Recent trends in maternal and child health inequalities in Latin America and the Caribbean: analysis of repeated national surveys

The 2030 Agenda for Sustainable Development includes 17 Goals (SDGs) [1], of which the third, or SDG3, consists of “ensuring healthy lives and promoting well-being for all at all ages” [2]. The wording of this goal, in particular the mention of “for all at all ages” echoes the overarching SDGs motto of “leaving no one behind” [3]. Consistently with the social determinants of health framework, the SDGs recognize that the pace of improvement will depend not only on changes in health services but also on broader societal changes to address inequalities related to wealth, education, residence, ethnicity and other drivers. The recent publication of the report of the Commission of the Pan American Health Organization on Equity and Health Inequalities in the Americas has highlighted the importance of social determinants in driving levels and distribution of health in the region and provided specific recommendations for multisectoral actions aimed at improving health status [4]. In consonance with the report, the Every Woman Every Child initiative for Latin America and the Caribbean (EWEC-LAC) has been providing further resources and technical guidance for countries to measure health equity and implement a multisectoral approach to women’s, children’s and adolescents’ health [5].

Although several Latin America and Caribbean (LAC) countries made important progress towards the achievement of the 2015 Millennium Development Goals (MDGs) [6], progress towards the SDGs is scarcely documented, as these goals were only defined in 2015. A special challenge faced by LAC is that it remains as one of the world regions with the highest levels of socioeconomic inequality [6], with large proportion of the population facing poverty and poor health outcomes [7‐9].

The MDGs were rightfully criticized for measuring progress solely at the national level, without considering whether specific population subgroups – for example, the poor, rural residents or ethnic minorities – were being left behind [10]. This concern was addressed by SDG 17.18, which claims for disaggregated analyses of national statistics in order to monitor inequalities related to wealth, education, ethnicity and other relevant dimensions. Disaggregation is particularly important because inequalities affect access to health services and health outcomes among women, adolescents, and children [11, 12]. Accordingly, the EWEC-LAC initiative defined a set of core indicators to monitor health inequalities in the region [13] and the Pan American Health Organization has pushed the boundaries of the SDGs by proposing equity-specific targets [14]. Regular and frequent monitoring of inequalities is even more relevant in the context of the COVID-19 pandemic that may lead to the reversal of recent progress [15‐18].

Two recent publications provided information on the baseline levels of SDG3-related inequalities in the LAC region by summarizing results on women, adolescents, and children from 21 countries with national surveys from 2011–2016 [19, 20]. The objective of the present analyses is to assess changes in the reported baseline levels, relying on data from recent surveys made available in 2018 or later for eight countries in the LAC region.

Eight LAC countries had a national survey carried out in 2018 or later and an earlier survey between 2010 and 2014 (Table 1). In total, 221,989 women and 152,983 children were sampled in the 16 surveys used for the analysis. The 8 countries studied account for aproximately 20% of all women aged 15 to 49 years and 19% of all children aged less than 5 years in the LAC region by 2020. The Human Development Index (HDI) in 2020 ranged from 0.621 in Honduras to 0.840 in Argentina; on average, HDI was 0.756 among the eight countries (0.755 for LAC region). Compared to the recent SDG baseline medians from 21 LAC countries [19], the 8 countries studied had statistically significant lower levels for SAB (P < 0.0001) and higher levels for NMR (P = 0.0282), and U5MR (P = 0.0428); likewise, they had lower SII for mDFPS (P = 0.0019), and higher SII for ANC4 (P = 0.0002), SAB and PNM (P =  < 0.0001), and AFR (P = 0.0434). Therefore, the 8 countries studied tended to show worse health indicator and wider inequalities than the rest of the region.

Table 2 shows the values of the seven coverage indicators and their corresponding SIIs in the baseline and endline survey, as well as the AARC for coverage and its 95% confidence intervals throughout the period of the study. When describing the results, we focus on countries for which the 95% intervals of the baseline and endline levels did not overlap, which are highlighted in green color when the change was significantly positive (increase in coverage or reduction in inequality) and in salmon color when the change was significantly negative (reduction in coverage or increase in inequality). Figure 1 shows coverage levels by wealth quintile (known as “equiplots”), in the most recent survey in each country. Corresponding P values are provided in Supplementary Table 2.

Compared to most other low- and middle-income countries in the world, those from the LAC region are characterized by relatively good performance in terms of women’s, adolescents’ and children’s health [29]. Our analyses of recent trends in eight countries, however, show that the performance of LAC countries is heterogeneous, and that important inequalities remain within several of them.

Three of the countries studied – Argentina, Costa Rica and Cuba – showed high baseline and endline coverage levels for most indicators and relatively narrow social inequalities with the exception of coverage with eight or more antenatal visits. For the two countries where it was possible to study time trends for most indicators, Cuba showed significant coverage reductions in CCI and in both antenatal care indicators, and Costa Rica in family planning and full immunization coverage. Data on the latter were not available for Cuba or Argentina. Results for these three countries must be interpreted in light of their relatively high baseline levels and therefore limited scope for further improvements, but the declines in coverage are noteworthy.

At the other extreme, four countries showed below-average baseline coverage and wider inequalities for most indicators: Guyana, Honduras, Peru and Suriname. For most indicators, these countries had substantial room for improvement. Of the eight countries included in our analyses, Peru was the top performer in terms of increasing coverage, reducing stunting and mortality, and narrowing down socioeconomic inequalities. With positive trends over time observed for 10 of the 12 indicators, Peru was also the only country to show improvements in full immunization (increased coverage) as well as decreases in tobacco use by women and in adolescent fertility. The second positive outlier was Honduras where there were significant improvements in levels for six of the 12 indicators. Progress in Suriname and Guyana was not so clearcut, and the progress in Peru had already been highlighted in the literature as associated to an antipoverty political agenda [30]. On the negative side, Peru remained as a country with wider socioeconomic inequalities at endline than most countries in our analyses, as summarized by the SII.

Finally, the Dominican Republic did not quite fit in either of the two groups of countries described above. It had the highest under-five mortality rate of the eight countries, which is consistent with international data sources [31], yet prevalence of stunting was the second lowest after Cuba. Coverage was reasonably high for most indicators, except for full immunization and antenatal care (eight or more visits). Socioeconomic inequalities were also rather small except for adolescent fertility and antenatal care. In terms of progress over time, there were noticeable reductions in family planning and immunization coverage, the latter having peaked in 2006 at 59% and declining to 32% by 2019 (data not shown).

We now turn to a discussion of the most remarkable findings for each indicator. The two coverage indicators with the worst performance over time were family planning (with significant declines in three out of six countries with data) and full immunization (with declines in two out of five countries) [32, 33]. Earlier analyses of family planning in the region pointed to the need to revigorate national programs, particularly the use of long-acting reversible methods [34].

On the positive side, skilled attendance at birth increased in four out of seven countries with data and antenatal care (8 + visits) in three out of seven. There was hardly any progress in terms of the CCI, for which the only significant yet adverse change was a decline in Cuba.

Regarding outcomes other than coverage, stunting prevalence fell significantly in two countries, tobacco smoking by women and adolescent fertility only fell in Peru, whereas under-five mortality fell in three out of five countries with data [35].

The most remarkable socioeconomic inequalities were observed for adolescent fertility with all five countries with data showing five-fold or higher gaps between the richest and poorest quintiles. Earlier publications had already highlighted similar findings throughout the LAC region [36]. In most countries, inequality in coverage with eight or more antenatal care visits also were wider than those for four or more visits. This is not surprising given that, according to the literature, more demanding indicators, requiring a larger number of contacts with health workers, tend to be more inequitable as poorer women are likely to have less access [37, 38]. It is important to recall that WHO recommends at least eight prenatal care visits to reduce perinatal mortality and improve women’s experience of care [26].

Our analyses have limitations. Data for the analyses were only available for eight of the 33 LAC countries that account for only 19.6% of the regional population, and only four countries had complete trend data on the selected indicators. Using these data, it is not possible to provide a comprehensive picture of regional trends, particularly when the three most populous countries in the region – Brazil, Mexico, and Colombia – could not be included due to lack of comparable data in the 2 periods of analysis. Even among the eight countries, surveys from Argentina, Cuba and Suriname lacked several indicators. Time trend analyses are also affected by different time periods between the endline and baseline surveys; countries with only two surveys that are close in time such as Cuba and Guyana may provide less reliable time trend estimates than countries with three or more surveys spread throughout the whole decade. Peru has had annual surveys from 2010 to 2020 and this may be one of the reasons why changes over time were so marked, even in spite of a pandemic year. Furthermore, any non-linear evolution in the average or the inequality of particular health indicators experienced by countries over time cannot be captured by an analysis based on two points in time, as presented in our study.

Other limitations are related to the indicators under analysis. Data on antenatal, delivery and postnatal care refer to births that took place during the five (DHS) or two (MICS) years prior to the survey, thus failing to pick up recent progress. Maternal recall of events that took place up to five years before the interview may also be affected by long recall periods. Lack of precision may result from small sample sizes in some surveys for rare events (such as neonatal mortality), particularly when stratified by wealth quintiles. For maternal schooling, some groups (such as no education or primary only) are small in countries where most women completed secondary education, as in Cuba.

The strengths of the analyses include reliance on nationally representative, population-based surveys with consistent methodology and use of standardized indicator definitions. Results were based on reanalysis of the raw data carried out by a single team, and different dimensions of inequality were used.

We recognize at least two other limitations in data availability in our study. Firstly, there is a lack of standardized data collected on the quality of care in the region, with some studies estimating that poor quality of care may account for over half of treatable condition-related deaths [39]. Inequalities in the quality of care received by people in different strata could be important drivers of poor health outcomes. Secondly, most health surveys in the region only collect data from women, which limits comparisons across genders, as well as analysis of the implications of gender and health. Analyzing gender disparities in health is still a significant challenge in the region [40, 41] and is especially important for addressing risk factors such as, for instance, tobacco use and physical activity. These two relevant dimensions of inequality cannot be ascertained in our study.

Our paper is about a multicountry study aimed at describing recent inequality trends. Explaining the findings described here, and understanding the underlying mechanisms and drivers of success (or lack thereof) deserves in-depth studies, with different and specific study designs and the use of quantitative and qualitative methods to explore both the overall political and socioeconomic context as well as the effectiveness and potential impact of specific pro-equity policies put in place. Such in-depth studies are outside the scope of our paper, but we hope that our findings will drive further studies on those important questions.

Our findings, and the limitations of our analyses, highlight the pressing need for more frequent collection of data using population-based surveys to meet the requirement of SDG 17.18 for disaggregated analyses. The situation is even more pressing giving the fact that survey activities were interrupted in many countries starting in 2020 and in 2021 due to the COVID-19 pandemic. A new round of surveys in all 35 countries in the region is urgently needed in order to document the impact of COVID-19 on service coverage, nutrition and mortality, and to provide essential information for planning the recovery efforts.

Although LAC countries are well placed in terms of current levels of health indicators compared to most low- and middle-income countries, in those for which survey data for trend analysis is available important inequalities remain, and reversals are being observed in some areas. More targeted efforts and actions on the proximal and distal determinants of maternal and child health are needed in order to leave no one behind by 2030 and beyond. Monitoring progress with an equity lens is essential, but this will require further investment in conducting surveys routinely,including collecting standardized and comparable quality of care data, as well as in strengthening national and subnational health information systems in the region.