Citizens' Views on Health Insurance in Croatia

43(4):417-424,2002

PUBLIC HEALTH

Citizens’ Views on Health Insurance in Croatia

Miroslav Mastilica, Sanja Babiæ-Bosanac Andrija Štampar School of Public Health, Zagreb University School of Medicine, Zagreb, Croatia

Aim. To examine the citizens’ attitudes toward health insurance and its reform in Croatia, and their views on private payments for health care services. Method. In 1999 and 2000, we surveyed 500 randomly selected adults from all regions of Croatia, aged 40 years and over. The questionnaire included questions on social health insurance, private payments for health care, and background information. The net response was 393 (79%). The analysis of the data collected included univariate and multivariate analyses to test the differences in the attitudes among sociodemographic and socioeconomic groups. Results. Most interviewed Croatian citizens (83.2%) expressed the opinion that everybody should have access to health care services, irrespective of the health insurance contributions. However, 32.1% agreed that the utilization of services should depend on the payment of contribution; 39.1% believed that the money they contributed to health insurance corresponded to health care services they received; 60.1% agreed that insurance rate should increase propor- tionally to income. When asked about reforms, more than half (53.4%) thought that the current health insurance covered less benefits than 10 years earlier, whereas more than a third believed that changes offered more choice (36.9%) but less equity (37.7%), and 46% disagreed with the introduction of the basic package of health care benefits and supplementary insurance. About the same percentage of respondents thought that they had already been paying too much for health care out of their own pockets. Conclusion. Citizens in Croatia do not hold a positive opinion on health insurance reform. They fear the changes would bring about limitations in their social rights and increase their financial burden.

Key words: Croatia; financing, government; health care reform; insurance, health; public opinion

Over the last 10 years, social health insurance in 2000. In that situation, only 33% of the population Croatia has been continually changed, arousing fur- paid insurance contributions or, in other words, one ther public, professional, and political debates on the insured person contributing to insurance fund cov- need for a more radical approach. Eventually, a new ered the insurance for himself of herself and two other strategy for the reform of the health care system and persons who did not pay contributions. health insurance was adopted (1). In the last decade, the health care reform in The government decided to change the social Croatia, as elsewhere in Europe (3-5), was very much health insurance primarily because of the increasing a response to the issue of how to contain health care health care expenditure and constant lack of re- costs in a situation of scarce resources (6-9). A num- sources for financing health care services. Health care ber of measures aimed at cost containment were in- expenditure (expenditure of the national health insur- troduced in the Croatian health care system over the ance fund) increased from 5.6 billion Croatian Kunas last 10 years, such as rationing of services, limitation (HRK) or US$1.1 billion (US$1=HRK5.2) in 1994 to of services provided, penalties for excessive over-pre- HRK14.0 billion (or US$2.69 billion, calculated in scribing or referrals, limited positive drug list, reduc- 1994 US$) in 2000, or 2.5 times in five years (2). tions in health care budgets, and increase in co-pay- According to the Croatian Institute for Health In- ments. However, they had limited success and gained surance, the only national health insurance fund, the little acceptance from both providers and the public main reason for the current deficit are the insufficient (10-13). contributions to the Fund. The number of insured em- Macro financial goals of health care reform in ployees (those who pay contributions) is decreasing Croatia suppressed the issues of the quality of health and the number of those who do not contribute to in- care services and their response to the needs and ex- surance (dependants, retired, unemployed, and so- pectations of consumers. The reform measures were cially vulnerable) is increasing. In 1990, there were mostly directed to the supply side (providers), and the 38 retired per 100 employed, and in 1999 there were demand side measures were not much pronounced, 74 retired per 100 employed persons (2). Unemploy- although they were directly or indirectly affecting the ment rate increased to 22% of active population in consumers. Thus, the growing reduction of public

www.cmj.hr 417 Mastilica and Babiæ-Bosanac: Citizens‘ Views on Health Insurance in Croatia Croat Med J 2002;43:417-424 health care resources under the reforming health care household (address) from the initial address by standard random system, privatization and marketization of services, procedure. In each household, the respondent was drawn at random. The selected adults were face-to-face interviewed in their and increased cost sharing have been shifting a pro- homes by trained interviewers (medical students). The interviews portion of health care costs on to the health care us- took 15 minutes on average to be completed. The net response ers, with a significantly negative impact on low in- rate was 79%. Major reasons for non-response were absence come patients (14,15). from home and refusal to be interviewed due to lack of time. The new changes in the health insurance system The final sample consisted of 393 respondents, 39.2% men and 60.8% women, with the mean age of 51 years (±9.8 SD). in Croatia, primarily those aimed to restrict patients’ Most respondents (81.8%) were younger than 60 years, and rights and benefits and to increase private payments, 18.2% were aged 60 years and over. Majority were married are showing serious impact on the users of health care (76.5%), with a family of four persons on average. The average services. For that reason, it is important to analyze the education level of the respondents was secondary school. The views of the public on health insurance, to see how education distribution showed that 54.7% of the respondents had elementary or secondary education, and 45.3% higher educa- the new reform measures correspond to the needs tion. Most were employed (66.2%), and 25.2% were retired. The and expectations of the citizens. occupational status distribution showed that 30.2% were agricul- Our aim was to examine the citizens’ attitudes tural workers (farmers), unskilled, and skilled workers; 39.2% were routine non-manual employees in administration and com- toward reforms in health care and health insurance in merce; and 30.5% were professionals. The average family in- Croatia. Specifically, we wanted to assess the public come of the respondents was HRK4,000 (or US$500). To analyze opinion on health insurance contributions, changes the income group differences, the sample was divided into two in the health insurance, and the out-of-pocket pay- income groups. The low-income group comprised those with ments for health care services. We assumed that the family income of HRK5,000 (US$625) or less (49.6% of the respondents), and the medium-income group had income of people in Croatia, having enjoyed high and unlimited HRK5,001 or higher (50.4% of the respondents). Most respon- benefits of social health insurance for a long time and dents lived in urban areas (57.3%), 23.4% lived in semi-urban, having traditionally a strong perception of health care and 19.3% in rural areas. as a social good, “something you get for free”, did not The results on the citizens’ views on health insurance con- have positive attitudes towards the ongoing reforms, tributions, reforms, and out-of-pocket payments for health care which reduce public financing and health care bene- were analyzed by sex, age, education, income, occupation, and place of residence. To determine the statistical significance of the fits and significantly increase private payments for differences in the sociodemographic and socioeconomic categor- health care services and pharmaceuticals. We also ical groups, a chi-square test was used. analyzed views of different demographic and socio- To analyze the latent dimensions in the citizens’ views, the economic groups on health insurance contributions, principal component model of factor analysis with a VARIMAX health insurance reforms, and private payments for rotation was used. For this purpose, only the respondents who health care, assuming that differences in their opin- expressed positive opinion (“completely agree” or “mostly ions would be significant. agree”) or negative opinion (“mostly disagree” or “completely disagree”) to the given statements were included in the analysis. The answers to the statements included in the factor analysis Subjects and Methods ranged from “complete agreement” to “complete disagreement”. Differences between sex, age, education, income, occupation, Data were collected through a sample survey carried out in and place of residence were analyzed by comparing the mean 1999 and 2000 in all regions of Croatia. The questionnaire ap- values of the factor scores, using t-tests or analysis of variance plied in the study included originally constructed questions about (ANOVA). The statistical analysis was performed with the citizens’ views on compulsory health insurance and its reform. STATISTICA data analysis software system (Version 6; StatSoft Inc., It also included questions on direct payments for health care ser- Tulsa, OK, USA). vices and socio-demographic characteristics of the person surveyed. The respondents were asked to express their agreement or disagree- Results ment with the list of statements. The answers were based on a four- degree scale: “completely agree”, “mostly agree”, “mostly dis- Opinions on Contributions for Health Insurance agree”, and “completely disagree”. The multistage sample was composed of 500 adults aged To analyze the citizens’ views on health insur- 40 and over, randomly selected from households in all regions ance, the respondents were asked questions related and from all types of residential areas (metropolitan, ie, Zagreb; to their knowledge on and attitudes toward health in- urban; semi-urban; and rural). As the first step, in each of the five surance contributions (Table 1). Great majority (83.2 main regions in Croatia (North-West, North-East, South-West, %) of the respondents completely or mostly agreed South-East, and Zagreb) a number of sampling places was drawn at random from the lists of the places of residence. In each of 69 that everybody should have the right to health care ir- selected sampling places, a starting household was selected at respective of contributions. The low income group random and further households were selected as every Nth significantly more frequently (59.7%) than the high

Table 1. Croatian citizens’ attitudes on their contributions for health insurance Response (%) completely mostly mostly completely don't know/ How do you feel about the following statement: agree agree disagree disagree no answer Everybody should have the right to health care irrespective of contribution 53.2 30.0 9.9 2.5 4.4 to health insurance Utilization of health care services should depend on health insurance 8.4 23.7 27.0 30.3 10.7 contribution The money you contribute to health insurance corresponds to health care 5.3 33.8 18.3 12.0 30.6 services you receive Persons with higher income should pay higher health insurance rate 28.0 32.1 14.8 13.5 11.7 (contribute more), and persons with lower income should pay lower health insurance rate (contribute less)

418 Mastilica and Babiæ-Bosanac: Citizens‘ Views on Health Insurance in Croatia Croat Med J 2002;43:417-424 income group (47.4%) expressed their strong agree- As the health insurance financing in Croatia is ment with the universal right to health care irrespec- based on the income-proportional contribution rate, tive of payments of contributions (chi-square=5.8, we wanted to know what the respondents thought p=0.016). about the idea of a progressive health insurance rate. One of the problems of the Croatian health insur- A vast majority (60.1%) agreed that the persons with ance fund at the time of the survey was related to ir- higher income should pay higher health insurance regular payments of contributions. When the respon- rate (contribute more), and persons with lower in- dents were asked whether the actual use of the ser- come should pay lower health insurance rate (con- vices should be linked to the payments of contribu- tribute less). Distribution by sociodemographic and tions, about one third of the respondents (32.1%) socioeconomic groups did not show significant differ- agreed with the statement. Opinion distribution by ences, although the respondents with higher educa- sociodemographic and socioeconomic groups showed tion and higher income were more likely to agree that there were significantly more men (chi-square with that statement. =4.0, p=0.047); those in high income group (chi- Reform of Health Insurance square=12.6, p=0.001); and professionals (chi- Respondents were asked about their attitudes to- square=6.5, p=0.011) among respondents who ward changes implemented in the social health insur- mostly or completely agreed with that statement (Ta- ance system since 1993, when the new law on health ble 2). Those with higher education and from urban insurance was introduced in Croatia. More than half areas were also more likely to agree, although not sig- (53.4%) of the interviewed population reported that nificantly. However, when asked whether higher use the current health insurance covered less medical of health services should be related to the increase in rights (benefits) than in the previous health care sys- health insurance contributions, only 17.2% agreed tem. Only 4.8% of the respondents thought that the that people who were more frequent users of health present health insurance covered more benefits than care services should pay higher health insurance rate. 10 years earlier, and 15.8% thought that it was the same as before. Table 2. Citizens’ attitudes on the statement that utilization of Analysis by sociodemographic and socioeco- health care services should depend on payments of health in- nomic groups revealed significant differences in the surance contributions agreement that the health insurance today covered Social group % of agreementa chi-square p less benefits than before (Table 3). In general, the re- Sex: spondents with higher education, higher income, and women 29 4.0 0.047 men 39 higher occupation status were significantly more Age: likely to share that view. <60 years 37 0.3 0.362 ³60 years 33 Education: Table 3. Citizens’ attitudes on the statement that current com- lower 29 2.6 0.067 pulsory health insurance covers less medical benefits com- higher 37 pared to 10 years ago Income: Social group % of agreementa chi-square p lower 24 12.6 0.001 Sex: higher 41 women 80 12.1 0.001 Occupation: men 61 workers 27 6.5 0.011 Education: professionals 41 lower 62 10.4 0.001 Place of residence: higher 80 urban 35 1.0 0.188 Income: rural 30 low 66 5.2 0.022 aPercentage of those who completely or mostly agred. high 78 Occupation: workers 69 4.8 0.029 We also asked the respondents whether the professionals 81 money they contributed to health insurance corre- Place of residence: urban 77 5.2 0.022 sponded to health care services they received, and rural 65 about 40% gave positive answer. Approximately the aPercentage of those who completely or mostly agreed. same proportion (30.3%) disagreed or did not know (30.6%). But when asked to estimate how much they The respondents were further asked about the contributed monthly from the salary to the compul- changes that occurred in the health care and compulsory health insurance, only 28.1% could estimate the sory health insurance system since the introduction of amount of contribution and most respondents the new health care and health insurance laws (refer- (71.9%) did not know how much their contribution to ence period was five years earlier, Table 4). A greater health insurance was. Significantly more respondents percentage of the respondents (36.9%) agreed that younger than 60 (33.9%) than those older (17.1%) the changes brought more choices, compared with (chi-square=7.5, p=0.006); and significantly more 30.8% of those who disagreed and 32.3% who did those with high income (36.0%) than those with low not know. High percentage of the respondents (37.7%) income (20.7%) (chi-square=6.1, p=0.048) dis- also agreed that changes in the health care system and agreed that money they paid for health insurance cor- health insurance brought less equity, as compared responded to health care services they received. with those who disagreed (20.8%) with that state-

419 Mastilica and Babiæ-Bosanac: Citizens‘ Views on Health Insurance in Croatia Croat Med J 2002;43:417-424

Table 4. Citizens’ attitudes on health insurance reform in Croatia Response (%) completely mostly mostly completely don't know/ How do you feel about the following statement: agree agree disagree disagree no answer Recent changes in health insurance and health care system 7.1 29.8 17.8 13.0 32.3 have brought more choice Recent changes in health insurance and health care system 15.8 21.9 17.0 3.8 41.5 have brought less equity The government should cover health insurance for pensioners, 67.4 23.4 3.3 1.5 4.3 socially vulnerable groups and children through budget revenues It is necessary to define the basic package of health services and 13.7 28.8 23.9 22.1 11.5 to introduce supplementary health insurance ment. Significantly higher percentage of respondents and from rural areas were more likely to disagree, but who agreed with the opinion that changes during the not significantly (Table 5). On the contrary, those last five years brought less equity was found among who agreed with the limitation of basic package and women (42.8% vs 30.7% men, chi-square=5.7, introduction of supplementary insurance were signifi- p=0.017), those with higher education (64.5% vs cantly more likely to be younger, with high income, 42.3% with low education, chi-square=7.6, p=0.022), and from higher occupational group. those with higher income (65.2% vs 48.1% with low Only a small proportion of the respondents re- income, chi-square=9.3, p=0.001), those with ported having a supplementary (privately paid) health higher occupational status (74.7% vs 57.4% with insurance (6.9%) although the legal provision for that lower occupational status, chi-square=5.8, p=0.053), kind of voluntary insurance had been introduced as and those from urban areas (42.2% vs 32.5% from ru- early as 1993. ral areas, chi-square=3.7, p=0.053). The demand that larger share of the state budget Table 5. Disagreement of citizens with the introduction of be spent on health care services, specifically for those basic package of health services and supplementary insur- financially most vulnerable, was examined through ance agreement with the statement “the government Social group % of disagreementa chi-square p Sex: should cover health insurance for pensioners, socially women 47 0.4 0.914 vulnerable groups, and children through budget reve- men 44 nues”. A large majority of the respondents (90.8%) Age: agreed that the government should spend more from <60 years 50 5.8 0.053 ³60 years 59 budget for non-paying groups of population. When Education: those who strongly agreed with the statement were lower 55 0.4 0.387 analyzed, it was found that significant majority be- higher 51 longed to the low-income group (75.3% vs 60.4% in Income: lower 59 5.7 0.017 the high-income group, chi-square=5.8, p=0.053) higher 46 and group with lower occupational status (74.8% vs Occupation: 59.8% with higher status, chi-square=3.8, p=0.052). workers 58 6.6 0.010 professionals 43 Women (72%) and persons over 60 years of age Place of residence: (75%) were also more likely, but not significantly, to urban 51 0.2 0.647 agree with the statement (data not shown). rural 54 In the following phase of the health insurance re- aPercentage of those who completely or mostly disagreed. form, the Croatian Government decided to define a basic health insurance package to be covered by Private Payments and Inequalities in Access compulsory health insurance, and introduce voluntary, supplemental, and private health insurance To examine the public views on private pay- scheme. The respondents were asked what they ments for health care, the respondents were asked thought would be necessary to define and limit the how much out-of-pocket money they spent for health proportions of the health care services costs to be cov- care during the previous 12 months (Table 6). More ered by compulsory health insurance (the basic than half (56.5%) of the respondents reported having health insurance package) and leave the difference in small expenses and 35.5% reported having large or costs and all other excluded services or the amenities very large expenses. There were no significant differ- to be covered by supplementary insurance. The per- ences between social groups although those with centage of those who disagreed (46.0%) was higher lower education (37.5%), lower income (38.2%), and than those who agreed (42.5%) with the limitation of lower occupational status (36.2%) were more likely health care service covered by compulsory health in- to report having large or very large expenses, com- surance. Analysis by sociodemographic and socio- pared with those with high education, high income, economic groups showed that those who disagreed and high occupational status. with the restriction of basic package were respon- When asked about their perception of private dents aged 60 years and over (chi-square=5.8, payments, a large proportion of respondents (45.8%) p=0.053), those with low income (chi-square=5.7, held the opinion that they were paying too much p=0.017), and low occupational status (chi-square= from their own pocket for health care. Analysis of the 6.6, p=0.010). Women, those with low education, sociodemographic and socioeconomic group differ-

420 Mastilica and Babiæ-Bosanac: Citizens‘ Views on Health Insurance in Croatia Croat Med J 2002;43:417-424

Table 6. Citizens’ views on private payments for health care in Croatia Question Response % What would you say, how much did you spend out of your pocket I did not have any expenses 8.0 for health care during the past 12 months? I had small expenses 56.5 I had large expenses 28.9 I had very large expenses 6.6 What do you think, do the citizens pay for health care too much Yes, too much has to be paid from one's 45.8 out of pocket or not? own pocket for health care It is tolerable 37.8 A little 6.0 Don't know 10.4 If you need immediate health care services, as for example surgery or special I should pay out of my pocket 35.1 diagnostic procedures, can you get it without paying out of your own pocket? I should not pay 29.1 Don't know 35.8 How do you agree with the statement that "some people in our country have Completely agree 44.3 easier access to health care services and receive better quality of care than others"? Mostly agree 38.0 Mostly disagree 5.5 Completely disagree 2.6 Don't know 9.6 ences showed that those who agreed were mostly respondents from low-income group (51.8% vs 35.0% Table 7. Croatian citizens who believe they should pay out of with high income, chi-square=9.7, p=0.008), those the pocket when in need of immediate health care services with lower occupations (59.6% vs 35.7% with higher Social group % chi-square p Sex: occupations, chi-square=12.7, p=0.002), and more women 49 4.7 0.030 likely, but not at a significant level, those with low ed- men 62 ucation and from rural areas. Age: <60 years 54 0.3 0.598 The respondents were further asked whether ³60 years 58 they could receive immediately needed health care Education: services without paying out of their own pocket, for lower 64 6.6 0.010 higher 46 example, surgery or special diagnostic procedures, Income: and 35.1% of the respondents reported that if they lower 65 10.4 <0.001 needed health care services without delay, they higher 44 would have to pay out of their pocket. When ana- Occupation: workers 60 5.1 0.024 lyzed by social group, it was found that significantly professionals 44 more respondents with lower education (chi- Place of residence: square=4.7, p=0.030), lower income (chi-square= urban 42 21.4 <0.001 10.4, p=0.001), lower occupational status (chi- rural 72 square=5.1, p=0.024), and from rural areas (chi- square=21.4, p<0.001) believed that when in need ute to health insurance corresponds to health care ser- of prompt health care services they would have to pay vices you receive; A4 – Compared to 10 years ago, for it (Table 7). current compulsory health insurance covers less medical benefits; A5 – Recent changes in health in- The respondents were asked to assess the social surance and health care system have brought more inequalities regarding access to health care services. choice; A6 – Recent changes in health insurance and A large majority of respondents (82.3%) agreed that health care system have brought less equity; A7 – It is some people had easier access to health care services necessary to define the basic package of health ser- and received better care than others. Significantly vices and to introduce supplementary health insur- more likely to strongly agree that there existed social ance, A8 – Assessment of own expenses for health inequalities in access to health care services were re- care; A9 – Citizens pay for health care too much out spondents with lower education (51.5% vs 36.2% of pocket; and A10 – Some people have easier access with higher education, chi-square=8.5, p=0.004), to health care services and receive better quality of low income (49.5%, chi-square=4.5, p=0.034), and care than others. from rural areas (50.6% vs 39.5% urban, chi- square=4.7, p=0.031). Answer categories ranged from “complete agreement” to “complete disagreement”, except for the “as- Underlying Dimensions in Citizens’ Opinions sessment of own health care expenses” where the an- on Health Insurance (Factor Analysis) swers were “no expenses”, “small”, “tolerable”, and To identify the underlying dimensions in citi- “large or very large expenses”. zens’ opinions on health insurance and payments for Three latent dimensions with eigenvalue above health care services, the factor analysis was per- 1.0 were obtained by factor analysis (Table 8). They formed. Ten variables reflecting the attitudes of the re- accounted for 51.7% of the total variance. Commu- spondents were used for the factor analysis: A1 – Ev- nalities of all three principal components were >0.4. erybody should have the right to health care, irrespec- The first factor (eigenvalue 2.651) was recognized as tive of contribution to health insurance; A2 – Utiliza- “negative opinion on changes in health insurance”, tion of health care services should depend on health with the highest loadings of 0.523-0.795 for attitudes insurance contribution; A3 – The money you contrib- A6, A4, A5, and A8. This factor was predominantly

421 Mastilica and Babiæ-Bosanac: Citizens‘ Views on Health Insurance in Croatia Croat Med J 2002;43:417-424 described by agreement that recent changes brought group was responsible for the differences in the third less equity (A6) and that current health insurance cov- factor. The occupational groups significantly differed ers less benefits than 10 years ago (A4), disagreement only in the first factor (F=8.063, p=0.001), with the that recent changes brought more choice (A5), and as- low occupational group responsible for the difference. sessment of own direct expenses for health care as large or very large (A8). The second factor (eigenvalue Discussion 1.290) could be interpreted as the “affirmative assessment of changes in health insurance”, with the high- This study showed that Croatian citizens highly est loadings of 0.400-0.680 for attitudes A10, A9, A3, supported the universal health services, regardless of and A5. The second factor was predominantly de- insurance contribution payment. They strongly be- fined by disagreement that some people have easier lieved that health care is a social good to which every- access to health care services and receive better qual- body have a right, and not a market commodity that ity of care than others (A10), followed by assessment could be sold and bought. This may be seen as a re- that citizens pay small amount out of pocket for sult of a long tradition of the former socialist, state- health care (A9), and opinion that health insurance regulated social health insurance. However, it seems contributions correspond to health care received that although the majority thought that health care (A3). It could also be defined by the opinion that re- should be universally accessible, about one third cent changes brought more choice (A5). The third fac- agreed that the use of services should depend on pay- tor (eigenvalue 1.231) could be recognized as “uni- ments. More citizens believed that health care ser- versal health care services”, with the highest loadings vices they received corresponded to their contribu- of 0.616-0.789 for attitudes A2, A1, and A7. That fac- tions, and a large majority supported the idea of pro- tor was described by disagreement that use should gressive contribution rates. depend on payments of contributions (A2), followed The reform of the health care system that has by agreement that everybody should have right to been taking place in Croatia over the past 10 years health care irrespective of paying contributions (A1), has significantly affected the position of citizens as and disagreement that it is necessary to introduce ba- patients and increased their dissatisfaction (13-15). sic package of health services and supplementary According to the opinion of the majority of citizens health insurance (A7). included in this study, the patients’ benefits in the ex- The differences in principal components regard- isting health insurance were reduced in comparison ing sex, two age groups, and place of residence were to 10 years ago. More than one third believed that analyzed by a t-test. There were no significant differ- changes brought more choice, and a similarly large ences between men and women, younger and older but different social group believed that changes (over 60 years), and urban and rural respondents in brought less equity. The interviewed Croatian citi- any of the three factors. The differences between edu- zens strongly supported the demand that government cation, income, and occupational groups were ana- should increase spending for health care services lyzed by ANOVA. The analysis revealed that three through state budget revenues and thus take financial education groups were not significantly different. The responsibility for health care of non-paying groups of income groups differed significantly in all three fac- population, such as the retired, unemployed, chil- tors: “negative opinion on changes in health insur- dren, and other socially vulnerable groups. The most ance” (F=4.30, p=0.016), “affirmative assessment of important change in the recent health insurance re- changes in health insurance” (F=3.29, p=0.041) and form, the introduction of the basic health insurance “universal health care services” (F=3.99, p=0.021). package with limited value of benefits and implemen- Scheffe’s multiple comparison tests (p<0.05 signifi- tation of the supplementary insurance scheme, di- cance level) revealed that low- and medium-income vided the citizens, who more disagreed than agreed groups were responsible for differences in the first fac- with this change. tor, the high income group was responsible for differ- Privatization and marketization of services under ences in the second factor, and the medium income the health care reform significantly increased citi-

Table 8. Correlation coefficients among the survey variables and underlying factors – Varimax rotation with Kaiser normalization Variable Factor symbol description 1a 2b 3c A6 recent changes brought less equity -0.795 0.046 0.018 A4 current health insurance covers less benefits than before 0.773 0.036 0.050 A5 recent changes brought more choice 0.534 -0.400 0.113 A8 assessment of own expenses for health care 0.523 -0.171 0.314 A10 some people have easier access to health care services 0.067 0.680 0.044 A9 citizens pay for health care too much out of pocket -0.108 0.667 -0.071 A3 contributions correspond to received health care 0.185 -0.634 0.096 A2 utilization should depend on contribution -0.076 0.182 0.789 A1 right to health care irrespective of contribution -0.079 0.368 -0.635 A7 necessary to introduce basic package of health services and supplementary health insurance 0.323 -0.093 0.616 Eigenvalue 2.651 1.290 1.231 Variance (%) 26.5 12.9 12.3 Cumulative (%) 26.5 39.4 51.7 aNegative opinion on changes in health insurance. bAffirmative assessment of changes in health insurance. cUniversal health care services.

422 Mastilica and Babiæ-Bosanac: Citizens‘ Views on Health Insurance in Croatia Croat Med J 2002;43:417-424 zens’ out-of-pocket spending on health care, as evi- of co-payments, resulted in high dissatisfaction of denced in this study. There is a strong public belief consumers and increased inequalities in health care that health care services are not equally accessible to financing (13,15). Until recently, however, the all and that out-of-pocket payments can make the ac- changes have not significantly affected the health in- cess easier. In general, the higher socioeconomic surance benefits of the insured, which remained com- groups, those under 60 years of age, men, and re- prehensive and fully covered. spondents from urban areas expressed more criticism With the new Health Insurance Act of October in their views (less benefits, less equity) and agreed with 2001 (22), the government decided to solve the finan- the cost-containment measures. The lower socioeco- cial problems in health insurance fund by introducing nomic groups, those over 60 years of age, women and basic health insurance as compulsory, and supple- rural respondents were positively oriented towards uni- mentary and private health insurance scheme. In the versal access, state health care financing, and mostly basic health insurance package, financing is reduced disagreed with the introduction of basic health insur- from 85% to 50% of the total service price in special- ance package. However, more research is necessary ist and hospital care, and from 75% to 25% of costs to analyze the impact of health insurance reform from for medicaments. Preventive and primary health care the perspective of different social groups. services are fully covered. In addition to basic, sup- Three underlying dimensions in citizens’ views plementary insurance is established to cover the dif- revealed by the factor analysis were “negative opin- ference between the basic insurance and the total ser- ion on changes in health insurance”, significant for vice cost, and for amenities or services not covered by those with low- and medium-income and low occu- basic insurance. If not through voluntary supplemen- pational status; “affirmative assessment of changes in tary insurance, patients have to pay the uncovered health insurance”, significant for the high-income proportion of the total service price by cost sharing group; and “universal health coverage”, significant (co-insurance). Exemptions are made in regulations for the medium-income group. for those with the defined minimal income per per- Using survey data, we examined the public person. Under the new health insurance law, the financ- ception of the existing social health insurance in ing has been reduced to the basic health insurance Croatia and of the changes taking place under the re- package, which covers benefits limited in money form. In the study sample, the adults over 40 years of value. The insured are required to pay privately signif- age were selected, with the idea to address those who icantly higher proportions of the cost of services used health care services more frequently and were (co-insurance) or to purchase supplementary insur- better informed. There were certain limitations con- ance to cover expenses not included in the basic in- cerning the sample, with over-represented sociode- surance package. Against this background, we as- mographic group of women, those with higher edu- sumed that it was important to analyze the views of cation, higher family income, and from urban areas. the public, since the new health insurance reform had Previous studies showed that there was no consistent serious impact on the citizens. association between socioeconomic characteristics To take a proper account of the views and feel- and attitudes towards health care services (16,17), al- ings of the general public in the process of developing though it was found that older people tended to have and implementing health policy is particularly impor- more positive attitudes (less expectations), and that tant, having in mind the main objectives of the re- higher-education and higher-income people were formers to meet the citizens’ needs (23). The study of more ready to criticize and have negative attitudes the public opinion on the health reform is a valuable (higher expectations) (18,19). Accordingly, it could tool for testing how the implemented measures corre- be assumed that a more representative sample in this spond with the citizens’ expectations. It is also an in- study would show more positive attitudes and higher put for developing future health policy and establish- degree of agreement. Other limitations that could ing priorities in health care system. have affected our results concerned the problems oc- Although there were some attempts to analyze curring in public opinion surveys in general, such as health insurance and its changes in the countries in the structure of questions, cognitive bias, survey dura- transition (24-29), the reforms were rarely evaluated tion, or media influences (20,21). from the citizens’ perspective (13-15). The reasons The survey was carried out in 1999 and 2000, may be found partly in the skepticism about the use- two years before the actual change of health insur- fulness of opinion surveys for health policy and in ance legislation, at the time when the discussion on methodological and other survey-related difficulties insurance reform started. Different policy measures (30,31). However, the evidence on citizens’ views on were discussed in professional associations and in the health care reforms in the European Union countries public, yet not all were tested in this opinion survey. suggests a positive impact of the public evaluation ap- Some of those evaluated, however, did not happen to proach to reform’s issues (32-34). be implemented by the new law. The citizens in Croatia are presently expressing Most of the reforms that have been taking place their concerns with the undergoing health insurance in the Croatian health care system in the last 10 years reform. They mostly support the principle of the uni- were directed towards the supply side, targeting first versal health care services, with the government re- of all providers of services. Those reform measures sponsibility for providing it to all citizens, and do not that aimed to reduce the demand of services with the agree with rationing of benefits and implementation influence on the consumers, such as the introduction of market mechanisms into the Croatian social health

423 Mastilica and Babiæ-Bosanac: Citizens‘ Views on Health Insurance in Croatia Croat Med J 2002;43:417-424 insurance. At the time of significant changes, to have 17 Fitzpatrick R. Survey of patient satisfaction: I – impor- an insight into the views of the public can be valuable tant general considerations. BMJ 1991;302:887-9. for policy-makers, providing them with citizens’ per- 18 Hall JA, Dornan MC. Patient sociodemographic charac- ceptions of the implemented measures and policy op- teristics as predictors of satisfaction with medical care: a tions. If the goal of the health reform is to create a sys- meta-analysis. Soc Sci Med 1990;30:811-8. tem more responsive to citizens’ needs and expecta- 19 Sitzia J, Wood N. Patient satisfaction: a review of issues tions, knowledge of the public’s demands should be and concepts. Soc Sci Med 1997;45:1829-43. taken into account. 20 Blalock HM, Blalock AB, editors. Methodology in social research. 2nd ed. London: McGraw-Hill Inc.; 1971. Acknowledgments 21 Judge K, Solomon M, Miller D, Philo G. Public opinion, This research was a part of the project Evaluation of Transi- the NHS, and the media: changing patterns and per- tion in Health Care, supported by the Central European Network spectives. BMJ 1992;304:892-5. collaborative program for research and education in health sec- tor. We thank the medical students who served as interviewers. 22 Croatians Parliament. Health insurance act [in Cro- atian]. Narodne novine 94/2001. References 23 Mossialos E. Citizens and health systems: main results from a Eurobarometer survey. Luxembourg: European 1 Ministry of Health, Republic of Croatia. Health care re- Commission; 1998. form. Strategy and plan for the reform of the health care 24 Kersnik J. Determinants of customer satisfaction with system and health insurance of the Republic of Croatia the health care system, with the possibility to choose a [in Croatian]. Zagreb: Croatian Ministry of Health; 2000. personal physician and with a family doctor in a transi- 2 Croatian Institute for Health Insurance. Annual audit re- tion country. Health Policy 2001;57:155-64. port of Croatian Institute for Health Insurance and Cro- 25 Turek S. Reform of health insurance in Croatia. Croat atian Health System, 2000 – financial plan and basic Med J 1999;40:143-51. management propositions for 2001. Zagreb: Croatian Institute for Health Insurance; 2001. 26 Ljubiæ B, Hrabaè B, Rebac Z. Reform of health insur- 3 Organization for Economic Co-operation and Develop- ance in the Federation of Bosnia and Herzegovina. ment. The reform of health care systems: a review of Croat Med J 1999;40:160-5. seventeen OECD countries. Washington (DC): OECD; 27 Donev D. Health insurance system in the Republic of 1994. Macedonia. Croat Med J 1999;40:175-80. 4 Vienonen M. Overview of the principles in health care 28 Markota M, Švab I, Sara´in-Klemenèiæ K, Albreht T. reforms on the European scene. Antidotum Supplement Slovenian experience on health care reform. Croat Med 1995;1:107-10. J 1999;40:190-4. 5 World Health Organization, Regional Office for Eu- 29 Field MG. Reflections on a painful transition: from so- rope. European health care reforms: analysis of current cialized to insurance medicine in Russia. Croat Med J strategies. Copenhagen: WHO; 1996. 1999;40:202-9. 6 Saltman RB, Figueras J, editors. European health care re- 30 Mechanic D. Public perceptions of medicine. N Engl J form. Analysis of current strategies. Copenhagen (Den- Med 1985;312:181-3. mark): WHO Regional Publication; European Series, 31 Al-Swwailem AF. Should opinion polls be more widely No. 72; 1997. used in developing countries? Health Policy Plan 7 Abel-Smith B, Mossialos E. Cost containment and 1991;6:185-7. health care reform: a study of the European Union. 32 Pescosolido BA, Boyer CA, Tsui WY. Medical care in Health Policy 1994;28;89-132. the welfare state: a cross-national study of public evalu- 8 Maarse H. Cost containment in health care: a brief over- ations. J Health Soc Behav 1985;26:276-97. view. Antidotum Supplement 1995;1:56-65. 33 Blendon RJ, Leitman R, Morrison I, Donelan K. Satisfac- 9 Mossialos E, Le Grand J, editors. Health care and cost tion with health systems in ten nations. Health Aff containment in the European Union. Aldershot (UK): (Millwood) 1990;9:185-92. Ashgate; 1999. 34 Mossialos E. Citizens’ views on health care systems in 10 Hebrang A. Reorganization of the Croatian health care the 15 member states of the European Union. Health system. Croat Med J 1994;35:130-6. Econ 1997;6:109-16. 11 Oreškoviæ S. Health system reorganization in Croatia in the light of major reform tendencies in OECD countries. Croat Med J 1995;36:47-54. Received: March 6, 2002 12 Kovaèiæ L, Šošiæ Z. Organization of health care in Croatia: Accepted: June 12, 2002 needs and priorities. Croat Med J 1998;39:249-55. 13 Mastilica M, Chen M. Health care reform in Croatia: the Correspondence to: consumers’ perspective. Croat Med J 1998;39:256-66. Miroslav Mastilica 14 Chen M, Mastilica M. Health care reform in Croatia: for better or for worse? Am J Public Health 1998;88:1156-60. Department of Medical Sociology 15 Mastilica M, Bo´ikov J. Out-of-pocket payments for Andrija Štampar School of Public Health health care in Croatia: implications for equity. Croat Zagreb University School of Medicine Med J 1999;40:152-9. Rockefellerova 4 16 Fox JG, Storms DM. A different approach to sociodemographic predictors of satisfaction with health care. Soc 10000 Zagreb, Croatia Sci Med [A] 1981;15:557-64. [email protected]

424