Abstract

Introduction

Antenatal care (ANC) is essential for improving maternal and neonatal health outcomes, and its utilization is influenced by socioeconomic factors. This study aims to assess disparities in ANC service utilization by wealth, caste/ethnicity, and province among Nepalese women, using data from the nationally representative Nepal Multiple Indicator Cluster Surveys (MICS) 2014 and 2019.

Methods

We assessed the association of ANC service utilization with the household wealth index quintiles, caste/ethnicity and province using multivariable logistic and negative binomial regression models. We also measured wealth-related inequality using concentration curves and concentration indices.

Results

The proportion of women who received four or more ANC visits increased from 60.6% in 2014 to 77.8% in 2019. However, in both 2014 and 2019, women in the highest wealth quintile were over five times more likely to receive recommended ANC visits than those in the lowest quintiles. Similarly, the expected number of ANC visits was 70% higher in 2014 and 35% higher in 2019 for women in the highest wealth quintile compared to the lowest quintile. Concentration curves showed a decrease in pro-rich inequality in ANC utilization by 2019 relative to 2014, though notable inequality remained. Geographic disparities were evident: Karnali and Madhesh provinces had significantly lower utilization of recommended ANC visits and fewer ANC visits than Koshi Province. Ethnic disparities were also prominent, with women from Dalit, Disadvantaged Janajati, and Other Madheshi caste groups being significantly less likely to complete recommended ANC visits.

Conclusion

Our findings reveal persistent wealth-related inequality in ANC service utilization in Nepal, with women from wealthier households, advantaged caste/ethnic groups and better-off provinces having higher odds of receiving recommended ANC visits and a higher number of ANC visits. Concerted efforts are needed to address these equity gaps in ANC service utilization, particularly for women from low-income households, disadvantaged caste/ethnic groups, and underserved provinces.

Introduction

The importance of ensuring equitable access to high-quality maternal and neonatal health services in achieving universal health coverage (UHC) and meeting sustainable development goals (SDGs) by 2030 cannot be overstated [1]. Despite global progress in maternal and child health outcomes, maternal health indicators remain poor in many low- and middle-income countries (LMICs), including Nepal [2]. Nepal鈥檚 maternal mortality ratio of 174 per 100,000 live births is one of the highest in the world [2], making it imperative to enhance the quality and coverage of maternal healthcare services to meet the SDG targets [3,4,5].

Antenatal care (ANC) is an integral component of safe motherhood, providing pregnant women with health-promoting, preventive, and curative interventions while serving as a gateway to a continuum of care for mothers and newborns [3,4,5,6]. Early detection and treatment of potential obstetric complications during ANC can reduce maternal mortality and morbidity [3, 4, 7]. Moreover, ANC directly reduces stillbirths and newborn deaths by ensuring optimal fetal growth and development [8].

In Nepal, ANC services are widely available free of cost at all levels of government health institutions, and cash incentives are provided to encourage the completion of four or more ANC visits [9,10,11]. Nepal has implemented demand-side financing schemes since 2005, including the Aama Program, which was introduced in 2009. The Aama Program eliminated fees for institutional delivery and ANC services and provided cash incentives for women who completed the recommended ANC visits and delivered in a health facility [11]. These policies align with the National Health Policy 2014, the recent National Health Policy 2019, and the Nepal Health Sector Strategy 2015鈥�2020, which aim to improve maternal and newborn health by enhancing the effectiveness, sustainability, and equity of public health activities [9, 12]. Moreover, Nepal updated its ANC guidelines in 2022, aligning with the World Health Organization (WHO) recommendations, by changing the standard schedule from four focused ANC visits to eight ANC contacts to promote better service coverage [4, 13, 14].

Despite these efforts, studies have shown that disparities in access to and utilization of ANC services persist between different socioeconomic groups and geographic regions in Nepal [15,16,17,18,19]. Geographic inequities, ethnic disparities and wealth-related inequalities have been identified as major markers of inequity, not only in Nepal but also in other LMICs [20, 21]. Notably, between 2014 and 2019, Nepal experienced notable socioeconomic development, exemplified by a reduction in multidimensional poverty index from 30.1% to 17.4% [22]. Additionally, Nepal adopted a new constitution in 2015, establishing a federal structure and designating the provision of basic health services as the responsibility of local governments. To understand if these socioeconomic developments and health sector decentralization are translated into improved utilization of ANC services and a reduced inequalities across wealth, geographic, and ethnic groups, it is necessary to study the relationship between the socioeconomic status (SES) of women and their ANC service utilization in Nepal during this period.

Although previous studies have assessed the general socioeconomic predictors of ANC visits and their quality in Nepal [15, 23], none have comprehensively analyzed wealth-related inequality alongside geographic and ethnic dimensions using consecutive Nepal Multiple Indicator Cluster Surveys (MICS) data. By addressing these gaps, this study contributes insights into the intersection of socioeconomic development and ANC service utilization. Specifically, this study aims to assess the relationship between wealth, geography, and caste/ethnicity with ANC service utilization among Nepalese women of reproductive age using data from the 2014 and 2019 Nepal MICS. In addition, this study descriptively assessed the coverage and quality of ANC services received. By shedding light on these relationships and observing changes in ANC service utilization alongside changes in the inequality, this study provides insights into addressing disparities in ANC service utilization, thereby promoting equitable access to and utilization of ANC services in Nepal.

Methods

Data source

We conducted a secondary data analysis using data from two consecutive nationally representative surveys, the MICS 2014 and MICS 2019, carried out by the Central Bureau of Statistics (CBS) in collaboration with the United Nations Children鈥檚 Fund (UNICEF) as a part of the Global MICS program. Both surveys used a two-stage cluster sampling design to select a nationally representative sample of women of reproductive age (15鈥�49 years). In the MICS 2014, 28 strata were used based on geographic region and urban鈥抮ural classification, whereas MICS 2019 used 15 strata based on provincial and urban鈥抮ural classifications. Enumeration areas (EAs) or primary sampling units (PSUs) were identified in each stratum and data were collected from randomly selected PSUs. In MICS 2014, 500 PSUs were sampled, whereas in MICS 2019, 504 PSUs were sampled. In both surveys, systematic random sampling was used to select households within the sampled PSUs. Questionnaire-based interviews were conducted with women who had given birth in the two years preceding the survey.

A total of 2,086 women and 2,519 women who gave birth to a child two years preceding the survey participated in the MICS 2014 and MICS 2019, respectively. After excluding observations with missing values on ANC visits, the final analysis included 2,045 (weighted sample size =鈥�2,009) women from the MICS 2014 and 2,518 (weighted sample size =鈥�1,949) women from the MICS 2019 (Fig.听1). Details about the survey methodology are available in the full survey reports of Nepal MICS 2014 [24] and Nepal MICS 2019 [25].

Fig. 1

Flowchart of selection of study participants for analysis

Variables

Outcome variables and their measurement

To assess the utilization of ANC services, we used two primary outcome variables as described below.

• I) Receipt of recommended ANC visits

The first outcome, receipt of recommended ANC visits, was defined as a binary variable indicating whether a woman received at least four antenatal care visits during her last听pregnancy (Yes =鈥�1 vs No =鈥�0).

• II) The number of ANC visits

The second outcome variable, number of ANC visits, was a discrete count outcome variable defined as the total number of ANC visits that a woman received during her last pregnancy.

Independent variables and their measurement

In this study, we selected the household wealth index, caste/ethnicity and province of residence as the key independent variables. We utilized the household wealth index as a measure of long-term living standards because data on more preferred methods such as direct measures of income, expenditure, or consumption were not available in the surveys [26]. The household wealth index was constructed by the MICS team using a principal component analysis of information on access to electricity, type of fuel used for cooking, ownership of consumer goods and durables, dwelling characteristics, and the status of water, sanitation, and hygiene. The wealth index was used to rank households from the poorest to the richest and reflects long-term household SES, although it does not capture absolute income or current consumption. In the descriptive and regression analysis, we grouped the household wealth index into quintiles; for the concentration curves and concentration indices, we used raw household wealth index. Further details on the construction of the wealth index can be found in the Nepal MICS 2014 [24] and Nepal MICS 2019 survey reports [25].

Caste/ethnicity was another key independent variable, reflecting the social stratification prevalent in Nepal. These categories were aligned with Nepal's social structure, grouping populations into broad groups such as Brahmin (Hill Brahmin and Madheshi Brahmin), Chhetri (including Sanyasi and Dasnami), Dalit (Hill Dalit, Madheshi Dalit, and Other Dalit), relatively advantaged Janajati (Newar, Gurung and Thakali) and relatively disadvantaged Janajati (Hill Janajati, Madheshi Janajati, and Other Janajati), Other Madheshi caste, and Muslim/unspecified groups [27]. This categorization was intended to capture the socio-cultural and historical disparities among groups and their potential associations with health and socioeconomic outcomes.

Province of residence was included as an independent variable only for the 2019 survey, as Nepal adopted a federal structure in 2015 with the promulgation of the new constitution. Due to the lack of GIS data in the 2014 survey, it was not feasible to retrospectively assign 2014听survey participants to provinces.

Covariates and their measurement

We selected covariates based on their theoretical relevance and the availability of data in the MICS datasets. The major covariates included women鈥檚 sociodemographic characteristics, household characteristics, and type of residence.

Women's sociodemographic characteristics included age, education, child鈥檚 birth order, and media exposure. Age was measured in completed years and grouped into four categories: 15鈥�19, 20鈥�24, 25鈥�29, and 30 +鈥墆ears. Education was classified into three categories according to the categorization by the Ministry of Education: none, basic (grade 1鈥�8), and secondary or higher (grade 9 or higher) [28]. The birth order of the child was classified into two categories: first child or not. Mass media exposure was measured by whether a woman reported listening to radio, reading newspapers, or watching television at least once a week; women were classified as exposed to media or not.

Household characteristics included household size and the sex of the household head. Household size was the number of individuals in the household, classified into two categories: 鈮も��4 and 鈮モ��5. The sex of the household head was recorded as male or female.

Settlement of residence was classified into two categories: urban or rural, based on the administrative classification.

We included the survey year as a covariate in the pooled data analysis. The survey year indicated the year of the survey (2014 or 2019).

Statistical analyses

Data preparation and descriptive analysis

We merged the household dataset with the women's dataset using R software for statistical computing and prepared the dataset for the final analysis. For descriptive analysis, we calculated the mean (and standard deviation) and median, [with 25 th percentile (P₂₅) and 75 th percentile (P₇₅)] for continuous variables, and the number and percentage for categorical variables. Differences in proportions between 2014 and 2019 were tested using two-sample z-tests. Differences in the median number of ANC visits between survey years (within categories of the independent variables and covariates) were assessed using survey-weighted Wilcoxon rank-sum tests.

Regression analysis

We conducted separate cross-sectional regression analyses using the MICS 2014 and MICS 2019 datasets, in addition to the regression analyses using the data pooled (appended) from both surveys. We used multivariable logistic regression analysis with the survey package in R to explore the association between receipt of recommended ANC visits and wealth index quintiles, caste/ethnic group, and provinces [29]. For the count outcome, the number of ANC visits, we used multivariable negative binomial regression analysis in Stata version 14.1. Negative binomial regression is appropriate for count data, particularly when overdispersion occurs, where variance exceeds the mean because this model includes an additional dispersion parameter, ensuring robust standard errors and valid inference [30]. We exponentiated the regression coefficients and interpreted those as odds ratios and incidence rate ratios for logistic regression and negative binomial regression, respectively. We also presented their corresponding 95% confidence intervals (CIs) along with p-values. A two-sided p-value of less than 0.05 was considered statistically significant in all the analyses. All analyses accounted for survey design elements including survey weights, stratification, and clustering. For the pooled analysis, we adjusted survey weights by de-normalizing them to reflect the actual population distribution of women of reproductive age (15鈥�49 years) during the corresponding survey years, using data from the United Nations Population Division [31, 32].

Concentration curves and concentration indices

We examined wealth-related inequality in ANC service utilization in Nepal using concentration curves and concentration indices. Concentration curves were used to illustrate the extent to which ANC service utilization was unequally distributed across household wealth in 2014 and 2019. The concentration curve plots the cumulative percentage of the health variable (y-axis) against the cumulative percentage of the population, ranked by living standards, beginning with the poorest and ending with the richest (x-axis) [26]. We plotted separate concentration curves to visualize the distribution of the receipt of recommended ANC visits and the number of ANC visits by wealth index. A concentration curve below the diagonal line, also called the line of perfect equality, indicates that ANC visits are concentrated among the rich, also called pro-rich inequality, whereas a curve above the diagonal line indicates that they are concentrated among the poor, also called pro-poor inequality [26].

Furthermore, we calculated the concentration indices as summary measures of of wealth-related inequality. The concentration index ranges from 鈭�1 to 1, with 0 indicating perfect equality, negative values indicating that ANC service utilization is concentrated among the poor, and positive values indicating that it is concentrated among the rich [26]. To calculate the concentration index for receipt of recommended ANC visits, which is a binary outcome, we used the Erreygers-corrected method [33]. Similarly, for the number of ANC visits, we used the generalized concentration index proposed by Wagstaff et al. [34]. The concentration indices were calculated using the conindex command in Stata version 14.1 [35]听for both outcomes. The concentration indices were computed for the receipt of recommended ANC visits and the number of ANC visits in 2014 and 2019.

Ethical considerations

Ethical review and approval were waived for this study because the survey protocols for MICS 2014 and MICS 2019 were approved according to the Statistical Act of Nepal, 1958. The Statistical Act enables the CBS to conduct surveys according to the government鈥檚 ethics protocol without involving an institutional review board. The datasets obtained from the MICS did not include any personal identifiers.

Patients and public involvement

As this study is based on the analysis of secondary data provided by MICS, it was not possible to directly involve patients or the public in the design, conduct, or reporting of our research.

Results

Comparison of recommended ANC visits between 2014 and 2019

Between 2014 and 2019, the percentage of women receiving the recommended ANC visits increased across all sociodemographic groups. Younger women (15鈥�19 years) showed the largest improvement, rising from 50.4% in 2014 to 78.9% in 2019 (p鈥�< 0.001). Similar upward trends were observed across all age groups. Women with higher educational attainment were more likely to receive recommended ANC, with the percentage increasing from 79.4% to 89.3% among those with secondary or higher education (p鈥�< 0.001). Recommended ANC visits also increased significantly among women who gave birth to their first child with coverage rising from 68.2% to 84.1% (p鈥�= 0.014). A similar significant increase was observed in women with and without exposure to mass media (Table听1).

Table 1 Percentage distribution of women of reproductive age who had recommended ANC visits during the pregnancy of the most recent live birth in Nepal in 2014 and 2019

Caste/ethnic disparities persisted, though improvements were evident across groups. The highest coverage was observed among Brahmin women, increasing from 76.0% to 93.7% (p鈥�< 0.001), while Dalit women also experienced a significant increase from 52.8% to 72.2% (p鈥�< 0.001). Moreover, wealthier women had consistently higher ANC coverage, with the highest quintile reaching 94.5% in 2019 (p鈥�= 0.094), whereas the lowest quintile showed the greatest improvement, from 41.8% to 69.6% (p鈥�< 0.001) (Table听1, Fig.听2).

Fig. 2

Distribution of recommended ANC visits and number of ANC visits by wealth index quintile and caste/ethnic group in Nepal, 2014 and 2019

Women in smaller households (鈮� 4 members) had higher ANC coverage compared to those in larger households in 2019, though both groups experienced significant increases over time. Women in male-headed households also showed improvement (60.2% to 76.6%, p鈥�< 0.001). Rural鈥搖rban disparities persisted, with rural women showing a notable increase from 57.0% to 72.4% (p鈥�= 0.019), while urban coverage remained high but slightly declined from 85.0% in 2014 to 80.7% in 2019 (p鈥�= 0.015) (Table听1).

Overall, the percentage of women receiving the recommended ANC visits significantly increased from 60.6% (95% CI: 56.9, 64.3) in 2014 to 77.8% (95% CI: 75.5, 80.1) in 2019 (p鈥�< 0.001), indicating substantial progress in ANC utilization over time (Table听1).

The proportion of women attending at least one ANC visit was universally high across provinces in 2019, exceeding 90% in all regions. Madhesh and Karnali provinces had the lowest coverage for at least one ANC visit with 90.8% and 91.5% of women attending at least one ANC. Gandaki reported the highest coverage, with 98.4% of women receiving at least one ANC visit, closely followed by Sudoorpashchim (98.3%) and Bagmati (97.3%). However, stark disparities emerged in the proportion of women achieving the recommended ANC visits. Gandaki had the highest coverage for recommended ANC visits with 91.2%, while Madhesh and Karnali lagged significantly at 56.8% and 72.3%, respectively (Fig.听3).

Fig. 3

Distribution of at least one ANC visit and recommended ANC visits by province in Nepal in 2019; A Percentage of women with at least one ANC visits; B Percentage of women with recommended ANC visits

The mean number of ANC visits increased across all wealth quintiles and caste/ethnic groups between 2014 and 2019. Women from the highest wealth quintile had the highest mean ANC visits in both years (5.67 in 2014 and 5.74 in 2019), while those in the lowest wealth quintile had the lowest (2.6 in 2014, increasing to 3.59 in 2019). Among caste/ethnic groups, Brahmin and Advantaged Janajati women had the highest mean ANC visits (4.6 and 5.42 in 2014, increasing to 5.25 and 5.55 in 2019, respectively), while Dalit women had the lowest, increasing from 3.39 in 2014 to 3.81 in 2019. In 2019, Bagmati had the highest mean ANC visits (5.51), whereas Karnali had the lowest (3.61). Nationally, the mean ANC visits increased from 3.8 in 2014 to 4.33 in 2019 (Table听2).

Table 2 Mean and Median number of ANC visits by pregnant women of reproductive age in 2014 and 2019 in Nepal

Components of ANC services

A substantial increase in access to ANC tests was observed across wealth quintiles from 2014 to 2019, with the highest improvements noted in the lowest wealth quintile for all three tests (blood pressure, urine test, and blood test). For example, in the lowest wealth quintile, the proportion of women receiving blood pressure checks increased from 87.3% (95% CI: 83.1, 90.6) in 2014 to 90.4% (95% CI: 86.6, 93.2) in 2019, while blood test coverage rose from 50.5% (95% CI: 44.3, 56.8) to 79.8% (95% CI: 75.5, 83.6). The most pronounced improvement was seen in the proportion receiving all three recommended ANC tests, where coverage increased from 46.5% (95% CI: 40.5鈥�52.5) in 2014 to 76.2% (95% CI: 71.6鈥�80.3) in 2019.

Brahmin women had near-universal coverage for all ANC tests in 2019 (blood pressure: 99.4%, urine test: 98.1%, blood test: 98.3%, and all three: 97.2%). In contrast, Dalit women had significant improvements, particularly in receiving a blood test (62.8% in 2014 to 83.7% in 2019), although they still lagged behind other groups. Janajati (Advantaged) women had the highest coverage for all tests in 2019, with 98.8% (95% CI: 96.2鈥�99.6) receiving blood pressure checks and 96.8% (95% CI: 93.3鈥�98.5) receiving all three recommended tests.

Similarly, Gandaki showed the highest overall ANC test coverage in 2019 among provinces, with blood pressure checks at 97.1% (95% CI: 90.6鈥�99.2) and all three tests at 95.1% (95% CI: 89.7鈥�97.8). Conversely, Karnali had the lowest coverage, with only 83.6% (95% CI: 76.6鈥�88.8) of women receiving blood pressure checks, and just 68.8% (95% CI: 61.3鈥�75.5) completing all three tests (Supplementary table S1).

Multivariable regression analysis of the receipt of recommended ANC visits

The logistic regression models showed that the odds of receiving recommended antenatal care visits increased progressively across wealth quintiles in both survey years and in the pooled analysis. In 2014, compared to women in the lowest wealth quintile, the adjusted odds ratio (aOR) for receiving recommended ANC visits was 1.71 (95% CI: 1.15, 2.52) in the second quintile, 2.37 (95% CI: 1.47, 3.83) in the middle quintile, 3.10 (95% CI: 1.87, 5.14) in the fourth quintile, and 5.80 (95% CI: 3.08, 10.9) in the highest quintile. Similar patterns were observed in 2019, with the aORs ranging from 1.89 (95% CI: 1.27, 2.82) in the second quintile to 5.58 (95% CI: 2.69, 11.6) in the highest quintile. Pooled analysis revealed consistent associations, with the odds increasing from 1.48 (95% CI: 1.14, 1.92) in the second quintile to 5.23 (95% CI: 3.25, 8.42) in the highest quintile.

Disparities in ANC utilization were evident across ethnic groups, as听Dalit women, women from the Disadvantaged Janajati group, Other Madheshi caste group and Muslim and other group showed consistently lower odds of receiving recommended ANC visits. In 2014, women from the Disadvantaged Janajati group (aOR: 0.70; 95% CI: 0.42, 1.16) and women from Other Madheshi caste group (aOR: 0.48; 95% CI: 0.25, 0.89) had reduced odds compared to Brahmin women. In 2019, Dalit women (aOR: 0.51; 95% CI: 0.27, 0.94), Muslim and others (aOR: 0.43, 95% CI: 0.22, 0.85) and Other Madheshi caste groups (aOR: 0.36, 95% CI: 0.19, 0.66) had significantly lower odds of receiving recommend ANC visits. Pooled analysis revealed similar trends, with Dalit women, Disadvantaged Janajati women, women from Muslim and other and Other Madheshi caste groups persistently underutilizing recommended ANC services.

Substantial provincial differences were also observed in ANC utilization in 2019. Compared to Koshi Province, women in Madhesh Province were significantly less likely to receive recommended ANC visits (aOR: 0.43; 95% CI: 0.28, 0.66). In contrast, women in Sudoorpaschim Province exhibited the highest odds (aOR: 3.08; 95% CI: 1.82鈥�5.22), followed by Gandaki Province (aOR: 1.65; 95% CI: 0.96鈥�2.84), though the latter association was not statistically significant (Table听3).

Table 3 Association of household wealth index quintile, province of residence and caste/ethnicity with receipt of recommended ANC visits among women aged 15鈥�49 years in Nepal

Multivariable regression analysis of the number of ANC visits

The incidence rate ratios (IRR) from negative binomial regression models for the number of听ANC visits revealed substantial disparities across wealth quintiles, caste/ethnic groups, and provinces.

Women in higher wealth quintiles exhibited progressively higher adjusted incidence rate ratios (aIRR) for receiving ANC visits. In 2014, compared to women in the lowest quintile, the aIRR was 1.26 (95% CI: 1.11, 1.43) in the second quintile, 1.46 (95% CI: 1.28, 1.67) in the middle quintile, 1.49 (95% CI: 1.31, 1.69) in the fourth quintile, and 1.70 (95% CI: 1.48, 1.96) in the highest quintile. Similar trends were observed in 2019, with the aIRR increasing from 1.08 (95% CI: 1.01, 1.16) in the second quintile to 1.35 (95% CI: 1.22, 1.49) in the highest quintile. The pooled analysis demonstrated consistent patterns, with the highest quintile achieving an aIRR of 1.55 (95% CI: 1.42, 1.69).

Marked disparities were evident among caste/ethnic groups. In 2014, Disadvantaged Janajati women (aIRR: 0.92; 95% CI: 0.83, 1.02) and women from Other Madheshi caste groups (aIRR: 0.88; 95% CI: 0.76, 1.00) were less likely to receive recommended ANC visits compared to Brahmin women. By 2019, Dalit women (aIRR: 0.91; 95% CI: 0.84, 0.99) and women from Muslim and other ethnic groups (aIRR: 0.88; 95% CI: 0.79, 0.98) continued to demonstrate reduced number of ANC visits. Pooled analysis reinforced these findings, with other Madheshi caste groups exhibiting an aIRR of 0.86 (95% CI: 0.79, 0.93), highlighting persistent inequities. Advantaged Janajati group had the higher aIRR compared to Brahmin group in all the regression models.

Provincial disparities in the number of听ANC visits were pronounced in 2019. Compared to women in Koshi Province, those in Madhesh Province had a reduced aIRR of 0.93 (95% CI: 0.84, 1.01), while women in Bagmati Province exhibited higher IRR (aIRR: 1.11; 95% CI: 1.02, 1.20) (Table听4).

Table 4 Association of household wealth index quintile, province of residence and caste/ethnicity with the number of ANC visits among women aged 15鈥�49 years in Nepal

Wealth-related inequality in the receipt of recommended ANC visits

The concentration curves shown in the left panel of Fig.听4 illustrate the distribution of the receipt of recommended ANC visits for 2014 and 2019. In both years, the concentration curve was below the line of equality. This indicates that the distribution of receiving recommended ANC visits was unevenly skewed toward wealthier population groups. However, a noticeable shift toward the line of equality was observed in 2019 compared to 2014. This finding suggests a reduction in the equity gap for receiving recommended ANC visits between the two years.

Fig. 4

Concentration curves for recommended ANC visits and number of ANC visits

To quantify this trend, we utilized the Erreygers normalized concentration index (CIx), which provides a measure of the extent of pro-rich or pro-poor inequality. The results indicate that for 2014 and 2019, the CIx values were 0.352 (standard error:0.031; p鈥�< 0.001) and 0.175 (standard error:0.234; p鈥�< 0.001), respectively. This suggests that there is a pro-rich inequality in receiving recommended ANC visits, indicating that women with higher SES are more likely to receive recommended ANC visits. Nevertheless, the observed reduction in the CIx for 2019 indicates a reduction in the magnitude of the pro-rich inequality and an improvement in equity in the distribution of receiving recommended ANC visits.

Wealth-related inequality in the number of ANC visits

The concentration curves shown in the right panel of Fig.听4 illustrate the distribution of the number of ANC visits for 2014 and 2019. It is clear that in both years, the concentration curve lies below the line of equality. This indicates that the distribution of the number of ANC visits is unevenly skewed towards the wealthier population groups. However, a shift toward the line of equality was observed in 2019 compared to 2014. This suggests a reduction in the equity gap for the number of ANC visits between the two years.

The CIx for the number of ANC visits in 2014 and 2019 were 0.145 (standard error:0.011; p鈥�< 0.001) and 0.091 (standard error:0.009; p鈥�< 0.001), respectively. This suggests that there is a pro-rich inequality in the number of ANC visits, which implies that women with higher SES are more likely to receive a higher number of ANC visits. Nevertheless, the observed reduction in the CIx in 2019 compared to 2014 indicates a reduction in the magnitude of the pro-rich inequality in the number of ANC visits.

Discussion

We measured the wealth-related, caste/ethnic and provincial inequalities in ANC service utilization among women of reproductive age in Nepal using two outcome variables: receipt of recommended ANC visits and the number of ANC visits. The findings of this study reveal that service coverage rates and quality of ANC service increased between 2014 and 2019, although marked wealth鈥�, caste/ethnic-, and region-based disparities were observed in the utilization of ANC services in both years. We found that women from poorer households, disadvantaged caste/ethnic groups and historically underserved provinces (Madhesh and Karnali) were less likely to receive recommended ANC services, and have a lower number of ANC visits. Encouragingly, the equity gap narrowed over time: the concentration index became less pro-rich in 2019 compared to 2014, indicating that ANC utilization has become more equitable. Nevertheless, an appreciable pro-rich pattern persists, meaning wealthier women still enjoy higher ANC coverage than poor. These results are consistent with previous studies showing a positive association between household wealth and ANC service utilization in LMICs [36,37,38].

Several interrelated factors help explain why wealthier and听higher-caste women have better ANC uptake. Women from wealthier households are more likely to have financial resources, higher education, exposure to mass media, and autonomy in healthcare decisions, all of which facilitate the use of maternal health services. They also tend to live in areas with better health infrastructure and transportation. In contrast, poorer and marginalized women often face geographical barriers (remote terrain, lack of roads), indirect costs (transport, lost wages), and social obstacles (lower decision-making power, potentially even exposure to intimate partner violence) that hinder access to care [19, 23, 39,40,41]鈥�. Moreover, wealthier women are more likely to benefit from government efforts, such as the Aama program [10].

Nepal is a lower-middle-income country and its health system has been characterized as weak and under resourced [42, 43]. Access to healthcare is particularly challenging for individuals residing in remote areas, and the quality of healthcare services varies widely across different geographical locations [42]. Despite these challenges, Nepal has made remarkable progress in improving maternal and child health status in recent decades. The Nepal Demographic and Health Survey (NDHS) 2011 showed that the proportion of women who received 鈮モ��4 ANC visits during the five years prior to the survey was 50.1% [44]. Our study shows that receipt of recommended ANC visits increased to 60.6% in 2014, which indicates an increase in 鈮モ��4 ANC visits between 2011 and 2014. Similarly, in our study, the percentage of women who received recommended ANC visits increased to 77.8% in 2019, an increase from 69.4% of women who had 鈮モ��4 ANC visits in 2016, according to NDHS 2016 [45]. Moreover, the proportion of women who had no ANC visits in 2011 was 15.2% [44], which decreased to 12.3% in 2014 based on our analysis (results not shown in the text). This proportion continued to decrease to 4.5% in 2016 [45] and 2.3% in 2019, according to our analysis, reflecting a sustained decreasing trend.

The observed improvement in the utilization of ANC services from 2014 to 2019 can be attributed partly to several concerted efforts by government and development partners, alongside broad improvements in living standards. During 2014鈥�2019, Nepal experienced rapid socioeconomic progress, with the multidimensional poverty rate falling from about 30% to 17% [22]. These findings are consistent with previous studies conducted in other LMICs, which have also demonstrated an association between improved SES and increased utilization of maternal health services [21].

Our findings on the positive association between household wealth, relatively advantaged ethnicity, and better-off provinces with the receipt of recommended antenatal visits and the frequency of ANC visits in Nepal are consistent with several previous studies. For instance, a study using the 2016 NDHS data revealed that women in the upper 60% of the wealth group had significantly higher rates of maternal care visits, including ANC, institutional delivery, and postnatal visits than women in the lower 40% group [46]. Likewise, a separate study analyzing the 2016 NDHS data showed that women from the middle and richest tertiles were 1.3 and 1.5 times more likely to complete 鈮モ��4 antenatal visits than women from the poorest wealth tertile [47]. A similar pattern emerged in the analysis of a nationally representative survey in India, where women from the richest wealth tertile were 2.99 times more likely to receive 鈮モ��4 antenatal visits than their counterparts in the poorest wealth tertile [48]. An analysis of the 2016 NDHS data established the household wealth index as a significant predictor of ANC visit numbers in Nepal, with the highest wealth tertile exhibiting an adjusted prevalence ratio of 1.27 in comparison to the lowest wealth tertile [23]. Likewise, a comprehensive study across 36 Sub-Saharan countries corroborated this trend, revealing that women in the middle and richest wealth tertiles were 1.32 and 1.38 times more likely, respectively, to have 鈮モ��8 ANC visits than their counterparts in the poorest wealth tertile [37]. Consistent with our findings, previous studies have consistently shown that women from Brahmin, Chhetri and advantaged Janajati ethnic groups have significantly better health outcomes, including ANC coverage, compared to Dalit, disadvantaged Janajati groups and other caste groups in Madhesh region of Nepal [16, 40]. Madhesh and Karnali Provinces of Nepal have the lowest Human Development Indices (HDIs), at 0.51 and 0.538, respectively, while Bagmati and Gandaki Provinces boast the highest HDIs at 0.661 and 0.618, surpassing the national average [49]. Madhesh and Karnali provinces face persistent challenges in education, particularly among women, and in ensuring equitable access to health services [49]. Moreover, these provinces also have lower availability of health service providers, lower adherence to standard/protocols for antenatal services, and lower availability of basic client services in general, potentially contributing to lower utilization of ANC services [50]. Karnali Province is additionally severely affected by rugged geographical terrain, which hinders geographical access to health facilities.

Our study suggests a pro-rich inequality in ANC service utilization, as indicated by the concentration curves and concentration indices. It is concerning that women from poorer households have lower utilization of ANC visits despite having a greater need for services. However, we note that the concentration curves are progressively approaching the line of equality, which offers some encouragement. Our findings are consistent with the results from several previous studies. A study that used concentration curves to analyze the wealth-related disparities in maternal health service utilization from 2001 to 2016 in Nepal found that the coverage of 鈮モ��4 ANC visits is disproportionately higher among women in richer households [15]. The same study also reported a consistent decrease in the concentration index from 2001 to 2016 in 鈮モ��4 ANC visits, with the most substantial gains occurring between 2011 (CIx: 0.22) and 2016 (CIx: 0.08) [15]. Furthermore, a study encompassing 12 countries in the Southern African Development Community region, which utilized the concentration index, revealed that 11 out of 12 countries had wealth-related inequality in the number of ANC visits, including 鈮モ��4 ANC visits [51].

Strengths and limitations

A strength of this study is the use of recent, nationally representative survey data from two time points with sufficiently large sample size, which enhances the generalizability of the findings. Additionally, we used appropriate statistical tools including concentration curves and concentration indices to quantify the wealth-related inequality in ANC service utilization. However, our study has some limitations. The self-reported data on ANC service use used in this study may have been subject to recall bias or social desirability bias. Moreover, the associations of independent variables with ANC service utilization may not be interpreted as causal. Finally, our focus was on quantitative service coverage; aspects of care quality were not assessed in depth, though they are undoubtedly an important component of effective ANC.

Policy implications

Our findings have important policy implications for Nepal鈥檚 national health goals and international commitments. The persistent inequalities in ANC uptake identified by this study indicate that Nepal鈥檚 National Health Policy 2019 goal of听"universal and equitable access to health services",听particularly for underserved populations, is not yet fully realized [12]. To bridge these gaps, government must strengthen financial support programs like the Aama initiative and enhance their effectiveness among poorer and marginalized women through targeted awareness campaigns and active community engagement. Furthermore, with the recent shift to the World Health Organization's recommendation of eight ANC contacts per pregnancy under the Antenatal service and postnatal service continuum听guidelines (2022) [4], Nepal must urgently address barriers that disadvantaged communities face in accessing these increased service contacts. Without additional support, disparities may worsen, given the greater ease with which wealthier urban women can access frequent visits compared to poorer rural counterparts.

These equity considerations are directly tied to Nepal鈥檚 broader commitment to achieving听UHC听and the SDGs by 2030. While recent statistics鈥�94% of women receiving at least one ANC visit and 80% achieving four or more visits鈥攊ndicate progress toward these targets [52], intensified efforts are needed to prevent the exclusion of the most vulnerable groups. Equity-oriented interventions, including targeted resource allocation to low-performing provinces such as Madhesh and Karnali, increased outreach ANC clinics, improved transportation infrastructure, telehealth options, and strengthened health system capacities in rural areas, are essential. Policymakers should also foster multisectoral collaboration, encompassing education and women's empowerment initiatives, to effectively address underlying socioeconomic barriers to ANC utilization. By embedding these targeted interventions within the existing policy frameworks of the National Health Policy 2019 and SDG-driven national plans, Nepal can significantly reduce maternal and neonatal health inequities, fulfilling its national and international health commitments.

Conclusion

The study reveals an increase in antenatal care (ANC) service utilization and a reduction in wealth-related inequality among women of reproductive age in Nepal between 2014 and 2019. However, significant disparities persist across caste, ethnic groups, and provinces. Women from wealthier households, Brahmin/Chhetri, and Advantaged Janajati groups, as well as those in Bagmati and Gandaki provinces were more likely to receive the recommended ANC visits and had higher overall utilization rates compared to women from disadvantaged households and provinces. These findings underscore the disproportionate concentration of ANC services among women from higher socioeconomic strata, emphasizing the urgent need for targeted interventions to enhance access to and utilization of ANC services among women from disadvantaged households in Nepal.