Magnitude and determinants for place of postnatal care utilization among mothers who delivered at home in Ethiopia: a multinomial analysis from the 2016 Ethiopian demographic health survey

Introduction Above half of mothers in Ethiopia give birth at home. Home based care within the first week after birth as a complementary strategy to facility-based postnatal care service is critical to increase the survival of both mothers and newborns. However, evidence on utilization of postnatal care and location of service among mothers who delivered at home in Ethiopia is insufficiently documented. Therefore, this study assessed the magnitude and determinants for place of postnatal care service utilization among mothers who delivered at home in Ethiopia. Methods We used the 2016 Ethiopian Demographic and Health Survey, and extracted data from 4491 mothers who delivered at home during 5 years preceding the survey. A multinomial logistic regression model was applied to examine the determinants of both facility and home -based postnatal care service utilization. Likelihood ratio test was used to see the model fitness and p-value of < 0.05 was used to determine statistical significance at 95% confidence interval. Results From the total 4491 mothers who delivered at home, only 130(2.9%) and 236(5.3%) of them utilized postnatal service at home and at a health facility respectively. Being from an urban region (AOR = 0.378, 95%CI: 0.193–0.740), ever using the calendar method to delay pregnancy (AOR = 0.528, 95%CI: 0.337–0.826), receiving four and above antenatal care visits (AOR = 0.245, 95%CI: 0.145–0.413) and having a bank account (AOR = 0.479, 95%CI: 0.243–0.943) were the factors associated with utilizing home- based postnatal care. Similarly being a follower of the orthodox religion (AOR = 1.698, 95%CI: 1.137–2.536), being in the rich wealth index (AOR = 0.608, 95%CI: 0.424–0.873), ever using the calendar method to delay pregnancy (AOR = 0.694, 95%CI: 0.499–0.966), wantedness of the pregnancy (AOR = 0.264, 95%CI: 0.352–0.953), receiving four and above antenatal care visits (AOR = 0.264, 95%CI: 0.184–0.380) and listening to radio at least once a week (AOR = 0.652, 95%CI: 0.432–0.984) were the determinants of facility-based postnatal care utilization. Conclusion The coverage of postnatal care service utilization among mothers who delivered at home was very low. Living in urban region, following the Orthodox religion, having higher wealth index, having a bank account, ever using calendar method to delay pregnancy, wantedness of the pregnancy, receiving four and above antenatal care visit and listening to radio at least weakly were associated with postnatal care service utilization. Therefore, targeted measures to improve socio-economic status, strengthen the continuum of care, and increase health literacy communication are critically important to increase postnatal care service utilization among women who deliver at home in Ethiopia.


Introduction
The early postnatal period, particularly the first hours of life, extending into the first 2-3 days of life is a period of high risk for both mothers and newborns [1][2][3]. The risk is more severe in communities where the overwhelming majority of births occur at home [4,5]. In such communities unless safe household practices are followed and care is provided by health workers, undoubtedly catastrophic adverse outcomes will persist [1,6]. Based on evidence, the World Health Organization (WHO) recommends homebased postnatal care (PNC) care for mothers and newborns in the first week of life as a complementary strategy to facility-based PNC in order to improve mothers and newborns survival [1,7,8].
Studies had shown that home-based newborn care interventions can prevent 30-61% of newborn deaths in high mortality settings and can lower the odds of maternal postpartum distress [9].
A home-based care strategy to promote an integrated package of preventive and curative newborn care is effective in reducing neonatal mortality in communities, even within a weak health system, especially if it is given by trained health professionals such as midwives [6,9]. However, providing this service and ensuring optimal practices is not straightforward for many reasons such as staff shortages, and inconvenient transportation, which results in low to moderate home visitation coverage for PNC even with intensive support [1,7]. Though they do not replace antenatal care (ANC), home visits by the community health workers during pregnancy were higher than the visits after birth [1,[10][11][12].
For institutional births, opportunities already exist to provide PNC to mothers and newborns before discharge [1]. However, even at health facility level, in Sub-Saharan Africa women are often discharged before 24 hrs, which limits them from receiving the WHO's recommended services. Additionally visits at 72 hrs and seventh day are rare [13]. For home births, some mothers and babies access early postnatal care through a visit to a health facility. However, many cannot or do not receive facility-based PNC following home deliveries, leaving home visitation as a potentially attractive way to make such care available [1].
In Ethiopia, the maternal mortality ratio is still high; 412 maternal deaths out of 100,000 live births. Maternal mortality ratio has been reduced over the past two decades though still it remains high, with most maternal deaths happening during the postpartum period [14,15]. The country has a significant proportion (above half, 52%) of home deliveries, and has been implementing PNC through home visitation as one of the interventions to reduce maternal and child morbidity and mortality [1,16]. In Ethiopia, the Health Management Information System (HMIS) has indicators used to monitor the implementation of PNC for mothers and babies, which includes both health facility and home visits [7]. However, PNC coverage is very low with 42-48% of mothers delivering in a health facility receiving PNC, and only 1-2% of women delivering at home receiving PNC in the first 2 days of postpartum period [1,14].
The main reasons listed for low PNC utilization in studies of different countries range from low awareness about the need to fear of health facilities [17,18]. Furthermore mothers who deliver at home may experience discrimination in receiving PNC due to some locally practiced measures like financial penalties for home births [18,19]. Moreover; the perspective of women, their families and community on the quality of service could influence their decision whether or not to seek care [20,21]. Maternal, newborn and child health issues are national priorities of many countries. However, specific policies for postpartum care are weak, and there is a very little evidence of effective PNC implementation [7,13]. The ideal way to provide maternal and child health services such as PNC is through health services delivered by skilled personnel in health facilities. However, in developing countries like Ethiopia, a more realistic model requires working hard to strengthen the health system and improve access to facilitybased care, while also enabling access to PNC at the community level through home visit [13,22].
Comprehensive studies on utilization of PNC through facility-based and home-based care are rare. However, having such inclusive evidence is critically important to develop appropriate strategies to help improve the low service utilization and reduce the maternal and newborn morbidity and mortality, especially if the evidence is derived from country specific data like the Demographic Health Surveys. Therefore, this study aimed to assess the utilization of PNC and its determinants among mothers who delivered at home in Ethiopia.

Data source and study design
This was a cross-sectional secondary analysis of data collected in the 2016 Ethiopian Demographic Health Survey (EDHS). It is the fourth EDHS conducted by the Central Statistical Agency as part of the International Demographic and Health Survey program. The data for mothers who delivered at home were extracted from the pregnancy and PNC dataset of the EDHS 2016 and included. We only included the most recent child of the women so as to avoid mix-ups in the recall and reporting of mothers experiences, especially for mothers who had more than one birth in the previous 5 year period. Additionally, mothers who did not remember the PNC they received for either the mother herself or her newborn were excluded from analysis. Therefore, a total of 4491 mothers (between 15 and 49 years) were included in this study.

Outcome variable
To develop the outcome variable from the dataset the question "where did you give birth?" was used as a starting point for this study. According to this question the dataset is categorized in to two subgroups i.e. "1=health facility delivery" and "2 = home delivery". Accordingly, from the total (N = 7188) mothers, 37.5% (n = 2697) delivered at a health facility and 62.5% (n = 4491) of them gave birth at home. The latter subgroup was used as our study participants in this study. Using the responses to the questions "Did anyone check on your health after you gave birth?" and "In the two months after delivery, did anyone check the health of your child?" we determined how many mothers and newborns received postnatal care services. According to the WHO, PNC service is defined as service given for both the mother and her newborn within 42 days of delivery [2]. However, the EDHS 2016 dataset contained data for care provided to newborns extending up-to 2 months. This does not contradict to the WHO's definition rather it is better to give comprehensive understanding on the community based newborn care (CBNC) program which is being implemented as an intervention to decrease the high ratio of newborn deaths in the country [23].
For mothers who indicated that they had been "Checked by anyone", we used the questions "Where did your first health check take place?" and "Where did your child's first health check take place?" Responses categorized "at health facility" included those who were checked at governmental, private and NGO health facilities and "at home" was used to define those receiving PNC at home. Since PNC includes services for both the newborn and the mother, we consider checked at home and health facility for either of the two (mother or her newborn) or both for those who were checked by Doctor, Nurse, Midwife, Health Officer or Health Extension Worker using the data on the questions "who checked your health at that time?" and "who checked your child's at that time?". Respondents who were checked by traditional birth attendants and other family members were considered as not receiving PNC.
The outcome variable was "place of PNC utilization" which was categorized in to three levels: 1) At home when either the mother or newborn or both were checked at home by a health personnel after giving birth at home, 2) At health facility when either of mother or newborn or both were checked at health facility after giving birth at home and 3) Not utilized at all when neither the mother nor newborn were checked by a trained personnel either at home or at a health facility a giving birth at home.

Independent variables
The independent variables included socio-demographic factors (age, marital status, educational status, place of residence, region type, religion, number of family members and sex of family head), socio-economic factors (wealth status, respondent's and husband's occupation, having a bank account and autonomy for health service utilization) and maternal health service, and health literacy communication related factors (wantedness of the last child, ever using anything to delay pregnancy, ANC utilization and frequency, awareness on fistula, owning a mobile phone and frequency of listening to radio).

Data analysis
Frequencies and proportions were used to describe categorical variables using cross tabulation. Multi-collinearity was checked using Variance Inflation Factor (VIF) test and variables with value of > 5 were excluded from the model. Multinomial logistic regression model was applied to identify determinants of place of postnatal care service utilization. Likelihood ratio test was used to see the model fitness and the value was 0.000 which showed the model was well-fitted. The adjusted odds ratio (AOR) was reported with their 95% confidence interval, and a p-value of < 0.05 was considered to declare statistical significance. The overall correct predicted classification of the model in this analysis was 91.9%.

Socio-demographic characteristics
Among young mothers (15-24 years of age) who delivered at home, only 4.5% (n = 48) utilized PNC at a health facility and more than 90% (n = 980, 92.9%) did not utilize PNC at all. Among urban and rural residents, 14% (n = 36) and 7.8% (n = 330) utilized PNC service at a health facility or home respectively. A slightly higher percentage of mothers with 1-4 family members utilized PNC either at home or at a health facility (8.4%) compared to mothers with > 8 number of family members (6.7%) ( Table 1).

Socio-economic characteristics
A higher proportion of mothers in the highest wealth index utilized PNC either at home or health facility than participants in the lowest wealth index (14.4 and 5.5% respectively). Furthermore, mothers with a bank account were more likely to utilize PNC either at health facility or home compared to participants without bank account (20.8% (n = 38) and 7.6% (n = 328) respectively) ( Table 2).

Health service and health literacy communication related characteristics
About 12% (n = 105) of mothers who had received 4+ ANC visits utilized PNC at health facility, while only 2.4% (n = 55) of mothers who did not receive ANC utilized PNC at health facility. Furthermore 19% (n = 148) of mothers who were told about pregnancy danger signs during their ANC visit utilized PNC at home or a health facility, compared to only 10% (n = 139) of mothers who were not counseled on danger signs during ANC. A higher proportion of mothers who were aware of fistula utilized PNC at home or health facility compared to those who were not aware of fistula (12.1 and 6.7% respectively) ( Table 3).

Magnitude of PNC utilization
From the total 4491 mothers who delivered at home included in the study, 130 (2.9%) utilized postnatal service at home, while 236 (5.3%) of them utilized PNC at a health facility (Fig. 1).    (Table 4). Similar to the findings on factors associated with receiving PNC at home, there were a number of factors associated with receiving PNC at a facility among women who delivered at home. Religion, wealth index, ever using anything to delay pregnancy, wantedness of the pregnancy, ANC visits and listening to radio were associated with facility-based PNC utilization. Women who follow the Orthodox religion were 69% more likely (AOR = 1.698, 95%CI 1.137-2.539) to utilize PNC at a health facility than not utilizing at all when compared to women whose religion was "others"(not orthodox or Muslim). Mothers in the low wealth index were 40% less likely (AOR = 0.608, 95%CI 0.424-0.873) to utilize PNC at a health facility than not utilizing at all when compared to women in the upper wealth index. Women who wanted their pregnancy were 42% less likely (AOR = 0.579, 95%CI 0.352-0.953) to utilize PNC at a health facility than not utilizing at all when compared with women who had not want to became pregnant. Lastly, listening to radio was associated with PNC utilization in a health facility; women who did not listen to radio at all were 34.8% less likely (AOR = 0.652, 95%CI 0.432-0.984) to utilize PNC at a health facility than not utilizing at all when compared with women who listened to radio at least once a week when other variables were held constant (Table 4).

Discussion
This study used data from the 2016 DHS assessment in Ethiopia-the fourth survey to capture detailed information on maternal and newborn health in general and PNC services in particular at national level. In this paper we have explored the coverage of home-based and facility-based PNC utilization among mothers who delivered at home, and factors associated with place of PNC service utilization in Ethiopia. Below we will discuss the implications of the study findings for program, policy and future research in Ethiopia.
This study shows that only 8.2% of mothers who delivered at home utilized PNC services (2.9% of mothers received at home and 5.3% a health facility). Analysis of the factors associated with place of PNC utilization in the study show that region type, religion, wealth index, having a bank account, ever using anything to delay pregnancy, wantedness of pregnancy, antenatal care visits, and radio listening appear to be important factors for PNC utilization either at home or facility in Ethiopia.
In this study, the magnitude of PNC utilization either at home or at a health facility was similar with findings from some studies [1,[24][25][26] though lower than the findings from other studies [11,13,[27][28][29][30]. This difference could be due to differences in the schedule of visits considered in the different studies, which could affect the comparability and results between the different studies [1]. Additionally, it could reflect differences in the socio-demographic and socio-economic characteristics of the study populations. Furthermore, these studies may have included PNC given at a health facility to women who delivered in health facilities, which would be expected to be higher compared to a study like ours', which only considered mothers who delivered at home.
In our study, mothers from urban regions were more likely to utilize PNC at home compared to participants from agrarian regions while there was no statistically significant variation in PNC utilization in a facility by region type. Being from urban regions (Addis Ababa, Dire-Dawa, and Harrar) could expose women to improved awareness and good access to trained professionals [24,31] even to mobile home-based care. Utilization of services that precede PNC, including ever using anything to delay pregnancy and receiving ANC were two other factors which were significantly associated with PNC service utilization, both at home and at a health facility. This is not surprising, as many other studies [28][29][30][32][33][34][35][36][37] found similar findings. It is plausible that if mothers had previous contact with health professionals (especially for ANC), they would be more likely to seek PNC at a health facility or to have contact with a community health worker after giving birth to receive PNC at home [1,32,38].  Surprisingly, although the vast majority of women wanted the pregnancy at the time they became pregnant (80%, n = 3605), women who did not wish to become pregnant then were more likely to utilize PNC at health facilities compared to those who wanted the pregnancy when they become pregnant. This may be related a higher incidence of complications developing after birth among those with unwanted pregnancies. Additionally, findings from other studies have revealed that mothers may not seek care at health facilities if they did not develop complications [1,18,39,40], and those who did not want the pregnancy may develop more complications than those who wanted it.
Media exposure was also associated with PNC service utilization, women who listened to radio at least once a weak were more likely to utilize PNC at a health facility compared to those who did not listen to radio at all. Media exposure could improve the awareness of women of the value of seeking care [41], including PNC, even among mothers who gave birth at home.
Furthermore, mothers with a bank account were more likely to utilize PNC at home than those without bank account, and mothers from the highest wealth index were more likely to receive PNC at a health facility compared to women in the lowest wealth index. These findings are similar with other findings [31,[42][43][44]. A possible pathway of the influence of these factors could be that mothers with the lowest income could face financial hardship, including in obtaining transport to a health facility, while those with the highest income may struggle less with transport coast. In this study, mothers who follow the Orthodox Christian religion were more likely to utilize PNC at a health facility when compared with women of other religions, excluding Muslim women. Though we could not find any reference to explain this finding, a possible reason could be that there may be differences in health facility access among the Orthodox Christian followers compared to other religious groups (excluding Muslim women). Furthermore, there could be differences in the involvement of religious leaders in advising followers to utilize modern health services. However, this needs further research to explore the differences in PNC utilization among followers of different religions.

Limitations and strengths of the study
This study made use of cross-sectional data from the 2016 Ethiopian Demographic and Health Survey. The data relies on women's self-reported care utilization, and may be influenced by recall bias, given that the study events took place within the 5 years preceding the survey. However, the study has a number of strengths. The data is national survey data, and the sample size is powered to be generalizable at national and regional level. This study was unique in that conducted advanced analysis of PNC coverage and location, and factors associated with place of PNC service utilization among mothers who delivered at home. These findings will provide useful information for practitioners and policy makers for increasing utilization of PNC among the substantial population of mothers who give birth at home. It also highlights research areas that need further studies in the future to better understand the barriers and facilitators of place of PNC.

Conclusion
The utilization of postnatal care service among mothers who delivered at home is low in Ethiopia. Living in urban regions, having a bank account, and ever using anything to delay pregnancy were associated with receiving PNC at home. Furthermore being in the upper wealth index, having unwanted pregnancy, receiving 4+ ANC visits and listening to radio weekly were associated with PNC utilization in health facilities among women who delivered at home. Therefore, targeted measures to improve socioeconomic status, strengthen the continuum of care and improve health literacy communication are critically important to mitigate the gaps on postnatal care service utilization among women who deliver at home.