Introduction

Strong efforts to tackle health inequalities can be seen at international and national level since the 1980s. In early 2008, the World Health Organization’s (WHO) Global Commission on Social Determinants of Health called for action on the social determinants of health, the conditions in which persons are born, grow, work, live, and age, to “close the gap in a generation” [1]. In late 2008, the Spanish Public Health General Direction (Dirección General de Salud Pública) and the Foreign Health of Health Ministry and Social Policy (Sanidad Exterior del Ministerio de Sanidad y Política Social) requested the constitution of the Commission for the Reduction of Social and Health Inequalities (Comisión para Reducir las Desigualdades Sociales en Salud en España (CRDSS-E) [2]. The mission of CRDSS-E was to elaborate on a proposal of intervention measures to reduce health inequalities. The CRDSS-E published two documents: one analysing health inequalities in the Spanish context [3], and another describing some policy proposals to tackle them [4]. In 2011, a total of 125 countries, Spain being one of them, developed and signed the Rio Political Declaration on Social Determinants of Health [5]. The declaration recommended interventions from governments and international organisations [6].

At a regional level in Catalonia, tackling health inequalities is one of the main goals of both the Catalan Health Plan 2016–2020 (led by the Health Department of the Catalan Government) [7] and the Interdepartmental and Intersectorial Public Health Plan 2017–2020 (PINSAP) [8, 9]. During the past years, various reports and peer-reviewed papers about the health effects of the economic crisis on the population of Catalonia were published by the Catalan Health System Observatory [10,11,12,13,14,15,16,17].

Overall, much effort has been devoted to monitoring and tackling health inequalities at regional, national, and international levels. Even so, OECD countries continue to present large disparities in health, including, for example, significant differences in life expectancy between people with the highest and lowest levels of education [18]. The selection of topics represents the first step in monitoring health inequalities according to many conceptual frameworks and is highly relevant, as these topics will potentially limit the detection of health inequalities within the population, hence playing a key role in providing evidence for posterior decision-making [19, 20]. Yet some difficulties may arise during the selection of relevant topics, as well as their health indicators. A wide diversity of indicators for monitoring health inequalities have been used across different countries and organisations; this is due to the high variety of contextual factors that may have an influence on it, such as the study goals or the information resources available.

In order to help accomplish this task successfully, the objective of this study is to perform a systematic review of the most common topics and indicators used for measuring and monitoring health inequalities in the reports, projects, and databases of international, national, and regional governmental organizations.

The main purpose of this study is to provide a broad overview of health inequalities topics considered relevant by different public health organizations. Nevertheless, the focus of this review is on countries/regions with similar socioeconomic and political status to Catalonia. It may also be useful for other organizations who decide to study or monitor health inequalities to accomplish its very first step: topic selection. In addition, gaining some insights about which health issues are being prioritized, as well as which indicators were used, are considered secondary goals.

Material and methods

First, a bibliographic search was performed using PubMed, Google Scholar, and Google search engine with the terms “health inequalities”, “health observatories” and “health inequalities indicators”. Occasionally, names of concrete regions, countries, or organisations were added to these terms (i.e., “Andalucía health observatory” or “Canada health inequalities”). The search was performed from March to June of 2019. Once finished, a set of inclusion criteria was applied; studies included in the review had to:

  1. 1.

    Include health inequalities indicators: All the reports that contained no health indicators were automatically discarded (i.e., policy frameworks [21]).

  2. 2.

    Have been carried out by a governmental organisation or a related entity, whether at an international, national, or regional level: the reports not published by governmental (or government-related) organisations were discarded.

  3. 3.

    Have a socioeconomic and political status similar (or highly related) to Catalonia: some reports were discarded due to significant differences in the socioeconomic profile of the countries they were studying in comparison to Catalonia or Spain.

Once the reports were selected, the authors performed a quality control check of the indicators shown in the reports and databases. The indicators had to match the basic anatomy of an indicator as a minimum requirement to be considered an indicator. This basic anatomy consists of containing data, i.e. the numerical data input; and containing good metadata, like a title and an explanation of how an indicator is defined and calculated [22]. In addition, the different reports found were classified according to the geo-political region they were studying: 1. international, 2. national, and 3. regional (Table 1).

After this selection process, the indicators were grouped into topics by semantic matching of their definition as well as by the area of knowledge there are intended to measure. Most of the topics were supported by references of relevant organisations like the WHO or the United Nations (UN) (see Table 2). Each topic was uniquely named in accordance with the area of knowledge that instruments were intended to measure. Indicators from different sources were often merged due to high similarities between them (most commonly, the only differences were stratifiers such as age, gender or region). Every topic had to be formed by at least five indicators in order to be considered relevant enough; the topics with less than five indicators were discarded.

The search was performed by the two researchers, and the results shared in order to agree on any discrepancies. All the data was organised in spreadsheets to identify common indicators, and then sorted by the amount of indicators they included. The names of the indicators in the spreadsheets were those given in the original reports or their metadata information.

Results

In total, 21 reports, projects, and databases were identified and classified into three categories: 1. international [8], 2. national [10], and 3. regional [3]. In the first category, international, all the projects selected were carried out or funded by the European Commission [23,24,25,26,27], the WHO [28, 29] or the World Bank [30]. In the following category, national, studies were conducted by health agencies or governments of countries such as Andorra [31], Australia [32], Canada [33, 34], England [35, 36], Scotland [37], Slovenia [38], Spain [39] or Portugal [40]. In the last category, regional, some reports published by Spanish regions were included (Andalucía [41], Barcelona [42], Valencia [43]). A total of 691 health indicators were identified (Table 1).

Table 1 Reports and datasets identified in the search
Full size table

Following an iterative process of evaluation, we identified a core set of 120 candidate topics, of which 34 were finally selected (Fig. 1). Table 2 describes a complete list of 34 topics with the corresponding definition of each topic. The ten most commonly used were: “Life expectancy”, “Infant mortality”, “Obesity and overweight (BMI)”, “Mortality rate”, “Regular smokers/tobacco consumption”, “Self-perceived health”, “Unemployment”, “Mental well-being”, “Cardiovascular disease/hypertension”, “Socioeconomic status (SES)/material deprivation”. However, some topics that ranked below these were closely related with some of the most common topics; for example, “Perinatal, neonatal and stillbirths mortality” might be considered as a subtype of “Infant mortality”; and “Perceived mental health” is similar to both “Mental well-being” and “Self-perceived health”. Furthermore, some indicators may represent antithetically the same area of knowledge; that is the case for the indicators in the topic “Long-term limitations/chronic illnesses” and the indicator “Healthy Life Years (HLY)” [within the topic “Life expectancy”], where the former (long-term limitations or chronic illnesses) is used to determine the latter (the end of a healthy life condition). Nevertheless, the metadata of the health indicators within each topic was highly homogeneous: they all had similar definitions and methodology.

Fig. 1
figure 1

Flowchart of the processes undertaken to review the most common topics in the study of health inequalities

Full size image
Table 2 List of topics ranked by the number of health indicators grouped within
Full size table

In general, the indicators within each topic were very similar. In fact, often the differences among them were related to the different stratifiers (such as sex, age or region) used for their calculation. For example, in the first topic “Life expectancy”, indicators have different variations: “Life expectancy at birth” [23,24,25, 27,28,29, 32, 33, 37, 40,41,42], “Life expectancy at birth by sex” [22, 24,25,26,27, 33, 34], “Life expectancy at a certain age” [25, 33, 35, 40], “Life expectancy by educational attainment level” [25], and “Life expectancy at birth by socioeconomic status” [25]. Complex measures of inequality, such as “Slope index of inequality (SII) for male and female life expectancy” [34] may be considered another (more advanced) variation.

Furthermore, in the case of a health outcomes or diagnostics (such as mortality or cancer) the concrete disease or cause may also play an important role in the heterogeneity found among the health indicators. Indicators are focused on different aspects, such as prevalence and incidence, mortality, preventive measures, or treatments. For example, for the topic “HIV”: “HIV incidence” [23, 26, 28, 33, 40], “Prevalence of HIV, male/female, by ages” [29, 39] and “AIDS-related mortality rate” [28, 39] are the most common, yet “Antiretroviral therapy (ART) coverage” [28] and “HIV test results for TB patients (positive results)” [28] can also be relevant.

The results of the study show that the most common topics are related to:

  • Mortality/life expectancy: “Life expectancy” [44, 45], “Infant mortality” [46] or “Mortality rate” [47] are widely used to study health inequalities.

  • Incidence/mortality rates of specific diseases: “Cardiovascular disease/hypertension”, “Cancer (incidence or mortality)” [48], “Diabetes/insulin resistance” [49], “HIV” [50], “Tuberculosis (TB)” [51] or “Respiratory diseases”.

  • Social determinants of health [52]:

    • ◦ “Living and working conditions” where this could be studied at an individual level, was highly ranked: “Unemployment” [53] and “Primary studies/illiteracy”. Otherwise, these indicators were at the bottom of the list.

    • ◦ This was similar for “Individual lifestyle factors and social and community networks” topics, which can also be studied at an individual level: “Obesity and overweight (BMI)” [54, 55], “Regular smokers/tobacco consumption” [56], “Alcohol consumption” [57], “Hazardous alcohol consumption”, “Physical activity” [58], and “Food consumption (vegetables, fruit, salt)”.

    • ◦ Socioeconomic level: “Socioeconomic status (SES)/material deprivation” [59, 60].

    • ◦ Healthcare system: “Healthcare resources” and “Policy and Legislation”.

Discussion

Main results of the study

The results of the study showed that the most common topics were related to mortality/life expectancy, incidence/mortality rates of specific diseases (i.e., TB or HIV), and social determinants of health, such as living and working conditions, and individual lifestyle factors and social and community networks, according to Dahlgren and Whitehead’s model of the social determinants of health [52]. The indicators that can be studied at an individual level tended to be highly ranked, in comparison to those that are studied at different levels (such as hospital or region), which tended to be at the bottom. Many methodological differences between indicators were due to stratifiers in their calculation.

Study selection criteria

To include health inequalities indicators was a fundamental requirement for any report in order to be included in the review. Hence, although some reports provided in-depth insights about tackling health inequalities (i.e., [21, 61]) they were not selected due to lack of health indicators monitoring.

In addition, to be carried out by a governmental or a government-related organisation was also an important requirement, as many academic and/or private institutions carry out studies and reviews of health inequalities, but their policy-making influence is limited. Their reports tend to be focused on concrete knowledge fields, such as gender influence [62], the effects of economic crises [63], or access to healthcare [64]; which are also relevant for the study of health inequalities but whose authorship does not fit the selection criteria, as the main interest of this paper is identifying the health inequalities indicators used by health agencies or similar government-related entities. The reports not produced by this kind of organisation were rejected.

Lastly, to have a socioeconomic and politic status similar or highly related to Catalonia criteria was intended to exclude reports whose health indicators were adapted to least developed/developing countries where, for example, access to treatments of diarrhea for infants may still be an issue [65]. Hence, some reports were discarded due to significant differences in the socioeconomic profile of the countries they are studying in comparison to Catalonia or Spain.

These criteria were applied to all the reports and studies found after an extensive search. The addition of country/region names in the search responded to the need of knowing how particular regions of interest were dealing with health inequalities. Interest in regions was mainly based on previous knowledge of concrete public health organizations studying those regions, as well as interest in looking for other organizations in charge of tackling health inequalities in regions similar to Catalonia. Nevertheless, as in any review, it is not possible to ensure that absolutely all the reports suitable for this study were found during the search, nor that they were selected after applying the selection criteria.

Grouping health indicators into topics

As stated above, the most frequent health indicators were grouped into topics according to the health domain they were measuring (with each indicator related to only one topic). For example, although “Percentage of 15-year-olds who were overweight in 2009–10, EU Member States by sex” [25] and “Obesity rate by body mass index (BMI) (sdg_02_10)” [23, 66] are different indicators per se, they are both intended to measure the same health issue and, hence, were grouped under the same topic “Obesity and overweight (BMI)” under the indicator name “Obesity and/or overweight (total, by sex, age, or educational level)”.

Most of the selected health indicators were taken from the official statistics of different countries or international organisations, whose development and methodology has been closely consolidated over many years and respond to international standards. In addition, most of these indicators are related to relevant knowledge areas for the study of health inequalities, such as lifestyle habits, deprivation, and mortality.

To prioritise the most relevant topics, all the groups with less than five health indicators were deleted. This meant that, unfortunately, interesting topics such as the “Years of potential life lost” [34, 36, 42], “Unmet health needs” [23, 26, 39] and “Passive smokers” [33, 34] were not taken into consideration. Nevertheless, this selection does not imply per se a periodic monitoring of the selected health indicators, as specific topics not present in this list may be studied according to ultimate needs.

Relevance of the most common topics

All the topics aim to measure and study the relation between determinants of health and health outcomes. Interestingly, three of the top five topics in the list (see Table 2) are related to mortality: life expectancy, infant mortality, and mortality rate. Life expectancy at birth is an indicator of mortality conditions and, by proxy, of health conditions [44]. Hence, life expectancy as well as other mortality-related topics are widely used in the study of health inequalities.

Living and working conditions, such as BMI or smoking, are also key factors in the study of health inequalities as both share a strong socioeconomic gradient [54, 56]. Therefore, as may be expected, they appeared among the top 10 positions in the list of topics (see Table 2). According to our review, the most common way to measure socioeconomic status is to analyse the unemployment rates and material deprivation level of the population.

In Spain, some regions use less common health indicators for studying health inequalities that may be interesting for particular knowledge areas. For example, the Valencian Observatory of Health uses the “Caregiver profile”, which they report to be mostly women, without primary studies, 57-years-old on average, and reporting bad self-perceived health. In addition, they also use “Reasons why contraceptive methods are not used by age and nationality of women” as well as a “Sexual-health information resources (school, parents, friends, etc.)”, which may help to understand possible sexual health inequalities [43]. In the Andalusian School of Public Health, the indicator “Psychosis and mental illnesses due to drugs or alcohol abuse” may be helpful to estimate various negative health outcomes of alcohol and substance abuse that the healthcare system will need to address [41]. Even so, more than a half of topics in Table 2 appear in their reports.

As may be expected, the health indicators used in other reports produced by the Catalan Health System Observatory, such as the “Community Health Indicators”, match the implicit measurement concept behind many of the topics: obesity and overweight, mortality by age (including infant), self-perceived health, and population with primary studies are some of them [67].

Health indicators

As can be seen in Table 3, health indicators were combined within each topic if stratifiers such as population sex, age, or region were the only difference in their calculation methodologies. Some indicators were also merged if they were formerly different in the way they expressed the same data (i.e., raw number, rate per 1000 or 100,000).

Table 3 List of indicators by topic
Full size table

The health indicators within each topic often cover a different aspect relevant to health inequalities. For example, in the topic “Tuberculosis” [23, 29, 33, 34, 40, 42, 43] indicators about incidence, prevalence, or mortality can be observed. In addition, health indicators about treatment coverage or vaccination are also included in this topic. Overall, in most topics, indicators try to measure every relevant (and measurable) aspect of the topic.

Common stratifiers are sex, age, and studied region, something that is coherent with the determinants of health perspective and the focus on inequalities. However, the stratifiers found are highly heterogeneous and may also include socioeconomic status, educational level, or nationality/country of origin, among many others.

Research fitted to monitor health inequalities

This comprehensive review was carried out to help accomplish the first step in the process for tackling and monitoring health inequalities: selecting high-impact issues and health indicators [19, 20]. The next steps in the analysis will be to carefully take into account stratifiers such as area of residence, gender, age, and nationality. Lastly, after the identification of key health inequalities, decision-making stakeholders will need to play a role during the last steps: determining priorities of action and implementing changes. The time variable will play a key role in the monitoring, as it will indicate the possible health consequences of policy-making decisions [19, 20].

Conclusions

Reviewing the most common health indicators and topics used in the study of health inequalities may help research teams in different ways. First, having an overview of what is being done by their neighbouring countries or regions may highlight issues that should not be missed when selecting relevant health topics to study. Second, even if some topics might not ultimately be chosen as a priority of action, having a complete list of key issues will provide an overview of what is relevant in the study of health inequalities, as well as some interesting insights. Lastly, knowing what other research institutions are working on will promote potential collaborations between organizations, creating synergies and bonds that may lead to better understanding and monitoring of health inequalities.

At a regional level, these results are highly valuable for the first stages of health inequalities monitoring cycles in Catalonia. This study provided the basis for choosing health topics to study as well as helped gain insights about which indicators should be used. In addition, regions with similar socioeconomic status and goals in tackling health inequalities may benefit from this research. Similarly, at a national and international level these results may help organizations shift the focus towards undermined health inequalities topics or explore new areas of knowledge (yet unstudied or with a different perspective).