If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Adequate availability of public health facilities leads to better utilization of MCH services at district level in India.
•
Geospatial techniques are instrumental in identifying spatial clusters of underperforming districts for MCH services.
•
Along with the availability of health facilities, female literacy plays an essential role in utilizing MCH services.
•
The shortage of doctors and paramedical staff hinders the utilization of MCH services.
Abstract
Objective
This study aims to address the spatial variation in the availability of public health facilities and utilization of Maternal and Child Health (MCH) services at district level in India.
Methods
Two nationally representative secondary data sources such as the National Family Health Survey 2015-16 and Census of India, 2011 were used to identify districts, which are lagging behind in terms of the availability and utilization of MCH services. Mapping and spatial clustering analyses was performed to identify underperforming districts in India. Further, Ordinary least square and spatial autoregressive models were used to determine the predictor for utilization of MCH services at the district level.
Results
Significantly, less availability of public health facilities and utilization of MCH services were observed in districts of Jharkhand, Bihar, Uttar Pradesh, Madhya Pradesh, Chhattisgarh, and North-Eastern states as compared to the national average. A shortage of 72,198 (21.9%) doctors and 165,791 (17.5%) paramedical staff found at the country level. After controlling for socio-economic factors, we observed that availability of sub-centre, dispensary, and population hospital bed ratio were significant positive predictors for Antenatal Care (ANC) service utilization whereas, shortage of paramedical staff was a negative predictor for full immunization at district level in India.
Conclusions
Adequate availability of health facilities, and human resources are needed in districts, which are underperforming in terms of availability and utilization of MCH services to improve utilization of MCH services.
Better maternal and child health is the reflection of the entire spectrum of socio-economic development of any country. Over the last two-decades, Maternal and Child Health (MCH) has improved a lot. The global Maternal Mortality Ratio (MMR) reduced from 342 per 100,000 livebirth in 2000 to 211 per 100,000 live birth in 2017.
On the other hand, the Under-5 Mortality Rate (U5MR) has declined from 75.8 per 1000 live birth (9.6 million) in 2000 to 37.7 (5 million) per 1000 live birth in 2019.
Global age-sex-specific fertility, mortality, healthy life expectancy (HALE), and population estimates in 204 countries and territories, 1950–2019: a comprehensive demographic analysis for the Global Burden of Disease Study 2019.
However, improvement in MCH is not uniform across the globe. Many developing countries still lag from universal coverage of most essential interventions for reproductive, maternal and child health, which leads to high maternal and child deaths.
The majority of these preventable maternal and child deaths occurs in developing countries. A developing country like India has achieved significant progress in MCH during the last two decades,
which may be the result of global, national, and local efforts such as government policies towards universal health coverage, immunization program, and financial support for the poor and marginalized population. However, India missed achieving millennium development goals (MDG) targets of maternal and child health. Growing evidence suggests that a significant inequality exists in maternal and child health with demographic and geographical diversity across states and districts of India.
Neonatal and under-five mortality rate in Indian districts with reference to sustainable development goal 3: an analysis of the national family health survey of India (NFHS), 2015–2016.
With global set target by UN to achieve the target 3.1; reduce MMR to 70 per 100,000 live birth and target 3.2; to reduce U5MR to 25 per 1000 live birth under Sustainable Development Goals (SDG) by 2030, India also set targets in its National Health Policy (NHP) 2017 to achieve U5MR below 23 per 1000 live births by 2025 and MMR below 100 per 100,000 live birth by 2020. NHP (2017) also targets to increase coverage of health services such as increasing utilization of public health facilities by 50% from the current level by 2025, increasing coverage of ANC, skilled birth, and full immunization above 90% by 2025.
There are ample studies examining factors that affect the utilization of MCH services at the individual level. Still, supply side factors such as the availability of health facilities (doctors, hospital beds, nurses and paramedical staffs, etc.) are narrow in many studies. Another lacking in available literature is the identification of local level of administrative units (districts) within the state that are underperforming in utilization of MCH services and lack in availability of health facilities. Districts are the important local administrative units that are used for implementation of decentralized planning and policies within states. The utilization of health services is governed by many factors, which are explained in earlier models and frameworks, such as the three delays model.
In three delays model, first delay is explained as delay in decision making for seeking care which is governed by individual and household factors. Second delay is to reach health facility that is affected by availability, accessibility and affordability of health facilities and services. Third delay in getting proper care which is determined by quality of services provides in health facilities. The multidimensional framework highlights five dimensions: availability, accessibility, affordability, acceptability, and accommodation.
developed a behavioural model in 1950 and modified it in the 1990s, which deals with three dimensions: predisposing, enabling, and need factors, which were well adopted in many past researches and continue in the present to understand the utilization of health facilities.
The popular approach to assessing the utilization of health services in health economics is demand-side and supply-side.
Women's education increases the capacity to take advantage of better access to power and resources, resulting in better utilization of health services.
affect MCH utilization. On the other hand, supply-side factors such as availability, accessibility, quality, and affordability of health services directly impact the utilization of health facilities.
A study highlighted that improvement in overall health infrastructures such as increased availability of Accredited Social Health Activists (ASHA), free ambulance, treatment facilities and focused efforts on training and logistic could increase utilization of MCH services.
Several studies highlighted that human resources at the primary level of health care service could be increased by involvement of community health workers by basic training.
Comprehensive review of the evidence regarding the effectiveness of community–based primary health care in improving maternal, neonatal and child health: 8. summary and recommendations of the Expert Panel.
Use of mobile technology by frontline health workers to promote reproductive, maternal, newborn and child health and nutrition: a cluster randomized controlled Trial in Bihar, India.
Studies conducted in underperforming states of Bihar found that utilization of mobile health technology by frontline health workers (Angan Wadi Workers and ASHA) promise better utilization of MCH services.
Use of mobile technology by frontline health workers to promote reproductive, maternal, newborn and child health and nutrition: a cluster randomized controlled Trial in Bihar, India.
Public programs such as Janani Suraksha Yojana (JSY), Janani Shishu Suraksha Karyakarm (JSSK), and Rashtriya Bal Swasthya Karyakram (RBSK) lead better coverage of ANC, institutional deliveries, and full immunization.
A recent study observed that the availability of female medical officers in primary health care leads to higher utilization of ANC, skilled birth attendance, PNC, and immunization at the district level.
However, none of the studies has focused on the availability and spatial aspect of health facilities in all 640 districts of India. In this context, the main objectives of this paper are threefold. First, to find out states and districts that lag in terms of availability of health facilities and utilization of health services by types of health facilities. Second, to examine the association and contribution of the availability of health facilities for the utilization of MCH services at the district level. Third, to bring out the factors that affect the utilization of public health facilities at the district level while considering spatial components in models.
2. Data and methods
2.1 Data
We used two secondary data sources, namely the Census of India, 2011 and the fourth round of the National Family Health Survey (NFHS-4). Census of India provides the most credible and complete information on population characteristics, economic activities, education, housing, amenities and health facilities, fertility, mortality, migration, and urbanization on decadal intervals. On the other hand, NFHS is a large-scale multi-round survey that provides crucial information about population, maternal and child health, nutrition, socio-economic development, family planning, diseases, domestic violence, etc.
Information on the availability of health facilities such as hospitals, beds, doctors, and paramedical staff were collected from the village directory (597,607 inhabited villages) and town directory (7935 towns) of Census of India 2011. Maternal and child health indicators such as at least four antenatal coverage (ANC4), institutional delivery, and full immunization coverage are used from NHFS-4 to assess health services utilization. The sample size for this study was 259,627 women aged 15–49 years who delivered birth in the last five years before the survey, and 48,928 children aged 12–23 months. Detailed sampling design, data collection, and survey weights are provided in the national report.
The unit of analysis of this paper is district, which is suitable for examine spatial effects of availability of public health facilities for the utilization of MCH services. As earlier discussed that districts are essential administrative units for assessing decentralized planning and policies beyond sub-national level. The district-level analysis is widespread among researchers and planners as it highlights coverage beyond the state level and brings out inequality within the state. Many past researches highlighted that better-performing states have considerable disparities in the utilization of health services at the district level.
Neonatal and under-five mortality rate in Indian districts with reference to sustainable development goal 3: an analysis of the national family health survey of India (NFHS), 2015–2016.
The importance of decentralized assessment of health performance at district and lower administrative levels was also recommended in the Bhore committee report of 1946.
For the availability of health facilities, we computed information such as the total number of health facilities, population health facility ratio (per 10,000 population), population hospital bed ratio, population doctor ratio, population paramedical staff ratio (per 10,000 female population age group 15–49 years), percent shortage of doctor and paramedical staffs in health facilities at the district level. The shortage of doctors and paramedical staffs in each health facilities were provided in census data. We clubbed the information on shortage of doctors and paramedical staff from all public health facilities at district level and further, percent shortage were computed where denominator is considered as total number available position.
To access the utilization of maternal and child health services, we used three crucial MCH indicators at the district level by types of health facilities (public and private). These indicators are at least four antenatal care (ANC4), institutional delivery, and full immunization. At least four ANC visits were considered for the women who had last birth in five year prior to the survey. Institutional delivery is considered for all birth occurred in health facilities during five years prior to survey. Full immunization is defined as those children aged 12–23 months receiving one dose of Bacille Calmette Guerin (BCG), three doses of Diphtheria Pertussis, and Tetanus (DPT), and Oral Polio Vaccine (OPV), and one dose of measles vaccine.
Socio-economic and demographic indicators included in this study are percent distribution of poor, scheduled caste/scheduled tribes (ST/SC), female-headed household, female literacy, percentage coverage of Janani Suraksha Yojana (JSY), percent of households that live in an urban area, mean household wealth index at the district level.
2.3.1 Statistical methods and analytical approach
We applied relative measures namely population health facility ratio (per 10,000 population), population bed ratio, population doctor ratio, population paramedical staff ratio (per 10,000 female population aged 15–49 years). Descriptive measures such as percent shortage of doctor and paramedical staff and percent coverage of health care services were used to map at the district level and find spatial inequality in terms of availability and utilization of health services. Between and within-state variation in availability and utilization of health facilities were analyzed using a box plot. Before proceeding to the spatial analysis, we examined the association among indicators of availability of health facilities and utilization of health services using correlation matrix. In addition, spatial statistics such as univariate and bivariate Global Moran's Index were used to find out spatial dependence and heterogeneity in availability and utilization of health facilities and services.
The Queen Contiguity first-order spatial weight matrix was applied for spatial analysis as we tested other spatial weights matrices such as rook and inverse distance were not promised better fit. The equation for Moran's I presented below.
Eq. 1
where observation at district i, is expressed as Zi = Xi - , where is the mean of variable x. Wij as the element of spatial weights matrix, So = as the row standardization of spatial weights, and n is the number of observations.
Further, the ordinary least square (OLS) and spatial autoregressive (SAR) models were used to assess factors affecting maternal and child health services at the district level in India. Both models are explained below.
Eq. 2
Eq. 3
where Y is expressed as the outcome variable, X is the independent variable, Wy is the standardized lag value of the outcome variable for its neighbour districts, and is the error terms of the model.
The spatial autoregressive model (SAR) provides better scope over OLS to understand the direct, indirect, and total effect of independent variables for the outcome variable.
Further, it helps to understand the influence of spatial components in outcome variables. All statistical analysis was done using Stata 16.1 software, data visualization was done using R version 4.0.2 software, and GeoDa 1.18.0 software was used for spatial analysis.
3. Findings
3.1 Availability of health facilities
The availability of public and private (non-government) health facilities by rural-urban in India presented in Table 1. There were 312,138 public and 435,378 private health facilities available during the 2011 census. Around 6.2 health facilities were available per 10,000 people in India. However, a significant difference was observed in the availability of public (2.6) and private (3.6) health facilities, and further, this difference was elevated by place of residence. Public health facilities were mainly available in rural areas, whereas private health facilities were available in urban areas. Only 2.4 private health facilities per 10,000 population were available in rural areas, whereas 7.3 private health facilities were available for 10,000 people in urban areas. To better understand the availability of health facilities, we computed ratio measures with women aged 15–49 years and found that 8 doctors, 24 hospital beds, and 25 paramedical staff were available for 10,000 women in India. Further, shortage of 72,198 (21.9%) doctor and 165,791 (17.5%) paramedical staffs found at the national level.
Table 1Total number of health facilities by urban-rural in India, 2011.
Total
Rural
Urban
Total population (2011 census)
1,210,193,422
833,087,662
377,105,760
Community Health Centres
5048
5048
–
Primary Health Centres
24,074
24,074
–
Sub Centres
114,655
114,655
–
Maternal Child welfare Centres
41,398
35,512
5886
Allopathic Hospitals
11,361
5732
5629
Alternative Hospitals
10,819
7891
2928
Dispensary
40,783
27,684
13,099
Mobile Health Clinic
16,847
15,721
1126
Family welfare Centres
38,061
31,616
6445
Maternity hospital
9092
–
9092
Total public health facilities
312,138
267,933
44,205
Non-government Charity Hospital
37,434
17,608
19,826
Non-government Medical Shops
361,194
149,431
211,763
Non-government Others Health Facilities
36,750
35,493
1257
Total Non-government health facilities
435,378
202,532
232,846
Total health facilities
747,516
470,465
277,051
Population health facility ratio (per 10,000 pop.)
3.2 Availability of health facilities at state and district level
A significant variation in availability of public health facilities was observed between and within states. States/union territories (UTs) such as Bihar, Madhya Pradesh, Uttar Pradesh, Chandigarh, Daman and Diu, and Delhi had low less than three public health facilities per 10,000 population. Further, low availability of hospital beds, doctors, and paramedical staff was observed in Bihar, Uttar Pradesh, Chhattisgarh, Andaman and Nicobar Island, Arunachal Pradesh, Gujarat, Madhya Pradesh, West Bengal. On the contrary, a few states/UTs such as Karnataka, Goa, Sikkim, Lakshadweep, Chandigarh and Puducherry, Arunachal Pradesh, Himachal Pradesh, had better availability of hospital beds, doctors, and paramedical staff (see Fig. 1). Detailed mapping of district-level variation in the availability of health facilities is provided in appendix 1. For example, district-level variation within the state in availability of public health facilities was found higher in districts of Bihar, Uttar Pradesh, Madhya Pradesh, Odisha, Andhra Pradesh, Assam, and Maharashtra (appendix 1: map 1.3). District such as Varanasi from Uttar Pradesh has higher public health facilities from its surrounding districts. Percent distribution of shortage of doctors and paramedical staff in health facilities at state level provided in Fig. 2 depicted that there is more than 50% shortage of doctors in states namely Meghalaya, Assam, Jharkhand Chhattisgarh. A shortage of more than 35% paramedical staff are found in states of Odisha and Rajasthan. District level map for the shortage of doctor and paramedical staff (appendix 1: map 1.8 and 1.9) revealed that districts of Jharkhand, Odisha, Chhattisgarh, Gujarat, Rajasthan, Assam, Uttarakhand, and Uttar Pradesh had a high shortage of doctors and paramedical staffs. Significant variations in the shortage of doctors were also found within these states. For instance, high (more than 48%) shortage of doctors found in Bargarh, Kalahandi, Malkangiri, and Jagatsinghapur districts on the other side comparably low (nearly 10%) shortage was observed in Angul, Cuttack, Khordha districts of Odisha.
Fig. 1Availability of health facilities (per 10,000 population) in states and districts of India, 2011.
The utilization of maternal and child health services by type of health facilities is presented in Fig. 3. More than half of the women visited at least four ANC services, Seventy nine percent of women delivered their children in health facilities, and 62% children aged 12–23 months received full immunization in India. Further, we observed a considerable difference in the utilization of these health services by type of health facilities. Among women who received at least four ANC services, more than half of them received it from public health facilities. Similarly, among women who delivered their children in health facilities, around 66% of them delivered in public health facilities. However, in case of full immunization services; a vast majority of children (nine out of 10 children) who received full immunization services were from public health facilities.
Fig. 3Percent utilization of maternal and child health services by type of health facilities in India, 2015–16.
3.4 Utilization of MCH services at state and district level
The utilization of maternal and child health (MCH) services in states and districts of India is presented in Fig. 4. Significant spatial variation in utilization of all three services (ANC4, Institutional delivery, Full immunization) was observed between and within states. Utilization of all three services was comparatively high in states such as Kerala, Jammu and Kashmir, Goa, Tamil Nadu, and Karnataka. At the same time, states such as Bihar, Nagaland, Arunachal Pradesh, Uttar Pradesh, Uttarakhand, Jharkhand, Madhya Pradesh, and Chhattisgarh had low utilization of these health services. District-level variation in utilization of health services can be observed in north-eastern states. For instance, Jorhat district from Assam and Aizawl district from Mizoram have high (more than 95%) institutional delivery in the entire region (appendix 2: map 2.3).
Fig. 4Utilization of maternal and child health services in states and districts of India, 2015–16.
Further, we explore the utilization of health services by types of health facilities presented in Fig. 5. Utilization of ANC from public health facilities was found higher in states/UTs like Sikkim, Andaman, and Nicobar Island, Chandigarh, Jammu and Kashmir, Mizoram, Odisha, Assam. In contrast, a fewer public health facilities were utilized for at least four ANC services in states such as Nagaland, Telangana, Bihar, Andhra Pradesh, Gujarat, and Jharkhand. The same pattern can be observed for both institutional delivery and full immunization, which can be better visualized from appendix 2 at the district level.
Fig. 5Utilization of maternal and child health services in public health facilities in states and districts of India, 2015–16.
3.5 Spatial dependence of availability of health facilities and utilization of MCH services
The correlation matrix of availability of health facilities, utilization of health services, and socio-economic indicators is presented in Fig. 6. Positive and weak associations between the availability of health facilities and health care utilization were observed at the district level. A negative association was found among utilization of health services and shortage of doctors and paramedical staff. Positive and robust associations were observed among utilization of MCH services, for instance, ANC was positively associated with institutional delivery and full immunization. Household wealth index, percent urban, percent literacy were positively related to the utilization of health services. However, percent ST/SC, percent poor, percent female-headed household were negatively associated with utilization of health services.
Fig. 6Correlation matrix of availability of health facility, utilization of MCH services and socio-economic indicators in India.
Further, we examined spatial dependence and carried out Morans'I univariate and bivariate spatial analysis presented in Fig. 7. Univariate Morans'I values for all selected indicators were positive, which indicates spatial dependence at the district level. Bivariate Morans'I revealed a negative and weak spatial association with the utilization of MCH services, health facilities, and socio-economic indicators. Janani Suraksha Yojana (JSY), which is a financial incentive program was positively associated with MCH services received from public health facilities.
Fig. 7Univariate and bivariate spatial association test of availability of health facilities, utilization of health services and socio-economic indicators in districts of India.
3.6 Predictors for MCH service utilization at the district level
To understand predictors of maternal and child health service utilization at the district level. We performed two types of models (Ordinary Least Square (OLS) and Spatial Autoregressive (SAR) for utilization of antenatal care, institutional delivery, and full immunization by types of health facilities. Predictors were selected from four domains, e.g., health services, health facilities, socio-economic status, and demographic components. At the first stage, we performed stepwise OLS to identify significant predictors (at p-value ≤ 0.05) from 50 summary indicators (for details, see appendix 3) at the district level for utilization of health services. Further, same indicators were used for spatial autoregressive models to determine the spatial effect of lagged outcome variables. Results from OLS and SAR models are presented in Table 2. Post estimation of SAR model was performed to find out indirect effects of spatial lag and error terms of outcome variable presented in appendix 3.
Table 2Beta coefficients (95% CI) of stepwise OLS & SAR model for selected maternal and child health indicators at district level in India.
Dimension
Variables
Antenatal care
Institutional delivery
Full immunization
OLS model
SAR Model
OLS model
SAR model
OLS model
SAR model
Health Services
Antenatal care
0.38 (0.34–0.42)
0.38 (0.33–0.44)
0.23 (0.16–0.31)
0.25 (0.15–0.35)
Antenatal care public
0.22 (0.11–0.33)
0.34 (0.2–0.48)
Institutional delivery
0.3 (0.17–0.43)
0.41 (0.24–0.59)
Institutional delivery public
−0.15 (-0.26--0.03)
−0.19 (-0.35--0.04)
Health Facilities
Total doctor
−0.01 (-0.01–0)
0 (−0.01–0)
Population Hospital bed ratio
0.24 (0.04–0.45)
0.24 (−0.02–0.51)
Paramedical staff
0 (0–0)
0 (0–0)
0 (0–0)
0 (0–0)
Percentage paramedical staff shortage
−0.11 (-0.16--0.07)
−0.1 (-0.15--0.05)
Sub centre
0.04 (0.03–0.06)
0.03 (0.02–0.05)
Maternity hospital
0.07 (0.05–0.09)
0.07 (0.04–0.09)
Allopathy hospital
−0.1 (-0.17--0.02)
−0.05 (−0.16–0.06)
Alternative hospital
−0.15 (-0.2--0.09)
−0.09 (-0.17--0.01)
−0.06 (-0.1--0.01)
−0.07 (-0.13--0.01)
Dispensary
0.05 (0.03–0.07)
0.03 (0–0.06)
−0.02 (-0.03--0.01)
−0.02 (-0.03–0)
0.04 (0.03–0.06)
0.03 (0.01–0.05)
Mobile clinic
−0.01 (-0.02–0)
0 (−0.01–0)
Family welfare centre
−0.04 (-0.06--0.02)
−0.05 (-0.07--0.02)
Non.gov. charity
0 (-0.01–0)
0 (−0.01–0)
Non.gov. others
0 (0–0)
0 (0–0)
Total Non.gov health facility
0 (0–0)
0 (0–0)
Socio-Economic status
Percent poor
−0.37 (-0.45--0.29)
−0.4 (-0.53--0.27)
−0.23 (-0.28--0.17)
−0.26 (-0.33--0.2)
−0.08 (-0.15--0.02)
−0.08 (−0.16–0.01)
Percent STSC
−0.1 (-0.17–0.03)
−0.15 (-0.25--0.04)
−0.16 (-0.2--0.12)
−0.13 (-0.17--0.08)
−0.07 (-0.12--0.01)
−0.04 (−0.12–0.03)
Percent female headed household
−0.17 (-0.31--0.04)
0 (−0.16–0.16)
0.32 (0.15–0.49)
0.44 (0.2–0.68)
Percent female literate
0.46 (0.32–0.6)
0.67 (0.44–0.91)
Percent received JSY
0.16 (0.11–0.2)
0.15 (0.09–0.2)
0.2 (0.12–0.27)
0.17 (0.08–0.26)
Women Population
Female population1549
0 (0–0)
0 (0–0)
Rural female population1549
0 (0–0)
0 (0–0)
Adj. R-squared
0.53
0.53
0.65
0.64
0.49
0.49
Spatial autoregressive model
Lagged dependent variable
0.44 (0.35–0.52)***
0.08 (0.01–0.15)***
0.26 (0.17–0.36)***
Error term of dependent variable
0.5 (0.36–0.64)***
0.69 (0.61–0.77)***
0.42 (0.29–0.55)***
Note: coefficients in bold are significant at p < 0.05 and values in parenthesis () are lower and upper limits of CI at 95%.
After controlling effects of socio-economic and other factors we observed availability of sub-centre (β = 0.04; 95% CI 0.03, 0.06), dispensary (β = 0.05; 95% CI 0.03, 0.07) and population hospital bed ratio (β = 0.24; 95% CI 0.04, 0.45) are significant positive predictors for at least four ANC service visits. Results from the SAR model highlight that lagged and error terms of ANC are significant predictors that have a spatial indirect effect on ANC coverage at district level. We observed a slightly reduced beta coefficient of sub-centre (β = 0.03; 95% CI 0.02, 0.05) for ANC services at the district level.
The effects of female literacy was significantly positive (β = 0.46; 95% CI 0.32, 0.6) for ANC coverage. Other factors such as percent poor (β = −0.37; 95% CI -0.45, −0.29), percent ST/SC (β = −0.1; 95% CI -0.17, −0.03) was negative predictor for utilization of ANC at district level. The indirect effect of spatial lag and error terms for ANC by neighbouring districts further increases the beta coefficient of female literacy rate (β = 0.67; 95% CI 0.44, 0.91) in the SAR model. The same pattern can be observed for negative predictors like percent poor and percent STSC.
For institutional delivery, ANC service (β = 0.38; 95% CI 0.34, 042) and percent received JSY (β = 0.6; 95% CI 0.11, 0.2) were significant positive predictors whereas, percent poor (β = −0.23; 95%CI -0.28, −0.17), percent ST/SC (β = −0.16; 95% CI -0.2, −0.12), percent female headed household (β = −17; 95% CI -0.31, −0.04) were negative predictors at district level in India. Result from SAR model for institutional delivery found significant, however, beta coefficient remain similar to OLS model.
After adjusting other factors we observed in both models that availability of maternity hospital (β = 0.07; 95% CI 0.05, 0.09), dispensary (β = 0.04; 95% CI 0.03, 0.06) and percent received JSY (β = 0.2; 95% CI 0.12, 0.27) play positive role in getting full immunization at district level. On contrary shortage of paramedical staff (β = −0.11; 95% CI -0.16, −0.07) have a negative effect on coverage of full immunization. MCH services like at least four ANC visits by women (β = 0.23; 95% CI 0.16, 0.31) and institutional delivery (β = 0.3; 95% CI 0.17, 0.43) were positive predictors for full immunization. Beta coefficient from SAR model provide slightly higher value compared to OLS model for full immunization.
Separate OLS and SAR models were applied for MCH services utilized from public health facilities are presented in Table 3. We found dispensary (β = 0.03; 95% CI 0.02, 0.05) and population health facility ratio (β = 1.47; 95% CI 1.02, 1.92) are positive predictors for ANC services from public health facilities in both OLS and SAR models after controlling for socio-economic factors. The lagged term of ANC was not significant, but the error term of ANC was significant, which has a significant effect on the beta coefficient of predictors. The overall effect of population health facility ratio in the SAR model was found slightly lower than the OLS model. Antenatal care from public health facility (β = 0.62; 95% CI 0.56, 0.68), population paramedical staff ratio (β = 0.18; 95% CI 0.04, 0.33), percent received JSY (β = 0.12; 95% CI 0.08, 0.16) were significant positive covariates for institutional delivery in public health facility. ANC (β = 0.08; 95% CI 0.03, 0.12) and institutional delivery (β = 0.25; 95% CI 0.2, 0.31) services from public health facility were positive covariates for full immunization services in public health facilities in both models. Percent shortage of paramedical staff (β = −0.03; 95% CI -0.05, 0.02) in public health facilities was a negative factor for full immunization in public health facilities in districts of India.
Table 3Beta coefficients (95% CI) of stepwise OLS & SAR model for selected maternal and child health services at public health facilities in districts of India.
Dimension
Variables
Antenatal care
Institutional delivery
Full immunization
OLS model
SAR Model
OLS model
SAR model
OLS model
SAR model
Health Services
Antenatal care public
0.62 (0.56–0.68)
0.61 (0.52–0.7)
0.08 (0.03–0.12)
0.09 (0.04–0.14)
Institutional delivery
−0.19 (-0.23--0.14)
−0.19 (-0.25--0.13)
Institutional delivery public
0.25 (0.2–0.31)
0.25 (0.19–0.31)
Paramedical staff
0 (0–0)
0 (0–0)
Health Facilities
Population paramedical staff ratio
0.18 (0.04–0.33)
0.08 (0.05–0.2)
Percentage paramedical staff shortage
−0.03 (-0.05--0.02)
−0.03 (-0.05--0.01)
Sub centre
0.01 (0.01–0.02)
0.01 (0–0.01)
Allopathy hospital
−0.1 (-0.16--0.04)
−0.06 (-0.12–0)
Dispensary
0.03 (0.02–0.05)
0.03 (0.01–0.04)
Mobile clinic
0.01 (0–0.01)
0.01 (0–0.01)
Population health facility ratio
1.47 (1.02–1.92)
0.9 (0.45–1.34)
Socio-Economic status
Percent urban
−0.06 (-0.09--0.03)
−0.09 (-0.12--0.05)
Percent STSC
−0.15 (-0.19--0.11)
−0.09 (0.14--0.04)
Percent female headed household
−0.72 (-0.9--0.54)
−0.47 (-0.67--0.26)
−0.15 (-0.3–0.01)
−0.04 (0.21–0.14)
Percent female literate
0.51 (0.43–0.6)
0.46 (0.36–0.57)
−0.09 (-0.14--0.05)
−0.08 (-0.13--0.02)
Percent received JSY
0.39 (0.34–0.45)
0.41 (0.34–0.48)
0.12 (0.08–0.16)
0.07 (0.02–0.13)
−0.07 (-0.1--0.03)
−0.07 (-0.11--0.03)
Women, Child Population
Populatio05
0 (0–0)
0 (0–0)
Female population1549
0 (0–0)
0 (0–0)
Female paramedical staff ratio
−0.03 (-0.05–0)
−0.02 (−0.05–0)
Adj. R-squared
0.45
0.45
0.54
0.54
0.45
0.45
Spatial autoregressive model
Lagged dependent variable
0.01 (−0.11–0.13)
0 (0.1–0.11)
0.02 (−0.03–0.08)
Error term of dependent variable
0.63 (0.5–0.76)***
0.72 (0.63–0.82)***
0.41 (0.3–0.51)***
Note: coefficients in bold are significant at p < 0.05 and values in parenthesis () are lower and upper limits of CI at 95%.
The present paper attempts to provide insights on gaps in the availability of public health facilities and the underutilization of MCH services at the state and district level in India. We identified factors that affect the utilization of MCH services at the district level considering spatial components in models. We analyzed two important secondary data sources Census of India 2011 and fourth round of National Family Health Survey (2015-16) to accomplish the objectives of this study. A number of important findings are highlighted below, which can further help to reduce the gap in coverage of MCH services and increase the availability of health facilities across 640 districts of India.
A significant sub-national variation in the availability of public-private health facilities and huge rural-urban gaps were observed. Most of the public health facilities are clustered in rural areas as compared to urban areas. At the same time, private health facilities are clustered in urban centres. Inadequate, less than three public health facilities per 10,000 people were observed in states/UTs such as NCT of Delhi, Uttar Pradesh, Chandigarh, Madhya Pradesh, Bihar, Haryana, and Gujarat. Whereas less populated states/UTs such as Sikkim, Andaman & Nicobar Island, Mizoram, and Uttarakhand have little better (>5 per 10,000 pop.) population to public health facility ratio. Regarding female population (aged 15–49 years) to hospital beds, doctor, paramedical staffs ratio we observed low availability in the states of Bihar, Uttar Pradesh, Haryana, Delhi, Arunachal Pradesh Chhattisgarh, Gujarat, Telangana, West Bengal, Odisha, and Madhya Pradesh. The possible reasons for this gap among these states/UTs may be more population pressure and shortage of doctors and paramedical staff in public health facilities, which can be observed in appendix 1. Similar findings of huge urban-rural gaps in health workforce availability and the vast difference in qualification of health workforce were observed in National Sample Survey-based studies.
Rural health statistics (2019-20) also revealed that over the period, human resources had increased, still 6.8% of allopathic doctors shortfall in PHCs whereas a considerable shortfall of surgeon (78.9%), obstetricians & gynaecologists (69.7%), physicians (78.2%), and paediatricians (78.2%) reported in CHCs as per current requirement.
The utilization of maternal and child health services by types of health facility highlights significant differences at the national level. All three selected MCH services (Antenatal care, Institutional delivery, Full immunization) were primarily utilized from public health facilities. Further, we observed significant spatial variation in coverage of MCH services in states and districts of India. States such as Nagaland, Bihar, Arunachal Pradesh, Uttar Pradesh Uttarakhand, Jharkhand, Madhya Pradesh, Rajasthan, Meghalaya, Assam, and Gujarat were low performing states in terms of utilization of MCH services. On the contrary, states/UTs such as Kerala, Jammu and Kashmir, Goa, Tamil Nadu, Karnataka, Puducherry, Andaman and Nicobar Island, and Lakshadweep have better coverage of MCH services. Another concern that emerges from the findings is that many states such as Bihar, Jharkhand, Uttar Pradesh, Arunachal Pradesh, Nagaland, Manipur are lagging behind in utilizing MCH services from both (public and private) health systems.
The correlation coefficient depicts a positive association between the availability of health facilities and MCH service utilization at the district level in India while shortage of doctors and paramedical staff was negatively associated with MCH service utilization. Similar findings were observed by Bhatia and Dwivedi
in empowered action groups (EAG) states of India. Another critical finding was that ANC, institutional delivery, and full immunization are strongly associated which indicates that once a woman enters in the continuum of care for MCH services during pre-pregnancy; she may continue services afterward. Incentive program such as Janani Suraksha Yojana (JSY) is positively associated with utilization of MCH services received from public health facilities.
Univariate Moran's I test suggested that all indicators at the district level are not randomly distributed. There is a strong spatial association with outcome variable and their average lag value of neighbour districts. Positive and high Moran's I value indicate that percentage coverage of MCH services at the district level are not independent of each other, which further guides to add the effects of space while predicting factors that affect utilization of MCH services at the district level.
After controlling socio-economic, demographic, and health service factors in the OLS model, we observed that the availability of sub-centre, dispensary, and population hospital bed ratios are significant positive predictors for ANC coverage at the district level. The analysis further indicates that a percent increase in ANC leads to a rise of 0.62% in institutional delivery at the district level. An increase in population paramedical staff ratio and percent received JSY leads to an increase in institutional delivery from public health facilities. After adjusting for socio-economic and other factors, we observed that availability of maternity hospital, dispensary, and percent received JSY leads to increase coverage of full immunization at the district level. On the contrary, the shortage of paramedical staff has a negative effect on coverage of full immunization. MCH services like ANC and institutional delivery were positive predictors for full immunization. A study conducted on full immunization gaps in all 640 districts of India observed that districts which has poor availability of public health facilities have higher coverage gap of full immunization.
We tried to control the effect of space by using the spatial autoregressive (SAR) model and found a significant indirect effect (neighbourhood effect) of lag and error terms of dependent variables. The beta coefficient of the independent variable slightly improved in all three SAR models, which indicates that the impact of the neighbourhood should be considered while planning or forming policies for maternal and child health services.
To reduce inequality and increase coverage of MCH services, there is a need to examine factors that hinder coverage of MCH services at the individual level as well as local level (district level). Many problems can be identified and solved at the local level of the administrative division. In this regard, studies at the district level are essential; the government of India launched the Aspirational Districts Program (2018) to quickly and effectively transform 115 most under-developed districts from 28 states of India. Many recent studies also highlighted the importance of district-level analysis to identify poor performing districts within states and reduce inter-district inequality.
Neonatal and under-five mortality rate in Indian districts with reference to sustainable development goal 3: an analysis of the national family health survey of India (NFHS), 2015–2016.
The primary strength of the study is that it covers all 640 districts of India as per Census of India (2011) and highlights the relationship between the availability of health facilities and the utilization of MCH services. Another strength of this study is mapping health facilities and utilization at district level and highlighted within and between variations in terms of availability and utilization of public health services. It also addresses spatial components while predicting factors for MCH service utilization at the district level. However, there are a few limitations as well, which is explained further. As this study is based on cross-sectional secondary data sources, this study has an inherent limitation of causal inferences between the availability of health facilities and the utilization of MCH services. Another limitation of this study is that we tried to associate two datasets from different times; census data was published in 2011, and NFHS in 2015-16 which may affect our results. Nevertheless, as NFHS collected information on MCH five years preceding the survey, the effect of time difference may be reduced. The unit of analysis is a district; hence, one should be very careful while generalizing the results of this study.
6. Conclusion
This study highlighted that the adequate availability of health facilities improves the utilization of maternal and child health services at the districts level. On the other hand, the shortage of doctors and paramedical staff hinders the utilization of MCH services. There is a dire need to fulfil the shortage of human resources in health facilities and increase health facilities in unserved areas of the country. States such as Nagaland, Assam, Arunachal Pradesh, Uttar Pradesh, Bihar, Madhya Pradesh, Jharkhand, Chhattisgarh, Rajasthan need more focused plans for maternal and child health to achieve SDG targets and reduce inequality in availability and utilization of MCH services from Public health facilities. There is also a need to identify women who are out of continuum of MCH care. This study also revealed that once women enter into continuum of MCH care, they continue for further health services and therefore, effort should be made to provide availability and accessibility of health care services. Socio-economic factors such as female literacy, poverty are crucial factors for MCH care utilization, which need further strengthening in districts that lag behind in female education. Districts with higher marginalized population groups (SC/ST and the poorer) need special attention to improve the utilization of MCH services.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Ethical approval
As the study is based on the secondary data and available in the public domain, it needs no prior ethical approval.
Declaration of competing interest
The authors declare that they do not have any competing interest.
Data sharing statement
The DHS and Census of India websites provides access to data. The corresponding author have full access to the data.
Acknowledgement
The authors express their gratitude to the reviewers and the editorial board of the Journal for their valuable suggestion and comments which help us to strengthen this paper.
Abbreviations
MCH
Maternal and Child Health
ANC
Antenatal Care
MMR
Maternal Mortality Ratio
U5MR
Under Five-Mortality Rate
MDG
Millennium Development Goals
SDG
Sustainable Development Goals
UN
United Nation
NHP
National Health Policies
ASHA
Accredited Social Health Activist
JSY
Janani Suraksha Yojana
JSSK
Janani Shishu Suraksha Karyakaram
RBSK
Rashtriya Bal Swasthya Karyakram
PNC
Postnatal Care
NFHS
National Family Health Survey
ANC4
At Least 4 ANC Visits
BCG
Bacillus Calmette Guerin
DPT
Diphtheria, Pertussis & Tetanus
OPV
Oral Polio Vaccine
ST/SC
Scheduled Tribes/Scheduled Caste
OLS
Ordinary Least Squared
SAR
Spatial Autoregressive
EAG
Empowered Action groups
Appendix A. Supplementary data
The following are the Supplementary data to this article:
Global age-sex-specific fertility, mortality, healthy life expectancy (HALE), and population estimates in 204 countries and territories, 1950–2019: a comprehensive demographic analysis for the Global Burden of Disease Study 2019.
Neonatal and under-five mortality rate in Indian districts with reference to sustainable development goal 3: an analysis of the national family health survey of India (NFHS), 2015–2016.
Comprehensive review of the evidence regarding the effectiveness of community–based primary health care in improving maternal, neonatal and child health: 8. summary and recommendations of the Expert Panel.
Use of mobile technology by frontline health workers to promote reproductive, maternal, newborn and child health and nutrition: a cluster randomized controlled Trial in Bihar, India.
00112-9&id=gr1.jpg){kind=link}
00112-9&id=gr2.jpg){kind=link}
00112-9&id=gr3.jpg){kind=link}
00112-9&id=gr4.jpg){kind=link}
00112-9&id=gr5.jpg){kind=link}
00112-9&id=gr6.jpg){kind=link}
00112-9&id=gr7.jpg){kind=link}