Health inequalities have their origin in the political, economic and social context of a society.1 These inequalities affect not only the health status of individuals and communities, but also the use of health care services. Several studies have demonstrated that health services utilization differs by social group. Primary care (PC) is mainly used by the lowest social classes, while secondary care (SC) utilization is more frequent in high classes.2,3 The existence of access barriers to health services, mainly in SC, such as increasing waiting times, lack of health literacy and reduction of available resources, are some of the main factors involved in the existence of a social gradient in the use of health care services.4,5

From 2008 to 2014 almost all European countries were affected by an economic recession.6 Public debt rose rapidly, as did unemployment.7 European governments implemented austerity policies in different areas, such as health care.6,8 This led to situations, such as having to decide in a family between spending on health care or food, prioritizing the latter.7,8 In Spain, one of the countries most affected by the Great Recession, austerity policies such as pharmaceutical co-payment or the closure of hospital beds, among others,7,9 were implemented to limit spending on health.7,9 This economic context had a direct and indirect impact on both health and the use of health systems, especially in some vulnerable groups10 such as ethnic minorities, children,11 the young and the elderly,2 among others.12 Now, it is widely asserted that the economic recession is over,13 but there remains doubt as to whether it has really ended or, on the contrary, the population has learned to live in a negative context.14 The population that became accustomed to living with fewer resources gave rise to the phenomenon of a group known as “the children of the crisis”. The economic status and health of these people are adversely affected throughout their lives, aspects which worsen due to a lack of related governmental policies.5,15

Several studies have been carried out to analyze the effect of the Great Recession on the population's health.16,17 There are also some studies that explore the impact of economic recession on the use of health services.7,12,18 In this context, the objective of our study is to analyze the evolution of inequalities in health care service utilization in Spain and the impact of the Great Recession on these inequalities.

We used a repeated cross-sectional study design. Data were obtained from Spanish National Health Surveys (NHS) conducted in 2001, 2003, 2006, 2011, and 2017 and Spanish European Health Interview Survey (EHIS) in 2014. The economic recession period is split into three, with three surveys before (2001,2003, 2006), two during (2011, 2014) and one post-recession period (2017). NHS and EHIS are representative surveys, with a stratified multistage design. They were performed by means of personal interview among a non-institutionalized population from 15 years old. Therefore, in order to homogenize the sample in this study, only those over 16 were included in the analysis. Seasonal effect was avoided by including autumnal months in the sample collection.19,20 The methodology applied allows comparability between surveys. Sample sizes ranged from 21,067 in 2001, 22,575 in 2003, 29,478 in 2006, 20,884 in 2011, 22,659 in 2014 and 22,903 in 2017, with a total sample size of 139,566 subjects. The 2009 EHIS was not included in this study because of the different collection methods of some key variables, such as social class and level of education which made it incomparable with the other surveys. More information about data sources can be consulted in the reference section.21

In Spain, from 2001 to 2017, there was a change in the use of health services. The increase observed in the pre-recession period led to a plateau during the recession and a decrease in the post-recession period, which was more marked in SC. This trend existed for both sexes, with a higher utilization in women and an increase in the gender gap. Socioeconomic differences in the use of PC and SC were observed for all the socioeconomic indicators used, and persisted even after adjusting for factors of need. These socioeconomic differences increased for SC during the recession period.

Spanish social structure improved for the period considered in terms of education. Nonetheless, unlike other surrounding countries,9 it has not been linked with a change in social class, which could be explained by different factors. First, Spain is one of The Organisation for Economic Co-operation and Development (OECD) countries where economic inequalities have increased more markedly in recent years, mainly because of unemployment and the lack of social policies to protect the most vulnerable groups.27 Other reasons may be related to the type of manufacturing work in Spain and the “brain drain” to other European countries with better job opportunities during the Great Recession. Because of that, the 2011 data reflected a negative migration balance with a significant migration rate amongst young Spanish workers with a high educational level.28

In relation to health care use patterns during the study period, our findings are consistent with the literature.11,29 The prevalence of health care utilization was higher in women, which has been justified by the social roles assumed by women and men. Women use health services not only to cover their own health needs, but also those of their family members. On the contrary, there are still men who consider that going to the doctor as negatively related to their masculinity.30,31 Authors29,32 have associated the increasing trend in health services utilization for various reasons. Firstly, the increase in SC may be due to changes in clinical practice. Lostao et al.33 related this with an increase in the rate of patient referral from general practitioners to specialist physicians. Secondly, this increase could be due to the worsening of health status and changes in the perception of health needs, mainly in the elderly. In this sense, some health indicators have worsened during this period, such as the deceleration of improvement in mortality rates since 2008, mainly in winter and in women, and increases in the prevalence of psychosocial stress.5

The plateau in the use of health services during the economic recession has also been found in other studies, and has been associated to an increased number of households that have given up or postponed medical care for financial reasons.12 In addition, other measures to control health care demand, such as an increase in waiting times, a reduction in service supply or in pharmaceutical co-payments34 could explain this trend.

There are socioeconomic inequalities in Spain in the use of healthcare services.10,29,34 The probability of PC use is higher for those in vulnerable socioeconomic conditions. Lostao et al.29,33 found that lower income categories used mainly PC, and Córdoba-Doña et al.35 observed a pro-low social class in the use of PC in Andalusia. These inequalities cannot be explained by health or sociodemographic characteristics. As we have observed for PC, after adjusting by the Andersen's Model, inequalities persist according to socioeconomic conditions, such as social class. This fact has been associated with different factors, such as low use of preventive services or low frequency of private health insurance, which forces disadvantage social classes to use PC.35 The characteristics of PC, in which access barriers are low and medical attention is supply on demand, could also impact the use of PC by disadvantaged people. With regard to the most advantaged groups, their high access to SC could be related with their better knowledge of the health system, as well as a higher use of private insurance.29 In contrast, the higher utilization observed amongst the unemployed for both sexes in the pre-recession period was lost during and after the recession. This reduction in use among low income groups has been attributed to a new pharmaceutical co-payment system as well as other measures. The pharmaceutical co-payment system led to difficulty for some in accessing certain medications that had been subsidized before the recession, and had been introduced due to difficulty in covering the cost of the previous arrangements.12,34,36

SC utilization was more frequent in people with positive socioeconomic indicators, and these inequalities increased during the recession period. People with medium and low educational level showed, during the economic recession, a lower probability of SC use than those with high education, even after adjusting for factors of need. In Spain, patients cannot use SC unless they are referred from PC.7 So, a deficit in health literacy amongst these educational groups could explain, at least partly, their lower utilization.6,29 Regarding social class, García-Subirats et al.12 reported that high social classes use SC with higher frequency than low social classes, as was observed in this study. This could be associated, also, to health literacy, which facilitates SC access to more advantaged social groups.

Finally, on both PC and SC, it is possible to observe a high probability of utilization by male pensioners during the recession that is not associated with factors of need. This result is consistent with other studies33,34 and has been related to a high frequency of health insurance plans in this group.

This study has some limitations associated with the use of health surveys as an information source, such as the inability to establish causality and the lack of temporality of the exposure-effect association. Health care use information was obtained from direct statements of those interviewed. However, health results obtained by survey are similar to those registered by electronic medical records.37 In addition, we could not use income level due to the high percentage of absent observations and the different criteria used among surveys. Instead of income, we have used three different socioeconomic indicators as proxies of socioeconomic level, which also provide different perspectives of the socioeconomic context. On the other hand, the surveys used are representative at a national level with a substantial sample available and a low number of absent values in the variables included (less than 5%).

The results of our study show that, in Spain, for the period 2001-2017, the use of healthcare services has increased, showing a plateau during the economic recession and a decrease when the recession ended. There are socioeconomic inequalities in the use of health services, with a higher use of PC by disadvantaged socioeconomic classes and a lower SC utilization. These inequalities increased for SC during the recession, and cannot be attributed to factors of need. Explaining the use of health care services based on socioeconomic indicators should be part of explanatory models in order to achieve health equity38 as well as to explore how these socioeconomic differences translate into different health results. Socioeconomic inequalities must be taken into account in order to control the effect of the subject's economic context on our health systems.

María-Victoria Zunzunegui.

The corresponding author on behalf of the other authors guarantee the accuracy, transparency and honesty of the data and information contained in the study, that no relevant information has been omitted and that all discrepancies between authors have been adequately resolved and described.

R. Sánchez-Recio, J.P. Alonso-Pérez de Ágreda and I. Aguilar Palacio conceived and designed the study. R. Sánchez-Recio, J.P. Alonso-Pérez de Ágreda and I. Aguilar Palacio participated in the study, obtaining the data, interpreting the results, and writing and critically reviewing the manuscript. All signatories have approved the final version of the text.

This study was carried out with the financial support of the Grupo de Investigación en Servicios Sanitarios de Aragón (GRISSA) (B09-17R). This research group is funded by Gobierno de Aragón and Fondo Europeo de Desarrollo Regional (FEDER).

None.