Article Text
PDF
Abstract
Background To perform the first European overview of educational inequalities in the use of blood pressure and cholesterol screening.
Methods Data were obtained on the use of screening services according to educational level from nationally representative cross-sectional surveys in Belgium, Czech Republic, Denmark, Estonia, Finland, Hungary, Italy, Latvia and Lithuania. Screening rates were examined in the preceding 12 months and 5 years, for respondents 35+ years (45+ for women). ORs comparing low- to high-educated respondents were estimated using logistic regression controlling for age.
Results Inequalities in cholesterol screening favouring higher socioeconomic groups were demonstrated with statistical significance among men in four countries, whereby men with higher education were more likely to receive screening, with 1.22 as the highest OR. Among women, a similar pattern was found. Inequalities in blood pressure screening were even smaller and less often statistically significant. Hungary was the only country with higher rates of both types of screening in the low-educated group. In other countries, pro-high inequalities were slightly increased after controlling for self-rated health.
Conclusions All European countries in this study had small educational inequalities in the utilisation of blood pressure and cholesterol screening. These inequalities are smaller than those previously observed in the USA. Further comparative studies need to distinguish between screening for preventive purposes and screening for treatment and control.
- Access to health services
- blood pressure
- epidemiology
- preventive medicine
- social inequalities
- access to healthcare
- blood pressure
- epidemiology
- inequalities
- public health
- statistics
- study design
- health behaviour
- demography
- Eastern Europe
- epidemiology
- geography
- inequalities
- public health
- social epidemiology
- health expectancy
Statistics from Altmetric.com
- Access to health services
- blood pressure
- epidemiology
- preventive medicine
- social inequalities
- access to healthcare
- blood pressure
- epidemiology
- inequalities
- public health
- statistics
- study design
- health behaviour
- demography
- Eastern Europe
- epidemiology
- geography
- inequalities
- public health
- social epidemiology
- health expectancy
Introduction
Large educational inequalities in cardiovascular disease (CVD) mortality have been documented across Europe.1 Narrowing these inequalities requires an understanding of the role of access to preventive services.2 Since early treatment of hypertension and cholesterolemia may prevent or slow cardiovascular complications, international and national guidelines have all prescribed regular screening.3
Although lower educational groups may be in most need of screening, American studies reveal that individuals of higher socioeconomic status are more likely to receive testing for blood pressure or cholesterol.4–6 Healthcare insurance and access to a usual source of care are important contributors to the likelihood of screening in the American context, although they did not fully explain the socioeconomic inequalities.3 ,6 No association has been found between cholesterol screening and income or education in Canada, where a national health service exists.7 In another Canadian study on blood pressure screening, those with lower education were less likely to be screened and there was a positive, but weaker, association for income.8
Much less research on socioeconomic inequalities in blood pressure and cholesterol screening has been conducted in Europe than in North America. In Belgium, small socioeconomic inequalities, according to Lorant et al,9 were found for cholesterol screening, with people in the two lowest income quintiles being less likely to receive screening (ORs of 0.7 and 0.8, respectively). A starker gradient was found in Spain, where the likelihood of cholesterol screening was reduced with decreasing education.10 Another Spanish study found that lower occupational classes less often received blood pressure screening.11 In Italy, high-socioeconomic groups were found to be more likely to have their blood pressure and cholesterol checked, with the largest inequalities observed among men in relation to education level.12
The present study is the first to assess inequalities in cardiovascular preventive services from a broader European perspective. Its specific aim was to assess the direction and magnitude of educational inequalities in the use of screening for hypertension and cholesterolemia in nine European countries. A generalised pattern of small inequalities might perhaps be expected in European healthcare systems, given the relatively small levels of inequalities in access to primary care.13 Systems with nationalised healthcare and/or universal insurance, such as in Europe, are generally designed to reduce administrative, geographical and financial barriers. At the same time, preventive services are organised differently in different countries, which could influence the magnitude of equality in their use.2
Methods
Study population
Data were obtained on cholesterol and blood pressure screening according to educational level from nationally representative cross-sectional surveys of individuals aged 16 years and older in the following European countries: Belgium, Czech Republic, Denmark, Estonia, Finland, Hungary, Italy, Latvia and Lithuania. There is no upper age limit to those surveyed in most countries, but Estonia, Latvia, Finland and Lithuania have an upper age limit of 64. Further details on the surveys are given elsewhere.1
This study relied on micro-data files constructed on the basis of national data files. These files were created in accordance with an extensive set of detailed specifications to enhance intersurvey comparability. This study did not need ethics approval, as only anonymised micro-data available at national statistical offices were used. Table 1 presents the composition of the study population.
Sample composition
Measures
Education level was harmonised between the different countries on the basis of the International Standard Classification of Education. The national codes were classified into four categories of the International Standard Classification of Education scheme and later dichotomised into ‘high’ and ‘low’ education. Low education referred to no completed education, only primary education or lower secondary education (up to approximately 9 years). Higher education included upper secondary, post-secondary non-tertiary and tertiary education.
The simple distinction between low and high education does not take into account differences within these two broad groups. Furthermore, the meaning of ‘low’ and ‘high’ varies between European countries, partly because of differences in the educational systems. However, a more sophisticated measure of educational inequalities, the Relative Index of Inequality,1 generated similar patterns to those reported in the Results section below.
The information on blood pressure screening was derived from survey questions such as, “When was the last time that you had your blood pressure checked?” For cholesterol screening, respondents were asked questions such as, “When was the last time that you had your cholesterol checked?” Survey questions were phrased similarly, with minor differences not expected to affect the comparability of estimates between countries.
Responses to these questions were categorised in the same way among the countries, using three classes: (1) screening performed for the last time within the last 12 months; (2) longer than 12 months ago, but within the last 5 years and (3) last time >5 years ago or never. In Italy and the Czech Republic, the recall period is only 12 months. In Belgium, only data for the 5-year recall period are available. In Denmark, the extended recall period referred to the last 3 years, instead of the last 5 years.
The variable age refers to the age of the respondent at the time of the survey, wherein age is grouped in 5-year intervals. Screening rates were examined among the general population, which includes individuals aged 16 years and older, as well as women aged 45 years and over and men aged 35 years and over. These age groups comprise the recommended target groups according to 1998 recommendations of a number of European Societies, which were the most recent version at the time of the surveys.3
The different target age groups for men and women led to the decision to present the data separately for men and women. In addition, previous studies have found gender differences in the receipt of preventive services.14
As the use of preventive services may be influenced by the prevalence of health problems, socioeconomic inequalities may contribute to inequalities in screening, secondary to their effect on inequalities in health. To evaluate the possible role of health, each respondent's self-rated health was assessed. Respondents' general assessment of health was assessed by questions similar to, “How do you judge your general state of health?” The most important difference was in the wording of the answer categories, with some questions starting with ‘excellent’, while others started with ‘very good’. Other answer categories were ‘good, fair, poor, very poor’.
No internationally comparable data were available for control variables, with the exception of smoking, body mass index and self-reported diabetes. Control for these factors did not substantially change the results as presented below. In order to maintain a relatively simple model, a decision was made to present results only controlled for age and self-rated health.
Statistical analyses
Analyses were conducted for each country and gender separately. The percentage of the population screened in both the lowest and the highest educational level was computed. Age-standardised rates in 5-year intervals were calculated using direct method with the European Standard Population as the reference population.
The association between education and screening outcomes was measured using ORs and their 95% CIs on SPSS version 15.0.1. Logistic regression was applied in two steps. First, control was made for 5-year age group. In a second step, control was made for both age and self-rated health. Self-rated health was introduced into the model as a dichotomous variable (good or very good health vs poorer health), as more detailed control for this variable yielded results similar to those reported below.
Results
Table 2 presents the percent distributions of cholesterol screening by education level, stratified by gender. Screening rates tend to be slightly greater among women than among men in some countries. Differences in overall screening between countries are notable, with only Italy, Finland and the Czech Republic achieving screening rates of >50% in the 12-month recall period. Screening rates are approximately 50% and are expectedly higher when the recall period is extended to 5 years. Overall, educational differences in screening rates were small.
Age-standardised rates of cholesterol screening utilisation by country, sex and educational level (educ.) in target population group*
Table 3 shows OR for cholesterol screening among men over 5-year and 12-month recall periods. ORs higher than 1.00 imply higher screening rates for those with high education. Over a 5-year recall period, inequalities in cholesterol screening could be demonstrated with statistical significance for Belgium, Denmark, Finland and Hungary, with OR ranging from 1.10 (95% CI 1.03 to 1.18) in Belgium to 1.21 (95% CI 1.07 to 1.36) in Denmark. In these countries, men with higher education were more likely to be screened. When the recall period was truncated at 12 months, inequalities remained minimal and were statistically significant only in Italy (OR=1.10, 95% CI 1.07 to 1.14), Denmark (OR=1.22, 95% CI 1.05 to 1.41) and Finland (OR=1.22, 95% CI 1.09 to 1.37). In Hungary, the pattern of inequalities reversed and the OR jumped from 0.72 (95% CI 0.65 to 0.79) to 1.15 (95% CI 1.10 to 1.20) for the 12-month recall period. When controlling for self-rated health, slight increases in inequalities were observed in Italy (OR=1.14, 95% CI 1.11 to 1.18), Denmark (OR=1.30, 95% CI 1.12 to 1.50) and Latvia (OR=1.26, 95% CI 1.02 to 1.55), while inequalities elsewhere changed little. Controlling for self-rated health eliminated the statistically significant associations in Hungary.
OR (95% CI)* comparing prevalence of cholesterol screening among low-educated to high-educated persons
The results for cholesterol screening in women are presented in table 3. Inequalities were markedly smaller in women compared with men. The reverse pattern of inequalities observed in Hungary in men was not observed among Hungarian women. Over both 5-year and a 12-month periods, statistically significant inequalities were found only in Finland (OR=1.15, 95% CI 1.02 to 1.30), whereby the higher socioeconomic groups received greater screening. After controlling for self-rated health, inequalities became statistically significant in Italy (OR=1.07, 95% CI 1.03 to 1.11) and in Hungary (OR=1.22, 95% CI 1.12 to 1.34).
Turning to blood pressure screening, table 4 shows absolute prevalence rates for men and women over both a 12-month and a 5-year time span. Overall, screening rates of over 70% have been achieved in men in all countries, with Belgium, Estonia, Finland, Italy and Hungary achieving screening rates of over 90% among both educational groups over a 5-year period. Women achieved slightly higher rates over both 12-month and 5-year recall periods. Overall, rates between educational groups are quite comparable.
Age-standardised rates of blood pressure screening utilisation by country, sex and educational level (educ.) in target population group*
Results of the logistic regression for blood pressure screening in men are presented in table 5. No statistically significant inequalities were observed in screening over the 5-year recall period. Over the 12-month period, small inequalities with OR >1 were demonstrated with statistical significance in Italy (OR=1.10, 95% CI 1.06 to 1.13), Lithuania (OR=1.10, 95% CI 1.00 to 1.21) and Finland (OR=1.15, 95% CI 1.04 to 1.26). Here, too, Hungary remained an outlier, where those with low education were more likely to receive screening (OR=0.62, 95% CI 0.58 to 0.67). About half of this inverse relationship could be explained by controlling for self-rated health (OR=0.81, 95% CI 0.75 to 0.88). Controlling for self-rated health revealed slightly elevated inequalities in Italy, Latvia and Lithuania, where lower socioeconomic groups were less likely to be screened.
OR (95%CI)* comparing prevalence of blood pressure screening among low-educated to high-educated persons
Table 5 presents the ORs for blood pressure screening in women. For the 5-year recall period, no inequalities could be demonstrated. All ORs are within a narrow range (1.00–1.05). For the 12-month recall period, inverse inequalities with OR <1.00 were found in Hungary (OR=0.78, 95% CI 0.73 to 0.84). Most other countries showed small inequalities in screening favouring higher socioeconomic groups over the 12-month recall period. After controlling for self-rated health, the OR for some countries became statistically significant. In Denmark, Latvia and Lithuania, the odds of a higher educated woman receiving screening were found to be 1.11 or 1.12, after controlling for self-rated health.
Discussion
This study yielded a generally positive view of the distribution of cardiovascular preventive screening in nine European countries, with representation from both Eastern and Western Europe. Socioeconomic inequalities in screening for both hypertension and cholesterolemia were found to be generally small. The use of screening services tended to be somewhat higher among respondents with high education. Slightly greater inequalities favouring the higher socioeconomic groups were found in cholesterol than in blood pressure screening and greater inequalities were observed among men than among women. These inequalities were slightly more pronounced when inequalities in general health were controlled.
The available data had some limitations. First, international differences in data collection methods may have affected the comparability of estimates for different countries. However, small variations between countries in the magnitude of inequality estimates suggest that any such differences in data collection methods did not affect the results in a meaningful way.
A second potential limitation is that non-response rates, which ranged from 13% to 42% (table 1), could have biased inequality. However, even if this bias was present, it is unlikely to explain the small magnitude of inequalities observed, as large inequalities were observed with the same data source for many other health outcomes.1
Third, for the sake of brevity and international comparability, educational levels were grouped into two broad categories. Using a more refined educational classification, slightly larger educational differences in screening rates were observed (results not shown).
Hungary was the only country where those with lower education were more likely to have received screening in the past 12 months. This may be attributed to the effect of general health: screening rates are higher among those with poorer health, and poor health is more common among lower socioeconomic groups. Furthermore, at the time of the data collection, Hungary was undergoing drastic political and economic changes, which may have undermined the general health of lower socioeconomic groups in particular.15 This may also account for the relatively large inequalities in cardiovascular mortality in Hungary, compared with most other European countries.1
In this study, no distinction could be made between screening for preventive purposes and the measurement of blood pressure and cholesterol in patients with CVD. Since CVD is more prevalent among lower socioeconomic groups,16 measurements for treatment and control are likely to be more prevalent among these groups. If so, the results in this study may have concealed inequalities in the reverse direction in measurements undertaken for preventive purposes.
On a related point, inequalities in the use of preventive healthcare were assessed without taking into account inequalities in need. This approach may be justified on the grounds that, according to international guidelines, all men and women at specific ages should be regularly screened for cholesterolemia and hypertension.3 Nevertheless, if lower socioeconomic groups are at higher risk for developing these conditions, one might argue that preventive services should be used more frequently by these groups. Thus, despite the small ‘inequalities’ observed, the situation may be less favourable in terms of ‘inequities’, that is, after adjusting for need.9 In a recent Italian study, larger educational inequalities were observed for cholesterol screening, with ORs ranging from about 1.15 to 1.63, after controlling for self-reports of hypertension, diabetes, smoking status and body mass index.12 The finding in this study that the small inequalities were slightly widened after controlling for general health supports this possibility. Therefore, the possibility cannot be excluded that detailed control for the occurrence and severity of CVD would have shown larger inequalities, disadvantaging the lower socioeconomic groups.
Interestingly, Finland, a northern European country with universal access to healthcare services, had the highest rates of inequalities in blood pressure and cholesterol screening. However, in the 1990s, equal access to services among different regions and different socioeconomic groups in that country was adversely affected by government cuts to municipal subsidies.17 ,18 Decentralisation of health service funding at that time made individual municipalities responsible for funding their health services through local taxation, perhaps thereby increasing the risk of social and geographic inequalities in care.
American studies that have assessed educational disparities in blood pressure screening report that the odds of being screened over a 5-year period for higher educated people, compared with lower educated, are as high as 2.82 (95% CI 1.57 to 5.07).19 In addition, inequalities by race, insurance status and having regular access to primary care have been documented extensively. Inequalities in cholesterol screening are similarly striking, as people with higher education have 2.11 (95% CI 1.31 to 3.40) times higher odds of being screened.19 ,20 While it is not possible to compare directly the inequality measures reported in different studies, the education-based disparity in European screening rates seems to be at similar levels to those in Canada and smaller than in the USA.
Despite the greater equality, overall prevalence rates of screening are much lower in European countries (and in Canada) than in the USA (see tables 2 and 4). A main challenge for European countries is to increase national rates of preventive screening up to the levels recommended by the European Joint Task Force, without increasing socioeconomic disparities.3
The relatively small socioeconomic disparity in screening in European countries suggests that low education is not necessarily a barrier. The larger inequalities in cardiovascular screening in the USA suggest that the structure of the healthcare system, rather than personal attributes of individuals, may be of key importance. A US study suggested that higher screening utilisation by people with higher education was related to their greater access to health insurance reimbursement, and consequently, usual care.19 Compared with the USA, European countries have achieved greater equality in access to healthcare, in part due to less variability in insurance status and user fees.21 For example, general practitioners' services are universally accessible and relatively often used by lower socioeconomic groups in Europe.2
Different patterns of inequalities have been observed in Europe for breast and cervical cancer screening. Pro-rich inequalities, in which higher socioeconomic groups have greater access to screening, have been observed in several European countries for both types of cancer screening.2 These larger inequalities might be due to a larger effect of patient-related demand in the case of screening for cancer. However, while pro-rich inequalities in breast cancer screening were large in European countries with opportunistic screening, these inequalities were small or even absent in countries where pro-active national screening programmes were available.22 This suggested that supply factors, such as the way screening services are delivered, play a key role.23
To conclude, all European countries in this overview have small inequalities in the utilisation of blood pressure and cholesterol screening. Comparisons to the USA suggest that the size of inequalities mostly depends on the structure of the healthcare system and on service delivery. Small levels of inequality in screening are important in reducing CVD across all socioeconomic groups. However, screening for secondary prevention is only one step in the sequence of events needed to reduce morbidity and mortality. National primary prevention strategies to address obesity, living conditions and physical fitness are equally important goals.
What is already known on this subject
-
Although American studies have revealed large socioeconomic inequalities in access to blood pressure and cholesterol screening, much less is known about access to these preventive services in Europe.
-
The present study is the first to examine educational inequalities in the use of screening for hypertension and cholesterolemia from a broad European perspective
What this study adds
-
This study found overall low rates of inequalities in screening, although the use of screening services was somewhat higher among respondents with high education, particularly when inequalities in general health were controlled.
-
The evidence suggests that the size of the inequalities mostly depends on the structure of the healthcare system and on service delivery.
Acknowledgments
We thank Stefaan Demarest, Herman van Oyen, Ritva Prättälä, Mall Leinsalu and Csilla Kaposvari for preparing and making available the micro-data sets on the basis of national health surveys from Belgium, Finland, Estonia, Latvia, Lithuania and Hungary.
References
Footnotes
-
Funding This study was carried out as part of the Eurothine project, which is funded by the public health programme of the SANCO Directorate General of the European Commission (grant number 2003125).
-
Competing interests None.
-
Provenance and peer review Not commissioned; externally peer reviewed.