Care recipients’ and family members’ perceptions of quality of older people care: a comparison of home-based care and nursing homes

Authors

  • Henna Hasson,

    1. Authors:Henna Hasson, PhD, Postdoctoral Researcher, Department of Business Administration, Lund University School of Economics and Management, Lund; The Vårdal Institute, The Swedish Institute for Health Sciences, Lund; Karolinska Institutet, The Department of Learning, Informatics, Management and Ethics (LIME), Medical management Centre (MMC), Stockholm, Sweden; Judith E Arnetz, PhD, MPH, PT, Associate Professor, Department of Family Medicine, Division of Occupational and Environmental Health, Wayne State University, Detroit, Michigan USA and Department of Public Health and Caring Sciences, Section for Social Medicine, Uppsala University, Uppsala, Sweden
    Search for more papers by this author
  • Judith E Arnetz

    1. Authors:Henna Hasson, PhD, Postdoctoral Researcher, Department of Business Administration, Lund University School of Economics and Management, Lund; The Vårdal Institute, The Swedish Institute for Health Sciences, Lund; Karolinska Institutet, The Department of Learning, Informatics, Management and Ethics (LIME), Medical management Centre (MMC), Stockholm, Sweden; Judith E Arnetz, PhD, MPH, PT, Associate Professor, Department of Family Medicine, Division of Occupational and Environmental Health, Wayne State University, Detroit, Michigan USA and Department of Public Health and Caring Sciences, Section for Social Medicine, Uppsala University, Uppsala, Sweden
    Search for more papers by this author

Henna Hasson, Postdoctoral Researcher, Department of Business Administration, Lund University School of Economics and Management, Box 7080, SE-220 07 Lund, Sweden. Telephone: +46 73 6340730.
E-mail:henna.hasson@fek.lu.se

Abstract

Aims.  To compare care recipients’ and their relatives’ perceptions of quality of care in nursing homes and home-based care.

Background.  Older people care is increasingly being provided in community care facilities and private homes instead of hospitals. A few studies have compared care recipients’ and relatives’ perceptions of care quality in nursing homes as opposed to home-based care.

Design.  Cross-sectional surveys.

Methods.  Care recipients’ and relatives’ perceptions of quality of care were measured by questionnaire in 2003 in two older people care organisations. Quality measures were compared between care settings. Multiple regression was used to determine the predictors of overall quality ratings for each group, and possible interactions between quality measures and care settings were tested.

Results.  Care recipients rated their opportunities for activities significantly lower in home care than in nursing homes. Relatives of care recipients in home-based care rated several aspects of care quality significantly lower than relatives of nursing home residents. No significant interaction effects regarding predictors of overall quality ratings between the care settings were found. Staff behaviour was the strongest predictor of care recipients’ overall quality rating, and staff professional skills were the strongest predictor of relatives’ overall quality rating.

Conclusions.  Compared with nursing homes, home-based older people care seems to be in greater need of development regarding staff competence, staff interaction with relatives and activities offered to older people. In both settings, nursing staff behaviour influences these stakeholders’ satisfaction with care.

Relevance to clinical practice.  These results point to a need to improve services in both care settings, but especially in home-based care. It is suggested that care recipients’ preferences for social and physical activities be investigated on a regular basis in both care settings. Efforts should also be made to improve communication and interaction between family members and older people care staff, especially in home-based care.

Introduction

In recent years, several European countries (Kerkstra & Hutten 1996, Oberski et al. 1999, Meijer et al. 2000, Paasivaara et al. 2003), as well as countries in the Organisation for Economic Co-operation and Development (OECD 2005), have made efforts to build up community-based services for older people to replace care in hospitals. Communities have been given the responsibility for management of care and services for older people (Thorslund et al. 1997, Higgs et al. 1998). In Sweden, the aim has been to enable older people to live in their own homes as long as possible (Thorslund et al. 1997). The Swedish national goals for care of older people emphasise that older people should maintain their independence, have opportunities for an active life with influence over their daily activities and have access to good health and nursing care (Government bill 1997/98:113). However, a few studies have systematically compared how different stakeholder groups perceive the quality of care offered in different types of community-based older people care settings. Specifically, little is known as to how care recipients and their relatives perceive the quality of services offered in home care as opposed to nursing homes. Such comparisons could offer concrete information regarding quality aspects in need of improvement in each respective setting.

Background

Concerns regarding poor quality of care in older people services have been reported in several countries, such as the UK (Fahey et al. 2003), Sweden (The National Board of Health and Welfare 2005), the USA (Castle 2003) and Australia (Chenoweth & Kilstoff 2002). Reports of poor quality of care have included improper medication, late or incorrect diagnoses (The National Board of Health and Welfare 2006c), inadequate treatment of chronic pain, pressure ulcers and inappropriate use of chemical or physical restraints (OECD 2005). Studies investigating older people’s (Wellwood et al. 1995, McCartan-Quinn et al. 1996, Wressle et al. 2006) and their relatives’ (Maas et al. 1991, Lubart et al. 2004) perceptions of older people care have in general found high satisfaction with care. However, older care recipients (Wressle et al. 2006) and their relatives (Curry & Stark 2000) have been dissatisfied with specific aspects of care, such as opportunities to influence care. Older people have also expressed dissatisfaction with information received from staff members (Wellwood et al. 1995, McCartan-Quinn et al. 1996) and with social activities available (McCartan-Quinn et al. 1996). Relatives of nursing home residents have reported communication difficulties with nursing staff (Hertzberg et al. 2001). These studies have been conducted either in home-based care or in nursing homes. To our knowledge, no prior studies have compared the quality of community nursing home care with home-based care.

Quality of care has in prior research been associated with staff competence (McAiney 1998), stress (Cohen-Mansfield & Rosenthal 1989) and work satisfaction (Redfern et al. 2002, Chou et al. 2003), i.e. lower quality of care has been related to lower competence, less satisfaction and higher stress levels among nursing staff. High stress levels have been reported among staff in nursing homes (Proctor et al. 1999, Morgan et al. 2002, van den Berg et al. 2006, Ejaz et al. 2008) as well as in home care (Laamanen et al. 1999, Denton et al. 2002). Practical nurses and nurses’ aides comprise the largest professional groups in older people care, and both these groups are less skilled and less educated than other health care providers (D’Eramo et al. 2001, Fahlstrom & Kamwendo 2003, Pennington et al. 2003). In Sweden, the formal requirement for working as a licensed practical nurse or nurse’s aide is a three-year upper secondary school (high school) programme with a focus on nursing. This can be compared to the training for registered nurses, who complete a three-year university education, which may be complemented by additional courses towards specialisation. However, only 56% of working practical nurses and nurses’ aides have completed the formal upper secondary education (The National Board of Health and Welfare 2004). In addition, staff in home-based older people care have reported to have less formal education than staff in nursing homes (The National Board of Health and Welfare 2004). The fact that care recipients today have more complex medical and psychiatric needs than previously has placed new demands on nursing staff competence (Robertson & Cummings 1996, Proctor et al. 1999, Sung et al. 2005). A recent study (Hasson & Arnetz 2008b) found that staff in home-based care had significantly less sufficient self-rated knowledge regarding several work-related subject areas compared to staff in nursing homes. It was, therefore, of interest to investigate whether quality of care would differ between these care settings.

Aims

This study aimed to:

  • 1 Compare perceptions of quality of care in home-based care and nursing homes, respectively, among care recipients and their family members in two municipal older people care organisations;
  • 2 Examine determinants of overall quality ratings and possible interactions between the determinants and the care setting (home care vs. nursing homes) among care recipients and their family members, regardless of municipality.

Methods

Study design

The study was a cross-sectional survey using a previously validated questionnaire to measure care recipients’ and their relatives’ perceptions of quality of care in two older people care organisations.

Settings

This study was conducted in two municipal older people care organisations in Sweden. Swedish municipalities are responsible for the social welfare and health care of older people, both in private homes, service apartments and nursing homes. Social welfare includes services such as food distribution, administration of security alarms, organisation of social activities, assistance with eating, personal hygiene and general house cleaning. Service apartments function as private homes and services and care required are offered by home care personnel. Nursing homes have 24-hour nursing surveillance, i.e. staff members are always available for older care recipients. In both care settings, practical nurses and nurses’ aides comprise the largest professional groups. Registered nurses are also employed in both home care and in nursing homes, but in much smaller numbers.

In the municipalities involved in this study, approximately half the older care recipients receive services from home care staff while half live in nursing homes. Municipality 1 has approximately 11,000 inhabitants, of which 3·5% (n = 390) are care recipients in the older people care organisation. Municipality 2 has 49,500 inhabitants, 3% (n = 1530) of whom receive older people care.

Participants

All the older people receiving home help services, home care or living in nursing homes in these two municipalities were considered eligible for the study. However, older people not capable of expressing their perceptions because of severe dementia did not receive the questionnaire. The questionnaires were distributed by the nursing staff who were instructed to discuss the capability of each care recipient to reply to the questionnaire with their supervisors. Thus, the only reason for not receiving the questionnaire was severe dementia.

All the individuals who were registered as a primary family member to care recipients were included in the study and are referred to as care recipients’ relatives. One relative per care recipient was registered as primary family member and therefore one relative per care recipient was invited to participate in the study.

The questionnaires

Care recipients’ and relatives’ ratings of the quality of care were measured using previously validated Pyramid questionnaires (Arnetz & Arnetz 1996, Ygge & Arnetz 2001, Verho & Arnetz 2003, Hasson & Arnetz 2008a). Both instruments were based on questionnaires that were developed for use in hospitals among older patients and their relatives, respectively, and adapted to municipal older people care with the help of focus groups (Verho & Arnetz 2003, Hasson & Arnetz 2008a). In short, this process included focus group discussions with patients in municipal care of older people and their relatives to investigate the relevance of the questions in the hospital questionnaire for these respondent groups. Many of the items were confirmed to be relevant even in municipal care of older people. However, questions regarding hospital treatment and information about treatment were excluded, while questions regarding physical and social activity were added to the community questionnaires (Verho & Arnetz 2003).

The care recipients’ questionnaire consisted of 35 questions. The questions were formulated so that they would be relevant for care recipients in nursing homes as well as in home-based care. The questionnaire consisted of four quality of care indices, information, staff behaviour, care and activity. An overview of the indices with their component items and Cronbach’s alpha values for internal homogeneity is presented in Table 1.

Table 1.   Measurement indices (Cronbach’s alpha) and component items in care recipient and family members questionnaires
Scale (Cronbach’s alpha)Component items (n)
Care recipients’ questionnaire
 Information (0·69)Information received concerning medication, staff routines, which nurse is responsible for care (3)
 Staff behaviour (0·71)Staff has sufficient time for me, staff introduce themselves, staff offer emotional support, staff are open for my needs/requests (4)
 Care (0·80)I am well-cared for regarding my personal hygiene, meals and physical transfers; confidence in staff competence, confidence in staff skill (3)
 Activity (0·77)Access to physical training/physiotherapy, opportunities to participate in social activities, opportunities to participate in excursions (3)
Relatives’ questionnaire
 Information (0·85)Information received concerning care recipient’s medication, rehabilitation, staff routines, which nurse is responsible for care (4)
 Staff professional skills (0·83)Confidence in staff competence, confidence in staff skills, staff spend adequate time with care recipient (3)
 Care (0·70)Care recipient is well cared for regarding personal hygiene, meals and is treated with respect (3)
 Activity (0·71)Care recipient is well cared for regarding physical transfers, physical training/physiotherapy, care recipient opportunities to participate in social activities (3)
 Contact (0·82)Opportunities to ask staff about care recipient’s situation, staff take an initiative to discuss care recipient’s situation, relatives contact with staff as desired (3)
 Social support (0·88)Staff introduce themselves to relative, staff give emotional support to relative, staff are open for relatives’ needs/requests, staff treat relative with respect (4)
 Relative participation (0·69)Staff make use of relatives’ knowledge about care recipient, relative participates in the care of care recipient, opportunities to discuss the goals of care recipient’s care with the staff (3)

The relatives’ questionnaire was developed in a similar way to the care recipient questionnaire and has been described previously (Verho & Arnetz 2003). It consisted of 50 items encompassing seven quality of care indices: information, staff professional skills, care, activity, contact, social support and relative participation (Table 1).

Response alternatives in both questionnaires were typically a four-point Likert-type scale: ‘Yes, to a great degree’, ‘Yes, somewhat’, ‘No, not especially’ or ‘No, not at all’. When necessary, for instance regarding medication, a ‘not relevant’ response alternative was added. Values for the quality indices were calculated for each respondent by totalling the scores on the component items and converting that sum to a percentage of the maximum possible score. For all indices, higher scores indicate higher quality ratings.

In addition, care recipients and relatives were each asked to give an overall rating to the quality of care on a visual analogue scale from 1 (very negative) – 10 (very positive): ‘How would you rate the overall care/service you receive?’ and ‘How would you rate the overall care/service your relative receives?’ Mean values for the overall ratings were also converted to percentages and treated as continuous variables.

Study procedure

Care recipients were permitted to receive assistance reading the questions and/or writing their responses from a friend or relative. This was stated on the first page of the questionnaire. In addition, some assistants, for instance from pensioner organisations, were recruited to help the care recipients with the questionnaire. These persons were directed to read the questions and response alternatives and to complete the care recipient’s response, but not to interpret or answer the questions themselves. Staff members were instructed not to help any care recipients with the questionnaire. Those care recipients who requested help with reading and/or filling out the questionnaires preferred their relatives or friends to help them. Approximately 10% of the older care recipients requested help from the recruited assistants. In municipality 1, two hundred and twenty-one care recipients received a questionnaire and in municipality 2946 questionnaires were distributed.

Relatives received the questionnaire by mail. The researchers received a list of relatives’ addresses in municipality 1 and posted the questionnaires. In municipality 2, questionnaires were mailed to relatives by local administrators. In both municipalities, university envelopes were used for mailing the relatives’ questionnaires. A total of 350 relative questionnaires were posted in municipality 1 and 1214 questionnaires were mailed in municipality 2.

All the questionnaires were accompanied by an introductory letter and postage-paid return envelope. The envelopes were sent directly to data registration, and staff or supervisors at care units did not see the completed questionnaires. Responses were anonymous, and no follow-up letters were sent.

Statistical methods

The SPSS statistical software package (version 14.0 for Windows) was used for all statistical analyses, with statistical significance set at 0·05 (two-tailed) for all the analyses. Chi-square statistics were used to compare discrete background variables of the respondents between care settings as well as to compare respondents with the total population in each municipality, respectively.

All the continuous variables (measurement indices and overall quality grades) were assessed for normality using Kolmogorov–Smirnov tests. Not all the variables were normally distributed, therefore, non-parametric tests were used to examine possible differences in ratings between the care settings. Thus, Mann–Whitney rank sum test was used to examine possible differences in care recipients’ and relatives’ ratings of the quality indices and the overall quality grade between home-based care and nursing homes.

Stepwise multiple regression analysis was used to determine the predictors of the overall quality grade among care recipients and relatives, respectively. Because the aim was to predict overall quality regardless of municipality, a data set comprising both municipalities was used. In addition, we wanted to examine any statistically different effects of specific quality indices in nursing homes and home care. We therefore combined the data from both care settings, to test for possible interaction effects of specific quality indices. All the continuous independent variables were plotted separately against the dependent variable to check for linear relationships. All the relationships were found to be linear. In the care recipient data set, the four measurement indices information, staff behaviour, care and activity were entered into the model as independent variables, along with respondents’ age, gender, years receiving care in the organisation and municipality. Dummy variables were created for each of the categorical control variables (age, gender, years of care, municipality). The probability for the entry of an independent variable was set at <0·05 and for removal >0·1. In the next step, we forced the significant main effects that resulted from the first step into the regression model using the enter command. Into this fixed model of significant main effects, interaction variables for each independent quality of care index (index × care setting) were entered in a stepwise regression. All the four interaction terms were included in the final regression model, to examine possible different effects of an index on the overall quality grade in each care setting, respectively. Corresponding regression analyses were conducted on the relative data set, using the overall quality grade as the dependent variable. The seven indices information, staff professional skills, care, activity, contact, social support and relative participation were entered into the model as independent variables, along with dummy variables for respondents’ age, gender, years of contact with the organisation, relation to care recipient and municipality. Following the procedure outlined for the care recipient data, the significant main effects resulting from the first step were forced into the regression model using the enter command. Interaction variables were created for each quality index (index × care setting) and included in the final regression analysis. All the data sets were examined for multicollinearity.

Power calculation

The power to detect a 10% point difference in mean values of the overall quality grade in home care vs. nursing homes was estimated at 83% (alpha = 0·05, two-tailed) for the care recipient study with a group size of 76 respondents [standard deviation (SD) 21%] and at 77% (alpha = 0·05, two-tailed) for the relatives’ study with a group size of 72 respondents (SD 22%) in each setting in municipality 1.

Ethical approval

Ethical approval for this project was given by the research ethics committee of the medical faculty of Uppsala University, as well as its local affiliate at Gothenburg University, dossier number 00–206.

Results

Characteristics of respondents

The response rates, number of respondents and characteristics of respondents in the care recipient survey in home-based care and nursing homes are presented in Table 2. A total of 541 care recipients responded to the survey for an overall response rate of 46%. Response rates were higher, over 60%, in municipality 1 compared with municipality 2, where response rates were around 40%. There were significantly less care recipients in nursing homes in the respondent population in municipality 2 (p < 0·01), compared to the total care recipient populations in that municipality. Otherwise, the respondents did not differ significantly from the total population of care recipients regarding age, gender and care setting in either of the municipalities. No information was available regarding the functional status of care recipients’ in either home-based care or nursing homes. The respondents to the care recipient surveys did not differ significantly between the care settings regarding age, gender or years receiving services or care. The majority of respondents were women who had received older people care for less than five years in both care settings.

Table 2.   Number of respondents and background characteristics of care recipient respondents. p-values are based on Chi-square statistics
 MunicipalityHome careNursing homesp-value within municipalityHome careNursing homes
n (%)n (%)p-value between municipalitiesp-value between municipalities
  1. ns, non-significant.

  2. Two responses in municipality 2 lacked information about care setting (home care or nursing home) and were excluded from this analyses.

n (response rate)176/119 (64)89/102 (87)   
2266/666 (40)110/280 (39)
Age (years)1  ns0·015ns
2  ns
  ≤85131 (41)35 (41)   
2151 (57)57 (53)
  >85144 (59)51 (59)   
2113 (43)50 (47)
Years in older people care organisation1  nsnsns
2  ns
 <5155 (72)49 (65)   
2189 (72)66 (63)
 ≥5121 (28)26 (34)   
272 (28)39 (37)
Sex1  nsns0·006
2  ns
 Male124 (33)36 (40)   
275 (29)24 (22)
 Female149 (67)53 (60)   
2183 (71)83 (78)

The response rates, number of respondents and characteristics of respondents in the relatives’ surveys in home-based care and nursing homes are presented in Table 3. A total of 780 relatives responded to the survey for an overall response rate of 50%. Response rates were approximately 60% in both care settings in both municipalities, expect for home-based care in municipality 2 where the response rate was 40%. There were no significant differences in gender distribution of the respondents compared to the municipality’s register of primary family members in municipality 1. No other data regarding primary family members were available. Approximately 60% of the respondents to the relatives’ questionnaire were women in both care settings. A significantly greater proportion of relatives’ of care recipients in home-based care had had contact with the older people care organisation for a shorter time and had contact with the care recipient daily, compared to relatives of care recipients in nursing homes. In addition, a larger proportion of relatives of care recipients in home-based care in municipality 2 were younger or equal to 55 years, compared to relatives of care recipients in nursing homes.

Table 3.   Number of respondents and background characteristics of care recipient relative respondents. p-values are based on Chi-square statistics
 MunicipalityHome careNursing homesp-value within municipalityHome careNursing homes
n (%)n (%)p-value between municipalitiesp-value between municipalities
  1. ns, non-significant.

n (response rate)172/123 (59)126/227 (56)   
2294/739 (40)288/475 (61)   
Age (years)1  nsnsns
2  <0·001
 ≤55128 (41)44 (36)   
2130 (47)82 (29) 
 ≥56141 (59)77 (64)   
2145 (53)202 (71) 
Years of contact with older people care organisation1  0·005nsns
2  0·015
 <5145 (68)57 (47)   
2183 (68)163 (58)
 ≥5121 (32)65 (53)   
286 (32)118 (42)
Sex1  nsnsns
2  ns
 Male128 (41)47 (39)   
299 (35)114 (41)
 Female140 (59)74 (61)   
2182 (65)167 (59)
Relation to care recipient1  nsnsns
2  ns
 Spouse13 (4)10 (8)   
226 (9)44 (15)
 Son/daughter153 (76)77 (62)   
2205 (71)179 (63)
 Other family relative16 (9)11 (9)   
220 (7)18 (6)
 Other relation18 (11)26 (21)   
239 (13)44 (16)
Frequency of contact with care recipient1  <0·001ns0·007
2  <0·001
 Daily139 (56)26 (21)   
2149 (52)80 (28)
 Once or twice a week129 (42)61 (49)   
2117 (41)160 (56)
 Once or twice a month11 (1)32 (25)   
219 (7)39 (14)
 A few times a year/never11 (1)6 (5)   
20 (0)6 (2)

Quality of care ratings between home-based care and nursing homes

Care recipients’ ratings of the quality of care indices in home-based care and in nursing homes are presented in Table 4. Care recipients in home-based care rated the quality of care index activity significantly lower compared to care recipients in nursing homes in both municipalities. In addition, care recipients in home-based care rated the index staff behaviour significantly higher compared to residents in nursing homes in municipality 2. The indices staff behaviour and care received the highest ratings from care recipients in both care settings in both municipalities. Ratings of care recipients who had received assistance in responding to the questionnaire did not differ significantly from those who had responded on their own in either setting.

Table 4.   Care recipients’ and relatives’ ratings of the quality indices, means (SD, Standard deviation) in home-based care and in nursing homes. p-values are based on Mann–Whitney rank sum tests
 MunicipalityHome careNursing homesp-value within municipalityHome careNursing homes
Mean (SD)Mean (SD)p-value between municipalitiesp-value between municipalities
  1. ns, non-significant; SD, standard deviation.

Care recipients’ indices
 Information170·6 (29·8)61·1 (26·1)nsnsns
267·1 (23·2)68·0 (25·8)ns
 Staff behaviour184·1 (20·6)82·1 (18·8)nsnsns
281·1 (20·2)77·3 (19·4)0·037
 Care185·0 (21·2)91·5 (14·6)ns0·028<0·001
279·8 (21·5)82·7 (16·3)ns
 Activity157·2 (31·3)78·2 (21·7)0·017ns0·002
246·9 (30·1)56·7 (31·1)0·044
 Overall quality grade180·6 (23·1)80·1 (18·8)nsnsns
276·0 (21·9)74·8 (21·5)ns
Relatives’ indices
 Information144·0 (29·3)54·6 (30·5)nsnsns
252·3 (29·3)61·8 (27·3)0·003
 Staff professional skills162·5 (21·7)70·1 (21·0)0·008nsns
262·2 (21·3)71·4 (21·3)<0·001
 Care182·2 (16·9)87·0 (14·6)nsnsns
276·3 (17·6)86·7 (16·4)<0·001
 Activity133·9 (23·7)56·1 (22·1)<0·001nsns
246·2 (25·2)58·0 (24·7)<0·001
 Contact162·8 (23·5)76·7 (21·5)<0·001nsns
261·8 (25·3)78·7 (21·0)<0·001
 Social support168·6 (24·5)75·5 (24·1)0·044nsns
271·2 (26·2)77·7 (21·5)0·010
 Relative participation151·1 (27·9)57·9 (26·6)nsnsns
259·7 (27·0)60·0 (26·0)ns
 Overall quality grade167·9 (23·3)75·7 (20·7)0·016nsns
266·3 (22·0)73·1 (22·2)<0·001

Relatives of home care recipients rated five of the quality of care indices, as well as the overall quality grade, significantly lower than relatives of nursing home residents in both municipalities (Table 4). In addition, relatives of home care recipients in municipality 2 rated the indices information and care significantly lower compared to relatives of nursing home residents. The index care received the highest ratings from relatives to residents in both care settings.

Predictors of the overall quality grade

The final regression analyses revealed that the indices staff behaviour and activity were significant predictors of care recipients’ overall quality ratings, explaining 51% of the variance in overall quality ratings (Table 5). Staff professional skills, contact and care were significant predictors of relatives’ overall quality ratings. These variables explained 60% of the variance (Table 5). There were no significant interaction effects between any of the quality indices and care setting, indicating no differences in the effect of the quality indices on the overall quality grade between home care and nursing homes. For all the regression analyses, collinearity statistics were within acceptable range, with VIF <3·0 (Neter et al. 1996).

Table 5.   Regression analysis: predictors of care recipients’ and relatives’ overall quality ratings
VariableUnstandardised coefficientsStandard errorSignificance level
Care recipients
 Constant20·6444·524 
 Staff behaviour0·5940·065<0·001
 Activity0·1490·050<0·01
 R20·514  
Relatives
 Constant−3·2083·449 
 Staff professional skills0·3260·048<0·001
 Contact0·2360·043<0·001
 Care0·4120·058<0·001
 R20·601  

Discussion

Care recipients’ ratings of opportunities for activities were significantly lower in home care than in nursing homes. Relatives of home care recipients rated several aspects of care quality significantly lower compared to relatives of nursing home residents. Both respondent groups in both care settings gave highest ratings to the basic nursing elements of care and lower ratings to information received from staff and for the activities offered to care recipients. Adjusted regression analyses indicated that the effects of the quality indices on the overall quality ratings did not differ between the care settings. Staff behaviour and activity were significant predictors of care recipients’ overall quality rating and staff professional skills, contact and care were significant predictors of relatives’ overall quality ratings. Few significant differences were found between the municipalities for any of the care recipients’ or relatives’ outcome variables, indicating similar ratings among these respondent groups in both care settings in both organisations.

Relatives of home care recipients rated the quality of care indices staff professional skills, contact, social support and activity significantly lower in both municipalities compared with relatives of nursing home residents. The index staff professional skills included questions regarding relatives’ confidence in staff competence and skills. Similar questions were included in the care recipients’ index care, which in one of the municipalities was rated significantly lower by home care recipients when compared to nursing home residents. Previous studies of staff competence have shown that both the formal level of education (The National Board of Health and Welfare 2004) and the staff’s own perceptions of their competence (Hasson & Arnetz 2008b) were lower for staff in home care when compared to staff in nursing homes. Thus, it is possible that care recipient relatives and care recipients were sensitive to differences in staff competence between the care settings. Several prior education programmes for older people care staff have been designed for nursing home staff (Beck et al. 1999, Ross et al. 2001, Aylward et al. 2003); however, perhaps there is a greater need for such programmes among staff in home-based care. Further research is needed to identify effective ways of improving competence among home care nursing staff. The indices contact and social support reflect relatives’ interactions with nursing staff. It is possible that relatives of home care recipients have a harder time establishing contact with care staff. The relatives of older people receiving home-based care do not necessarily meet care staff when visiting their older family member, as is the case in nursing homes. Thus, relatives’ opportunities to interact with care staff differ greatly between the care settings. In addition, increased demands for efficiency among home care staff have been reported to have decreased their opportunities to offer psychosocial support for clients (Arts et al. 1998). Most of the prior studies investigating relatives’ contact with care staff were conducted in nursing homes. These studies have shown that, despite recognition that family members might be important in patient care (Astedt-Kurki et al. 2001, Weman et al. 2004), staff still give low priority to working with family members (Hertzberg et al. 2003). Staff have also reported that communication with relatives is too time-consuming (Hertzberg et al. 2003). In the light of the increasing trend towards home-based care (Thorslund et al. 1997), it is suggested that further studies are needed to understand potential hindrances for home care nursing staff to interact with family members. It is important to improve family members’ and staff interactions, because family members in home-based care often have an active role in care provision (Blomberg et al. 2000). More structured ways to inform and involve the relatives are needed to maintain and improve the involvement and potential help of relatives of care recipients. At the same time, relatives who are actively involved in caring for their next-of-kin may be extra critical of the quality of care/services provided, because their ratings may be influenced by their own care-giving burden and possible social isolation (Stoltz et al. 2004).

Activities offered to older people was one of the aspects with lowest ratings by care recipients and their relatives in both care settings. In addition, both respondent groups rated activities significantly lower in home care when compared to nursing homes. Previous studies have also found that activities were given low ratings by care recipients (Wellwood et al. 1995, Berglund 2007) and families (Maas et al. 1991, Ryan & Scullion 2000, Finnema et al. 2001), but these studies did not compare different care settings. Lower ratings in home-based care may reflect older people’s anxiety for going outside their homes, especially those in poorer physical health with difficulties walking. It is also possible that not all the care recipients in home-based care were entitled to activities organised by the municipality. Although all the nursing home residents are entitled to these activities, a decision from municipal authorities is required for care recipients’ participation in municipal older people activities (The Swedish Association of Local Authorities and Regions 2005). Nursing home residents might have limited interest in social and physical activities (Berglund 2007), because they often are in poorer health than those receiving home care; perhaps nursing home residents are satisfied with fewer activities than care recipients in home care. These findings are of particular importance, since opportunities for active life for older people are emphasised in the Swedish national goals for older people care (Government bill 1997/98:113). Social and physical activities have a great impact on older peoples’ health (Batty 2002, Menec 2003, Fratiglioni et al. 2004) and quality of life (Rejeski & Mihalko 2001, Warr et al. 2004). Furthermore, participating in activities may affect older people’s ability to continue to live in their own homes, which has been one of the main goals for communities responsible for older people care (Thorslund et al. 1997). These findings also have implications for managers and nursing staff involved in older people care. It is suggested that older peoples’ preferences for social and physical activities be investigated on a regular basis, along with factors that could enable older individuals’ participation in activities.

It is possible that the observed differences regarding relatives’ perceptions between the care settings reflect different relative groups. In both municipalities, a significantly greater percentage of relatives of home care recipients had contact with the care organisation for fewer than five years. Previous studies have shown that the length of time receiving services and care was associated with relatives’ satisfaction ratings (Grau et al. 1995, Chiu 1997). However, longer contact periods with the care organisation were associated with both lower and higher satisfaction. Home care relatives also had daily contact with care recipients to a greater extent than relatives of nursing home residents. It is likely that those with daily contact felt obliged to take on more responsibility in caring for their next-of-kin, which may have made them less satisfied with the services provided. However, our survey did not ask relatives to describe how much care they provided for their next-of-kin. The degree to which these differences in relatives’ characteristics explain the differences in quality ratings between care settings is thus unclear.

In general, care recipients and their relatives gave the same aspects of care the highest and lowest quality ratings, respectively, in these care settings. Care recipients gave highest ratings to the indices staff behaviour and care both in home-based care and in nursing homes. Relatives gave highest ratings to the index care in both care settings. Thus, it seems that these groups’ perceptions were that the older care recipients in both home care and nursing homes were well taken care of regarding the elementary aspects of care such as personal hygiene, meals and physical transfers. Previous studies of Swedish older people care have found that relatives and care recipients in general gave positive ratings to basic nursing care (The National Board of Health and Welfare 2006a). It has also been reported that staff prioritise the nursing tasks of their work instead of, for instance, the social aspects (The National Board of Health and Welfare 2006b), which could be the reason for high ratings regarding nursing care and lower ratings regarding information and activities.

Adjusted regression analyses including tests of interaction between the quality indices and the care settings did not reveal different effects of any indices on the overall quality ratings in home care and nursing homes. Staff professional skills was the strongest predictor of relatives’ overall quality ratings, and staff behaviour was the strongest predictor of care recipients’ overall quality rating. These findings are in line with a previous study showing that care recipients’ overall satisfaction with care quality was highly influenced by their satisfaction with nursing care staff (Chou et al. 2002). Our findings suggest that nursing staff behaviour and their competence are important factors influencing care recipients’ and relatives’ overall quality ratings, regardless of care setting.

Methodological considerations

Some methodological limitations, such as low response rates, need to be considered when interpreting the results. The care recipients deemed capable of expressing their views on the quality of care comprised <70% of all the care recipients. The response rate among those who received the questionnaires ranged between 22–86%, depending on the municipality and care setting. The response rates in the relative surveys were also low, which indicates that the results of this study were based on limited material. It is possible that different results could have been obtained if a larger proportion of care recipients and relatives had responded to the surveys. Non-respondents among older people have been found to be older and more disabled than those responding to a postal questionnaire (Victor 1988), which might affect quality of care evaluations (Curtis et al. 2005). Limited information about those who did not respond to the present surveys makes it difficult to determine whether the study samples were representative of the total populations of care recipients and relatives.

Selection bias was possible regarding the care recipients’ study, because supervisors’ and nursing staff decisions regarding which care recipients were able to reply to questions may not have been correct. Previous research has indicated that staff inappropriately excluded competent residents from quality evaluations when using subjective screening criteria (Simmons et al. 1997). However, the individuals who were recruited to help care recipients with the questionnaire remarked that a substantial proportion of those care recipients deemed capable by staff of responding to the questionnaire were only partially able to do so. This indicates that staff did not easily exclude care recipients from the survey. It is also possible that the help offered to care recipients in replying to the questionnaire biased their responses. However, in previous research with a similar sample of older people, no systematic differences in quality ratings between those who required assistance and those who filled out the questionnaire by themselves were found (Hasson & Arnetz 2008a). This is in line with another study (Chou et al. 2001) that used a similar procedure to assist older people with a quality of care questionnaire and did not find any differences in quality ratings between residents who answered without help and those who received assistance.

The current study encompassed only two municipal older people care organisations, which limits the generalisation of the results. Future studies encompassing a greater number of work organisations are needed. Similar studies in other countries would also reveal whether similar differences between care settings exist in other care systems. The two municipalities differed in sample size and response rates, but there were only a few significant differences between questionnaire respondents in the two municipalities. This indicates that the groups were similar with regard to several background variables. Similar results, i.e. lower quality of care ratings in home-based care, were found in both municipalities, despite different response rates.

Conclusion

Compared with nursing homes, home-based older people care seems to be in greater need of development regarding staff competence, staff interaction with relatives and activities offered to older people. In both settings, nursing staff behaviour and competence influence these stakeholders’ satisfaction with care.

Relevance to clinical practice

The findings of the study have implications for managers and nursing staff involved in older people care. It is suggested that older peoples’ preferences for social and physical activities, both in home-based care and in nursing homes, be investigated on a regular basis, along with factors that could enable older individuals’ participation in activities. In addition, opportunities for older peoples’ relatives to interact with care staff differ greatly between the care settings. Efforts should also be made to improve communication and interaction between family members and older people care staff, especially in home-based care.

Acknowledgements

The authors thank all the care recipients and their relatives who participated in the study. Special thanks to Joel Ager, PhD, Yun Wang, MAS and Claire Brown, PhD, for statistical advice. This study was financially supported by Swedish Agency for Innovation Systems VINNOVA, The Swedish Foundation for Health Care Sciences and Allergy Research (The Vårdal Foundation), Swedish Skandia Life (publ) Group, and the two municipalities involved in the project.

The first author was affiliated with Uppsala University, Sweden during the time that this project was carried out.

Contributions

Study design: HH, JA; data collection and analysis: HH, JA and manuscript preparation: HH, JA.

Conflict of interest

The authors have no conflict of interest.

The Journal of Clinical Nursing (JCN) is an international, peer reviewed journal that aims to promote a high standard of clinically related scholarship which supports the practice and discipline of nursing.

For further information and full author guidelines, please visit JCN on the Wiley Online Library website: http://wileyonlinelibrary.com/journal/jocn

Reasons to submit your paper to JCN:

High-impact forum: one of the world’s most cited nursing journals and with an impact factor of 1·194 – ranked 16 of 70 within Thomson Reuters Journal Citation Report (Social Science – Nursing) in 2009.

One of the most read nursing journals in the world: over 1 million articles downloaded online per year and accessible in over 7000 libraries worldwide (including over 4000 in developing countries with free or low cost access).

Fast and easy online submission: online submission at http://mc.manuscriptcentral.com/jcnur.

Early View: rapid online publication (with doi for referencing) for accepted articles in final form, and fully citable.

Positive publishing experience: rapid double-blind peer review with constructive feedback.

Online Open: the option to make your article freely and openly accessible to non-subscribers upon publication in Wiley Online Library, as well as the option to deposit the article in your preferred archive.

Ancillary