registration company. Supervisors or staff members had no knowledge of which individuals responded.

Future researchIt is suggested that an educational toolbox, consisting of practical instru-ments adapted to local needs, could be an effective intervention to improve staff competence and work environment. However, the toolbox concept should be further investigated in other organizations in order to further evaluate the impact on staff competence and work environment as well as on quality of care. A randomized controlled study design would allow a stronger evaluation of the impact.

Further analyses at the workplace level are also needed to investigate characteristics of intervention workplaces using the toolbox to a greater ex-tent. These analyses could contribute to further knowledge regarding

64

whether extensive toolbox use improves competence and work environment ratings more than less extensive use.

There is a need to further assess whether there is an association between staff competence and care recipients’ and relatives’ ratings of quality of care, and what the possible mechanisms are between these factors. It should also be further investigated whether staff change their working behavior after improving their knowledge and skills, and what circumstances would in-crease the chances of staff changing their behavior.

It is suggested that future studies should combine subjective and more ob-jective measures of quality of care in order to achieve a greater understand-ing of the possible effects of educational interventions on the quality of care.

Future studies should also investigate the psychosocial work environment and quality of care between home-based care and nursing homes. Perhaps there are more specific interventions in each environment that could be con-sidered in order to decrease negative staff outcomes such as exhaustion, and to improve quality of care.

Studies focusing on possible non-response bias when investigating care recipients’ and their family members’ perceptions of quality of care are also needed.

The preliminary findings that higher costs were apparently correlated with higher ratings by staff and relatives in some of the areas assessed need fur-ther exploration. Reliable cost measures that can be compared between dif-ferent care units and between care providers are crucial for researchers, but also for the care providers when investigating the efficiency of elderly care.

Conclusions A substantial need was found for competence development among elderly care nursing staff, both in home-based care and nursing homes. High work-related exhaustion levels were also found in both care settings and work-related exhaustion was found to the strongest negative predictor of staff rat-ings of work satisfaction i.e. higher ratings of exhaustion predicted lower ratings of work satisfaction. The findings also provided specific information about causes of emotional and physical work strain as well as about aspects of the psychosocial work environment that were rated as limited. These fac-tors should be taken into consideration in future efforts to improve the work situations of elderly care staff.

The basic elements of elderly care, such as personal hygiene, meals, and physical transfers, were rated high by care recipients and their relatives. Other aspects, such as information received from staff, activities provided to care recipients, and relatives’ opportunities to participate in care provision were rated lower by the respondents. However, there are several possible methodological limitations, such as a rather small proportion of care recipi-

65

ents and relatives responding to the surveys, which need to be taken into consideration when interpreting the findings with elderly care recipients and their family members. These methodological challenges should also be taken into consideration in future studies with care recipients and their relatives.

It is suggested that the educational intervention, the “toolbox” that is adapted to local needs, might be an important and sustainable intervention to improve nursing staff competence, psychosocial work environment, and work satisfaction in elderly care. However, despite the positive effects on staff’s skills and work environment, no significant impact on care recipients’ and their family members’ ratings of the quality of care were found. This indicates a need to further expand our understanding of the possible link between nursing staff competence, the psychosocial work environment and the quality of care provided to elderly people.

66

Acknowledgements

The work with this thesis has given me an incredible opportunity to meet and interact with interesting and talented people including study participants, municipality representatives and co-workers from different universities. A great number of people have contributed with their time and knowledge to the research project. I would like to extend my deepest gratitude to all of you and I hope that I haven’t accidentally forgotten any of you in the following paragraphs. If that would be the case, please be certain that your help has been appreciated. I would like to extend my deepest thanks to:

My honourable advisor, Associate professor, Judith Arnetz, for always thoroughly and carefully reading and commenting the papers and this thesis. It has been a true privilege to receive such a detailed and constructive com-ments on my work. You have introduced me to the world of science and I have learned a lot from you. Thank you for your patience, inspiration and support along the way.

My co-advisor Professor Bengt Arnetz for inspiration and wise advice along the research project. You have patiently followed me on this journey from the very beginning. You have to a great extent designed and recruited finances for the original TOOL- research project, which made this thesis possible. Thank you for always being available to answer to my questions and for proofreading the papers and this thesis.

My co-advisor, Professor Marianne Carlsson, for your good collabora-tion, support and immense knowledge regarding research and statistics. Thank you for your essential proofreading and invaluable comments on the papers and this thesis. I’m especially grateful that you always had time and patience to answer to my questions.

My co-advisor Dr Paul Nystedt for providing a “macro perspective” to my research. You have always managed to provide a wider societal perspec-tive and it has been inspiring to experience your excellent talent in making systematic interpretations. Your feedback on the papers always included nicely organized models, which have been appreciated! Thank you for gen-erously sharing your time and knowledge and for providing brilliant com-ments on the papers and this thesis.

The Vårdal Institute and its all co-workers for support and inspiration. It is not possible to mention how much the research institute has done for me professionally. I have felt extremely lucky to have had the opportunity to be connected to the institute and its co-workers during my years as a PhD stu-

67

dent. I would especially like to extend my sincere gratitude to Professor In-galill Rahm Hallberg for always being enthusiastic and supportive. Your work to develop The Vårdal Institute has showed me that everything is pos-sible –with hard work it is possible to achieve your goals! Professor SölveElmståhl for many interesting and inspiring conversations within the re-search platform for elderly care. Your kindness, humour and extensive knowledge about research have contributed to stimulating multidisciplinary scientific meetings that have taught me a lot. Dr Anna-Karin Edberg and Dr Synneve Dahlin-Ivanof for your wise advice and your great work to further develop the research platform. All PhD students at the Vårdal Insti-tute for inspiring conversations about science and other important things in life. I’m especially grateful to Ann-Helene Almborg, Neda Agahi, Anna-Karin Welmer, Ingela Steij Stålbrand and Elisabet Mattsson for support, friendship and good laughs. The great group of brave PhD students at the Vårdal Institute who I had the opportunity to write the “Etiska utmaningar i hälso- och sjukvården”: Kristofer Hansson, Harshida Patel, Anne Wen-nick and Henrik Levinsson. It was a stimulating project and I learned a lot from your constructive, multidisciplinary comments. My deepest gratitude also goes to Professor Björn Lindgren for the invaluable advice regarding my chapter in the book.

My employer Uppsala University for providing excellent working condi-tions and the colleagues at the Uppsala University for inspiration and warm support. My sincere gratitude to the active group of researchers at the “Tuesday seminars” for sharing their skills and knowledge. These seminars have been valuable and appreciated.

All the contact persons in Bengtsfors and Uddevalla who kindly pro-vided their help in data collection. I would especially like to thank SusanneErixon, Lillemor Johansson, Annika Simonsson, Sara Molander, Lynn Joel and Pär Levander for patiently answering to my, sometimes very complicated, questions.

Dr Sarah Thomsen who was part of the original TOOL-project group and was responsible for implementation and analysis of the pilot study; and Josepine Fernow, who was TOOL-project administrator during the initial phases of the project.

Inge Jonsson, Ulla Areskoug and Annika Bengtsson at Åmålshälsan for their help with the development of the original TOOL-project concep-tion. Agneta Andersson, also a part of the original project group, for the significant help during the whole study period.

Solna Stans for fast and professional services regarding data registration. All the administrative staff at the Department of Public Health and

Caring Sciences and at The Vårdal Institute for all your help and patience with the practical aspects of my work.

All talented people with computer related knowledge that have helped me with IT-based problems. I would especially like to extend my thanks to

68

Anna Fryczynska, Per Kjellsson, Håkan Jansson and Maxim Escurel for your help and patient.

Staff at the Library of Health and Caring Sciences at the Uppsala Uni-versity for excellent and professional help.

Margot Lundquist and Margit Robeson for fast and excellent language revision.

Inna Kinnunen, Mikko Olli and Jennie Norrman for valuable help with data collection.

Dr Vanja Blomkvist & Claes Blomkvist for warm friendship, great dis-cussions and exciting outdoor excursions. Dr Claire Brown & Dr Jen Sed-don for inspiring, humoristic and sometimes crazy conversations about dif-ferent aspects of life. I also would like to thank Claire for excellent courses in statistics. Ronnit Hasson for warm friendship and relaxing conversations that have provided me with great opportunities to lighten up. Family of Lindblad/Kujiper for encouragement and genuine support throughout the years. Annika & Mikael Hannus for kind friendship and inspiration.

Kaisa Alihanka, Leena Lampila, Maarika & Juha Vähä-Lassila, Ca-rita Ahonen, Pauliina Salminen, Sini & Petri Kieksi, Saija Kangasjärvi, Riina Rikalainen-Salmi & Antti Salmi, Hanna & Ari-Pekka Heikkinen, Liisa Tuohimaa and Petteri Löppönen for genuine long-lasting, long dis-tance friendship that has inspired and supported me.

The warmest appreciation to my parents-in-law Aviva & Eli Hasson for loving attitude and for your outstanding help in the everyday life. With your commitment and support it was possible for me to find necessary time and strength to concentrate on the work. Your help has been invaluable and I’m eternally grateful for your contribution.

The warmest appreciation to my amazing grandparents Liisa & Altti Ju-vonen as well as to Merja & Tuomo Korpela and Matti Jalonen for sup-port throughout my life. The entire Verho-family for showing genuine inter-est for my projects and always supporting me on my choices.

My lovely parents Anne Juvonen & Kalle Verho for your endless sup-port. I have got many gifts from you, and the most important for this thesis is probably your enthusiasm for books. You have inspired me and taught me the joy of reading, and without that I would never have accomplished this project. My most humble gratitude to you.

My lovely husband Dan and son Daniel, who are my true inspirations and joy in life. Dan, you have always provided a strengthening and loving atti-tude, which have supported me along this journey. You have believed in me and encouraged me with patience and love and I’m eternally grateful for having you in my life.

The studies in this thesis were financially supported by: The Swedish Agency for Innovation Systems VINNOVA

69

The Swedish Foundation for Health Care Sciences and Allergy Research (The Vårdal Foundation) The Vårdal InstituteSwedish Skandia Life (publ) Group Bengtsfors and Uddevalla municipalities Örebro University hospital

70

Sammanfattning (in Swedish)

Det övergripande syftet med denna avhandling var att utveckla, implemente-ra och utvärdera en utbildningsintervention för personal inom kommunal äldreomsorg. Tidigare studier har visat att personalen har låg formell kompe-tens och att möjligheterna till vidareutbildning är begränsade. Samtidigt har de kommunala resurserna allt mer koncentrerats till äldre med större vård- och omsorgsbehov, vilket har inneburit ett större krav på personalens kun-skaper kring medicinskt komplicerade sjukdomar och demens.

Bristande kompetens har i tidigare forskning associerats med högre stress, lägre arbetsstillfredställelse och lägre vårdkvalitet. Det är därför angeläget att studera utbildningsinsatsers effekter på personalens psykosociala arbets-miljö och vårdkvalitet. I denna avhandling utvärderades möjliga effekter av en utbildningsintervention på personalupplevd kompetens och psykosocial arbetsmiljö, samt på vårdtagarnas och deras anhörigas uppfattning av vård-kvalitet. Utvärderingen genomfördes i en prospektiv, kontrollerad studie, där resultat jämfördes mellan en interventionskommun och en referenskommun.

Den första studien i avhandlingen syftade till att testa och validera en en-kät som senare i studie IV användes för att mäta anhörigas upplevelser av vårdkvalitet. Anhörigas svar på enkäten på Örebro universitetssjukhuset och vid två mättillfällen i interventionskommunens äldreomsorg under år 2001 användes som datamaterial. Psykometrisk analys med bland annat faktorana-lys resulterade i åtta kvalitetsindex/skalor och ett övergripande kvalitetsbe-tyg. Det fanns relativt höga korrelationer inbördes mellan några index. En andra ordningens faktoranalys visade att indexen ytterligare kunde grupperas i tre domäner: personal, anhörigas roll och vårdens innehåll. Enkäten visade god reliabilitet och validitet men det behövs ytterligare studier för att bekräf-ta indexens validitet.

Personal som arbetade i hemvård respektive särskilda boenden ingick i in-terventionen. Studie II syftade till att jämföra skattningar av kompetens, arbetsbelastning, stress och arbetstillfredsställelse mellan personal som arbe-tar inom hemvård respektive på särskilda boenden. Personalens svar på en-kätstudierna i interventions- respektive referenskommunerna 2003 användes som datamaterial. Resultaten visade att en signifikant större andel personal i hemvården skattade sin kompetens som otillräcklig jämfört med personal som arbetade på särskilt boende. Hemvårdspersonal upplevde dock signifi-kant lägre arbetsbelastning, både fysisk och psykisk, jämfört med personal på särskilda boenden. Personalens skattningar av arbetsrelaterad utmattning,

71

psykisk energi och arbetstillfredsställelse skiljde sig inte signifikant mellan de olika vårdmiljöerna. Arbetsrelaterad utmattning var den starkaste (nega-tiv) förklaringsfaktorn för arbetstillfredsställelse i båda miljöerna.

Studierna III och IV syftade till att utvärdera möjliga effekter av utbild-ningsinterventionen för äldreomsorgens personal på deras skattningar av kompetens, psykosocial arbetsmiljö (studie III) samt på vårdtagares och an-hörigas skattningar av vårdkvalitet (studie IV). Studien var en prospektiv, icke-randomiserad, kontrollerad intervention. Interventionen baserades på resultaten från baslinjemätningar som genomfördes 2001 bland personal, vårdtagare och anhöriga. Resultaten från dessa enkätstudier användes sedan vid framtagning av en ”verktygslåda”, som bestod av olika instrument som personalen kunde använda för att förbättra och vidareutveckla verksamheten. Studiernas resultat visade att personalens skattningar av sin kompetens, psykosociala arbetsmiljö och arbetstillfredsställelse förbättrades signifikant i uppföljningsmätningar i 2003 och 2004 och jämfört med personalen i refe-renskommunen, där ingen verktygslåda introducerades. Interventionen visa-de ingen mätbar effekt på vårdtagarnas och anhörigas vårdkvalitetsskatt-ningar, som var oförändrade över tid i såväl interventions- som referens-kommunen.

Ett antal slutsatser kan dras utifrån resultaten i avhandlingen. Det fanns ett påtagligt kompetensutvecklingsbehov hos äldreomsorgens personal både i hemvården och i särskilda boenden. Personalens arbetsrelaterade utmattning var hög och arbetsrelaterad utmattning var den starkaste prediktorn (negativ) för personalens skattningar av arbetsstillfredställelse. Det innebär att högre skattningar av utmattning predicerade lägre skattningar av arbetsstillfredstäl-lelse. Framtida interventioner som syftar till att öka arbetstillfredsställelse bör fokusera på att stödja kompetensutveckling och motverka arbetsrelaterad utmattning både i hemvården och i särskilda boenden. Resultaten gav också tydlig information om vilka aspekter av arbetet som upplevdes mest psykiskt och fysiskt påfrestande samt vilka komponenter av den psykosociala arbets-miljön som uppvisade störst utvecklingspotential. Dessa fynd kan vara väg-ledande i framtida interventioner som syftar till att utveckla arbetssituationen för personal inom äldreomsorgen.

De grundläggande aspekterna av äldrevården, det vill säga personlig hy-gien, måltider och fysiska förflyttningar, skattades högt av både vårdtagare och deras anhöriga. Andra aspekter såsom information från personalen och aktiviteter som erbjuds till de äldre, skattades betydligt lägre. Anhöriga an-såg också att deras möjlighet att vara delaktig i vården av sin närstående var begränsad.

Sammanfattningsvis kan konstateras att en utbildningsintervention i form av en “verktygslåda” som är anpassad till arbetsplatsernas behov, kan vara en viktig och effektiv intervention för att utveckla personalens kompetens och psykosociala arbetsmiljö inom äldreomsorg. Däremot kunde inga effek-ter på vårdtagarnas eller anhörigas skattningar av vårdkvalitet noteras. Ytter-

72

ligare studier behövs för att undersöka om, och i sa fall hur, personalens kompetens är relaterad till vårdkvalitet. I sådana studier borde olika aktörers upplevelser av verksamheten kompletteras med objektiva indikatorer på kvalitet.

73

References

Aharony L & Strasser S (1993): Patient satisfaction: what we know about and what we still need to explore. Med Care Rev 50, 49-79.

Aiken LH, Smith HL & Lake ET (1994): Lower Medicare mortality among a set of hospitals known for good nursing care. Med Care 32, 771-787.

Ansvarskommittén (2003) Vård och omsorg om äldre – samhällsorganisation, an-svar och styrning. (Elderly care - societal organization, responsibility, and con-trol. In Swedish) SOU. Underlags-PM. Fi 2003-02.

Arbetarskyddsstyrelsen (2000) Stress och belastning i vård och omsorg: verksge-mensamt tillsynsprojekt 1998 och 1999 (Stress and work load in nursing: a sur-veillance project in 1998 and 1999. In Swedish). Arbetarskyddsstyrelsen, Solna.

Arbetsmiljöverket (1990) Omvårdnadsarbete i enskild hem. Arbetarskyddsstyrelsens kungörelse med föreskrifter om omvårdnadsarbete i enskilt hem samt allmänna råd om tillämpningen av föreskrifterna (Nursing care work in the private home. In Swedish) (AFS 1990:18).

Arbetsmiljöverket (2001) Systematisk arbetsmiljöarbete -Arbetsmiljöverkets före-skrifter om systematiskt arbetsmiljöarbete och allmänna råd om tillämpningen av föreskrifterna (A systematic work environment work. In Swedish) (AFS 2001:1).

Arnetz BB (1999): Staff perception of the impact of health care transformation on quality of care. Int J Qual Health Care 11, 345-351.

Arnetz BB & Wiholm C (1997): Technological stress: psychophysiological symp-toms in modern offices. J Psychosom Res 43, 35-42.

Arnetz JE & Arnetz BB (1996): The development and application of a patient satis-faction measurement system for hospital-wide quality improvement. Int J Qual Health Care 8, 555-566.

Astedt-Kurki P, Paavilainen E, Tammentie T & Paunonen-Ilmonen M (2001): Inter-action between adult patients' family members and nursing staff on a hospital ward. Scand J Caring Sci 15, 142-150.

Astrom S, Nilsson M, Norberg A, Sandman PO & Winblad B (1991): Staff burnout in dementia care-relations to empathy and attitudes. Int J Nurs Stud 28, 65-75.

Axelsson R (2000): The organizational pendulum-healthcare management in Swe-den 1865-1998. Scand J Public Health 28, 47-53.

Aylward S, Stolee P, Keat N & Johncox V (2003): Effectiveness of continuing edu-cation in long-term care: a literature review. Gerontologist 43, 259-271.

Baillon S, Scothern G & Vickery L (1999): Job satisfaction and stress in staff work-ing in a specialist psychiatric unit for the elderly following relocation from a traditional psychiatric hospital setting. J Nurs Manag 7, 207-214.

Bartoldus E, Gillery B & Sturges PJ (1989): Job-related stress and coping among home-care workers with elderly people. Health Soc Work 14, 204-210.

Beck C, Ortigara A, Mercer S & Shue V (1999): Enabling and empowering certified nursing assistants for quality dementia care. Int J Geriatr Psychiatry 14, 197-211; discussion 211-192.

74

Bengtsfors kommun (2003): Årsredovisning 2003 (Municipality annual report 2003. In Swedish).

Bengtsfors kommun (2004): Årsredovisning 2004 (Municipality annual report 2004. In Swedish).

Bergmark A, Parker MG & Thorslund M (2000): Priorities in care and services for elderly people: a path without guidelines? J Med Ethics 26, 312-318.

Bergmark Å (1997): From reforms to rationing? Current allocative trends in social services in Sweden. Scand J Soc Welfare 6, 74–81.

Blackwood B (2006): Methodological issues in evaluating complex healthcare inter-ventions. J Adv Nurs 54, 612-622.

Blegen MA (1993): Nurses' job satisfaction: a meta-analysis of related variables. Nurs Res 42, 36-41.

Blomberg S, Edebalk P & Peterson J (2000): The withdrawal of the welfare state: elderly care in Sweden in the 1990s. Eur J Soc Work 3, 151-163.

Boswell CA (1992): Work stress and job satisfaction for the community health nurse. J Community Health Nurs 9, 221-227.

Boumans N, Berkhout A & Landeweerd A (2005): Effects of resident-oriented care on quality of care, wellbeing and satisfaction with care. Scand J Caring Sci 19,240-250.

Bowers BJ (1988): Family perceptions of care in a nursing home. Gerontologist 28,361-368.

Bowers BJ, Fibich B & Jacobson N (2001a): Care-as-service, care-as-relating, care-as-comfort: understanding nursing home residents' definitions of quality. Geron-tologist 41, 539-545.

Bowers BJ, Lauring C & Jacobson N (2001b): How nurses manage time and work in long-term care. J Adv Nurs 33, 484-491.

Broad ML (1997): Overview of transfer of training: From learning to performance. Performance Improvement Quarterly 10, 7-21.

Brodaty H, Draper B & Low LF (2003): Nursing home staff attitudes towards resi-dents with dementia: strain and satisfaction with work. J Adv Nurs 44, 583-590.

Brown C, Arnetz B & Petersson O (2003): Downsizing within a hospital: cutting care or just costs? Soc Sci Med 57, 1539-1546.

Brulin C, Gerdle B, Granlund B, Hoog J, Knutson A & Sundelin G (1998): Physical and psychosocial work-related risk factors associated with musculoskeletal symptoms among home care personnel. Scand J Caring Sci 12, 104-110.

Brulin C, Winkvist A & Langendoen S (2000): Stress from working conditions among home care personnel with musculoskeletal symptoms. J Adv Nurs 31,181-189.

Bäckman O (2001) Med välfärdsstaten som arbetsgivare -arbetsmiljön och dess konsekvenser inom välfärdstjänsteområdet på 1990-talet (The welfare state as employer - work environment and its imlications for welfare services during the 1990s. In Swedish) In: Välfärdstjänster i omvandling: antologi från Kommittén Välfärdsbokslut SOU 2001:52. Edited by Szebehely, M., Stockholm.

Chang JT, Hays RD, Shekelle PG, MacLean CH, Solomon DH, Reuben DB, Roth CP, Kamberg CJ, Adams J, Young RT & Wenger NS (2006): Patients' global ratings of their health care are not associated with the technical quality of their care. Ann Intern Med 144, 665-672.

Chiu L (1997): Family caregiver's satisfaction with home care in the Taipei metro-politan area. Public Health Nurs 14, 42-50.

Chong AM (2003): Validating the accuracy of the resident satisfaction measure. JCross-Cult Ger 18, 203-226.

75

Chou SC, Boldy DP & Lee AH (2002): Staff satisfaction and its components in residential aged care. Int J Qual Health Care 14, 207-217.

Chou SC, Boldy DP & Lee AH (2003): Factors influencing residents' satisfaction in residential aged care. Gerontologist 43, 459-472.

Clarke SP, Sloane DM & Aiken LH (2002): Effects of hospital staffing and organ-izational climate on needlestick injuries to nurses. Am J Public Health 92, 1115-1119.

Cohen-Mansfield J & Rosenthal AS (1989): Absenteeism of nursing staff in a nurs-ing home. Int J Nurs Stud 26, 187-194.

Cohen J (1965) Some statistical issues in psychological research. In: Handbook of clinical psychology Edited by Wolman, B. B. McGraw-Hill New York, New York.

Cowan DT, Roberts JD, Fitzpatrick JM, While AE & Baldwin J (2004): The ap-proaches to learning of support workers employed in the care home sector: an evaluation study. Nurse Educ Today 24, 98-104.

Curry A & Stark S (2000): Quality of service in nursing homes. Health Serv Man-age Res 13, 205-215.

Curtis MP, Sales AE, Sullivan JH, Gray SL & Hedrick SC (2005): Satisfaction with care among community residential care residents. J Aging Health 17, 3-27.

D'Eramo A L, Papp KK & Rose JH (2001): A program on complementary therapies for long-term care nursing assistants. Geriatr Nurs 22, 201-207.

Dahl-Jorgensen C & Saksvik PO (2005): The impact of two organizational interven-tions on the health of service sector workers. Int J Health Serv 35, 529-549.

Daigh RD (1991): Financial implications of a quality improvement process. Top Health Care Financ 17, 42-52.

Davis MA (1991): On nursing home quality: a review and analysis. Med Care Rev48, 129-166.

Dellefield ME (2000): The relationship between nurse staffing in nursing homes and quality indicators. J Gerontol Nurs 26, 14-28.

Denton M, Zeytinoglu IU, Davies S & Lian J (2002a): Job stress and job dissatisfac-tion of home care workers in the context of health care restructuring. Int J Health Serv 32, 327-357.

Denton MA, Zeytinoglu IU & Davies S (2002b): Working in clients' homes: the impact on the mental health and well-being of visiting home care workers. Home Health Care Serv Q 21, 1-27.

Donabedian A (1966): Evaluating the quality of medical care. Milbank Mem Fund Q44, Suppl:166-206.

Donabedian A (1996): The effectiveness of quality assurance. Int J Qual Health Care 8, 401-407.

Dougall AL & Baum A (2001) Stress, health, and illness. In Handbook of Health Psychology Edited by Baum, A., Revenson, T., Singer, J. Lawrence Erlbaum As-sociates, Mahwah, N.J.; London, p. 321–337.

Duffy JA, Duffy M & Kilbourne WE (2001): A comparative study of resident, fam-ily, and administrator expectations for service quality in nursing homes. Health Care Manage Rev 26, 75-85.

Dunn LA, Rout U, Carson J & Ritter SA (1994): Occupational stress amongst care staff working in nursing homes: an empirical investigation. J Clin Nurs 3, 177-183.

Edebalk P & Lindgren B (1994) Social care and the elderly. In: Population, econ-omy, and welfare in Sweden. Edited by Bengtsson, T. Springer-Verlag, Heidel-berg.

76

Edvardsson B (1996) Kvalitet och tjänsteutveckling (Quality and service develop-ment. In Swedish). Studentlitteratur.

Edvardsson B (1998): Service quality improvement. Managing Service Quality 8,142-149.

Ejaz FK, Straker JK, Fox K & Swami S (2003): Developing a satisfaction survey for families of Ohio's nursing home residents. Gerontologist 43, 447-458.

Ersek M, Kraybill BM & Hansberry J (1999): Investigating the educational needs of licensed nursing staff and certified nursing assistants in nursing homes regard-ing end-of-life care. Am J Hosp Palliat Care 16, 573-582.

Fahlstrom G & Kamwendo K (2003): Increased physiotherapy in sheltered housing in Sweden: a study of structure and process in elderly care. Health Soc Care Community 11, 470-476.

Finnema E, de Lange J, Droes RM, Ribbe M & van Tilburg W (2001): The quality of nursing home care: do the opinions of family members change after imple-mentation of emotion-oriented care? J Adv Nurs 35, 728-740.

Fitzpatrick JM & Roberts JD (2004): Challenges for care homes: education and training of healthcare assistants. Br J Nurs 13, 1258-1261.

Fleming ST (1990): The relationship between the cost and quality of hospital care: a review of the literature. Med Care Rev 47, 487-502.

Flood AB (1994): The impact of organizational and managerial factors on the qual-ity of care in health care organizations. Med Care Rev 51, 381-428.

Forbes DA & Neufeld A (1997): Strategies to address the methodological challenges of client-satisfaction research in home care. Can J Nurs Res 29, 69-77.

Friedman SM, Daub C, Cresci K & Keyser R (1999): A comparison of job satisfac-tion among nursing assistants in nursing homes and the Program of All-inclusive Care for the Elderly (PACE). Gerontologist 39, 434-439.

Goodell TT & Van Ess Coeling H (1994): Outcomes of nurses' job satisfaction. JNurs Adm 24, 36-41.

Grau L, Chandler B & Saunders C (1995): Nursing home residents' perceptions of the quality of their care. J Psychosoc Nurs Ment Health Serv 33, 34-41.

Grieshaber LD, Parker P & Deering J (1995): Job satisfaction of nursing assistants in long-term care. Health Care Superv 13, 18-28.

Gummesson E (1992): Quality dimensions: what to measure in service organiza-tions. Advances in Services Marketing and Management 1, 177–205.

Gustafson DH, Sainfort F, Eichler M, Adams L, Bisognano M & Steudel H (2003): Developing and testing a model to predict outcomes of organizational change. Health Serv Res 38, 751-776.

Gustafsson RÅ & Szebehely M (2005) Arbetsvillkor och styrning i äldreomsorgens hierarki: - en enkätstudie bland personal och politiker (Work conditions and control in elderly care: - a questionnaire study for staff and politicians. In Swe-dish). Institutionen för socialt arbete - Socialhögskolan Stockholms universitet, Stockholm.

Haggstrom E, Skovdahl K, Flackman B, Kihlgren AL & Kihlgren M (2004): To feel betrayed and to feel that you are betraying the older residents: caregivers' ex-periences at a newly opened nursing home. J Clin Nurs 13, 687-696.

Hannan S, Norman I & Redfern S (2001): Care work and quality of care for older people: a review of the research literature. Reviews in Clinical Gerontology 11,189-203.

Hansson UW, Hallberg IR & Axelsson K (1995): Nurses' satisfaction with nursing care and work at three care units for severely demented people. J Psychiatr Ment Health Nurs 2, 151-158.

77

Hasson D (2005) Stress management interventions and predictors of long-term health: prospectively controlled studies on long-term pain patients and a healthy sample from IT-and media companies. Academic dissertation. Acta Universitatis Upsaliensis: Uppsala University Sweden.

Health and Medical Services Act (1982) Hälso- och sjukvårdslag (1982:763). Hertzberg A & Ekman SL (1996): How the relatives of elderly patients in institu-

tional care perceive the staff. Scand J Caring Sci 10, 205-211. Hertzberg A & Ekman SL (2000): 'We, not them and us?' Views on the relationships

and interactions between staff and relatives of older people permanently living in nursing homes. J Adv Nurs 31, 614-622.

Hertzberg A, Ekman SL & Axelsson K (2001): Staff activities and behaviour are the source of many feelings: relatives' interactions and relationships with staff in nursing homes. J Clin Nurs 10, 380-388.

Hertzberg A, Ekman SL & Axelsson K (2003): 'Relatives are a resource, but.' regis-tered nurses' views and experiences of relatives of residents in nursing homes. JClin Nurs 12, 431-441.

Higgs PF, MacDonald LD, MacDonald JS & Ward MC (1998): Home from home: residents' opinions of nursing homes and long-stay wards. Age Ageing 27, 199-205.

Hoff T, Jameson L, Hannan E & Flink E (2004): A review of the literature examin-ing linkages between organizational factors, medical errors, and patient safety. Med Care Res Rev 61, 3-37.

Hsu MC, Moyle W, Creedy D & Venturato L (2005): An investigation of aged care mental health knowledge of Queensland aged care nurses. Int J Ment Health Nurs 14, 16-23.

Hörstedt J, Prutz c, Wells C, Edebalk P & Lindgren B (1996) Hur kan man förstå variationer i kostnaderna för kommunernas äldre-och handikappomsorg? En explorativ förklaringsmodell (How can differences in municipalities' costs for elderly and disabled care be understood? An explorative model. In Swedish).Lunds Univ., Samhällsmedicinska institutionen.

Jansen PG, Kerkstra A, Abu-Saad HH & van der Zee J (1996): The effects of job characteristics and individual characteristics on job satisfaction and burnout in community nursing. Int J Nurs Stud 33, 407-421.

Jenkins H & Allen C (1998): The relationship between staff burnout/distress and interactions with residents in two residential homes for older people. Int J Geri-atr Psychiatry 13, 466-472.

Johansson JA (1995): Psychosocial work factors, physical work load and associated musculoskeletal symptoms among home care workers. Scand J Psychol 36, 113-129.

Karasek RA (1979): Job Demands, Job Decision Latitude, and Mental Strain: Impli-cations for Job Redesign. Administrative Science Quarterly 24, 285-308.

Karasek RA, Theorell T, Schwartz JE, Schnall PL, Pieper CF & Michela JL (1988): Job characteristics in relation to the prevalence of myocardial infarction in the US Health Examination Survey (HES) and the Health and Nutrition Examina-tion Survey (HANES). Am J Public Health 78, 910-918.

Karsh B, Booske BC & Sainfort F (2005): Job and organizational determinants of nursing home employee commitment, job satisfaction and intent to turnover. Ergonomics 48, 1260-1281.

Kazdin AE (2003) Research design in clinical psychology. 4th edition. Allyn and Bacon Boston, Boston.

Kerkstra A & Hutten JB (1996): Organization and financing of home nursing in the European Union. J Adv Nurs 24, 1023-1032.

78

Kommunstyrelsen (2001) Årsredovisning för kommunkoncernen Bengtsfors (Annu-al report for the municipality of Bengtsfors. In Swedish). Bengtsfors kommun.

Kvalitetsnätverk Västkust (2003) Kvalitets- och kostnadsjämförelse inom särskilt boende (Nursing home quality and cost comparison. In Swedish). Kvalitetsnät-verk Västkust.

Källestål C (2004) Hälsofrämjande arbete på arbetsplatser–Effekter av interventio-ner refererade i systematiska kunskapsöversikter och i svenska rapporter (Health-promoting at workplaces - Effects of health-promotion interventions in systematic litterature reviews and in Swedish reports. In Swedish with English summary). Statens Folkhälsoinstitut, Sandviken.

Laamanen R, Broms U, Happola A & Brommels M (1999): Changes in the work and motivation of staff delivering home care services in Finland. Public Health Nurs 16, 60-71.

Laitinen P, Merilainen P & Sinkkonen S (1996): Quality of elderly-patient care: an interrupted time series study. Int J Nurs Pract 2, 129-137.

Lang TA, Hodge M, Olson V, Romano PS & Kravitz RL (2004): Nurse-patient ratios: a systematic review on the effects of nurse staffing on patient, nurse em-ployee, and hospital outcomes. J Nurs Adm 34, 326-337.

Lu H, While AE & Barriball KL (2005): Job satisfaction among nurses: a literature review. Int J Nurs Stud 42, 211-227.

Lubart E, Leibovitz A & Habot B (2004a): Attitudes of relatives and nursing staff toward tuboenteral feeding in severely demented patients. Am J Alzheimers Dis Other Demen 19, 31-34.

Lubart E, Leibovitz A, Shapira A, Tischenko O, Peled Z, Baumoehl Y & Habot B (2004b): Satisfaction with the care of institutionalized psychogeriatric patients, as reflected by a survey of their relatives. Arch Gerontol Geriatr 38, 45-50.

Lundstrom T, Pugliese G, Bartley J, Cox J & Guither C (2002): Organizational and environmental factors that affect worker health and safety and patient outcomes. Am J Infect Control 30, 93-106.

Maas ML, Buckwalter KC & Kelley LS (1991): Family members' perceptions of care of institutionalized patients with Alzheimer's disease. Appl Nurs Res 4,135-138.

Maslin-Prothero SE (1997): A perspective on lifelong learning and its implications for nurses. Nurse Educ Today 17, 431-436.

McAiney CA (1998): The development of the empowered aide model. An interven-tion for long-term care staff who care for Alzheimer's residents. J Gerontol Nurs24, 17-22; quiz 60.

McCartan-Quinn D, McAleer E & Naqvi I (1996): Service quality: a satisfaction survey of the elderly. Int J Health Care Qual Assur 9, 4-10.

Meijer A, Van Campen C & Kerkstra A (2000): A comparative study of the financ-ing, provision and quality of care in nursing homes. The approach of four Euro-pean countries: Belgium, Denmark, Germany and the Netherlands. J Adv Nurs32, 554-561.

Mentes JC & Tripp-Reimer T (2002): Barriers and facilitators in nursing home in-tervention research. West J Nurs Res 24, 918-936.

Meyer JD & Muntaner C (1999): Injuries in home health care workers: an analysis of occupational morbidity from a state compensation database. Am J Ind Med35, 295-301.

Mitchell PH & Shortell SM (1997): Adverse outcomes and variations in organiza-tion of care delivery. Med Care 35, NS19-32.

Molloy DW, Guyatt GH, Russo R, Goeree R, O'Brien BJ, Bedard M, Willan A, Watson J, Patterson C, Harrison C, Standish T, Strang D, Darzins PJ, Smith S &

79

Dubois S (2000): Systematic implementation of an advance directive program in nursing homes: a randomized controlled trial. The Journal of the American Medical Association 283, 1437-1444.

Morgan DG, Semchuk KM, Stewart NJ & D'Arcy C (2002): Job strain among staff of rural nursing homes. A comparison of nurses, aides, and activity workers. JNurs Adm 32, 152-161.

Moyle W, Skinner J, Rowe G & Gork C (2003): Views of job satisfaction and dis-satisfaction in Australian long-term care. J Clin Nurs 12, 168-176.

Mukamel DB & Spector WD (2000): Nursing home costs and risk-adjusted outcome measures of quality. Med Care 38, 78-89.

Möller T (1996) Brukare och klienter i välfärdsstaten om missnöje och påverkans-möjligheter inom barn- och äldreomsorg (Clients in the welfare discuss dissatis-faction and opportunities to influence pediatric and elderly care. In Swedish).Norstedts juridik AB, Stockholm.

Navaie-Waliser M, Lincoln P, Karuturi M & Reisch K (2004): Increasing job satis-faction, quality care, and coordination in home health. J Nurs Adm 34, 88-92.

Nelson DL, Quick JC & Simmons BL (2001) Preventive management of work stress: Current themes and future challenges. In Handbook of health psychology. Edited by Baum, A., Revenson, T., Singer, J. Lawrence Erlbaum Associates, Mahwah, N.J; London, p. 349–363.

Nordgren L (2003) Från patient till kund: intåget av marknadstänkande i sjukvår-den och förskjutningen av patientens position. (From patient to customer: arri-val of market ideas in health care and the shift in the patient's position. In Swed-ish). Academic dissertation. Lund Business Press, Lund.

Novak M & Chappell NL (1996): The impact of cognitively impaired patients and shift on nursing assistant stress. Int J Aging Hum Dev 43, 235-248.

Nyman JA (1988): Improving the quality of nursing home outcomes. Are adequacy- or incentive-oriented policies more effective? Med Care 26, 1158-1171.

Oberski IM, Carter DE, Gray M & Ross J (1999): The community gerontological nurse: themes from a needs analysis. J Adv Nurs 29, 454-462.

OECD (2005) Long-term care for older people. Organisation for Economic Co-operation and Development.

Owens DJ & Batchelor C (1996): Patient satisfaction and the elderly. Soc Sci Med42, 1483-1491.

Paasivaara L, Nikkonen M, Backman K & Isola A (2003): Practical nurses in elder care and changes in Finnish society - a biographical study. J Adv Nurs 41, 170-178.

Parker MG (2000): Sweden and the United States: is the challenge of an aging soci-ety leading to a convergence of policy? J Aging Soc Policy 12, 73-90.

Pearson A, Hocking S, Mott S & Riggs A (1993): Quality of care in nursing homes: from the resident's perspective. J Adv Nurs 18, 20-24.

Petterson IL, Hertting A, Hagberg L & Theorell T (2005): Are trends in work and health conditions interrelated? A study of Swedish hospital employees in the 1990s. J Occup Health Psychol 10, 110-120.

Private communication Nils Andrée (6 October 2006), Bengtsfors municipality. Private communication Pär Levander (25 Oktober 2006), Uddevalla municipality. Private communication Sara Molander (9 October 2006), Bengtsfors municipality. Proctor R, Burns A, Powell HS, Tarrier N, Faragher B, Richardson G, Davies L &

South B (1999): Behavioural management in nursing and residential homes: a randomised controlled trial. Lancet 354, 26-29.

Rafnsdottir GL, Gunnarsdottir HK & Tomasson K (2004): Work organization, well-being and health in geriatric care. Work 22, 49-55.

80

Rantz MJ, Zwygart-Stauffacher M, Popejoy L, Grando VT, Mehr DR, Hicks LL, Conn VS, Wipke-Tevis D, Porter R, Bostick J, Maas M & Scott J (1999): Nurs-ing home care quality: a multidimensional theoretical model integrating the views of consumers and providers. J Nurs Care Qual 14, 16-37; quiz 85-17.

Raudonis BM, Kyba FC & Kinsey TA (2002): Long-term care nurses' knowledge of end-of-life care. Geriatr Nurs 23, 296-301.

Redfern S, Hannan S, Norman I & Martin F (2002): Work satisfaction, stress, qual-ity of care and morale of older people in a nursing home. Health Soc Care Community 10, 512-517.

Ribbe MW, Ljunggren G, Steel K, Topinkova E, Hawes C, Ikegami N, Henrard JC & Jonnson PV (1997): Nursing homes in 10 nations: a comparison between countries and settings. Age Ageing 26 Suppl 2, 3-12.

Riksbanken (2006) http://www.riksbank.com/templates/stat.aspx?id=17212. Ac-cessed 1 October 2006.

Riksdagens revisorer (2002) Nationella mål i kommunernas äldreomsorg (National goals for municipal elderly care. In Swedish). Riksdagens revisorer, Stockholm.

Riksrevisionsverket (2003) Vård och omsorg för äldre: analys av problem och för-slag till statliga åtgärder (Elderly care: analysis of problems and suggestions for goverment measures. In Swedish). Riksrevisionsverket, Stockholm.

Robertson A, Gilloran A, McGlew T, McKee K, McKinley A & Wight D (1995): Nurses' job satisfaction and the quality of care received by patients in psycho-geriatric wards. International Journal of Geriatric Psychiatry 10, 575-584.

Robertson JF & Cummings CC (1996): Attracting nurses to long-term care. JGerontol Nurs 22, 24-32.

Rondeau KV & Wagar TH (2006): Nurse and resident satisfaction in magnet long-term care organizations: do high involvement approaches matter? J Nurs Manag14, 244-250.

Ross MM, Carswell A, Dalziel WB & Aminzadeh F (2001): Continuing education for staff in long-term care facilities: corporate philosophies and approaches. JContin Educ Nurs 32, 68-76; quiz 94-65.

Rowlands G, Sims J & Kerry S (2005): A lesson learnt: the importance of modelling in randomized controlled trials for complex interventions in primary care. FamPract 22, 132-139.

Runesson I & Eliasson-Lappalainen R (2000) Att sörja för äldre: en översikt (To take care of older people: an outline. In Swedish). Svenska kommunförbundet, Stockholm & Uppsala.

Ryan AA & Scullion HF (2000): Family and staff perceptions of the role of families in nursing homes. J Adv Nurs 32, 626-634.

Sainfort F, Karsh BT, Booske BC & Smith MJ (2001): Applying quality improve-ment principles to achieve healthy work organizations. Jt Comm J Qual Improv27, 469-483.

Sainfort F, Ramsay JD, Ferreira PL & Mezghani L (1994): A first step in total qual-ity management of nursing facility care: development of an empirical causal model of structure, process and outcome dimensions. Am J Med Qual 9, 74-86.

Schirm V, Albanese T & Garland TN (1999): Understanding nursing home quality of care: incorporating caregivers' perceptions through structure, process, and outcome. Qual Manag Health Care 8, 55-63.

Seth N, Deshmukh SG & Vrat P (2005): Service quality models: a review. Interna-tional Journal of Quality & Reliability Management 22, 913-949.

Sikorska-Simmons E (2006): Linking resident satisfaction to staff perceptions of the work environment in assisted living: a multilevel analysis. Gerontologist 46,590-598.

81

Simmons SF & Schnelle JF (1999): Strategies to measure nursing home residents' satisfaction and preferences related to incontinence and mobility care: implica-tions for evaluating intervention effects. Gerontologist 39, 345-355.

Sitzia J & Wood N (1997): Patient satisfaction: a review of issues and concepts. Soc Sci Med 45, 1829-1843.

Sjogren K, Fochsen G, Josephson M & Lagerstrom M (2005): Reasons for leaving nursing care and improvements needed for considering a return: a study among Swedish nursing personnel. Int J Nurs Stud 42, 751-758.

Skovdahl K, Kihlgren AL & Kihlgren M (2003): Different attitudes when handling aggressive behaviour in dementia-narratives from two caregiver groups. Aging Ment Health 7, 277-286.

Social Services Act (2001) Socialtjänstlag (2001:453). Socialdepartementet (2002) Hur står det till med hälso- och sjukvården och

äldreomsorgen: vad säger brukarundersökningar och andra mått på kvaliteten (How is the quality of health care and elderly care: what do client evaluations and other measures of quality say. In Swedish). Socialdepartementet Regerings-kansliet: Fritzes offentliga publikationer, Stockholm.

Socialstyrelsen (1998) Socialstyrelsens allmänna råd om kvalitetssystem inom om-sorgerna om äldre och funktionshindrade. (The National Board of Health and Welfare’s recommendations for quality system in elderly and disabled care (SOSFS 1998:8) (S) (Kvalitetssystem socialtjänstlagen).

Socialstyrelsen (2001) Verksamhetsuppföljning inom vård och omsorg till äldre och funktionshindrade i kommunerna: redovisning av läget 2000-2001 (Evaluation of care for the elderly and disabled in municipalities: reporting the situation in 2000-2001. In Swedish). Socialstyrelsen, Stockholm.

Socialstyrelsen (2002) Äldre: vård och omsorg år 2001 kommunala insatser enligt socialtjänstlagen samt hälso- och sjukvårdslagen (Older people: care year 2001 municipality efforts according to the Social Services Act and the Health and Medical Services Act. In Swedish). Socialstyrelsen, Stockholm.

Socialstyrelsen (2003) Kvalitetsuppföljning inom vård och omsorg om äldre: kart-läggning och analys av aktuell litteratur (Quality evaluation in elderly care: re-view and analysis of current literature. In Swedish). Socialstyrelsen, Stockholm.

Socialstyrelsen (2004) Investera nu! handlingsplan för kompetensförsörjning inom vård och omsorg (Invest now! A plan of action for competencedevelopment in nursing. In Swedish). Socialstyrelsen, Stockholm.

Socialstyrelsen (2005a) Borta bra men hemma bäst? en nationell granskning av hemsjukvården i Sverige (There’s no place like home: a national investigation of home health in Sweden. In Swedish). Socialstyrelsen, Stockholm.

Socialstyrelsen (2005b) Social tillsyn 2004. Resultat av länsstyrelsernas tillsyn (So-cial surveillance 2004. Results of county administrative boards’ surveillance. In Swedish). Socialstyrelsen, Stockholm.

Socialstyrelsen (2005c) Socialstyrelsens föreskrifter om ledningssystem för kvalitet och patientsäkerhet i hälso- och sjukvården (National Board for Health and So-cial Welfare’s regulations for system of quality and patient safety in health care (SOSFS 2005:12).

Socialstyrelsen (2005d) Tid för vård och omsorg: hur använder personalen inom vården och omsorgen om äldre och funktionshindrade personer sin tid? (Time for nursing care: how staff who care for the elderly and disabled use their work time? In Swedish). Socialstyrelsen, Stockholm.

Socialstyrelsen (2005e) Vård och omsorg om äldre: lägesrapport 2004 (Current developments in care of the elderly in Sweden. In Swedish with English sum-mary). Socialstyrelsen, Stockholm.

82

Socialstyrelsen (2005f) Äldre: vård och omsorg år 2004 kommunala insatser enligt socialtjänstlagen samt hälso- och sjukvårdslagen (Older people: care year 2004 municipality efforts according to the Social Services Act and the Health and Medical Services Act. In Swedish). Socialstyrelsen, Stockholm.

Socialstyrelsen (2005g) Äldre: vård och omsorg år 2004 kommunala insatser enligt socialtjänstlagen samt hälso- och sjukvårdslagen (Statistics – Social Welfare. Care and services to elderly persons 2004. Official statistics of Sweden. In Swe-dish with English summary). Socialstyrelsen, Stockholm.

Socialstyrelsen (2006a) Brukare och anhöriga om kvalitet i äldreomsorgen. Kvalitetsbarometern 1998, 1999, 2002, 2004 och 2005 (Clients’ and relatives’ opinions of quality in elderly care. The quality barometer 1998, 1999, 2002, 2004 and 2005. In Swedish). Socialstyrelsen, Stockholm.

Socialstyrelsen (2006b) Omsorg människor emellan - En översikt av omsorgsgivan-de i den svenska befolkningen (Nursing care between people - an outline of ca-regiving in the Swedish population. In Swedish). Socialstyrelsen, Stockholm.

Socialstyrelsen (2006c) Social rapport (Social report. In Swedish with English summary). Socialstyrelsen, Stockholm.

Socialstyrelsen (2006d) Vård- och omsorgsassistenters kompetens – en litteratur-genomgång (Nursing assistants competence - a literature review. In Swedish). Socialstyrelsen, Stockholm.

Socialstyrelsen (2006e) Vård och omsorg om äldre: lägesrapport 2005 (Current developments in care of the elderly in Sweden. In Swedish). Socialstyrelsen, Stockholm.

Stolee P, Esbaugh J, Aylward S, Cathers T, Harvey DP, Hillier LM, Keat N & Feightner JW (2005): Factors associated with the effectiveness of continuing education in long-term care. Gerontologist 45, 399-409.

Sung HC, Chang SM & Tsai CS (2005): Working in long-term care settings for older people with dementia: nurses' aides. J Clin Nurs 14, 587-593.

Svensson M, Edebalk PG & Persson U (1996): Group living for elderly patients with dementia-a cost analysis. Health Policy 38, 83-100.

Sveriges Kommuner och Landsting (2005a) Aktuellt om äldreomsorgen (Current topics in elderly care. In Swedish). Sveriges Kommuner och Landsting, Stock-holm.

Sveriges Kommuner och Landsting (2005b) Ett gemensamt ansvar: strategier för utveckling av framtidens äldrevård (The common responsability: strategies for developing elderly care for the future. In Swedish). Sveriges kommuner och landsting, Stockholm.

Theorell T (2002) Kontroll över den egna situationen–en förutsättning för hantering av negativ stress. (Control over one's own situation - a condition for stress man-agement. In Swedish). In In: Stress. Molekylerna, individen, organisationen, samhället. Edited by Ekman, R. & Arnetz, B. Liber, Stockholm.

Theorell T (2003) Är ökat inflytande på arbetsplatsen bra för folkhälsan. Kun-skapssammanställning (Is increased work influence beneficial to public health? In Swedish with English summary). Statens folkhälsoinstitut, Stockholm.

Thorslund M, Bergmark Å & Parker MG (1997): Difficult decisions on care and services for elderly people: the dilemma of setting priorities in the welfare state. Scand J Soc Welfare 6, 197-206.

Uddevalla kommun (2001) Årsredovisning 2001 (Municipality annual report 2001. In Swedish).

Uddevalla kommun (2003): Årsredovisning 2003 (Municipality annual report 2003. In Swedish).

83

Uddevalla kommun (2004): Årsredovisning 2004 (Municipality annual report 2004. In Swedish).

Vagg PR & Spielberger CD (1998): Occupational stress: measuring job pressure and organizational support in the workplace. J Occup Health Psychol 3, 294-305.

Walcott-McQuigg JA & Ervin NE (1992): Stressors in the workplace: community health nurses. Public Health Nurs 9, 65-71.

Wall TD, Bolden RI, Borrill CS, Carter AJ, Golya DA, Hardy GE, Haynes CE, Rick JE, Shapiro DA & West MA (1997): Minor psychiatric disorder in NHS trust staff: occupational and gender differences. Br J Psychiatry 171, 519-523.

van den Berg TI, Landeweerd JA, Tummers GE & van Merode GG (2006): A com-parative study of organisational characteristics, work characteristics and nurses' psychological work reactions in a hospital and nursing home setting. Int J Nurs Stud 43, 491-505.

Weech-Maldonado R, Shea D & Mor V (2006): The relationship between quality of care and costs in nursing homes. Am J Med Qual 21, 40-48.

Wellwood I, Dennis M & Warlow C (1995): Patients' and carers' satisfaction with acute stroke management. Age Ageing 24, 519-524.

Weman K, Kihlgren M & Fagerberg I (2004): Older people living in nursing homes or other community care facilities: Registered Nurses' views of their working situation and co-operation with family members. J Clin Nurs 13, 617-626.

Westlund P & Edvardsson B (1998) Tjänsteutveckling och kvalitet i äldreomsorg (Service development and quality in elderly care. In Swedish). Studentlitteratur, Lund.

Wiener JM (2003): An assessment of strategies for improving quality of care in nursing homes. Gerontologist 43 Spec No 2, 19-27.

Wigren A, Engström A & Larsson A (2002) Kommuner i global omvandling. Om-världsanalys för Bengtsfors kommun (Municipalities in global change. A situa-tion analysis for the municipality of Bengtsfors. In Swedish). Ingredia AB.

Wilde B, Larsson G, Larsson M & Starrin B (1995): Quality of care from the elderly person's perspective: subjective importance and perceived reality. Aging (Mi-lano) 7, 140-149.

Wills W, Trieman N & Leff J (1998): The Taps Project 40: quality of care provi-sions for the elderly mentally ill--traditional vs alternative facilities. Int J Geri-atr Psychiatry 13, 225-234.

Work Environment Act (1977) Arbetsmiljölag (1977:1160). Wressle E, Eriksson L, Fahlander A, Rasmusson IM, Tedemalm U & Tangmark K

(2006): Patient perspective on quality of geriatric care and rehabilitation-development and psychometric testing of a questionnaire. Scand J Caring Sci20, 135-142.

Ygge BM & Arnetz JE (2001): Quality of paediatric care: application and validation of an instrument for measuring parent satisfaction with hospital care. Int J Qual Health Care 13, 33-43.

Yin JC & Yang KP (2002): Nursing turnover in Taiwan: a meta-analysis of related factors. Int J Nurs Stud 39, 573-581.

Acta Universitatis UpsaliensisDigital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Medicine 210

Editor: The Dean of the Faculty of Medicine

A doctoral dissertation from the Faculty of Medicine, UppsalaUniversity, is usually a summary of a number of papers. A fewcopies of the complete dissertation are kept at major Swedishresearch libraries, while the summary alone is distributedinternationally through the series Digital ComprehensiveSummaries of Uppsala Dissertations from the Faculty ofMedicine. (Prior to January, 2005, the series was publishedunder the title “Comprehensive Summaries of UppsalaDissertations from the Faculty of Medicine”.)

Distribution: publications.uu.seurn:nbn:se:uu:diva-7383

ACTAUNIVERSITATISUPSALIENSISUPPSALA2006

nursing staff competence, psychosocial work environment ...169399/fulltext01.pdfnursing staff...

Documents