Equity in antenatal care quality: an analysis of 91 national household surveys

Background: Emerging data show that many low-income and middle-income country (LMIC) health systems struggle to consistently provide good-quality care. Although monitoring of inequalities in access to health services has been the focus of major international efforts, inequalities in health-care quality have not been systematically examined. Methods: Using the most recent (2007–16) Demographic and Health Surveys and Multiple Indicator Cluster Surveys in 91 LMICs, we described antenatal care quality based on receipt of three essential services (blood pressure monitoring and urine and blood testing) among women who had at least one visit with a skilled antenatal-care provider. We compared quality across country income groups and quantified within-country wealth-related inequalities using the slope and relative indices of inequality. We summarised inequalities using random-effects meta-analyses and assessed the extent to which other geographical and sociodemographic factors could explain these inequalities. Findings: Globally, 72·9% (95% CI 69·1–76·8) of women who used antenatal care reported blood pressure monitoring and urine and blood testing; this number ranged from 6·3% in Burundi to 100·0% in Belarus. Antenatal care quality lagged behind antenatal care coverage the most in low-income countries, where 86·6% (83·4–89·7) of women accessed care but only 53·8% (44·3–63·3) reported receiving the three services. Receipt of the three services was correlated with gross domestic product per capita and was 40 percentage points higher in upper-middle-income countries compared with low-income countries. Within countries, the wealthiest women were on average four times more likely to report good quality care than the poorest (relative index of inequality 4·01, 95% CI 3·90–4·13). Substantial inequality remained after adjustment for subnational region, urban residence, maternal age, education, and number of antenatal care visits (3·20, 3·11–3·30). Interpretation: Many LMICs that have reached high levels of antenatal care coverage had much lower and inequitable levels of quality. Achieving ambitious maternal, newborn, and child health goals will require greater focus on the quality of health services and their equitable distribution. Equity in effective coverage should be used as the new metric to monitor progress towards universal health coverage. Funding: Bill & Melinda Gates Foundation. We selected all Demographic and Health Surveys (DHS) and Multiple Indicator Cluster Surveys (MICS) done in LMICs in the past 10 years and included the most recent survey for each country (as of Jan 15, 2018). The DHS and MICS, respectively funded by the US Agency for International Development and the United Nations Children’s Fund, gather data on a range of population health indicators with a strong focus on maternal and child health. Standardised questionnaires ensure that data collected are comparable across countries. Sampling strategies and methodology have been described previously.17, 18 Our population of interest included all women of reproductive age (15–49 years) who had at least one livebirth in the past 2 years (MICS) or 5 years (DHS). In each country, we estimated antenatal care coverage as the proportion of women who had at least one antenatal care visit with a skilled provider during their last pregnancy. We used country-specific definitions of skilled providers as defined in the DHS and MICS. These included doctors, nurses, midwives, and country-specific skilled providers (such as maternal and child health aides in Sierra Leone and health extension workers in Ethiopia). Country-specific definitions of skilled antenatal care providers are available in the appendix. Guided by the framework of the Lancet Global Health Commission on High-Quality Health Systems in the SDG Era2 and by the WHO recommendations3 on antenatal care for a positive pregnancy experience, we assessed the availability of potential indicators of antenatal care quality in household surveys. Quality antenatal care involves the provision of respectful, evidence-based care including appropriate patient assessments such as history questions, examinations, and diagnostic tests (eg, full blood count testing and urine culture); appropriate preventive and curative treatments (eg, tetanus toxoid vaccination and iron supplementation); and patient counselling and education (eg, on healthy eating and signs of complications). In the DHS and MICS, women who reported attending antenatal care were asked whether they received specific services during consultations. We found 13 indicators related to antenatal care quality: weight and height measurement, blood pressure monitoring, urine and blood samples taken, HIV testing and counselling, tetanus vaccination, iron supplements, malaria prophylaxis, drugs for intestinal worms, counselling on signs of complications, and counselling on where to go in case of complications. The number of items collected varied across countries and only three indicators remained consistently measured in all countries and relevant in all contexts: blood pressure monitoring and urine and blood testing. To obtain a comparable measure across the largest possible set of countries, we limited our estimate of antenatal care quality to these three indicators. Antenatal care quality was therefore included as a binary outcome measuring the proportion of women who reported having their blood pressure checked and giving a urine and blood sample at any point during pregnancy among those who sought care from skilled providers. These three services do not comprise the full range of necessary antenatal care services and offer a limited view of antenatal care quality. However, they are recommended by WHO as essential components of antenatal care and are crucial to the detection of several pregnancy risks including hypertension, pre-eclampsia, infections, anaemia, and nutritional deficiencies.3 Our quality measure is also limited by the fact that information on the specific urine and blood tests done is not available in the DHS and MICS; women were asked whether they gave urine and blood specimens but not what the tests were for. As a sensitivity analysis, and given the importance of counselling for the detection of pregnancy complications, we also estimated antenatal care quality by including a fourth indicator available in 55 (60%) of 91 countries. This indicator measured whether women were counselled on potential danger signs to look out for during pregnancy or where to go in case of a complication. The survey questions for these four indicators are shown in the appendix. At the country level, we included gross domestic product (GDP) per capita and country income groups specific to the survey year based on World Bank classification as independent variables. At the individual level, we used the wealth index constructed by the DHS and MICS as an estimate of socioeconomic position. The wealth index is based on a household’s ownership of selected assets, housing construction materials, and types of water access and sanitation facilities and is estimated by principal component analysis. As a further independent variable at the individual level, we also used the woman’s educational attainment based on country-specific categories. Most surveys contained a variable with six categories (no education, attended primary, completed primary, attended secondary, completed secondary, or attended higher education). A few surveys used between three and five education categories, making inequality measured by education groups less comparable across countries. We also used the woman’s age at childbirth categorised into three groups (15–19 years, 20–35 years, and 35–49 years); her place of residence (urban or rural); region, state, or province of residence; and the total number of antenatal care visits attended (modelled as a continuous variable). To assess inequalities between countries, we ranked countries by levels of antenatal care coverage and quality and compared results across country income groups (low income, lower-middle income, and upper-middle income). We also plotted antenatal care coverage and quality against GDP per capita. To assess inequalities within countries, we ranked women using the wealth index and assigned a relative ranking based on their position in the cumulative wealth distribution. We then measured inequalities in antenatal care coverage and quality using the slope index of inequality (SII) and the relative index of inequality (RII).19, 20 Inequalities have been found to differ substantially when measured according to dimensions of inequality other than wealth;12 therefore, we also measured the SII and RII using the woman’s education and show these results in the appendix. The SII expresses the absolute percentage point difference in antenatal care coverage or quality between the predicted poorest and richest in the wealth distribution, assuming a linear relation between social rank and the outcome.21 The RII expresses the ratio of the predicted outcomes between the two extremes of the wealth distribution, assuming a log-linear relation.21 We used logistic regression models to estimate the association between the woman’s relative rank and her antenatal care outcomes. The SII and RII were obtained using marginal effects and the lincom and nlcom post-estimation commands in Stata version 14.2. Individual-level sampling weights and robust SEs were used in all regressions.17, 18 To summarise coverage, quality, and inequalities across countries, we pooled the estimates using random-effects meta-analyses weighted by the inverse variance of the estimates.22 We assessed heterogeneity across countries using I2 statistics. Finally, to assess the extent to which wealth-related inequalities in antenatal care quality could be explained by other geographical and socioeconomic determinants, we sequentially added five variables to the wealth rank in the regression models used to estimate the SII and RII: the woman’s education and age group, urban versus rural residence, subnational region, and total number of antenatal care visits (because women who attend more visits might have more opportunities to receive the three services we assessed). The study sponsors did not have any role in study design, data analysis, data interpretation, writing of the report, or submission for publication. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.