What did India do?

A variety of factors contributed to India’s progress in reducing neonatal and maternal mortality. For this study, these interrelated factors were mapped to a conceptual framework that consolidated factors from existing evidence and frameworks which included categories of distal, intermediate, and proximate drivers, as shown in Figure 7. In this section, progress on key proximate drivers is highlighted—especially in regard to intervention coverage spanning the continuum of maternal and newborn care, including preconception, antenatal care, delivery care, and postnatal care.

Where each indicator is discussed, its progress is contextualized within India’s four major national health policy periods, shown in Figure 8. More detailed descriptions of each of these policy periods—which influenced intermediate and distal drivers from our conceptual framework—can be found in the How Did India Implement section. Additional details of how the higher and lower mortality state clusters specifically achieved progress can also be found within their respective narratives.

After reviewing trends in selected key maternal and newborn health intervention coverage indicators, quantitative analyses are presented that link this progress to reductions in neonatal mortality rate (NMR) and maternal mortality ratio (MMR). This includes an analysis that isolates the impact of fertility decline on mortality levels using Jain’s decomposition method and an analysis that leverages the Lives Saved Tool to assess the impact of specific interventions on mortality trends. It also includes an analysis that calculates the fraction of neonatal mortality attributable to a variety of risk factors. While each distinct analysis comes with its own limitations, each also offers unique insights. Together, this body of evidence tells a story about how proximate factors contributed to India’s reductions in neonatal and maternal mortality.

Fertility decline has contributed to mortality reductions, in part by influencing related risk factors

India’s decline in fertility has, in part, been linked to trends in proximate factors, such as contraception access and uptake, as well as age at marriage. Importantly, shifts in intermediate and distal factors such as women’s education, employment, socioeconomic status, and other social determinants have also influenced the country’s fertility rate.¹ These factors contributed to evolving social norms around having smaller family sizes, evidenced in reductions in the total fertility rate, defined as the number of children a woman would have if she were to experience the prevailing fertility rates at all ages and survived throughout her childbearing years.

As shown in Figure 9, between 1990 and 2019, the total fertility rate in India decreased from 3.8 to 2.1 births per woman. ² Despite differing baseline levels, the higher and lower mortality state clusters showed similar trajectories of decline, with respective fertility levels at 4.7 and 3.0 births per woman in 1990, which decreased to 2.6 and 1.6 births per woman by 2019.²

Declines in fertility rates among younger women are central to India’s overall reductions in fertility. In 2019, among women ages 15 to 19, the age-specific fertility rate was 11.9 births per 1,000 woman-years, which is 70.5% lower than it was one decade prior.² For women ages 20 to 24, there was a 39.6% reduction during the same period, with the age-specific fertility rate at 124.9 births per woman-year in 2019.² These two age groups saw the largest relative fertility declines during that decade, which is notable considering the higher health risks for younger women and girls in pregnancy and childbirth.

Declines in fertility rate also have consequences for the broader health system. While India’s fertility declined, the number of births has stayed relatively steady in recent decades as a result of population momentum. Consequently, maternal and newborn health programs have been able to focus on a relatively consistent number of annual births and do not need to keep pace with a growing number of births per year. This trend of fertility decline is, therefore, interconnected with several key proximate factors in our study and is a crucial contextual factor underlying India’s trajectory of progress. Fertility decline may also be associated with a shift in the age-parity distribution of births that reduced maternal and newborn mortality risk, since risks tend to be higher with younger maternal age and higher parity.

Assessing fertility decline as a contributor to NMR/MMR reduction

India has experienced substantial decreases in fertility, as noted earlier, with the total fertility rate declining from 3.8 to 2.1 births per woman between 1990 and 2019.² As described in Jain’s decomposition methodology, declines in fertility translate to fewer high-risk pregnancies by way of longer birth intervals, lower birth parity, as well as decreased birth rates, specifically among adolescent girls and older adult women.³ This analytical approach isolates the impact of fertility decline on neonatal and maternal mortality reduction. This analysis found that between 2000 and 2018, fertility decline was estimated to account for 29% of the reductions in NMR and MMR.

Declines in fertility levels were associated with 537,337 fewer neonatal deaths and 37,330 fewer maternal deaths in 2020 than would have been expected if fertility levels had remained constant since 2000. This decomposition approach attributes the remaining portion of mortality reductions to “safe motherhood” initiatives, a term used to collectively refer to intentional interventions that target maternal and newborn health. Together, this suggests that fertility decline and improved coverage of safe motherhood initiatives led to 1,179,460 neonatal lives and 97,790 maternal lives saved in 2020 compared with what would have been expected if fertility rate and intervention coverage levels had remained constant since 2000. This finding, shown in Figure 10, highlights the impact of fertility decline as a key driver of NMR and MMR decline, in tandem with other health care indicators commonly associated with mortality reductions.

Improvements in proximate reproductive, maternal, and newborn health coverage indicators

The proximate maternal and newborn health intervention coverage factors described in this section were found to contribute most directly to improved maternal and newborn health outcomes in India during recent decades. They were also in turn influenced by broader distal and intermediate factors in the conceptual framework. Throughout this section, we mention upstream intermediate and distal factors that fed into this progress, which are described in more depth in the How Did India Implement and Context sections.

Contraception

With each successive national health policy period—described in the How Did India Implement section—family planning policies became less siloed and increasingly integrated into maternal and child health programming. As a result, progress toward satisfying demand for family planning has accelerated since the onset of the 2012–2020 RMNCH+A/NHM period. The most common method of contraception in India remains female sterilization, as it has been for the last several decades, but use of other modern methods has increased in recent years.^4,5,6,7,8

Over the last three decades, satisfied demand for family planning has increased slightly from high baseline levels, as shown in Figure 11. The 1992–1993 National Family Health Survey (NFHS)⁴ found that nationally 63.8% of demand for family planning among married women was satisfied using modern methods, which rose to 83.9% by the 2019–2021 NFHS.⁸ This most recent evidence also indicates that demand satisfied was higher in the lower mortality state cluster, at 90.3%, compared with 77.8% in the higher mortality state cluster.⁸ Although the coverage gap between higher and lower mortality state clusters persists, it almost halved between the 1992–1992 and 2019–2021 NFHS surveys, from 21.9% to 12.5%.^4,8

Antenatal care

Over the last several decades, antenatal care (ANC) coverage in India has increased rapidly. Progress in any ANC coverage—meaning that a woman received at least one ANC visit was particularly pronounced during the RCH I policy period from 1997 to 2005 as shown in Figure 12. Subsequently, advancements in ANC4+ coverage—meaning that a woman received at least four ANC visits—increased during the RCH II/NRHM policy period from 2005 to 2012. In 1990, 64.3% of women received at least one ANC visit while 28.0% of women received at least four ANC visits⁴; by 2018 these coverage levels had respectively increased to 93.7% and 58.6%⁸. The most recent 2019–2021 NFHS found similar coverage levels of any ANC and ANC4+, at 92.6% and 58.5%, respectively.⁸ The gap between the higher and lower mortality state clusters has narrowed substantially for any ANC coverage, while a considerable gap remains for ANC4+ coverage.

Although many factors contributed to India’s increased ANC coverage, a new cadre of community health workers introduced in the mid-2000s—called accredited social health activists (ASHAs)—has been identified as a key factor. In one study, visits by ASHAs during pregnancy was associated with a 17% increase in any ANC coverage.⁹ ASHAs identify pregnant women in the community, promote antenatal care-seeking, and even escort women to ANC clinics. ASHAs are also compensated through a results-based financing scheme that includes incentives related to coordinating local ANC.¹⁰ In addition to the influence of ASHAs, efforts to mitigate financial barriers to care have also been crucial to improving ANC coverage. While previous policies had specifically attempted to mitigate costs associated with institutional delivery, the 2014 expansion of the Janani Shishu Suraksha Karyakaram (JSSK) program was also intended to lessen financial barriers associated with ANC.¹¹

In addition to improving ANC coverage, India also made remarkable progress in enhancing the quality of ANC provided. To approximate quality of ANC, we considered an index that included nine components that generally reflect whether ANC was administered in a timely manner by a skilled provider who conducted key screenings, tests, and examinations as appropriate—collectively representing a maximum score of 13 points. Using that index, quality ANC (ANCq) was defined as ANC that achieved at least nine of thirteen points. In 1997, 37.5% of women in India received ANCq^5,12, but because of remarkable progress—especially during the 2005–2012 RCH II/NRHM policy period—this value increased to 85.3% by 2018⁸. The gap in ANCq between higher and lower mortality state clusters narrowed particularly quickly during the 2005–2012 RCH II/NRHM policy period and decreased from 43.2% in 1997 to 16.1% in 2018^5,12,8.

Institutional delivery

One of India’s most impactful shifts in maternal and neonatal health over the last three decades occurred between 2005 and 2012 during the RCH II/NRHM period, when the government prioritized the promotion of institutional delivery. Previous national health policy periods had included efforts to strengthen the capacity of traditional birth attendants in an attempt to make home deliveries safer.¹³ However, during the NRHM period, skilled birth attendance training for providers, such as auxiliary nurse midwives and staff nurses, improved the capacity of health workers to manage uncomplicated deliveries in health facilities and refer cases with complications to higher-level facilities.¹⁴ The NRHM also introduced the Janani Suraksha Yojana (JSY) cash incentive scheme, which provided financial incentives for mothers who delivered in facilities, as well as the ASHAs who supported institutional delivery.^15,16 The same study that found contact with an ASHA increased the likelihood of a woman attending ANC also found that exposure to an ASHA was associated with a 28% increase in the likelihood of giving birth in a health facility.⁹

These initiatives catalyzed major progress on institutional delivery coverage from 26.0% in 1990 to 90.0% by 2018 (Figure 13).^4,8 The most recent 2019–2021 NFHS⁸ found a similar institutional delivery rate, at 88.6% nationally, with most of the deliveries occurring in public facilities. As mentioned earlier, improvements in institutional delivery were especially sharp during the RCH II/NRHM period—even more so for the higher mortality state cluster. During this policy period from 2005 to 2012, the institutional delivery rate among the higher mortality state cluster increased rapidly, from 27.4% to 68.9%—an AARC of 13.1%.^6,17,7 This period saw a particularly rapid reduction in the institutional delivery coverage gap between higher and lower mortality state clusters, which decreased from 29.0% in 1990 to 10.7% by 2018.^4,8 Although this gap has narrowed overall, the types of facilities that women access vary between higher and lower mortality state clusters. In the higher mortality state cluster, deliveries occur predominantly in lower-level facilities such as community health centers (CHCs) or primary health centers (PHCs), whereas in the lower mortality state cluster, deliveries occur more often at tertiary facilities such as district or private hospitals.⁸

As access to institutional delivery rapidly increased throughout the 2005–2012 RCH II/NRHM period, the government of India also began emphasizing the importance of improving quality of care. One development partner spoke to this shift by saying:

Establishment of rigorous facility standards such as the 2006 Indian Public Health Standards and 2013 National Quality Assurance Standards helped to set norms for facilities providing delivery care.^18,19 Subsequently, the 2017 Labour Room Quality Assurance Initiative (LaQshya) provided guidance for facilities that was specifically focused on delivery care, with an emphasis on improving the patient’s experience.²⁰ These standardized criteria set quality and preparedness expectations for facilities, which were complemented by efforts to train and build the capacity of health workers.

Cesarean section

In line with increases in institutional delivery coverage, C-section rates in India also increased steadily, with substantial growth especially during the 2005–2012 RCH II/NRHM policy period as shown in Figure 14. In 1990, India’s C-section rate was 2.4%, but by 2018 this had increased to 22.4%.^4,8 Specifically during the RCH II/NRHM policy period, the country’s C-section rate nearly doubled, from 8.9% to 16.3%.^6,17,7 Much of this increase was driven by increases in C-section rates at private hospitals.^4,8 C-section rates vary considerably across state clusters; for example, in 2018, 13.7% of all deliveries in the higher mortality state cluster were C-sections compared with 34.5% in the lower mortality state cluster.⁸ Private facilities in each state cluster have historically had fairly similar institutional C-section rates, although rates have been slightly higher in the lower mortality state cluster. In contrast, institutional C-section rates in public facilities have risen more markedly in the lower mortality state cluster compared with the higher mortality state cluster.

In the public sector, a subset of existing community health centers (CHCs) were upgraded to become first referral units (FRUs), which helped bring comprehensive emergency obstetric and neonatal care (CEmONC)—including blood transfusions and C-sections—to settings beyond hospitals. These were established in the late 1990s, with operations streamlined during the RCH II/NRHM policy period, corresponding to the next period when C-section rates saw marked increases.²¹ To help fill specialist provider positions at FRUs, India established skills training programs for generalist doctors specifically targeted at teaching CEmONC skills, including lifesaving anesthesia skills to enable C-sections.²²

Postnatal care

Between the 1998–1999 NFHS and 2019–2021 NFHS, coverage of postnatal care (PNC) for the mother or newborn within two days of delivery improved from 13.4% to 82.8%, as shown in Figure 15.^5,8 There has also been a substantial narrowing of PNC coverage gaps between the higher and lower mortality state clusters. Between the two surveys, PNC coverage improved from 2.9% to 79.7% in the higher mortality state cluster and 26.6% to 87.0% in the lower mortality state cluster.^5,8 This resulted in a coverage gap that shrank from 23.7% to 7.3%, with particularly rapid progress in the higher mortality state cluster during the 2005–2012 RCH II/NRHM period.^5,8

In addition to rising institutional delivery rates, two key factors have contributed to improvements in PNC coverage and quality in India over recent decades. First, the role of ASHAs was expanded to include provision of home-based newborn care. A 2014 revision of India’s guidelines on home-based newborn care emphasized the role of ASHAs in making postnatal home visits, with one study finding that ASHA visits after delivery were associated with positive health outcomes.⁹

Secondly, the establishment of special newborn care units (SNCUs) provided infrastructure with the capability to provide more complex forms of newborn care in case of complications.²⁴ SNCUs are generally located in district hospitals and were made increasingly accessible after the establishment of the 2014 Janani Shishu Suraksha Karyakaram (JSSK) scheme that covered costs of referrals.²⁴

Modeling the effect of increased service coverage on maternal and newborn mortality

Analysis of lives saved by interventions

The Lives Saved Tool uses data on intervention coverage, global data on effectiveness of interventions, as well as MMR and NMR data to model the contribution of each intervention towards mortality reductions.²⁵ This approach estimates the number of lives saved each year as a result of improved intervention coverage compared with baseline levels in 2000.

The number of maternal lives saved as a result of improved intervention coverage is shown in Figure 16. The specific interventions that underwent the most consequential improvements in terms of saving maternal lives were parenteral administration of uterotonics, contraceptive use, and parenteral administration of antibiotics. Between 2000 and 2018, improved coverage of these interventions compared to 2000 levels was estimated to have saved 115,360, 87,992, and 56,718 maternal lives, respectively. Many key interventions emerging from this analysis are administered at the time of delivery, with clean birth environment, C-section, and antibiotics for preterm or prolonged rupture of membranes also emerging as key interventions.

The number of neonatal lives saved as a result of improved intervention coverage is shown in Figure 17. The interventions with the greatest impact on neonatal survival include case management of neonatal sepsis/pneumonia, C-section, and thermal protection. Between 2000 and 2018, improved coverage of these interventions compared to 2000 levels was estimated to have saved 988,807, 702,696, and 493,379 neonatal lives, respectively. Improved coverage of several other interventions that are crucial immediately after birth were also found to have saved many neonatal lives, including case management of premature babies, clean cord care, and neonatal resuscitation.

Disaggregating risk factors for neonatal mortality yields insights into demographic groups that have experienced rapid progress

Increased coverage of proximate interventions and shifts in contextual factors, such as fertility, are linked to NMR and MMR reductions in India. Analyzing disaggregated neonatal mortality trends can also shed light on how equity gaps between socioeconomic and demographic groups have closed and how this has contributed to overall progress. By assessing relative risks and the prevalence of risk factors over time, we can assess where and for whom India has been most successful in reducing NMR. These insights can shed light into which populations have experienced the most rapid progress, and which risk factors have been key in driving progress.

In this analysis, we calculated the population attributable fraction (PAF) for a series of factors, representing the proportion of neonatal mortality that can be attributed to each specific risk factor. Trends in PAFs over time can be linked to both a change in the relative risk of that factor and a change in the composition of that factor. In the context of this study’s conceptual framework, the factors considered in this analysis include proximate indicators (e.g., ANCq coverage and place of delivery), contextual factors related to fertility (e.g., age at birth, previous birth interval, and birth order), and individual assessments of intermediate and distal factors (e.g., urban residence, religion, wealth, and education).

In the 2005–2006 NFHS, the factors with the highest levels of attributable neonatal mortality were identified as low household wealth (32.6% of neonatal deaths), residence in rural areas (26.7% of neonatal deaths), short birth interval (25.4% of neonatal deaths), low ANCq coverage (24.6% of neonatal deaths), and lower levels of maternal education (17.2%).⁶ Births in these categories were found to be, respectively, 1.9, 1.5, 2.2, 1.6, and 1.4 times more likely to result in neonatal deaths.

Between the 2005–2006 NFHS and 2019–2021 NFHS, the PAFs for several key factors experienced substantial declines (Figure 18). For example, the PAF for low household wealth decreased from 32.6% to 26.6% of neonatal deaths. This aligns with overall narrowing of the NMR gap between households in the wealthiest and poorest tertiles. Similarly, the PAF for short birth intervals declined from 25.4% to 16.9%, reflecting overall fertility rate declines that are associated with longer birth spacing. Additionally, the PAF for lower levels of maternal education decreased from 17.2% to 9.7% of neonatal deaths. This finding reflects increased educational attainment among women and girls in recent decades. The largest relative decline in PAF between the 2005–2006 NFHS and 2019–2021 NFHS was for lower levels of ANCq, whereas in the 2005–2006 NFHS, 24.6% of neonatal deaths could be attributed to the lower ANCq levels, by the 2019–2021 NFHS, only 8.7% of neonatal deaths could be attributed to the lower levels of ANCq. This is linked to a rapid improvement in ANCq coverage scores, which increased from 45.7% in the 2005–2006 NFHS to 85.7% in the 2019–2021 NFHS.^6,8

Other factors saw limited improvement over the study period, such as residence in rural areas. The percentage of neonatal deaths attributable to residence in rural areas increased slightly from 26.7% in the 2005–2006 NFHS to 27.7% in the 2019–2021 NFHS.^6,8 While NMR in rural areas declined from 41.8 to 27.4 neonatal deaths per 100,000 live births between the two surveys, this improvement was slightly slower than the NMR decline in urban areas, from 28.1 to 18.0 neonatal deaths per 100,000 live births.^6,8 As a result, the PAF for residence in rural areas increased slightly over time.

While multiple interrelated factors have contributed to NMR reduction in India, this analysis highlights that narrowing wealth gaps, declines in fertility, improvements in female education, and provision of key maternal health services, such as ANC and institutional delivery, are associated with declines in neonatal mortality over time.