Communication and equality in elderly care settings: perceptions of first- and second- generation immigrant and native Swedish healthcare workers

What is known on this subject . Overseas registered nurses report experiences of inequality and discrimination at work in their interactions with co-workers and patients. . Ethnic-minority healthcare workers in nursing home settings feel less respected by residents and their families than native healthcare workers, and report racism from both elderly patients and their family members. . Previous research has focused on the experiences of established ethnic-minority healthcare workers or recent immigrants such as overseas registered nurses. There is limited research comparing the experiences of majority/native healthcare workers with firstand second-generation ethnic-minority/immigrant healthcare workers.


Introduction
Sweden, along with other high-income countries, is facing a significant increase in the elderly population (European Migration Network, 2011), and it is estimated that, within 50 years, people over the age of 65 years will represent 25% of the population (Statistics Sweden, 2012a). The increase in the proportion of elderly people will place additional demands on healthcare services. A joint EU initiative to investigate labour market requirements across Europe identified a need to increase the number of immigrant healthcare workers to meet the future healthcare needs of an ageing population (European Migration Network, 2011). In Sweden the situation is compounded by the significant proportion of healthcare workers working in elderly care settings who are due to retire soon (Statistics Sweden, 2012b).
Sweden's population is ethnically diverse. The number of people from other countries has steadily increased since 2000, and now accounts for 15% of the total population of 9.5 million. The immigrant population, consisting of first generation (those born outside Sweden) and second generation (those born in Sweden with one or both parents born outside Sweden), is mainly located in large cities. In Stockholm, approximately 22% of citizens are first-or secondgeneration immigrants (Statistics Sweden, 2012c). The proportion of elderly people from immigrant backgrounds reflects that of the population at large, 88% of the elderly are from native Swedish backgrounds and 12% are from immigrant backgrounds (Statistics Sweden, 2011).
Approximately 9% of registered nurses and 20% of other healthcare professionals in the Swedish healthcare workforce are from an immigrant background, born outside Sweden, or have both parents from other countries (Organisation for Economic Co-operation and Development, 2007). Although an ethnically diverse healthcare workforce is seen to be beneficial in meeting the healthcare needs of an ethnically diverse population (Gerrish et al, 1996), it is not without challenges. Lack of equality at work among a diverse healthcare workforce, as well as difficulties in interpersonal relationships between co-workers and patients, are especially challenging in the UK (Culley and Mayer, 2001) and the USA (Dreachslin et al, 2002). However, little is known about the implications of diversity in the Swedish healthcare workforce. This paper presents the findings from a study examining healthcare workers' perceptions of equality and communication among an ethnically diverse workforce.

ABSTRACT
An ethnically diverse healthcare workforce is considered beneficial to meeting the needs of an ethnically diverse population. In the UK and the USA, lack of equality and difficulties in communication between co-workers and patients from different ethnic backgrounds is problematic. Little is known about the ethnically diverse healthcare workforce in elderly care settings in Sweden. This paper compares native Swedish and first-and secondgeneration immigrant healthcare workers' perceptions of diversity in relation to equality and communication in elderly care settings. The study used a cross-sectional design with a survey administered by self-completed questionnaire. The Assess Awareness and Acceptance of Diversity in Healthcare Organizations questionnaire was distributed to healthcare workers in elderly care settings in one municipality in Sweden. Responses from 643 healthcare workers were analysed. A factor analysis was performed on 26 items in the questionnaire. Reliability analysis on the subscales was conducted using Cronbach's alpha. Differences between native and first-and second-generation immigrants were analysed using ANOVA followed by post-hoc tests. The results showed that first-generation immigrant and native Swedish healthcare workers had different views on equality and communication in four of the five subscales, namely care of elderly patients from different backgrounds, equality in the workplace, communication with diverse co-workers, and treatment by family and significant others from a different ethnic background. Second-generation immigrants held similar views to native Swedish healthcare workers on two factors, namely equality in the workplace and communication between co-workers from different backgrounds. There were no differences between the groups with regard to their views on self-awareness in collaboration with co-workers. Differences in the experiences of first-and secondgeneration healthcare workers should be acknowledged, rather than assuming that they share similar experiences as immigrants. Managers need to promote equality and effective communication among an ethnically diverse workforce.

Background
Research from the 1990s that examined the experiences of immigrant nurses in the UK identified lack of equality in job promotion, but did not examine wider issues of equality, particularly interactions between co-workers and between immigrant nurses and patients and their significant others who were from the majority population (Beishon et al, 1995;Culley and Mayer, 2001).
More recently, several studies undertaken in the UK and the USA have examined the experiences of overseas registered nurses who migrated in response to nursing shortages in these countries. Immigrant nurses reported that co-workers from the majority population did not take their previous nursing experience into consideration, and questioned the immigrant nurses' professional knowledge (Gerrish and Griffith, 2004;Larsen, 2007;Smith et al, 2006). Immigrant nurses described difficulties in teamworking associated with situations where they were left alone without support, whereas co-workers from the majority population supported each other (Alexis, 2009;Larsen, 2007;Smith et al, 2006), and immigrant nurses were given tasks that were avoided by other nurses (Alexis and Vygdelingum, 2004;Larsen, 2007). Furthermore, immigrant nurses perceived inequality with regard to both career progression and their day-to-day interactions with co-workers (Alexis and Vygdelingum, 2004;Berdes and Eckert, 2001;Smith et al, 2006). Although not focusing specifically on the experiences of immigrant healthcare workers, Dreachslin et al (2002) identified that black healthcare workers in the USA were more likely to experience racism than white coworkers.
However, working in a diverse workforce is not only associated with difficulties. Gates and Mark (2012) observed that ethnic diversity contributed to increased job satisfaction among older healthcare workers in hospital settings. Other studies have explored interactions between healthcare workers and patients and their significant others from a different ethnic background. Research conducted in home care has identified the racism experienced by ethnic-minority healthcare workers from both elderly patients and their family members (Berdes and Eckert, 2001;Neysmith and Aronson, 1997). Similarly, Sloane et al (2010) identified that ethnic-minority healthcare workers in nursing home settings felt less respected by residents and their families than did native healthcare workers.
Strategies for coping with discrimination by patients and significant others in elderly care have been identified. Healthcare workers who experience discrimination may set limits to what is tolerated or choose to ignore it, thereby avoiding the situation by distancing themselves from patients and their families (Berdes and Eckert, 2001;Neysmith and Aronson, 1997). However, such strategies may have a negative impact on the quality of care provided (Berdes and Eckert, 2001).
Working in ethnically diverse healthcare settings requires culturally sensitive communication skills both between co-workers and between care workers and patients. Misunderstandings in communication can arise as a result of different cultural interpretations of the message (Dreachslin et al, 2002), differences in body language (Lundberg et al, 2005;Parker and Geron, 2007), and language barriers (Bourgeault et al, 2010).
Most studies examining communication in ethnically diverse healthcare settings have focused on the interactions between healthcare workers and patients and their significant others from a different ethnic background. For example, nurses working in acute care found that their inability to communicate with patients because of language difficulties was stressful, and they were concerned about the impact that this had on the quality of care delivered (Hultsjö and Hjelm, 2005). Language barriers also result in nurses feeling insecure in the caring relationship (Fatahi et al, 2010;Xu et al, 2008) and being afraid of making mistakes (Lundberg et al, 2005).
There has been relatively little research examining communication between ethnically diverse co-workers. The study by Beagan and Chacala (2012) identified that immigrant healthcare workers felt embarrassed by their language difficulties and avoided asking their co-workers questions. Moreover, lack of trust and respect between co-workers from different ethnic backgrounds resulted in communication difficulties among the workforce (Dreachslin et al, 2002).
Although several research studies have sought to examine the experiences of healthcare workers from different ethnic groups, it is acknowledged that ethnicity is a complex concept. Externally imposed ethnic categories may imply a shared experience between members of a particular ethnic group, but it is widely recognised that there is considerable variation between members of a particular group (Eriksen, 1993;Blakemore and Boneham, 1994). For example, Burholt (2004) draws attention to the considerable variation among South Asian immigrants living in a large city in England, in terms of their language abilities, education, occupational status and settlement patterns within the UK. Caution should therefore be exercised when associating ethnicity with particular outcomes or experiences, as a number of other factors may need to be considered.
In summary, previous studies conducted on ethnic diversity in the healthcare workforce have been undertaken predominantly in the UK and the USA, and have identified issues in relation to equality and communication. However, it cannot be assumed that the findings from these studies are applicable to a Swedish healthcare context. To date there has been a lack of research exploring these issues in Sweden. Previous research examining ethnic diversity in the healthcare workforce has focused on the experiences of established ethnic-minority healthcare workers such as black people in the USA (Abrahamson et al, 2011;Dreachslin et al, 2002), or recent immigrants such as overseas registered nurses in the UK and the USA (Allan et al, 2009;Gerrish and Griffith, 2004). Few studies have compared the experiences of majority/ native healthcare workers with those of ethnic-minority/immigrant healthcare workers. Moreover, no studies have considered whether there are differences between the experiences of recent immigrants (first generation) and those who are more established (second generation). The current study sought to address these gaps in knowledge relating to diversity in the workforce.

Aim of the study
The aim of the study was to compare native Swedish and first-and second-generation immigrant healthcare workers' perceptions of diversity in relation to equality and communication in elderly care settings.

Design
A cross-sectional design that employed a survey by self-completed questionnaire was used.

Sample and setting
The sample consisted of healthcare workers involved in direct patient care who worked in elderly care settings within one municipality in Stockholm. The respondents were drawn from nine nursing homes, six of which also provided home-based care. Eligible participants included registered nurses, nurse assistants, physiotherapists and occupational therapists.

Data collection
Nursing home managers distributed the questionnaires to potential respondents during one week in 2006. Respondents were asked to return the questionnaire anonymously by depositing it in a box in the workplace. Questionnaires were distributed to all employees in the participating organisations involved in direct patient care and non-patient care roles. Of the 1016 questionnaires that were distributed, 841 were returned (i.e. the response rate was 83%). A total of 678 questionnaires were returned by healthcare workers involved in direct patient care, and these questionnaires were included in the present study.
Questionnaire Data were collected using the self-completed questionnaire Assessing Awareness and Acceptance of Diversity in Healthcare Organisations (AAAD). This questionnaire explores employees' perceptions of interactions both between ethnically diverse co-workers and between healthcare workers and patients and their significant others in elderly care settings. The development and validation of the AAAD questionnaire have been described by Emami and Safipour (2013).
A brief version of the AAAD questionnaire was used to address the specific aim of the study, which was to examine communication and equality in the workforce. The brief version was divided into two sections and consisted of 26 of the 31 items included in the full questionnaire. Section I consisted of 18 items examining social interactions, interest in learning about people from different backgrounds, communication, and equality. Section II consisted of 8 items examining communication and equality both between staff and elderly patients and between staff and the patient's partner and/or family members.
Each item was scored on a 4-point Likert scale (where 4 = strongly agree, 3 = agree to some extent, 2 = disagree to some extent, and 1 = strongly disagree). Additional questions collected information on the characteristics of the respondents, including gender, age and workplace employment (including length of work experience, full-/part-time and permanent/temporary employment).

Data analysis
Data were analysed using SPSS Version 20. The data were entered by one researcher and subsequently checked for accuracy by two independent members of the research team. Descriptive statistics were used to analyse the characteristics of the sample, and the Chisquare test was used to look for differences between first-generation immigrants, second-generation immigrants and native Swedish healthcare workers. First-generation immigrants were those born in a country other than Sweden. Second-generation immigrants were those born in Sweden, with one or both parents born outside Sweden.
As the brief AAAD questionnaire drew upon selected items from the full AAAD questionnaire, a factor analysis was performed to assess the validity of the brief version in addressing the research aim of examining diversity in relation to equality and communication in elderly care settings. Missing data were imputed using the expectation maximisation (EM) algorithm (Tabachnick and Fidell, 2001). An exploratory factor analysis with varimax rotation was performed to find the underlying dimension of the items. The numbers of items were reduced through consecutive iterations based on communalities at 0.6. The Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy was estimated on the factorability by the variance explained by the subscales. According to Brace et al (2006), KMO values greater than 0.6 are acceptable, and values closer to 1.0 are regarded as more accurate.
When identifying the subscales, eigenvalues greater than 1.0 and a scree plot to confirm the subscales were used (Tabachnick and Fidell, 2001). Reliability analysis of the suggested subscales was performed using Cronbach's alpha. The mean of the items in each of the five subscales was calculated to five mean scores, one for each factor, for further analysis.
One-way analysis of variance (ANOVA) was performed to evaluate statistically significant differences between the three groups (first-generation immigrants, second-generation immigrants and native Swedish healthcare workers) with regard to the identified/extracted subscales of the AAAD. Follow-up post-hoc tests for the least significant difference (LSD) were performed to examine pairwise differences. To quantify the size of the difference between the groups, Cohen's d was calculated from the pooled standard deviation (Coe, 2002). In relation to Cohen's d test, a value of 0.2-0.5 is regarded as a medium-size effect (Becker, 2000). The level of statistical significance was set at P < 0.05.

Ethical considerations
Ethical approval was granted by the Research Ethics Committee at the Karolinska Institutet.
Following ethical approval, permission to distribute questionnaires was obtained from the managers of the nursing homes and home-care settings. A cover page to the questionnaire outlined the purpose of the study, and explained that participation was voluntary and that the findings would be published. All of the questionnaires were completed anonymously. Consent to participant in the study was assumed on the basis of the return of a completed questionnaire.

Results
Of the 678 questionnaires that were returned by healthcare workers, data on the respondents' ethnic background were missing from 35 questionnaires, and these were therefore excluded from the analysis. The final number of respondents was 643, consisting of 543 women (84.7%) and 98 men (15.3%); no information on gender was given for two respondents. The participants consisted of first-generation immigrants (n = 302), second-generation immigrants (n = 78) and native Swedish healthcare workers (n = 263).
The demographic details of the respondents are presented in Table 1. Statistically significant differences (Chi-squared test) were identified in relation to gender (P 0.001) and age (P 0.001). A larger proportion of men were first-generation immigrants (21.3%) compared with second-generation immigrants (11.5%) and native Swedish participants (9.5%). A larger proportion of native Swedish healthcare workers were over 55 years of age (23.0%) compared with first-generation immigrants (11.9%) and second-generation immigrants (10.3%). A higher proportion of first-generation immigrants were in the 35-55 years age group (59.9%) compared with secondgeneration immigrants (35.9%) and native Swedish healthcare workers (45.2%). The youngest group of healthcare workers (20-35 years of age) contained a much higher proportion of second-generation immigrants (53.8%) compared with first-generation immigrants (28.1%) and native Swedish healthcare workers (31.8%). Almost all of the respondents in both groups of immigrants rated their language proficiency in Swedish as good or very good (first-generation immigrants, 91%; second-generation immigrants, 92%).
In the factor analysis, the number of items was reduced from 26 to 14 through two consecutive iterations, based on a cut-off value for communalities of 0.6. The KMO value was estimated to be 0.729, and the anti-image correlation was in the range 0.541-0.852 in the last iteration of the subscales. The number of subscales retained for rotation was determined by the criteria of eigenvalues greater than 1.0 on the total variance explained, and the scree plot suggested five subscales. The five extracted subscales accounted for 69.6% of the total variance. Cronbach's alpha for the five subscales was in the range 0.646-0.800. The items and the labels of the subscales are presented in Table 2.
There were statistically significant differences between the three groups on four of the five subscales (see Table 3). With regard to subscale 1 (F = 14.70, P 0.001), native Swedish healthcare workers found it significantly more difficult (P 0.001, Cohen's d = 0.46) to provide 'Care for elderly patients from different backgrounds' than first-generation immigrants. With regard to subscale 2 (F = 20.70, P 0.001), second-generation immigrants (P = 0.002, Cohen's d = 0.38) and native Swedish healthcare workers (P 0.001, Cohen's d = 0.53) experienced 'Equality in the workplace' to a significantly greater extent than first-generation immigrants. With regard to subscale 3 (F = 21.02, P 0.001), 'Communication between co-workers with different backgrounds', second-generation immigrants (P = 0.001, Cohen's d = 0.43) and native Swedish healthcare workers (P 0.001, Cohen's d = 0.53) experienced significantly more difficulties in communication compared with firstgeneration immigrants. With regard to subscale 4, 'Self-awareness in collaboration with co-workers from different backgrounds', no significant differences were identified between the three groups. With regard to subscale 5 (F = 3.76, P = 0.024), 'Treatment from patient and family members from different backgrounds', native Swedish healthcare workers, when compared with first-generation immigrants, reported that patients and families were more likely to treat care workers from a different background to themselves less well than they treated care workers from their own background (P = 0.001, Cohen's d = 0.21).

Discussion
The findings from this study provide insight into how diversity in relation to equality and communication is experienced by healthcare workers from different ethnic backgrounds. Previous studies have drawn comparisons between the experiences of immigrant and native/ethnic majority members of the workforce (Alexis et al, 2007;Khatutsky et al, 2010). This study is unique in that it compares first-and secondgeneration immigrants and ethnic majority members of the workforce. The similarities in the responses of native Swedish healthcare workers and secondgeneration immigrants, compared with first-generation immigrants, that were identified in relation to two of the five subscales suggest that the experience of diversity in an ethnically diverse workforce is more than simply a matter of whether members of the workforce are from an immigrant or a native/ethnic majority background.
There were statistically significant differences between first-generation immigrants and native Swedish healthcare workers in their views on equality and communication on four of the five subscales, namely care of elderly patients from different backgrounds, equality in the workplace, communication with diverse co-workers, and treatment by family and sig-   nificant others from a different ethnic background. Second-generation immigrants held similar views to native Swedish healthcare workers on two factors, namely equality in the workplace and communication between co-workers from different backgrounds. There were no differences between the groups with regard to their views on self-awareness in collaboration with co-workers. The similarities in the perceptions of secondgeneration immigrants and native Swedish healthcare workers compared with first-generation immigrants may be due to the fact that second-generation immigrants have experienced a greater degree of acculturation. Acculturation refers to changes that occur as a result of an individual's encounters with other cultural groups (Sam and Berry, 2006). The process is complex and dynamic, and generates changes at both an individual and a group level. The acculturation process affects an individual's identity, values and behaviours, and can lead to cognitive changes (Sam and Berry, 2006). The second-generation immigrants in the study reported here would have been exposed to Swedish culture via their formal education as well as during their subsequent employment. Through immersion in the Swedish culture (alongside their immigrant culture) it may be that second-generation immigrants have developed workplace values, attitudes and behaviours that are similar to those of native Swedish healthcare workers.
No data were collected on length of residency of first-generation immigrants, although the majority had been employed in their current workplace for more than 3 years. However, by virtue of their immigrant status they would have experienced at first hand living in a different country and exposure to different cultural norms. First-generation immigrants' acculturation within the Swedish context may therefore be anticipated to be different from that of secondgeneration immigrants, and this may in part explain their different responses. Future research should examine the process of acculturation of first-and secondgeneration immigrants into the healthcare workforce, and the implications of this for their experience of equality and communication in the workplace.
Subscale 1 drew attention to native Swedish healthcare workers experiencing greater challenges than first-generation immigrants in caring for patients from a different ethnic background. There are several possible explanations for this. Although ethnic data were not collected for patients included in the present study, native Swedish people represent the largest proportion of older people requiring care in Sweden (Statistics Sweden, 2011). It is therefore likely that native Swedish healthcare workers had limited experience of caring for older people from a different ethnic background to their own. Kai et al (2007) have drawn attention to the professional uncertainty and dis-empowerment experienced by healthcare workers in the UK in responding to the needs of patients from different ethnic backgrounds. Healthcare workers were concerned about their perceived lack of knowledge about cultural differences, and anxious about appearing discriminatory in their interactions with patients. NkuluKalengayi et al (2012) have drawn attention to the tensions caused by differences in views about health and disease held by healthcare workers and immigrant patients in Sweden. Findings from the present study suggest that healthcare workers in elderly care in Sweden experience similar challenges.
Arguably healthcare workers from immigrant backgrounds will have more experience of caring for elderly patients from a different ethnic background to their own, and may be more confident in responding to the needs of ethnically diverse patients. Although there were no statistically significant differences between the experiences of first-and second-generation immigrants, or between those of second-generation immigrants and native Swedish healthcare workers, it is interesting that first-generation immigrants did not perceive these challenges to be as great as native Swedish healthcare workers did. In a study examining the experiences of ethnically diverse nursing students in Sweden, Jirwe et al (2010) identified how firstgeneration immigrant nursing students drew upon their wider experience of living in a multi-cultural society to give them confidence in caring for patients from different ethnic backgrounds, and were adept at using strategies to facilitate their understanding of patients' preferences. It may be that the first-generation immigrant healthcare workers in the present study were similarly well placed to draw upon personal resources associated with their immigrant position that gave them greater confidence.
The participants generally expressed positive views of equality in the workplace (subscale 2: first-generation immigrants, mean 3.11; second-generation immigrants, mean 3.37; native Swedish healthcare workers, mean 3.46). This may be indicative of Swedish legislation (Ministry of Integration and Gender Equality, 2008) requiring employers to ensure equality in the workplace irrespective of the ethnic background of employees. Moreover, the fact that over 50% of the respondents in each group had been employed in the same organisation for more than 3 years suggests a certain level of job satisfaction linked to a positive work environment. This view is supported by a study of Korean nurses in the USA, which reported that the length of residency in the country was positively correlated with greater job satisfaction (Ea et al, 2008). Nasrabadi (2007) andConte (2003) also propose that being an immigrant healthcare worker can be beneficial when caring for immigrant patients, although the reasons for this are unclear.
However, the first-generation immigrants were less positive about equality with regard to workload allocation, influence and respect. Statistically significant differences were noted between first-generation immigrants and both native Swedish healthcare workers and second-generation immigrants. Most of the research examining equality in the healthcare workforce has focused on the experiences of recently recruited overseas registered nurses post-migration. These nurses experienced lack of respect from co-workers, their expertise was not valued, and they were allocated lowlevel tasks that did not reflect their level of expertise (Alexis and Shillingford, 2011;Alexis et al, 2007;Omeri and Atkins, 2002). It is not clear from the present study whether the lower scores of the firstgeneration immigrants relate to any of these issues. This area warrants further investigation.
Much of the research examining communication in ethnically diverse healthcare contexts has focused on communication between healthcare workers and patients (Hultsjö and Hjelm, 2005;NkuluKalengayi et al, 2012). In contrast, the present study examined communication between healthcare workers from ethnically diverse backgrounds. The findings indicate that native Swedish healthcare workers and secondgeneration immigrants perceived greater difficulty than first-generation immigrants in communication between co-workers from different ethnic backgrounds. Areas of concern in subscale 3 related to perception of misunderstanding in communication between coworkers, concern about raising communication problems, and lack of opportunity to discuss these problems. This observation is interesting in view of the communication difficulties reported by overseas registered nurses as first-generation immigrants in their interactions with work colleagues (Allan et al, 2009), and warrants further investigation. Statistically significant differences were observed between native Swedish healthcare workers and firstgeneration immigrants with regard to how they perceived patients and their significant others reacted to healthcare workers from a different ethnic background. Native Swedish healthcare workers felt that patients and their significant others would respond more negatively to work colleagues who were from a different background to the patient. Other studies have highlighted the negative experiences of immigrant nurses and care workers in their interactions with patients from a different ethnic background. For example, immigrant nurses have experienced derogatory comments from patients and family members (Abrahamson et al, 2011), and patients and family members may refuse to receive care from these nurses (Alexis and Shillingford, 2011). It is interesting therefore that the first-generation immigrants in the present study did not perceive negative treatment of coworkers by patients and their significant others to be as problematic as native Swedish healthcare workers did. It is not clear why this should be so, but it may be influenced by the concept of 'otherness' whereby individuals differentiate between themselves ('us') and others ('them'). According to Gerrish (1998), a person's definition of the 'other' is part of what defines the self and is integral to a person's identity. People construct roles for themselves in relation to an 'other' as part of a process of reaction to people whom they perceive to be different. 'Othering' helps individuals to distinguish between the certain and the uncertain, and is not necessarily derogative, but may lead to heightened anxiety arising from interactions with those considered to be 'others.' It may be that native Swedish healthcare workers viewed immigrant patients as 'other' in contrast to native Swedish patients, and either experienced or anticipated more anxiety in their interactions with such patients.

Limitations of the study
This study was undertaken in one municipality in Sweden with an ethnically diverse healthcare workforce. It cannot be assumed that the findings are generalisable to other municipalities in Sweden where the ethnic composition of the healthcare workforce may be different. In addition, it may be that some of the more positive findings with regard to perceptions of equality in the workplace are a reflection of policy initiatives to address equality undertaken in the municipality in question. Further research is needed to ascertain the generalisability of the findings to other settings.
The overall response rate was 83%. The questionnaire was distributed to the entire healthcare workforce, not just those involved in direct patient care. It was not possible to ascertain the number of nonresponders who were involved in care delivery, and therefore an accurate response rate could not be determined. Nevertheless, the sample size of 678 is sufficient to allow a degree of confidence in the findings.
It is acknowledged that the participants' responses may have been influenced by the fact that the questionnaires were distributed by the managers. However, every attempt was made to minimise this potential effect by asking the respondents to return their anonymised questionnaires to a box located in the workplace, rather than to their manager.
It is recognised that the collection of additional demographic data from the respondents might have provided further insights into the similarities and differences that were observed between the three groups. No data were collected on the professional background of the respondents, although registered nurses, physiotherapists and occupational therapists, together with unqualified nursing assistants, were invited to participate.
It may be that different professional groups had different perspectives on equality and communication in the workplace. In addition, data were not collected on the educational background and professional qualifications of the respondents, and this may also have a bearing on their experiences in the workplace. Furthermore, data were not collected on the length of residency of the first-generation immigrants in Sweden. Most of the research on first-generation immigrant nurses has focused on recently recruited overseas registered nurses, and it is possible that the experiences of first-generation immigrants who migrated prior to gaining a professional qualification may be different, as they will have undergone professional training and in some cases general education in Sweden. Finally, no data were collected on the immigrant respondents' country of origin, ethnic identity and visible physical characteristics that might affect the ability of second-generation immigrants to fully integrate with the native Swedish population and thereby avoid overt racism. In a study of migrant care workers in elderly care in Ireland, Doyle and Timonen (2009) identified differences in the experiences of European, South Asian and African carers, and it is possible that similar differences might have been identified within the present study. Future research should consider the demographic variables identified above.
The brief version of the AAAD questionnaire that was used in this study consisted of 26 items selected from the full AAAD questionnaire that were considered relevant to the aim of the research. The factor analysis led to 14 of the original 26 items being included in the five subscales that were identified. Since, to our knowledge, no gold standard for communication and equality in the workforce exists, the magnitude of the criterion-related validity of this version of the AAAD could not be established. However, the results of the factor analysis suggest that sound construct validity for this questionnaire could be assumed. Furthermore, the reliability of the questionnaire was examined on the basis of internal consistency. Due to practical conditions inherent in the cross-sectional design of the study, it was not possible to obtain multiple measurements (Cook and Beckman, 2006). Consequently, other sources of reliability, such as the stability of scores or inter-rater reliability, were not deemed to be applicable. However, the adequate internal consistency of the scales indicates evidence of the reliability of our measurements (Cronbach, 1951). All in all, the shorter version of the AAAD questionnaire consisting of these 14 items seems to be appropriate for examining diversity in relation to equality and communication in the workplace.

Conclusion
This study provides insight into experiences of equality and communication among a diverse workforce in elderly care settings. An important finding is the difference in responses within immigrant groups, where second-generation immigrants' responses were more in line with those of native Swedish healthcare workers than with those of first-generation immigrants. This suggests the need to reconsider how healthcare workers from immigrant backgrounds are grouped. It is important to acknowledge the differences in perceptions between first-and secondgeneration immigrants, as failure to do so may lead to misunderstandings and create division instead of cohesion (Parker and Geron, 2007). As Parker and Geron (2007) have pointed out, minimising or avoiding cultural differences among immigrant groups can make staff feel unsupported, which in turn can lead to low job satisfaction and increased staff turnover.
Clearly the managers of elderly care settings have a responsibility to ensure that equality and communication among an ethnically diverse workforce are enhanced in order to ensure that patients receive high-quality care. In considering the challenges of diversity in the healthcare workforce, Dreachslin et al (2004) propose a three-pronged approach. First, public policy should encompass a legal and regulatory framework that will eliminate disparities. Secondly, healthcare professionals should be culturally competent to provide appropriate care. Thirdly, staff, leadership and the culture of healthcare organisations should represent the communities that they serve. Current policy in Sweden provides a legal and regulatory framework that is supportive of an ethnically diverse healthcare workforce. However, the findings of this study have identified the need to ensure that all healthcare workers feel confident about meeting the needs of immigrant as well as native Swedish patients, and that the culture of the organisation promotes equality and effective communication among co-workers. As Dreachslin et al (2004) have pointed out, this requires strong leadership and commitment from the managers of healthcare organisations.