Geographic Distribution and Determinants of Postnatal Care Utilization In Ethiopia: Spatial and A Multilevel Analysis of the Recent Demographic and Health Survey

Abstract
Introduction: Maternal and child health problems are very diverse and arises from maternal services like postnatal care. However, the low postnatal care being reported in middle and low-income countries are from the multiple factors, which change over geographic space and time and need timely assessments. Therefore, the current study was aimed to assess the prevalence, spatial variations and its associated factors from country representative data.
Methods: A secondary data analysis was conducted on the data from EMDHS 2019 on the 2,158 mothers with birth record in last 2 years. Global Moran’s I statistics and Getis Ord Gi* statistics tools were used to visualize the spatial pattern and identify hotspot/cold spot areas, respectively. Kulldorff SaTScan was used to detect purely significant spatial clusters. A multilevel binary logistic regression applied to examine the relationship between postnatal care and its determinants. Variables included into the model at p-value of <0.25. Descriptive statistics was presented using mean, numbers, standard deviation and proportion while, inferential statistics presented using AOR with 95% CI where association was declared at p-value of <0.05.
Results: From the analysis, the postnatal coverage from EMDHS 2019 was 11.91% (95% CI: 10.54%, 13.28%). There was significant geographical variation within the country; high regional prevalence of postnatal care utilization was detected in Tigray, Central Amhara and Oromia and Northern part of SNNP regions. Whereas, low prevalence was observed in Somali and Afar regions. Mothers who had ANC service for 1-3 times [AOR=4.08; 95% CI: 2.25, 7.43], ANC 4 or more times [AOR=4.87; 95% CI: 2.62, 9.05] and gave birth through caesarean section [AOR=1.72; 95% CI: 1.08, 2.73] are factors which had positive association with postnatal care. Whereas, delivery attended by other non-professional individuals [AOR=0.29; 95% CI: 0.12, 0.77], mothers who live in Afar 73% [AOR=0.27; 95% CI: 0.12, 0.64], Somali [AOR=0.33; 95% CI: 0.12, 0.86] and Benishangul Gumuz regions [AOR=0.42; 95% CI: 0.19, 0.91] were negatively associated factors with postnatal care service utilization.
Conclusion and recommendation: According to this analysis, postnatal care in the country was below the expected level recommended by the WHO. Although high proportion of uneducated mothers, low antenatal coverage, low institutional delivery and birth attendance need the major operation, the association beyond these with number of antenatal cares, delivery attendance and the pastoralist regions require government and international community commitments. Regional specific interventions should be launched by the government and other stakeholders, which is confirmed and supported by local academic research in regions with low prevalence, to tackle those determinant factors.

Keywords: Geographic distribution, postnatal care, determinants, maternal health services, Ethiopia

1. Introduction

The United Nations (UN) Sustainable Development Goals (SDG) planned to reduce maternal mortality rate to 70 per 100,000 live births in 2030. Yet, maternal mortality rate was 152 per 100,000 live births worldwide and it was 533 per 100,000 live births in sub-Saharan Africa (SSA) in 2020. In Ethiopia, the rate was 412 per 100,000 live births in 2020^1-3.

Similarly, infant and child mortality rates are among the basic indicators used to measure a country’s socioeconomic situation and quality of life. Globally, above one million newborns die every year during the first 24 hours after delivery and almost all of these deaths occur in the developing countries. The neonatal mortality rates were 28 deaths per 1,000 live births in SSA. Similarly, the 2019 Ethiopia Mini Demographic and Health Survey (EMDHS) report revealed that neonatal mortality rate was 30 deaths per 1,000 live births. The majority of causes for maternal mortality (75%) were related to preeclampsia, eclampsia, postpartum infections, prolonged labor and unsafe abortions^4-6.

Utilizing maternal health services during pregnancy, labor and delivery and the postpartum period highly prevent the maternal and newborn morbidity and mortality. Postnatal care (PNC) service is one of a fundamental element of the continuum of essential maternal health services given at the time from an hour after delivery of a newborn to 6 consecutive weeks. World Health Organization (WHO) defines PNC as a care given to the mother and her newborn child right after delivery up to 42 days of postpartum periods. Accordingly, WHO recommends a mother and her newborn baby to receive PNC within 24 hours of delivery, on day two to three, in the second to third week and 6 weeks after birth^7,8.

The 2019 report of UNICEF showed that PNC service utilization among mothers and newborns within the first 42 days of birth were 63% and 48% respectively⁹. In fact, the service is the most neglected, least utilized and with high rate of discontinuity among maternal and child survival interventions in the developing countries including Ethiopia. The study conducted in 36 SSA countries by using the most recent DHS data reported that 52.48% of mothers had utilized PNC service¹⁰. In Ethiopia, the PNC service coverage was very poor (6.9%), which is illustrated by the findings from secondary data analysis of the 2016 Ethiopian Demographic and Health Survey (EDHS)¹¹.

Previous studies conducted in Ethiopia and other developing countries has identified the predictors of PNC service utilization. The studies revealed maternal age, maternal education, paternal education, wealth quintile, religion, occupation, media exposure, distance to health facility, number of ANC visits, place of delivery, mode of delivery, parity, region and type of place of residence as determinants of PNC service utilization^10,12-18.

It is known that the determinants of PNC service utilization vary between populations with different socio-economic and demographic conditions, having different structures of health system, with diverse cultures and various geographical locations. Studies conducted at national level have prominent contributions in designing appropriate policies that help to intervene the gap shown in utilizing PNC service. Using multilevel mixed-effects logistic regression model help to consider variations observed due to clustering effect and to identify both the individual and community level variables, which cannot be handled by traditional logistic regression. Identifying factors that influence PNC service utilization in countries like Ethiopia is much important to overcome the problems and as a result, to reduce morbidity and mortality attributed to poor utilization of maternal health services in general and PNC service in particular.

2. Methods

2.1. Study settings and data source

This analysis used the 2019 Ethiopian Mini Demographic and Health Survey (EMDHS) data. Demographic and Health Survey (DHS) collects a countrywide representative data on basic health indicators in the five years’ time interval. It was used a two-stage stratified sampling technique to select respondents of the survey. The survey included all the 9 regions and the 2 city administrations of Ethiopia. The current study employed Individual Recode (IR) or Kids Recode (KR) file that covers all the collected data in the woman’s questionnaire. All Ethiopian women with a live birth 2 years prior the survey was the source population of the study.

2.2. Measurements of variables

2.2.1. Dependent variable: The response variable for this study was PNC service utilization. It was defined as having at least one visit that was provided by a health professional within 42 days of birth. Thus, the outcome variable was coded as “1” if mother visited at least once after delivery and “0” otherwise.

2.2.2. Independent variables:
2.2.2.1. Individual-level variables: This includes maternal education, maternal age, religion, marital status, wealth index, timing of 1st ANC check, number of ANC visits, place of delivery, mode of delivery, age of mother at 1st birth, birth order number, type of birth, delivery attendants, family size, parity and mentioned at least 3 pregnancy danger signs. Community-level variables were geographic regions where the mothers live and type of place of residence.