CONSUMERS’ PERCEPTIONS OF HEALTH AND FACTORS INFLUENCING FULFILMENT OF THE NEED FOR HEALTHCARE IN EU COUNTRIES

19 3, XXV, 2022 Economics 10.15240/tul/001/2022-3-002 CONSUMERS’ PERCEPTIONS OF HEALTH AND FACTORS INFLUENCING FULFILMENT OF THE NEED FOR HEALTHCARE IN EU COUNTRIES Irena Antošová1, Naďa Hazuchová2, Jana Stávková3 1 Mendel University in Brno, Faculty of Business and Economics, Department of Marketing and Trade, Czech Republic, ORCID: 0000-0002-4331-4187, [email protected]; 2 Mendel University in Brno, Faculty of Business and Economics, Department of Marketing and Trade, Czech Republic, ORCID: 0000-0002-5693-9872, [email protected]; 3 Mendel University in Brno, Faculty of Business and Economics, Department of Marketing and Trade, Czech Republic, ORCID: 0000-0002-0889-0218, [email protected]. Abstract: The paper deals with subjective perceptions of health by individuals. The research aimed at understanding socioeconomic and demographic factors influencing the fulfilment of healthcare needs and at finding out categories of factors that lead to the highest chances of meeting the need in consumer segments formed according to perceptions of their health status. The analyses were based on the EU-SILC database of primary data on the income situation and living conditions of households. In 2017, the database included extra questions on health. The method of cluster analysis was employed. As a result, three clusters of individuals representing EU countries formed depending on the perceived state of health – the authors named the clusters ‘optimistic’, ‘neutral’, and ‘pessimistic’. For each segment, the binary logistic regression was applied to determine categories of factors leading to the highest probability of meeting the healthcare need. The greatest influence over the fulfilment of the need for healthcare has been confirmed for the factor “Sector of economic activity”, followed by the type of economic activity. Some differences were revealed between segments. For example in the third segment, i.e., respondents who rated worst their health, a strong influence of education has been identified. The highest chances of meeting the need for health care are achieved in the first segment by executives, but in the second and the third segment by individuals active in education. On the other hand, craftsmen and workers have the lowest chances. In all segments, the influence of household composition was confirmed, with single households and single-parent households reporting lower chances of meeting their healthcare needs. Respondents who did not feel their healthcare need was met mostly said it was due to financial reasons, long waiting times, or fear of medical treatment. Keywords: Health, need for healthcare, consumer behaviour, income, household. JEL Classification: I31, P46. APA Style Citation: Antošová, I., Hazuchová, N., & Stávková, J. (2022). Consumers’ Perceptions of Health and Factors Influencing Fulfilment of the Need for Healthcare in EU Countries. E&M Economics and Management, 25(3), 19–34. https://doi.org/10.15240/ tul/001/2022-3-002 Introduction The health is defined as a state of a person’s physical, mental, and social well-being. Responsibility for health is determined not only by the healthcare system and genetic predispositions of individuals but also by one’s lifestyle and approach to achieving and keeping a good state of health (World Health EM_3_2022.indd 19 15.9.2022 13:48:54 20 2022, XXV, 3 Economics Organization, 2006). Consumer behaviour concerning healthcare differs from other areas, above all because it is “a question of life and death”. Therefore, this type of decision-making tends to get significantly affected by emotions (Cazacu, 2015). Another significant difference is that consumers get healthcare products and services through a third party, most often a physician, who recommends steps to be taken and makes the decisions (Radulescu et al., 2012). Kenkel (1990) states that physicians can create or reduce demand for their services. Meeting health care needs is not always a matter of consumer choice, but other factors also play a role. The main goal of the paper is to reveal socioeconomic and demographic factors influencing the fulfilment of EU consumers’ healthcare needs and to find out categories of factors that lead to the highest probability of meeting the need in consumer segments formed according to perceptions of their health status. How an individual’s health is perceived and, most importantly, whether healthcare needs are met when they occur have been the basic research questions of the paper. To learn about subjective views on health, the authors used the EU-SILC survey. The survey provides data on subjective perceptions of health as such, as well as information about meeting the need for healthcare and possible reasons for not meeting this need. The results of the analyses may represent a strong argument for implementing improvements in the healthcare systems. This means in particular improving access to healthcare services for the majority of consumers. 1. Theoretical Background People strive to meet their healthcare needs under the conditions set by the healthcare system and the financial resources they have available. The attitude of a household to their health and the use of healthcare services affects the household’s living standard (Callander et al., 2019). Khan and Ul Husnain (2019) demonstrated that healthcare expenditure and income are co-integrated and, therefore, there is a link between the standard of living, income situation, and health standard. Lenhart (2019) examined the effect of income on the state of health and found that higher income increased the chances of excellent or very good health being reported by households’ heads. The increase observed here ranged from 6.9 to 8.9 percentage points. According to Knaul et al. (2012), low-income households living near the risk-of-poverty threshold spent more on healthcare. It means their healthcare expenses accounted for a higher part of their disposable income. However, in absolute numbers, they could afford fewer healthcare services than households in higher-income categories. According to Blumberg et al. (2014), healthrelated expenditures are rising faster than incomes, both at the national and household levels. Shares of households’ disposable incomes spent on healthcare are increasing. In countries where parts of the population have no health insurance, the financial demands of healthcare could lead to personal bankruptcies. The subjective health in Central and Eastern European countries is influenced by a complex mix of determinants (Borisova, 2019). Differences in individuals’ socioeconomic statuses (stemming from different economic activities, education, or income categories) can contribute to health inequalities and to chances to meet the healthcare needs. Individuals with higher economic statuses are more influenced by behavioural and psychological factors in their approach to health than those with lower socioeconomic statuses (Atkinson & Marlier, 2010; Peretti-Watel et al., 2016). Socioeconomic status affects health-related quality of life (Puciato et al., 2020). Self-perceived health is influenced by income and labour status and by demographic factors such as gender or age in EU countries (Jindrová & Labudová, 2020). Chaupain-Guillot and Guillot (2015) state that demographic factors are other factors that influence consumers’ access to healthcare. Gender is one of the factors affecting approach to health-related questions (Socías et al., 2016; Roy & Chaudhuri, 2008). The results of the study by Roy and Chaudhuri (2008) showed that women tended to rate their health worse and used fewer health services – reportedly because of the lower socioeconomic status of women. Sonik et al. (2020) added that discrimination against women in access to healthcare was not necessarily the reason. Women and men simply often had different preferences as far as medical treatment is concerned. Next to gender, age is another significant factor, with preventive and aesthetic motives for medical treatment prevailing at younger ages. Another significant factor co-determining the EM_3_2022.indd 20 15.9.2022 13:48:55 21 3, XXV, 2022 Economics average number of doctor’s visits is education (Hoeck et al., 2011). Puciato et al. (2020) see education as important factor affecting an individual’s level of perceived health. According to Czibere et al. (2019) the level of the highest attained education influence the health status indirectly. They proved that the education have significant impact only on the age when an illness begun. Puciato et al. (2020) talks about a marital status as a determinant of health conditions. The marital status is closely related to the household composition. Radulescu et al. (2012) explain that family members and also friends can influence an approach of the individual to the health. Gender, age and other demographic and socioeconomic factors also affect an approach of individuals to health risky behaviour (Morkevičius et al., 2020; Kim et al., 2018). Kunzová and Hrubá (2013) point out that health is correlated with many factors and also with lifestyle. Failure to maintain a healthy lifestyle put consumers furtively at risk of ill health (Mlčochová & Papežová, 2012). The need for healthcare or medical treatment may not be met due to a variety of reasons. Kenkel (1990) explained the relationship between healthcare and individuals’ level of knowledge and available information. According to this author, poorly informed consumers tended to underestimate the importance of healthcare. Schmid (2015), on the other hand, found that information had a negative effect on the use of healthcare services. This was supposedly related to fears of being examined and diagnosed due to which people did not seek necessary medical care. According to Fiorillo (2020) and Popovic et al. (2017), the most common reasons for not seeking medical care were financial and time constraints and the distance to health facilities (in connection with the ‘wait-and-see’ approach used by the staff of medical facilities). It must be taken into account that meeting health-related needs has a strongly individual dimension and is not a matter of course for all individuals. Satisfaction rates are not the same for everyone under identical conditions (Banthin et al., 2008). There are plenty of objective indicators that tell us about the availability and quality of healthcare in individual countries. These objective data speak of the healthcare system as a whole in terms of its quality, new methods, and achieved results. However, the objective data do not address how healthcare services provided are perceived by individuals, whether healthcare is available at the time and quality needed, nor what are the reasons for any failure to meet the need for healthcare. The information on how individuals subjectively perceive health and healthcare services are of utmost importance for any responsible national healthcare system – hence the value of subjective variables in analyses in this area (Schokkaert et al., 2017). As explained by Borisova (2019), both subjective and objective indicators of health should be used wherever possible because they often interact with each other. Health policies should adopt a multidimensional approach and develop incentives to remove barriers that limit consumer access to health services (Popovic et al., 2017). 2. Research Methodology To learn about the behaviour of individuals in relation to their state of health, the authors used data obtained within the EU-SILC survey (European Union – Statistics on Income and Living Conditions), specifically, the EU-SILC 2017. In addition, the extensive EU-SILC microdata set provided detailed information on the income situations of households and individuals. The data also allowed for the identification of households and individuals in terms of various demographic and socioeconomic factors, as well as a description of households’ and individuals’ living conditions in different areas of life. The EU-SILC survey is mandatory in all EU countries and follows a uniform methodology published by Eurostat (Eurostat, 2019). Eurostat also publishes a uniform methodology for further processing of the results. In 2017, EU-SILC was conducted in a total of 256,468 European households and had a total of 515,880 individual respondents (this is the number of cases analysed in this paper). The EU-SILC microdata database originally included 7 indicators describing subjective perceptions of respondents concerning the need and the availability of healthcare services. The database has been extended in 2017 by an ad-hoc module of another 7 indicators describing the financial demandingness of healthcare, as perceived subjectively by households. This means, for example, the cost of medicines and dental care, or the number of EM_3_2022.indd 21 15.9.2022 13:48:55 22 2022, XXV, 3 Economics visits to medical specialists. The EU-SILC data contain a conversion factor which is used as a weight in the conversion of the sample data to the base population (i.e., the whole population of the country and the whole EU). A five-point scale (1 – very good state of health; 2 – good; 3 – fair; 4 – poor; 5 – very poor) was used for subjective state of health assessments. The authors used cluster analysis to identify segments of EU citizens that showed similarities in subjective perceptions of the state of health. Subjective assessment of health evaluated by consumers is the variable applied in the cluster analysis. The clusters are formed according to the proportion of individuals among respondents in each country who rate their health as very good, good, fair, poor and very poor. The goal of cluster analysis is to classify objects into a certain number of clusters. Objects within a cluster are similar to the greatest extent possible and objects within a cluster are the least possibly similar to objects from other clusters. Individual objects are gradually grouped into smaller clusters and these clusters are then merged to form larger clusters (Meloun & Militký, 2012). The authors used the K-means algorithm which identifies homogeneous groups of research objects based on selected characteristics. For each of the initial clusters, the authors determined the centroid value (centroid is a vector of the average values of each variable). Objects were assigned to clusters based on the centroid to which the object was closest. The optimal number of clusters is verified by applying ANOVA analysis showing significant difference between clusters. According to Hebák et al. (2015), K-means algorithm is an iterative procedure that minimizes the function of the following formula (1): , (1) where the uih ∈ {0,1} elements indicate whether the i-th object belongs (value 1) or does not belong (value 0) to the h-th cluster and is a vector of average values of the h-th cluster. The conditions of the following formula (2) must be met: (2) The chances of meeting the need for healthcare with respect to different categories of demographic and socioeconomic factors have been assessed by logistic regression analysis. The explained variable could take two values: unmet need for healthcare (0) and met need for healthcare (1). The following factors were used as explanatory variables: gender, education, economic status, sector of economic activity, and household income group. The binary logistic regression model can be expressed by the formula (3) showing the relationship between the probability of a phenomenon P(x) (Y = 1), i.e., meeting the need for healthcare, under conditions given by the values of the independent variables (x): . (3) The ln (P/(1 – P)) formula (called the logit of P), can be expressed as a weighted sum of the values of the independent variables. The logit of P is the logarithm of the probability of occurrence of the phenomenon under study. The model can be also expressed by the following formula (4): , (4) where the parameter estimates βi are obtained from the measurement matrix of x. If βi is equal to zero, then the parameter has no effect on the observed phenomenon (Hendl, 2006). The quality of the binary regression model is assessed by the Nagelkerke R-squared indicator, the significance of the model is verified by the Hosmer and Lemeshow test. The VIF indicator is used to verify a presence of multicollinearity in models. The VIF values higher than 10 indicates multicollinearity in the model (Hebák et al., 2015). The EU-SILC data have been processed by the IBM SPSS Statistics software. The algorithm of cluster analysis and the binary logistic regression have also been implemented in the SPSS software. 3. Research Results The authors took the opportunity to analyse data from the EU-SILC survey conducted in 2017. In that year, the survey was extended by an ad hoc module aimed at healthcare. The respondents commented on how they subjectively perceived EM_3_2022.indd 22 15.9.2022 13:48:56 23 3, XXV, 2022 Economics their states of health and whether their healthcare needs were met. If a respondent said their need for healthcare was not met, they were asked to give the reasons. The results of the survey provide important information on health-related behaviour of people and, given the representativeness of the population, are very useful for the implementation of corrective measures in the health sector. Given the size of the survey sample (covering 27 countries, i.e., about 515 thousand EU respondents and dozens of content questions), this paper could not cover all the values included in the survey, instead, the authors focused on typical and extreme values only. 3.1 Individual Perceptions of the State of Health The results of the subjective assessments of the state of health showed that there were countries where almost 50% of respondents rated their states of health as ‘very good’ (for example Cyprus and Greece). In most countries, a major part of respondents evaluated their states of health by the grade of ‘2’, i.e., ‘good’ (reported by about 50% of respondents), or grade ‘3’ – ‘fair’ (20–30% of respondents). However, there were countries where some respondents (up to 10% or in the order of tens of %) rated their states of health as ‘very poor’ or ‘poor’. The highest frequencies of negative health evaluations were found in the following countries (Tab. 1). In countries with negative health ratings (see Tab. 1), respondents also more frequently reported issues related to long-term illnesses that limited their everyday activities. In other EU countries (those not listed in Tab. 1), 2% or fewer respondents assessed their states of health as very poor. To provide an overall overview and summarize the subjective perception of the state of health in all EU countries, the authors employed cluster analysis and the K-means algorithm. As a result, three clusters of individuals were identified based on the perceived statuses of health. The proportions of individuals evaluating their health status as very good, good, fair, poor and very poor enabled the formation of three segments and sorted countries into segments according assessments by residents’ representatives (Tab. 2). Croatia Portugal Hungary Latvia Lithuania Poland Bulgaria Very poor state of health 3.9% 3.6% 3.2% 3.1% 3.1% 2.7% 2.5% Poor state of health 14.0% 11.0% 9.0% 13.8% 13.0% 10.0% 8.0% Source: EU-SILC microdata (Eurostat, 2021), own using IBM SPSS Statistics Tab. 1: EU countries with the highest proportions of individuals perceiving negatively their states of health Cluster 1 ‘optimistic’ Cluster 2 ‘neutral’ Cluster 3 ‘pessimistic’ EU countries in the cluster Austria, Cyprus, Greece, Croatia, Ireland Belgium, Bulgaria, Germany, Denmark, Spain, Finland, France, Italy, Luxembourg, Malta, Netherlands, Romania, Sweden, Slovakia Czech Republic, Estonia, Hungary, Lithuania, Latvia, Poland, Portugal, Slovenia Very good SH 40% 23% 13% Good SH 33% 48% 41% Fair SH 18% 21% 32% Poor SH 7% 6% 11% Very poor SH 2% 2% 3% Source: EU-SILC microdata (Eurostat, 2021), own using IBM SPSS Statistics Tab. 2: Subjective assessments of health (SH) in EU countries EM_3_2022.indd 23 15.9.2022 13:48:56 24 2022, XXV, 3 Economics Subsequently, ANOVA analysis confirmed the correct number of clusters identified (Tab. 3). Significance values are below the significance level α = 0.05. Clusters are significantly different. Also, there was a zero change according to the iteration history after three iterations during K-means algorithm process. If more cluster were formed, the difference between clusters was not confirmed (significance values were above the significance level). The K-means algorithm assigned individuals from five countries to the first cluster, of which almost three quarters rated their health as good and 40% as very good. Due to the positive health assessments, the cluster has been named as ‘optimistic’. In the second group, about half of the respondents rated their health as good. The second segment included the largest number of EU countries compared to the other segments. The third group has been more pessimistic about their health, with a higher number of respondents rating their health as fair. On average, 14% of respondents in this group evaluated their health as poor. Respondents from countries with more negative ratings were more likely to report problems related to long-term illness or health limitations. The average share of a country’s population reporting limitations in everyday activities due to poor health have amounted to units of per cent. Yet, there were countries where people did not perceive such limitations at all (Spain, Ireland, Malta, and Sweden). However, when drawing these conclusions, we need to take into account whether the conditions created by the state are so satisfactory that people can lead active lives without limitations, or whether the reported opinions were shaped by low awareness of the possibilities of improving living conditions. 3.2 PerceivedFulfilmentofHealthcare Needs When asked whether the medical assistance requested was actually received, there were countries where almost 100% of respondents answered positively. These were, for example, Spain, Austria, Malta, and Luxembourg. In some countries, on the other hand, significant amounts of respondents answered negatively, i.e., that they did not receive the treatment they needed. In Greece, for example, 25% of respondents gave negative answers, in Estonia, it was 13% of respondents, in Poland 12%, and in Latvia 10%. In other EU countries, unmet healthcare needs were reported by up to 10% of respondents. The authors used binary logistic regression to find out which factors influenced the fulfilment of the need for healthcare and which categories of demographic and socioeconomic factors increased the chances of the fulfilment of the need. The explained variable in the model has been the fulfilment of the need for healthcare. The variable could take two values: 0 indicating no satisfaction of the need (failure to meet the need for healthcare); and 1 indicating satisfaction of the need. The explanatory variables entering the regression model were Gender, Education, Household composition, Economic activity, Income quintile based on the household’s disposable income, and Sector of economic activity based on ISCO (International Standard Classification of Occupations). The authors have calculated the binary logistic regression for all three segments (created based on the subjective assessments of health by the respondents – see Tab. 2). This allowed for explanations of the results of binary logistic regressions in relation to optimistic and Cluster Error F Sig. Mean square df Mean square df Very good SH 1,124.076 236.397 24 30.884 0.000 Good SH 428.770 222.651 24 18.930 0.000 Fair SH 415.434 215.236 24 27.267 0.000 Poor SH 66.515 25.447 24 12.211 0.000 Very poor SH 3.953 20.626 24 6.315 0.006 Source: EU-SILC microdata (Eurostat, 2021), own using IBM SPSS Statistics Tab. 3: ANOVA in the cluster analysis EM_3_2022.indd 24 15.9.2022 13:48:57 25 3, XXV, 2022 Economics pessimistic assessments of health by individual respondents. Before interpreting the model, the presence of multicollinearity in three models for all segments was verified. The linear regression procedure with same predictors was used for this purpose and collinearity diagnostics were requested. All values of VIF indicators are below the value 10 (Tab. 4). Multicollinearity is not present in the models, as indicated by the low values of the Condition indexes implemented in the IBM SPSS Statistics. In all three logistic regressions, Hosmer and Lemeshow tests were used to prove the significance of the models (the resulting p-values had to be lower than 0.05). The Nagelkerke R-squared indicators proved the quality of the models as 83% of the variability in the dependent variable was explained for the first segment (Tab. 5), for the second model (Tab. 6) it was 89%, and for the third model (Tab. 7) it was 73% of the variability of the dependent variable. Categories with the highest chances of meeting the health for healthcare have been highlighted in bold in the tables. The results of the binary logistic regression for the first segment (Tab. 5) showed that all the explained variables influenced the fulfilment of the need for healthcare. The strongest influence has been identified in the ‘Sector’ variable, Segment 1 VIF Segment 2 VIF Segment 3 VIF Gender 9.446 8.677 9.347 Household composition 4.994 4.733 5.349 Education 9.654 9.847 9.822 Economic activity 3.549 3.585 3.705 Quintiles 5.791 6.041 6.184 Sector 4.496 5.096 5.435 Source: EU-SILC microdata (Eurostat, 2021), own using IBM SPSS Statistics Tab. 4: Collinearity statistics Estimate B Standard deviation Wald df Sig. Exp(B) Gender (males) 0.074 0.003 863.334 1 0.000 1.077 Household composition (other)a 30,666.113 4 0.000 Household composition (single) 0.024 0.003 48.946 1 0.000 1.024 Household composition (two adults) 0.433 0.003 20,608.297 1 0.000 1.542 Household composition (single parent) 0.228 0.009 692.013 1 0.000 1.256 Household composition (two adults and children) 0.378 0.004 10,978.219 1 0.000 1.460 Education (university)a 138,956.161 20.000 Education (basic) −0.382 0.004 11,072.918 1 0.000 0.682 Education (secondary/high school) 0.681 0.003 52,078.088 1 0.000 1.976 Tab. 5: Chances of meeting the need for healthcare for segment 1 – ‘optimistic’ – Part 1 EM_3_2022.indd 25 15.9.2022 13:48:57 26 2022, XXV, 3 Economics where legislators and executives were 11 times more likely to have their healthcare needs met than craftsmen and workers. The second most important variable in terms of influence significance was Economic activity, where employees were found to have the highest chances of the fulfilment of their need for healthcare. Similarly, the results of the logistic regression of the second segment data (Tab. 6) proved the significance of most of the factor categories except for the category of agriculture in the Sector variable and the category of single parents in the Household composition variable. The most significant factor in terms Estimate B Standard deviation Wald df Sig. Exp(B) Economic activity (other)a 152,401.000 4 0.000 Economic activity (employed) 1.225 0.004 115,136.806 1 0.000 3.403 Economic activity (self-employed) 0.510 0.005 11,126.210 1 0.000 1.666 Economic activity (unemployed) −0.035 0.004 66.343 1 0.000 0.965 Economic activity (old-age pensioner) 0.669 0.003 36,924.689 1 0.000 1.952 Quintiles (fifth)a 27,200.371 4 0.000 Quintiles (first) 0.467 0.004 16,704.470 1 0.000 1.595 Quintiles (second) 0.460 0.004 17,002.700 1 0.000 1.584 Quintiles (third) 0.411 0.004 13,717.567 1 0.000 1.508 Quintiles (fourth) 0.436 0.004 14,494.801 1 0.000 1.546 Sector (craftsmen and workers)a 358,990.333 10 0.000 Sector (legislators and executives) 2.430 0.006 158,988.056 1 0.000 11.354 Sector (science and technology) 1.892 0.007 68,520.486 1 0.000 6.630 Sector (healthcare) 1.816 0.008 46,283.147 1 0.000 6.147 Sector (education and training) 2.118 0.008 67,896.034 1 0.000 8.311 Sector (public administration) 2.051 0.007 83,857.200 1 0.000 7.777 Sector (information technology) 1.840 0.012 23,631.131 1 0.000 6.293 Sector (law, culture, sport) 1.445 0.008 33,479.194 1 0.000 4.243 Sector (officials) 1.385 0.005 79,697.080 1 0.000 3.995 Sector (services and sales) 0.986 0.003 95,813.287 1 0.000 2.681 Sector (agriculture, forestry, fishing) 0.488 0.003 20,737.801 1 0.000 1.630 Source: EU-SILC microdata (Eurostat, 2021), own using IBM SPSS Statistics Note: a This parameter has been set to zero because it is redundant. Tab. 5: Chances of meeting the need for healthcare for segment 1 – ‘optimistic’ – Part 2 EM_3_2022.indd 26 15.9.2022 13:48:58 27 3, XXV, 2022 Economics of increasing the likelihood of fulfilling the need for healthcare has been the Sector variable again, where the employees in the healthcare sector, the education and training sector had the highest chances of having their healthcare needs met. The chances of both categories were almost 7 times higher compared to craftsmen and workers. The results have also shown that people with primary education had the highest chances of the fulfilment of their healthcare needs (even three times higher compared to university graduates). This finding may be related to the fact that primary education (as the highest level of education attained) was reported largely by elderly respondents who were no longer economically active and had sufficient time for healthcare. Actually, time constraints were one of the main reasons for not fulfilling the need for healthcare. It is worth noting that the lowest chances were identified in the groups of single mothers and single households (the Household composition variable), in the first segment (Tab. 5). Estimate B Standard deviation Wald df Sig. Exp(B) Gender (males) 0.452 0.001 215,965.723 1 0.000 1.571 Household composition (other)a 462,436.433 4 0.000 Household composition (single) 0.132 0.001 11,210.770 1 0.000 1.141 Household composition (two adults) 0.620 0.001 262,190.359 1 0.000 1.859 Household composition (single parent) 0.003 0.003 1.207 1 0.272 1.003 Household composition (two adults and children) 0.690 0.001 255,848.438 1 0.000 1.994 Education (university)a 868,163.355 20.000 Education (basic) 1.194 0.002 542,042.619 1 0.000 3.299 Education (secondary/high school) 0.929 0.001 755,609.824 1 0.000 2.531 Economic activity (other)a 1,174,238.320 4 0.000 Economic activity (employed) 1.204 0.001 1,026,477.356 1 0.000 3.333 Economic activity (self-employed) 1.075 0.002 287,131.505 1 0.000 2.931 Economic activity (unemployed) 0.546 0.002 102,352.112 1 0.000 1.726 Economic activity (old-age pensioner) 1.009 0.001 588,650.365 1 0.000 2.744 Quintiles (fifth)a 320,935.300 4 0.000 Quintiles (first) 0.056 0.001 1,740.229 1 0.000 1.057 Quintiles (second) 0.361 0.001 69,696.199 1 0.000 1.434 Quintiles (third) 0.615 0.001 186,094.038 1 0.000 1.850 Quintiles (fourth) 0.554 0.001 150,418.704 1 0.000 1.741 Tab. 6: Chances of meeting the need for healthcare for segment 2 – ‘neutral’ – Part 1 EM_3_2022.indd 27 15.9.2022 13:48:59 34 2022, XXV, 3 Economics Status of the Unemployed. E&M Economics and Management, 23(3), 23–37. https://doi. org/10.15240/tul/001/2020-3-002 Radulescu, V., Cetina, I., & Orzan, G. (2012). Key Factors that Influence Behavior of Health Care Consumer, the Basis of Health Care Strategies. Contemporary Readings in Law, 4(2), 992–1001. Roy, K., & Chaudhuri, A. (2008). Influence of socioeconomic status, wealth and financial empowerment on gender differences in health and health care utilization in later life: evidence from India. Social Science & Medicine, 66(9), 1591–1962. https://doi.org/10.1016/j. socscimed.2008.01.015 Schmid, C. (2015). Consumer Health Information and the Demand for Physician Visits. Health Economics, 24(12), 1619–1631. Schokkaert, E., Steel, J., & Van de Voorde, C. (2017). Out-of-Pocket Payments and Subjective Unmet Need of Healthcare. Applied Health Economics and Health Policy, 15(5), 545–555. https://doi.org/10.1007/s40258-0170331-0 Socías, M. E., Koehoorn, M., & Shoveller, J. (2016). Gender Inequalities in Access to Health Care among Adults Livingin British Columbia, Canada. Women’s Health Issues, 26(1), 74–79. https://doi.org/10.1016/j.whi.2015.08.001 Sonik, R. A., Creedon, T. B., Progovac, A. M., Carson, N., Delman, J., Delman, D., & Cook, B. L. (2020). Depression treatment preferences by race/ethnicity and gender and associations between past healthcare discrimination experiences and present preferences in a nationally representative sample. Social Science & Medicine, 253, 112939. https://doi. org/10.1016/j.socscimed.2020.112939 World Health Organization. (2006). Constitution of the World Health Organization. Retrieved March 16, 2022, from https://www. who.int/governance/eb/who_constitution_en.pdf EM_3_2022.indd 34 15.9.2022 13:49:01