Introduction
Roma (so-called Gypsies) make up one of the largest minority populations in Europe. More than 400,000 Roma are estimated to be living in Slovakia, which represents 7.5% of Slovak population [
1]. Fewer than half of all Roma (46.5%) live scattered among the majority population; the rest live either in urban concentrations within or on the outskirts of towns or villages (36.7%) or in separated or segregated Roma settlements (17%) [
1]. The latter, in particular, are characterised by degrading living conditions. The poor housing of Roma often has no connection to a sewerage system and/or running water and is cut off from transport networks, etc. [
1]. This “circle of failure” is closed by a low educational level that leads to low employment rates and results in a high dependency on welfare benefits passed from one generation to the next [
2].
As mentioned above, Roma are the most deprived ethnic minority in Slovakia and suffer from discrimination, poverty and social exclusion [
2-
4]. These unfavourable conditions have a remarkable impact on the health of the Roma population compared with the majority population, as has been previously shown by several studies. Roma were found to have higher infant mortality rates, lower life expectancy and a higher frequency of health complaints [
5,
6]. Some studies have also shown poorer self-rated health in Roma compared with non-Roma [
7-
9]. Not only do Roma appear to have worse health, they also seem to face problems in accessing good quality health care [
10]. Moreover, there might be an association between these two factors [
11,
12], which have not been adequately studied [
10]. Roma seem to be systematically excluded from key aspects of health care, such as preventive, primary and specialised health services as well as pre- and postnatal health care [
10] and instead increasingly rely on emergency services [
10].
Roma appear to perceive their own health differently than the majority population [
13,
14]. The health of Roma is often perceived through the perspective of their families, which seems to be rather logical considering the tight social bonds within Roma families and communities [
15]. Whether the family is doing fine or not might have an influence on an individual’s perception of his or her own health and well-being [
15]. This is in line with the social capital perspective, which views networks of relationships as a valuable resource (social capital) for individuals. According to Putnam (2000) social capital consists of features such as networks, norms and trust that facilitate coordination and cooperation for mutual benefit and aid [
16]. Social capital can be conceptualized at both the collective and individual level [
17,
18]. Collective social capital refers to characteristics of communities, workplaces or neighbourhoods, whereas individual social capital concerns elements related to social relationships of individuals [
19]. Social network ties within family, friends and community provide social support and are thought to play a critical role in accessing and transferring resources, including material, information and knowledge [
20]. Social support has been categorised in various ways, but in general it covers social interactions that “lead the subject to believe that he or she is cared for, loved, esteemed and a member of a network of mutual obligations” [
21]. In line with this theoretical perspective, several studies have confirmed the role of social support in influencing health-care-seeking behaviour and treatment adherence [
22-
24]. Boateng et al. [
23] in their study explored possible enablers and barriers in access to the Dutch health care system among Ghanaians in Amsterdam. Among the most frequently mentioned enablers they listed availability/accessibility of family support and community initiatives/cohesion as important factors. Community was mentioned to provide enlightenment on health issues, and family members provided support in direct contact with health professionals. In addition, Gotay and Wilson [
24] in their summary of several studies focused on the association of social support with breast and cervical cancer screening in certain minority groups in the USA and found a connection between received social support and an increase in cancer screenings.
Problematic access to good quality health care as result of institutional and interpersonal discrimination of Roma affects their health [
25]; therefore, factors which affect the health care accessibility of Roma are of high importance for public health and policy makers.
Thus, the aim of this study was to explore the mediating as well as moderating effect of social support on the association between health care accessibility problems and ethnicity.
Results
There are significantly more women, those who reported health care accessibility problems and those with lower education among Roma in comparison with non-Roma (Table
1). Roma perceive significantly more PSS from their family and significantly less PSS from their friends in comparison with non-Roma (Table
1). We did not find any ethnic differences in regard to health care specific social support from the family (Table
1). Conversely, non-Roma perceive significantly more health care specific social support from their friends when compared with Roma (Table
1).
Roma have a 3.5-times higher chance of reporting health care accessibility problems in comparison with non-Roma (Table
2). There is also a slightly higher chance of reporting health care accessibility problems among women compared with men (Table
2). The likelihood of reporting health care accessibility problems increases with increasing age (Table
2). Social support (PSS) in both dimensions significantly decreased the likelihood of reporting health care accessibility problems (Table
2). Health care specific social support (HCSSS) from family significantly decreases the likelihood of reporting health care accessibility problems (Table
2), whereas the effect of HCSSS from friends is not significant.
Adding PSS from family and friends into the model decreased the association between health care accessibility problems and Roma ethnicity (Table
3, Model 2). Adding HCSSS from family into the model increased the association between health care accessibility problems and Roma ethnicity (Table
4, Model 3). The results from the final model indicate that family is a dimension of social support which partially decreases the likelihood of reporting health care accessibility problems (Table
3, Model 4). Adjustment for highest education hardly affected the observed associations (not shown), because there is a strong mutual association between ethnicity and highest level of education which showed effects of collinearity; thus, these two constructs measure practically the same variable.
Since we found statistically significant ethnic differences regarding perceived health care accessibility problems and social support, we also assessed whether the effect of PSS and HCSSS on the likelihood of reporting health care accessibility problems was modified by ethnicity (Table
4), and this showed a statistically significant interaction effect in regard to PSS from friends (Table
4, Model 2) and HCSSS from family (Table
4, Model 3). The significant interaction of ethnicity with PSS from friends indicates an increasing effect of ethnicity on health care accessibility problems with decreasing levels of PSS (Table
4, Model 2). Thus, non-Roma who perceive lower levels of social support from their friends have a higher chance of reporting health care accessibility problems in comparison with Roma. The significant interaction of ethnicity with HCSSS from family indicates an increasing effect of ethnicity on health care accessibility problems with decreasing levels of HCSSS (Table
4, Model 2). Thus, non-Roma who perceive lower levels of HCSSS from their family have a higher chance of reporting health care accessibility problems in comparison with Roma.
Discussion
The aim of this study was to explore the mediating as well as moderating effect of social support on the association between health care accessibility problems and ethnicity. Roma, in comparison with non-Roma, have a more than 3-times higher chance of reporting health care accessibility problems. Perceived social support (PSS) from family and friends decreases the likelihood of reporting health care accessibility problems in both Roma and non-Roma, whereas the family seems to be a more important factor. Health care specific social support (HCSSS) from family also decreases the likelihood of reporting health care accessibility problems, but only in non-Roma. There is no significant difference between Roma and non-Roma in how PSS from family affects health care accessibility problems. However, the lack of PSS from friends as well as HCSSS from family seems to increase the likelihood of health care accessibility problems more in non-Roma than in Roma.
We found no differences between Roma and non-Roma regarding HCSSS from family. However, the effect of HCSSS from family on the likelihood of reporting health care accessibility problems was modified by ethnicity, as the interaction was significant. These results suggest that the effect of HCSSS from family on health care accessibility problems differs between Roma and non-Roma. On the other hand, Roma and non-Roma differed in regard to PSS from family. The effect of PSS from family on the likelihood of reporting health care accessibility problems was not modified by ethnicity, as the interaction was not significant. Thus, it is likely that Roma and non-Roma do not differ in how general social support affects their health care accessibility problems. In line with this, several other studies have also confirmed the role of family support as an important factor in regard to influencing health care-seeking behaviour or treatment adherence [
22-
24].
We found statistically significant differences between Roma and non-Roma regarding HCSSS from friends. The effect of HCSSS from friends on the likelihood of reporting health care accessibility problems was not modified by ethnicity. Roma and non-Roma also differed in regard to PSS from friends. Moreover, the effect of PSS from friends on the likelihood of reporting health care accessibility problems was modified by ethnicity, as the interaction was significant. Thus it is likely that Roma and non-Roma differ in how general social support affects their health care accessibility problems.
We found that social support from family and friends has a positive/protective effect on health care accessibility problems, which means that people with higher levels of family or friend support are less likely to report problems. Of all the types of social support we explored, PSS from family is the most important factor for helping to overcome health care accessibility problems, regardless of ethnicity. However, it does not substantially decrease the ethnic differences in health care accessibility problems. Roma in Slovakia are not likely to participate in a wider network of relationships [
29]. They have limited access to structures beyond family ties and their local community, which are intertwined by close and distant family relationships [
29]. In our opinion, one of the possible explanations is that for the majority population the wider social network (e.g. friends, acquaintances, and community) serves as an additional support in order to find and arrange the best possible and most accessible care of a particular doctor in a particular field of medical care. This is in line with the informational function of social support (e.g. providing advice, suggestion, directives, information) as identified by House and Kahn (1985). On the other hand, we assume that Roma usually do not have such friends or acquaintances who can advise them about health care providers or solutions for their health related problems within the health care system.
The health of Roma is perceived through the perspective of their families, which seems to be rather logical considering the tight social bonds within Roma families and communities [
15]. Different social norms, values, health beliefs and behavioural differences [
13,
14,
30] may also lead to different perception of health as well as different ways of using or not using health care services [
12,
31]. Roma often prefer home treatment with their own curative methods and/or have had bad experiences with the health care system [
12,
32]. Moreover, Roma themselves often cite their own bad economic situation (lack of money for medication and transportation) as justification for the problematic use of health care services [
12]. Nevertheless, the worse health of Roma can be partially accounted for by worse access to health care services [
12,
32].
Strengths and limitations
The strength of our study is that it comprises a considerable representative sample of a hard-to-reach population living in separated or segregated Roma settlements. We were able to compare them with the majority population living in the same geographical area, although our results should be generalised with caution, as Roma are a very heterogeneous group in terms of living conditions and levels of integration; thus, our results can be generalised only to the Roma population living in Roma settlements.
A limitation of our study may be that data from the Roma were collected via an interview, and data from the non-Roma came via self-reported questionnaires. The reason for this was to cope with illiteracy among the Roma, which we considered to be a more serious source of non-response and bias than using two different types of administration. In Roma, we used the technique of assisted self-administration, which has shown good reliability of data [
27]. Moreover, using this technique turned out to be a good decision, as most of Roma were not able to fill in the questionnaires without assistance.
Implication
As suggested by our results, the worse accessibility of health care in the Roma population creates a severe public health concern. This issue can be partially solved through promotion of social support, which appears to partially decrease health care accessibility problems equally among the majority population as well as in the Roma population. As Roma do not seem to effectively use potential support from wider social networks beyond family ties, we need to look for other resources which might ease health care accessibility for Roma. Health-related behaviours of individuals within Roma communities seem to be influenced by privileged males and elders, or family members even when their advice or beliefs might be inappropriate [
33]. One of the possible options for benefiting from this is to use Roma health mediators and community workers to identify influential individuals who are able to change a community’s fear and distrust and persuade and teach Roma to seek out and appropriately use health care services [
34,
35]. Moreover, making access to health care easier for one Roma individual creates benefits for the whole community via social support. Social support might potentially promote population health beyond the targeted individuals by influencing the health of others through existing social networks [
36]. Future research should focus on a broader range of factors affecting Roma access to the health care system and proper validation of measures aimed at accessibility of health care being introduced into the practice.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
All authors (except of DK) contributed to the design of the study as well as to the protocol. Data collection was supervised by ZDV and AMG. DB, ZDV, and IB wrote the first version of the manuscript. DB, ZDV, IB, AMG and LC participated on statistical analyses, interpretation, and discussion of results. DK provided statistical support and supervision. All authors have substantially contributed to the manuscript as submitted. All authors have read and approved the final manuscript.
Daniela Bobakova, PhD., got her master degree in social work at the St. Elisabeth University of Health and Social Sciences in Bratislava. During her external studies she worked in temporary state service as an Independent Consultant in the Social and Family Bureau, Social Division, at the Department of Social and Legal Child Protection. She is currently a postdoctoral fellow at the Institute of Health Psychology, Medical Faculty, PJ Safarik University in Kosice. Her research is focused on health behaviour in school-aged children and the health of the Roma population.
Prof. Andrea Madarasova Geckova, PhD., received her PhD degree in medical sciences from the University of Groningen, the Netherlands. She is a professor in Social Psychology (Masaryk University in Brno, Czech Republic). Her research is at the crossroads of public health, health psychology and medical sociology. Since 1997 her international research projects have focused on inequalities in health among (Slovak) adolescents and young adults. Currently she is a Head of the Institute of Health Psychology.
Zuzana Dankulincová, PhD., received her Master degree in Psychology (Faculty of Arts, University of Presov) and PhD in Medical sciences (University of Groningen, The Netherlands). At present, she is the in the position of the senior researcher at the Institute of Health Psychology, Medical Faculty, PJ Safarik University in Kosice.
Dr. Ingrid Babinska, MPH, graduated as a medical doctor in 1995 at the Medical Faculty, University of PJ Safarik in Kosice. She received her specialization in the field of hygiene and epidemiology (1999), specialization in the field of public health (2009) and completed three years of study Master of Public Health (2002–2005). She has worked in the position of a physician, a public health professional and a regional health officer. She received her PhD degree at the Institute of Public Health, Medical Faculty of PJ Safarik University in Kosice. Her research focuses on risk factors for cardiovascular disease in the adult population with a particular focus on socioeconomically disadvantaged groups.
Daniel Klein, PhD., got his master’s degree in mathematics from PJ Safarik University in Kosice in 2004 and he received his PhD degree at the same University in 2008. After studies he began working at the Institute of Mathematics of PJ Safarik University as a researcher and from 2010 as an assistant professor. His research interests are linear models, multivariate statistics and computational statistics.
Prof. Lydia Cislakova, DVM, PhD., has been working at P. J. Šafárik University in Kosice, Faculty of Medicine, as university teacher in field of Epidemiology since 1971. From 1998 till 2015 she was a Head of Institute of Public Health. Currently she is a Head of the Institute of Epidemiology. She serves as a member of several scientific boards and on the editorial board of Helmithologia (SSCI journal). Her scientific work has been awarded several times by the Slovak Scientific Agency and other institutions. Her research work is focused on the epidemiology of infectious diseases. She participates in several research projects. She participated in the designing, preparation and delivery of curriculum for medical students in field of Public Health. She has a strong academic background.
HEPA-META team: Jarčuška Peter, Pella Daniel, Mareková Mária, Siegfried Leonard, Pastvová Lýdia, Fedačko Ján, Kollárová Jana, Dražilová Sylvia, Rosenberger Jaroslava, Schréter Ivan, Kristián Pavol, Veseliny Eduard, Janičko Martin, Virág Ladislav, Birková Anna, Kmeťová Marta, Halánová Monika, Cáriková Katarína, Petrášová Darina, Lovayová Viera, Merkovská Lucia, Jedličková Lucia, Valková Ivana