A SPATIAL ANALYSIS OF MATERNAL AND CHILD HEALTH IN UTTAR PRADESH, INDIA: EVIDENCE FROM NATIONAL FAMILY HEALTH SURVEY 4 (2015-16)

and Bundelkhand part of the state. PNC prevalence among the women was found to be highest in the western part of Uttar Pradesh. were represented in maps as spatial clusters. Those clusters were further characterized in terms of the relationships amid neighboring districts, high-high values, high-low values, low-high values, and low-low values. Low-low spatial association was found in Farrukhabad and Auraiya districts of central Uttar Pradesh and Jalaun, Banda and Chitrakoot districts of Bundelkhand region and Bahraich, Sarawasti, Balrampur and Gonda districts of Eastern Uttar Pradesh for ANC; Kanpur Dehat and Sitapur districts of Central Uttar Pradesh and Bahraich, Sarawasti, Balrampur and Gonda districts of Eastern Uttar Pradesh for PNC; Farrukhabad, Sitapur and Kheri districts of Central Uttar Pradesh and Bahraich, Sarawasti, Balrampur, Siddhartha Nagar and Gonda districts of Eastern Uttar Pradesh and Pilibhit, Shahajahanpur and Budaun districts of Western Uttar Pradesh for SBA; Sitapur district of central Uttar Pradesh and Bahraich, Sarawasti, Balrampur, Siddhartha Nagar and Gonda districts of Eastern Uttar Pradesh for full immunization; Jhansi, Jalaun, Hamirpur, Mahoba, Banda districts of Bundelkhand region and Kanpur Dehat, Barabanki, Sitapur, Raebareli districts of central Uttar Pradesh and Bahraich, Sarawasti, Balrampur, Siddhartha Nagar, Faizabad and Gonda districts of Eastern Uttar Pradesh for PNC among children aged 12-23 months. The high-high spatial association was found in Muzaffarnagar, Baghpat, Meerut, Ghaziabad districts of Western Uttar Pradesh and Mirzapur district of Eastern Uttar Pradesh for ANC; Muzaffarnagar, Baghpat, Meerut, Ghaziabad, Bijnor, Jyotiba Raophule Nagar, Rampur districts of Western Uttar Pradesh for PNC; Ghazipur, Azamgarh, Jaunpur and Sultanpur districts of Eastern Uttar Pradesh and Lalitpur, Jhansi, Mahoba, Jalaun, Hamirpur and Banda districts of Bundelkhand

Mother needs special medical care during pregnancy, delivery, and after delivery; as a mother, she is more prone to adverse health outcomes or death due to the unsafe and unhygienic methods of managing pregnancy and childbirth. According to the World Health Organization reports, globally, 0.53 million maternal deaths occur annually, out of which 0.12 million (22%) deaths occur only in India. A newborn child needs regular health check-ups as well as nutrition supplements to avoid deficiency diseases and illness. Child health is a foundation for adult health and well-being; therefore, it is imperative to certify good health. Healthy children assure healthy adults who, in turn, ensure good progress and development of the Country (Usmani and Ahmad, 2017). According to NHM, around 81% of under-five child death occurs in one year of birth that marks approximately 10.5 lakh newborn demises; however, 57% of under-five deaths occur in the first one month of life, constituting 7.3 lakh neo-natal deaths annually within the Country. Data and methods: data from National Family Health Survey-4 (2015-2016) on maternal and child health indicators for 75 districts of Uttar Pradesh state were used. Spatial analysis namely Moran's-I and LISA were applied to evaluate the maternal and child health indicators through all the districts of the state. Result: Each indicator portrayed prominent coverage variation across the Uttar Pradesh districts in this analysis. Among all the districts, the lowest ANC occurrence was observed mainly in the central part, PNC in the eastern region, SBA in 20 districts mainly of the western and eastern part. The prevalence of full immunization among the children was very high, primarily in the districts of the East region; high PNC among the children was perceived in the districts of eastern, central, Refining the quality of ANC, care at the time of giving birth, and postpartum care for mothers and their newborns are all crucial to avert these losses (WHO, 2017). According to the DHS, child health denotes the period between birth and five years old when children are chiefly susceptible to disease, illness, and death (Stallings and Rebecca, 2004). As per Census 2011, the children share (0-6 years) makes 13% of the entire population of India. An expected 12.7 lakh children decease each year before completing the age of five years (NRHM, 2007). However, 81% of under-five child mortality happens in one year of birth, which constitutes for approximately 10.5 lakh infant demises, although 57% of under-five deaths occur within the first one month of life constitutes for 7.3 lakh neo-natal deaths annually in the Nation (NHM, 2007). India is one of the highly significant contributors to the yearly worldwide tally of deaths of children under five years old -almost 1.1 million (WHO, 2012). The NFHS 2015-2016 spectacles that in India, 41 infants out of 1,000 deaths in the first year of being born (Vora, 2017). Globally, closely 20% of all deaths stirring annually amongst children under five years of age are vaccine-preventable ailments. Immunization has a vital part in attaining the aims stated in the Millennium Declaration (WHO, 2012).
The WHO recommends that expectant women obtain as a minimum four antenatal care visits (Wang, Alva, Wang, and Fort, 2011). Nevertheless, only 65% of mothers in developing nations utilized ANC, equated with 97% of mommies in developed countries (Dairo and Owoyokun, 2010). SBA deals with the proper medical and health worker's services during childbirth, essential for safe delivery. The type of aid during delivery has significant consequences for both the mother and the child. According to the reports of WHO, universally, 0.53 million maternal deaths occur every year, out of which 0.12 million (22%) deaths occur only in India (Fatema and Lariscy, 2020). Additionally, millions of women suffered pregnancy-related morbidity (Nair, 2011). Despite all efforts and planning like free maternal healthcare services 1008 introduced by the Government of India, such as Janani Shishu Suraksha Karyakaram (JSSK) and Janani Suraksha Yojna (JSY) schemes, maternal health remains a significant challenge for health care delivery organizations in the Country. According to the reports of UNICEF, India's maternal mortality rate reduced from 212 to 167 deaths per 100 thousand live births from 2007 to 2013. As 69 percent of India's entire population resides in rural areas, maternal healthcare is more worrisome in backward rural areas where the socio-economic and development activities lag (Shekhar, 2020). Brouwere (1998) documented that no country has succeeded in getting its maternal mortality ratio below 100 per 100 thousand live births without guaranteeing that an aptly skilled health staff attends all women during and after childbirth. Bloom et al. (1999), Ram et al. (2006), and Rani et al. (2008) reveal a strong positive association between the level of antepartum care, intrapartum care, and postpartum care. Chattopadhyay (2012) documented that over 75 percent of pregnant women were deprived of three ANC check-ups. The lack of pregnancy-related information amongst the husband and other family members and a robust patriarchal set-up, where the judgments about even the women's movement are in the hands of men, were the reasons behind this deprivation. Amit Kumar (2014) endorsed a significant association of socio-demographic indicators with maternal health services utilization in India. Also, the non-nuclear family and Male out-migration positively and highly considerably associated with the usage of ANC, institutional delivery, and PNC. Furthermore, the utilization of ANC increases with an increase in the birth interval and maternal education. In urban and rural areas, ANC utilization is higher in non-nuclear families than in nuclear families. Yadav and Dhillon (2015) found in their study that in low resources setting like Uttar Pradesh, health structure methods to improve PNC or ANC services should be prioritized with more effective counseling or advice on Family Planning to reduce unplanned births. Moreover, they showed that critical MH services (ANC, institutional delivery, PNC) utilization have effectively increased subsequent contraceptive use by 3.7%, 7.3%, and 6.8%, respectively, and have slightly reduced the unmet prerequisite for FP. Mohanty and Pathak (2009), in their study based on India and its states like Maharashtra and Uttar Pradesh, found that the rich-poor gap in child immunization, ANC, and child healthcare has widened while it has narrowed for contraception use over the years from 1992 to 2005 and the programs are barely reaching the poor sections of the society. Chauhan and Rai's (2015) study assesses the inequality in the coverage of Skilled Birth Attendance using data from NFHS-1 to NFHS-3 (1992-93, 1998-99, 2005-06). The study depicts a slow growth in SBA utilization in India and its state and urban and rural areas. The rural-Urban gap in SBA utilization, which contributes a lot in reducing maternal mortality, is more prominent in UP. The east, west, and south areas of UP have experienced a higher surge rate in SBA than the northern and central regions. Yadav and Kesarwani (2015) assessed the community factors that influence MHC services utilization in India. They found that women who obtained complete ANC were only 20.7%, whereas 41.7% obtained safe delivery care and 40.1% obtained postpartum care in two weeks of delivery. Yadav and Kesarwani (2015) also identified factors like age of women at birth, caste, parity, religion, females' education, autonomy index, family wealth index, and mass media at the individual level have a significant impact on maternal healthcare utilization. Sandhya (1991) suggested that socio-cultural factors-like caste, education, type of family, and occupation of parents, socio-economic status of the family, prenatal care, childbirth practices, and the type of medical consideration at the time of delivery define the level of infant and child mortality. Singh PK (2013) opined a decline in gender and urban-rural differences over time; however, children residing in rural parts and girls remained deprived. Furthermore, west, northeast, and south areas, which had the lowest gender inequality in immunization coverage in 1992, witnessed a rise in gender difference over time. Likewise, urban-rural inequality amplified in the west section during the year 1992-2006. This study tried to assess maternal and child health spatial patterns to comprehend the prevalence and patterns of several maternal and child health indicators.

Outcome Variable:
The occurrence of complete antenatal care, skilled birth assistance, mothers' and children's postnatal care, and children's full immunization are the outcome variable. The majority of these indicators are evaluated at the district level using the data of unit level.

Analytical Procedure:
We used Local Indicator and Moran's I index for Spatial Autocorrelation (LISA) to identify the presence of spatial autocorrelation at the global and local levels, respectively. Spatial autocorrelation depicts the extent to which data points are similar or dissimilar to their spatial neighbours ( Based on the four quadrants of Moran's I four types of spatial regions were generated. They are the "hot spots," i.e., the districts with high occurrence with similar neighbours also identified as "High-High," the "cold spots," i.e., the sections with low occurrence with similar neighbours also identified as "Low-Low," and the "spatial outliers," i.e., the districts with high occurrence having low occurrence neighbor districts and vice-versa. The Queen Contiguity spatial weights, which denote whether spatial units share the boundary or not (Sharma et al., 2020), are used. If the set of boundary points of unit I is specified by the band (i), then the Queen Contiguity Weight, W ij is demarcated by: However, this allows spatial units to share only a single boundary point (for instance, a shared corner point on a grid of spatial units). Hence, a more necessary condition requires sharing some positive portion of their boundary (Sharma et al., 2020).
Regression models were used to examine the significant correlates of complete vaccination and PNC among children in India. To see the extent of autocorrelation in the error term the spatial ordinary least square (OLS) regression model was applied (Anselin, 1995;. Since spatial autocorrelation in its error term for the dependent variable is confirmed by OLS, we additionally estimated the spatial lag model (SLM) and spatial error 1010 model (SEM) (Mishra et al., 2021). The underlying presumption of a spatial lag model is that the observations of the outcome variable are affected in the neighborhood areas. In contrast, the spatial error model is used to consider the effect of variables absent in the regression model but affect the outcome variable. The primary variation between the two models is that the spatial lag model, apart from the spatial error model, does not assess the spatial dependence of the error term (Srivastava et al., 2019). The OLS basic equation is as follows: Y=α + ΒX +Ɛ Y presents the outcome variable, while X is the vector of predictor variables, and α is the model intercept, and β is the corresponding coefficient vector.
The spatial lag model propounds that the units are spatially dependent and lagging in the nearby spatial locations. A typical spatial lag model is written as follows: Here denotes the complete vaccination and PNC among children for the ℎ district, δ is the spatial autoregressive coefficient, indicates the proximity's spatial weight between district i and j, is the complete vaccination and PNC among children in thej th district, while means the coefficient, is the predictor variable, and ε j is the residual.
On the other hand, the spatial error model considers the omitted variables' contribution that is not considered in the model but can significantly affect the analysis (Anselin, 1995;. A Spatial Error Model (SEM) is written as follows: Here, Y i denotes the complete vaccination and PNC among children for the i th district, λ shows the spatial autoregressive coefficient, W ij signifies the proximity's spatial weight between district i and j, Y j is complete vaccination and PNC among children in the j th district, while β j shows the coefficient, the predictor variable, and ε i is the residual (Anselin, 1995