Why are Madhya Pradesh, Odisha, Rajasthan, and Uttar Pradesh Exemplar states?

Exemplar countries for neonatal and maternal mortality are defined as those that have demonstrated exceptional progress at reducing neonatal mortality rate (NMR) and maternal mortality ratio (MMR)—beyond what can be attributable to their socioeconomic progress alone. More information about how India was identified as an Exemplar country is available in the national-level India report.

Within India, states were grouped into two clusters based on their baseline per capita income and mortality levels in 2000, according to India’s Sample Registration System (SRS).¹ The higher and lower mortality state clusters were then both assessed to identify which states within each cluster had achieved the most rapid mortality declines. As shown below in Table 1, the average annual rate of change (AARC) for both NMR and MMR decline were summed for each state, and states with the largest summed AARC values were chosen as Exemplar states. Ultimately, Rajasthan showed the most substantial summed AARC for all higher mortality cluster states, followed by Odisha, Uttar Pradesh, and Madhya Pradesh. In terms of MMR decline, Rajasthan experienced the fastest reduction of any higher mortality cluster state, whereas for NMR, Odisha displayed the quickest decline. Table 1 below also shows the rates of change in lower mortality cluster states for comparison.

Mortality trends for India’s higher mortality state cluster are described throughout this section, contextualized within India’s four major national health policy periods, shown below in Figure 2. These include the 1992–1997 Child Survival and Safe Motherhood (CSSM) program, the 1997–2005 Reproductive and Child Health I (RCH-I) period, the 2005–2012 Reproductive and Child Health II/National Rural Health Mission (RCH-II/NRHM) period, as well as the 2012–2020 National Health Mission/Reproductive, Maternal, Newborn, Child and Adolescent Health (NHM/RMNCH+A) period. Additional information and context for each of these periods—especially the influential RCH-II/NRHM period—can be found in the How did India implement section of the national-level India report. This report focuses on state-specific trends and factors within the higher mortality cluster Exemplar states.

Maternal mortality

From 2000 to 2018, India’s national-level MMR experienced a decrease from 327 to 103 maternal deaths per 100,000 live births.¹ In the higher mortality state cluster, MMR declined from 461 to 145 maternal deaths per 100,000 live births over the time span—as shown below in Figure 3.¹ This translated to a substantially narrowed gap in MMR between the higher mortality state cluster and India overall.

The rates of MMR progress in the higher mortality Exemplar states of Madhya Pradesh, Odisha, Rajasthan, and Uttar Pradesh outpaced average progress experienced across the higher mortality state cluster. In particular, the higher mortality Exemplar states generally achieved accelerated progress in the 2012–2020 NHM/RMNCH+A policy period. Odisha, Rajasthan, and Uttar Pradesh respectively experienced AARCs of -8.2%, -9.1%, and -8.9% for maternal mortality during this period.¹ Progress in Madhya Pradesh was relatively consistent over time, with less variation across policy periods.

Although Figure 3 indicates that Madhya Pradesh and Uttar Pradesh both experienced slight increases in MMR around 2015, this spike is known to reflect changes in the states’ geographic boundaries used to measure the mortality levels at different times by India’s Sample Registration System.

This study leveraged data from India’s Million Death Study (MDS) and the Institute for Health Metrics’ Global Burden of Disease (GBD) to examine how causes of maternal death differed across state clusters.^2,3 In the higher mortality state cluster, more maternal deaths resulted from hemorrhage and abortion-related complications, while fewer were due to sepsis, hypertensive disorders, and obstructed labor.

Nationally, India experienced a shift in recent decades from more direct causes of death such as hemorrhage and infection to indirect causes of death often associated with the underlying health status of the mother. This shift is described in more detail in the national-level India report. While the higher mortality state cluster has been found to have a higher proportion of maternal deaths attributable to direct causes, this percentage has declined in the higher mortality cluster over recent decades.

Neonatal mortality

India’s national NMR decreased from 43.2 to 22.7 neonatal deaths per 1,000 live births between 2000 and 2018.¹ During this period, NMR in the higher mortality state cluster declined from 48.6 to 28.3 neonatal deaths per 1,000 live births, as shown below in Figure 4.¹ The Exemplar states of Madhya Pradesh, Odisha, Rajasthan, and Uttar Pradesh were found to have particularly rapid NMR declines. Progress in higher mortality Exemplar states was most pronounced in the 2005–2012 RCH-II/NRHM and 2012–2018 NHM/RMNCH+A policy periods. As of 2018, neonatal mortality rates in Madhya Pradesh, Odisha, Rajasthan, and Uttar Pradesh were, respectively, 33.7, 31.0, 26.0, and 30.7 neonatal deaths per 1,000 live births.¹ Although these mortality levels are higher than India’s national mortality levels, the higher mortality cluster Exemplar states have achieved rapid progress in the past two decades and have narrowed the gap compared to the national NMR average.

As neonatal mortality rates declined in India’s higher morality state cluster, a greater portion of neonatal deaths occurred in the first two days of life. For example, the 2005–2006 NFHS survey found that 51% of neonatal deaths took place in the newborn’s first two days of life, but by the 2019–2021 NFHS the figure had increased to 67%.^4,5 This trend is similar in Madhya Pradesh, Odisha, and Rajasthan, where the percentages of neonatal deaths occurring in the first two days respectively increased from 53% to 61%, 49% to 59%, and 52% to 55% between the two surveys.^4,5

A related shift has occurred for neonatal causes of death across India’s higher mortality cluster. Across a subset of Indian states similar to the higher mortality state cluster defined by this study, India’s Million Death Study (MDS) examined cause-specific neonatal mortality.² The MDS found that from 2000 to 2015, the percentage of neonatal deaths caused by infections among these states halved, declining from 32% to 16%.² Birth trauma and asphyxia also declined substantially, constituting 20% of neonatal deaths in 2000, but 6% of neonatal deaths by 2015.² This shift was accompanied by a substantial increase in the percentage of neonatal deaths caused by preterm birth-related complications, which represented 22% of neonatal deaths in 2000 but 56% of neonatal deaths by 2015.²

Global Burden of Disease (GBD) estimates from the Institute for Health Metrics and Evaluation support these findings and indicate that the percentages of neonatal deaths due to infections in specific higher mortality Exemplar states have decreased markedly. For example, GBD estimates indicate that in Uttar Pradesh between 2000 and 2019, the percentage of neonatal deaths caused by infections decreased from 35% to 23%.³ This trend is mirrored in Madhya Pradesh, Odisha, and Rajasthan, where GBD estimates suggest that over this period the percentages of neonatal deaths caused by infections respectively decreased from 27% to 18%, 28% to 21%, and 33% to 25%.³ This trend in higher mortality Exemplar states mirrors broader shifts across India, in which a shrinking portion of neonatal deaths are the result of direct, often infectious, causes, and a growing portion are related to causes such as prematurity and low birthweight.

One key difference that emerged between the higher and lower mortality state clusters was progress with respect to the timing of neonatal death. In the higher mortality state cluster, neonatal mortality declined most quickly after the first days of life—days 3 to 27.¹ This evidence further suggests that causes like infections—which often result in mortality beyond the first days of life—have declined at rapid rates. In contrast, the lower mortality state cluster experienced more rapid progress during the first days of life, with more rapid progress in reducing intrapartum causes and prematurity.