BACKGROUND India has made substantial progress in improving child survival over the past few decades, but a comprehensive understanding of child mortality trends at disaggregated geographical levels is not available. We present a detailed analysis of subnational trends of child mortality to inform efforts aimed at meeting the India National Health Policy (NHP) and Sustainable Development Goal (SDG) targets for child mortality. METHODS We assessed the under-5 mortality rate (U5MR) and neonatal mortality rate (NMR) from 2000 to 2017 in 5 × 5 km grids across India, and for the districts and states of India, using all accessible data from various sources including surveys with subnational geographical information. The 31 states and groups of union territories were categorised into three groups using their Socio-demographic Index (SDI) level, calculated as part of the Global Burden of Diseases, Injuries, and Risk Factors Study on the basis of per-capita income, mean education, and total fertility rate in women younger than 25 years. Inequality between districts within the states was assessed using the coefficient of variation. We projected U5MR and NMR for the states and districts up to 2025 and 2030 on the basis of the trends from 2000 to 2017 and compared these projections with the NHP 2025 and SDG 2030 targets for U5MR (23 deaths and 25 deaths per 1000 livebirths, respectively) and NMR (16 deaths and 12 deaths per 1000 livebirths, respectively). We assessed the causes of child death and the contribution of risk factors to child deaths at the state level. FINDINGS U5MR in India decreased from 83·1 (95% uncertainty interval [UI] 76·7-90·1) in 2000 to 42·4 (36·5-50·0) per 1000 livebirths in 2017, and NMR from 38·0 (34·2-41·6) to 23·5 (20·1-27·8) per 1000 livebirths. U5MR varied 5·7 times between the states of India and 10·5 times between the 723 districts of India in 2017, whereas NMR varied 4·5 times and 8·0 times, respectively. In the low SDI states, 275 (88%) districts had a U5MR of 40 or more per 1000 livebirths and 291 (93%) districts had an NMR of 20 or more per 1000 livebirths in 2017. The annual rate of change from 2010 to 2017 varied among the districts from a 9·02% (95% UI 6·30-11·63) reduction to no significant change for U5MR and from an 8·05% (95% UI 5·34-10·74) reduction to no significant change for NMR. Inequality between districts within the states increased from 2000 to 2017 in 23 of the 31 states for U5MR and in 24 states for NMR, with the largest increases in Odisha and Assam among the low SDI states. If the trends observed up to 2017 were to continue, India would meet the SDG 2030 U5MR target but not the SDG 2030 NMR target or either of the NHP 2025 targets. To reach the SDG 2030 targets individually, 246 (34%) districts for U5MR and 430 (59%) districts for NMR would need a higher rate of improvement than they had up to 2017. For all major causes of under-5 death in India, the death rate decreased between 2000 and 2017, with the highest decline for infectious diseases, intermediate decline for neonatal disorders, and the smallest decline for congenital birth defects, although the magnitude of decline varied widely between the states. Child and maternal malnutrition was the predominant risk factor, to which 68·2% (65·8-70·7) of under-5 deaths and 83·0% (80·6-85·0) of neonatal deaths in India could be attributed in 2017; 10·8% (9·1-12·4) of under-5 deaths could be attributed to unsafe water and sanitation and 8·8% (7·0-10·3) to air pollution. INTERPRETATION India has made gains in child survival, but there are substantial variations between the states in the magnitude and rate of decline in mortality, and even higher variations between the districts of India. Inequality between districts within states has increased for the majority of the states. The district-level trends presented here can provide crucial guidance for targeted efforts needed in India to reduce child mortality to meet the Indian and global child survival targets. District-level mortality trends along with state-level trends in causes of under-5 and neonatal death and the risk factors in this Article provide a comprehensive reference for further planning of child mortality reduction in India. FUNDING Bill & Melinda Gates Foundation; and Indian Council of Medical Research, Department of Health Research, Ministry of Health and Family Welfare, Government of India.

中文翻译：

---

印度5岁以下儿童和新生儿死亡率趋势的国家以下地区制图：2000-17年全球疾病负担研究。

背景技术在过去的几十年中，印度在改善儿童生存方面取得了实质性进展，但是尚无法获得对各个地区儿童死亡率趋势的全面了解。我们将对儿童死亡率的国家以下趋势进行详细分析，以期为达到印度国家卫生政策（NHP）和儿童死亡率的可持续发展目标（SDG）目标而做出的努力。方法我们使用来自包括调查在内的各种来源的所有可访问数据，评估了印度各地以及印度各区和邦2000年至2017年5×5 km网格内的5岁以下儿童死亡率（U5MR）和新生儿死亡率（NMR）包含次国家级地理信息。根据其社会人口指数（SDI）等级，将31个州和工会领土组分为三类，根据25岁以下女性的人均收入，平均教育程度和总生育率计算，作为“全球疾病，伤害和危险因素负担研究”的一部分进行计算。使用变异系数评估了州内各州之间的不平等。我们根据2000年至2017年的趋势预测了到2025年和2030年各州和地区的U5MR和NMR，并将这些预测与NHP 2025和SDG 2030的U5MR目标进行了比较（分别为每1000例生命中23例死亡和25例死亡）。 ）和NMR（每1000个活产分别有16例死亡和12例死亡）。我们在州一级评估了儿童死亡的原因以及危险因素对儿童死亡的贡献。结果印度的U5MR从2000年的83·1（95％不确定区间[UI] 76·7-90·1）降低至2017年的每1000个活产42·4（36·5-50·0），而NMR从38每1000个活产儿·0（34·2-41·6）至23·5（20·1-27·8）。2017年，U5MR在印度各州之间的变化为5·7倍，在印度723个地区之间为10·5倍，而NMR分别为4·5倍和8·0倍。在SDI较低的州，2017年有275（88％）个区的U5MR为每千个活产40个或更多，而291（93％）个区的NMR为每千个活产20个或更多。2010年至2010年的年变化率2017年各区之间的差异从9·02％（95％UI 6·30-11·63）减少到U5MR没有明显变化，从8·05％（95％UI 5·34-10·74）减少NMR没有明显变化。从2000年到2017年，各州之间的地区不平等现象在U5MR的31个州中的23个州和NMR的24个州中有所增加，在低SDI州中，奥里萨邦和阿萨姆邦的增加幅度最大。如果到2017年观察到的趋势继续下去，印度将实现SDG 2030 U5MR目标，但不能达到SDG 2030 NMR目标或NHP 2025目标之一。为了单独实现SDG 2030的目标，U5MR的246（34％）个地区和NMR的430（59％）地区需要比2017年更高的改善率。对于印度5岁以下死亡的所有主要原因，死亡率在2000年至2017年之间有所下降，尽管各州之间的下降幅度差异很大，但传染病下降幅度最大，新生儿疾病下降幅度中等，先天性出生缺陷下降幅度最小。儿童和孕产妇营养不良是主要的危险因素，在印度，可能导致68·2％（65·8-70·7）的5岁以下儿童死亡和83·0％（80·6-85·0）的新生儿死亡归因于2017年; 5岁以下儿童死亡的10·8％（9·1-12·4）可归因于不安全的水和卫生设施，而8·8％（7·0-10·3）归因于空气污染。口译印度的儿童存活率有所提高，但各州之间死亡率下降的幅度和速度存在很大差异，印度各地区之间的差异甚至更大。对于大多数州来说，州内各地区之间的不平等现象有所加剧。这里介绍的地区级趋势可以为印度为降低儿童死亡率以实现印度和全球儿童生存目标所需的有针对性的努力提供重要指导。本条中的地区级死亡率趋势以及5岁以下和新生儿死亡原因的州级趋势以及危险因素，为进一步规划印度降低儿童死亡率提供了全面的参考。比尔和梅琳达·盖茨基金会 印度政府卫生与家庭福利部卫生研究部印度医学研究理事会。