This study provides initial evidence for social inequalities in the utilization of some specific health providers and institutions within a population of people with physical impairments due to SCI in Switzerland. However, we generally find that the basic health care provision is guaranteed in this wealthy country, and that social inequalities were only observed for specific care providers. Most prominent social inequalities were found for the visits to specialists, dentists, and dental hygienists, with persons from higher SES groups reporting higher likelihood for having visited those providers in the past 12 months. Moreover, we observed that persons with higher education reported more visits to pharmacists, persons with higher income more visits to natural healers, and persons with higher subjective social status more visits to chiropractors. Also, persons granted supplementary benefits were more likely to visit general practitioners and receiving support from home care services and persons with higher SES tended to visit a larger number of different health care providers. In contrast, we found statistically non-significant trends towards lower likelihood for inpatient stays, outpatient clinic, and emergency department visits and enhanced likelihood to visit a specialized SCI-center in higher SES groups. Effects of the perceived financial hardship on health service utilization was more volatile and mostly statistically insignificant. Although this study did not detect pronounced social inequalities other than in the utilization of specialists and dental care providers, findings nevertheless highlight the importance of including different SES indicators in research that aims to identify drivers of inequalities in health service utilization in people with SCI, as different SES dimensions relate to different resources important for health service utilization. SES-indicator specific findings are discussed and interpreted in the following paragraph.
In line with previous studies [3-5, 7, 28], we found an increased likelihood for specialist visits in higher SES groups. Earlier findings suggested that increased awareness for the importance of own health, enhanced cultural capital and larger social networks leading to better navigation in the complex health system, and differences in information policies of general practitioners might contribute to the inequalities in specialist visits . Increased awareness and higher health literacy in higher SES groups [29, 30], might present the critical arguments for the utilization of yearly routine medical check-ups, which are often performed by specialists in Switzerland and thus explain enhances specialist visits. Also, previous studies support the notion that preventive screenings are more prevalent in higher SES groups [31, 32]. Our finding that higher SES groups report higher odds for dentists and dental hygienist visits confirms a well-known phenomenon [11, 12, 33, 34]. Besides the fact that visits to dentists and dental hygienists are usually not covered by health insurance and thus reinforcing income inequalities in utilization, similar arguments can be used to explain the differences in utilization as other preventive medical treatments. It is however important to mention that the costs for dental care of persons with supplementary benefits are covered by the health insurance. The lower likelihood to visit dental hygienists can thus not be explained by the scarcity of financial resources, but is probably due to other factors, such as limited awareness for the importance of oral health or difficulties in overcoming environmental barriers (e.g., lack of transportation, no wheelchair accessible dental surgery). Further, it might even be the case that persons with supplementary benefits are not aware that the costs are overtaken by the insurance, or that bureaucratic barriers in the reimbursement of costs hinder people in utilizing this service. Persons with SCI are particularly vulnerable to poor oral health and poor oral health can have drastic effects on the immune system or the development of pressure ulcers [35, 36]. Therefore, specific interventions to enhance the utilization of oral health care services in lower SES groups are highly recommended.
The result showing that persons with higher education and higher subjective social status had higher odds to visit pharmacists is difficult to interpret. As many general practitioners in Switzerland are authorized to dispense medications, these findings possibly mirror the fact that persons from lower educational or subjective status groups more often obtain medications directly from general practitioners, and therefore do not need to visit pharmacists. This assumption is supported by the fact that we observed a slight tendency for more general practitioner visits in persons with lower education and lower subjective status. Further, the finding that persons with higher income reported a higher likelihood to visit natural healers might directly relate to financing issues. In Switzerland, only a selection of officially acknowledged alternative medical treatments (e.g., traditional Chinese medicine, homeopathy) are compensated by health insurances and any alternative treatment apart from these recognized treatments must be payed out of pocket, thus potentially limiting access from lower income groups. The trend that persons granted supplementary benefits were more likely to visit general practitioners has been observed for the general Swiss population  and might be explained by the fact that persons with supplementary benefits more often live in institutions, where basic health care is provided by the general practitioners. Moreover, elderly persons are at higher risk to be granted supplementary benefits, and elderly persons more often report long-term care relationships with general practitioners than younger persons . The higher support from home care services in persons receiving supplementary benefits might be explained by financial reasons, as health insurance fully covers the cost of home care for persons who are granted supplementary benefits, whereas others usually have to contribute a considerable part of the financing for home care services by themselves. It might also be the case that persons without supplementary benefits can more often rely on informal caregiving from family members and are thus less dependent on formal home care.
It is finally highly likely that persons in higher SES groups generally have more financial and knowledge-related resources to navigate the complex health system and use a combination of different treatments, as represented in the findings of higher number of different health care providers visited in higher SES groups. In contrast, our study did not provide evidence for social inequalities in the utilization of different health care institutions, as results on associations between any SES indicator and inpatient stays, visits to outpatient clinics, emergency departments and specialized SCI-centers were insignificant. Although we found non-significant trends showing slightly increased likelihood for higher emergency department visits in lower SES groups, previous general population findings  were not replicated in our sample of persons with SCI in Switzerland. This suggests that primary care provision only marginally follows social patterns and that access to basic health care is guaranteed for all persons with SCI, preventing persons from lower SES groups of high utilization of emergency departments. One reason for the lack of social inequalities in the utilization of those institutions might be that persons with SCI usually undergo long initial rehabilitation programs, where they are specifically educated in health management, irrespective of their socioeconomic conditions.
Strengths and limitations
The SwiSCI community survey provides a large population-based data base with a well-defined sampling frame and neglectable response bias related to known sociodemographic and lesion characteristics . Moreover, we adjusted final models for health states, allowing an understanding of SES-patterns in health service utilization that is not due to the different health care needs, but aims to understand 'true' differences between different SES groups. However, given that the SES of non-respondents was not assessed, we cannot evaluate whether we adequately included the most deprived groups or whether there is a selection bias towards over-representation of persons from higher SES groups. Results might be more pronounced if particularly disadvantaged individuals in lower SES groups, such as persons with low SES and language barriers, would have been included as well. It remains further unknown whether we missed some relevant, unmeasured confounders that impact on SES as well as on health service utilization. Moreover, the cross-sectional nature of the data prevents inferences about causal relationships. However, it seems reasonable to assume that SES indicators affect the health service utilization, and not vice versa. Also, the use of self-report data on the reporting of visits to the health care providers and institutions during the past 12 months might be subject to recall bias. Furthermore, the high amount of missing values in some of the constructs (e.g. household income) might limit the robustness of comparisons between analyses based on full case vs. imputed data.