Trends of delayed rst antenatal care visit among pregnant women in Ethiopia based on the EDHS 2000-2016 data: Multivariate decomposition analysis

Background: Timing of rst antenatal care visiting is crucial for women and fetus health. However, most of the women had delayed antenatal care visits and lacked early detection of existing complications. Hence, this study aimed to assess the trends and magnitude of delayed rst antenatal care visits among pregnant women in Ethiopia. Method: This study used cross-sectional data analysis on Ethiopian Demographic and Health Surveys. A total weighted sample of 2146 in 2000, 2051 in 2005, 3368 in 2011, and 4740 in 2016 women who received at least one antenatal care visit in the past ve years before each survey were included. Stata 14.1 software was used for data management and further analysis. Trends and multivariate logistic decomposition analyses were employed to examine the change in delayed rst antenatal care visits with sampling weighting for the survey data. The 95% condence interval was used for considering statistical signicance. Results: The proportion of delayed rst antenatal care visit was declined from 76.8 (95%CI: 75.1−78.6) in 2000 to 67.3 (95%CI: 65.9−68.6) in 2016. The decomposition analysis showed that 39% of the overall change in delayed rst antenatal care visit over time was due to differences in characteristics whereas about 61% was due to the effect of characteristics. The signicant contributing factors for the decline of delayed rst antenatal care visit were the place of residence, husband’s education, women occupation, told about pregnancy complications, caesarean delivery, and family size. Conclusion: The proportion of delayed rst antenatal care visits has been declined over time in Ethiopia, but still it remains high. More of the overall change in delayed rst antenatal care visit was attributed to a change in health behavior among pregnant women. Public health interventions or programs such as health education and behavioral change communication targeting the rural residents, lower economic status, and family size would be suggested to decline delayed rst antenatal care visit.

still it remains high. More of the overall change in delayed rst antenatal care visit was attributed to a change in health behavior among pregnant women. Public health interventions or programs such as health education and behavioral change communication targeting the rural residents, lower economic status, and family size would be suggested to decline delayed rst antenatal care visit.

Background
Antenatal care (ANC) is essential care that is given by professional health workers during pregnancy and childbearing to minimize and treat or manage existing and a leading cause of maternal and newborns mortality [1]. The World health organization recommended that every pregnant woman who is either with or without complication should have a minimum of four and above ANC visits, of which rst ANC visits take place in the rst trimesters to reduce any pregnancy and childbearing complications [1,2]. The recommended rst ANC visit has undertaken within 16 weeks or four months but not later in the federal ministry of health, Ethiopia [3]. Globally, about 86% of pregnant women have received at least once antenatal care visits of which 65% of women received at least four and above antenatal care visits while in Sub-Saharan Africa, 55% of pregnant women had received at least four and above antenatal care visits [4]. In Ethiopia, only 32% of the women received four and above antenatal care visits [5]. Timing of rst antenatal care visits can affect the quality of antenatal care and under-utilize of the recommended ANC visits that leading to adverse pregnancy outcomes, which are reduced through proper medication and treatments at risk of pregnancy [6]. Ethiopia is one of the countries that faced the highest maternal and infant mortality and morbidity. As the Ethiopian Demographic and Health Survey (EDHS) report stated that the pregnancy-related mortality ratio was 871 per 100,000 in the years 2000, it was 673 in 2005 per 100,000 live births, 676 in 2011, and 412 in 2016 per 100,000 live births [5]. Pregnancy complications are leading to a higher maternal and child disability as well as death in developing countries [7]. Globally, for every day, about 810 women die due to pregnancy complications or births [8]. Nearly one-third of million maternal death has occurred a year and about 94% of the death occurred in developing countries [9]. More than halves of these deaths of women have occurred in sub-Saharan countries [2] which included Ethiopia had a higher maternal mortality ratio due to lower utilization of ANC care services and they are postponed rst ANC visit in the second and third trimester [5,10].
In Ethiopia, the latest EDHS 2016 indicated that 37% of women had no ANC visits in the last ve years before the survey, and 20% of pregnant women made their rst ANC visit before the four months whereas urban women have made their ANC visits earlier than rural resident women [5]. When women delaying for their rst ANC visits, they have a higher risk of poor pregnancy outcomes, maternal and child mortality [11].
Consequently, pregnant women would have been missed the intended bene ts of ANC could include early detection and management of existing health conditions and complications of pregnancy to prevent the lifethreatening maternal and newborns condition. Thus, delaying for antenatal care makes it di cult to accomplish effectively the regular rst ANC visits that increase the well-being of maternal and pregnancy outcomes [2,12]. The low ANC coverage, failing to attend, and delayed rst antenatal care visits among pregnant women are a signi cant problem in Ethiopia [13]. Despite achieving improvement results in health indicators including the accessibility of maternal health care services and reduction of child mortality over time [10], Ethiopia is one of the countries with the highest rate of delayed rst antenatal care visiting in sub-Saharan Africa. Different studies have been identi ed determinants of maternal health care services utilization were included under predisposing, enabling, and needing factors [14,15]. The previous studies on the in uences of delayed initiation of ANC in local areas. But no study focuses on the trends and identify the contributing factors for the change in delayed rst antenatal care visit in Ethiopia over the studying period.
Studying the change in delayed rst ANC visit by using multivariate decomposition analysis to identify the signi cant factors contributing for the change in delayed rst ANC visit over time that have become important to targeting interventions on factors that decreasing delayed rst ANC visit and to forward policies and programs that focusing on reducing of delayed rst antenatal care visit in Ethiopia. Therefore, this study is aimed to address the magnitude and trends as well as contributing factors for change in delayed rst ANC information about sampling procedures was presented in the EDHS report [5,[16][17][18].

Source and study populations
The source population was all women who had received antenatal care visits in the past ve years for each survey in Ethiopia whereas the study population was all women who had received rst antenatal care visits in the past fteen years for each survey in the selected clusters of Ethiopia.

Outcome variable
The women were asked to report the timing of the rst antenatal care visits in each consecutive survey. The outcome variable was delayed rst antenatal care visits among women who received at least one ANC visit.
The response variable was a random variable and had two possible values. So, the response variable Y i for i th women was measured as a binary outcome variable and two possible values Yi=1, if women had received delayed ANC visit and Y i =0 if women had received early ANC visit.

Independent variables
The independent variables in this study were divided into predisposing, enabling, and need factors. The predisposing variables were the demographic and socio-cultural characteristics of the women that pre-exist to their health condition including maternal age, marital status, religion, women occupation, husband occupation, women education, husband education, parity, family size, and the number of living children. The enabling (economic) factors revealed that the means or facilitators needed to access maternal health care services that containing wealth index, head of household, distance to the health facility, source of information such as radio, television, and newspaper as well as the services. The third category of independent variables was need factors that re ect the perceived health status of the individual and were also the most potential causes leading to the health conditions. This includes contraceptive use, caesarean delivery, wantedness of pregnancy, told about pregnancy complication, and termination of pregnancy [19].

Data collection procedures
This study was done based on the Ethiopian demographic and health survey (EDHS) data by accessed from DHS program database website www.measuredhs.com after permission was granted through an online request by describing the objective of the study. The raw data were obtained from women who had received antenatal care visits in the selected clusters of a country through a structured and pre-tested questionnaire.

Data management and analysis
After clearing and editing, the raw data were weighted by using sampling weight, primary sampling unit, and strata to obtain the nationally representative sample of the surveys and the analysis was done using STATA The currently developed for the multivariate logistic decomposition analysis for the non-linear model for delayed rst antenatal care visit using mvdcmp STATA package [21].

Background characteristics of women
The following table showed that the individual characteristics of pregnant women in the ve years preceding  (Table 1).    (Fig. 1).

Decomposition analysis
This multivariate decomposition analysis was used to examine the trends of delayed rst antenatal care visit and identify factors either positively or negatively contributing to the change in the proportion of delayed rst antenatal care visit.

Decomposition analysis of delayed rst ANC visit in Ethiopia, 2000-2016
The decomposition of the change in the proportion of delayed rst antenatal care visit by considering the selected variables is used to understand the variations in the composition of characteristics or the change in health behaviors. The overall trend from 2000 to 2016 noted that there was a signi cant decline in the delayed rst antenatal care visit among pregnant women. The decline in the proportion of delayed rst antenatal care visits over the study period has been described by the differences in the women's characteristics and the differences in the effect of characteristics (Table 3).  The multivariate decomposition analysis result showed that about 39% of the overall change in the delayed rst antenatal care visit was due to a difference in characteristics or composition changes related to increased population access to antenatal care utilization (Table 3). Among the various important compositional factors, such as place of residence, listening to the radio, watching television, told about pregnancy complications, cesarean delivery and the number of family sizes in the household had a signi cant effect on the delayed rst antenatal care visit (Table 4). In general, the decomposition results indicated that changes in health behavior were the principal source of change in the proportion of delayed rst antenatal care visits among women between 2000 and 2016 regardless of the exposure variables.
An increase in the composition of women who reside in rural areas in the sample negatively contributes by 8.2% to the decline in the delayed rst antenatal care visit. Decreasing in the composition of women who had no exposure to listening to the radio in the sample revealed that a signi cant positive contribution to delaying the rst antenatal care visit, which contributes about 5.2% for the decline in delayed rst antenatal care visit (Table 4).
A decrease in the proportion of women who were not told about pregnancy complications in the sample contributes a positive 8% for the decline of the proportion of delayed the rst antenatal care visit. A decrease in the composition of women who had no experience of cesarean delivery in the sample revealed that a signi cant positive 0.3% contributes to the decline of delayed antenatal care visits among pregnant women (Table 4).
An increase in the composition of women who had three up to four-family sizes in the sample showed a signi cant negative 11.5% contributes to a decline in the proportion of delayed rst antenatal care visit. While a decrease in the composition of women who had ve and above family sizes in the house showed a signi cant positive contribution to the decline in the delayed the rst antenatal care visit by contributing 6.8% (Table 4).
Difference due to effects of the coe cient (effect of characteristics) Keeping the effect of change in compositional characteristics, about 61% of the overall change in the proportion of delayed rst antenatal care visit was due to the differences in the effect of characteristics (Table 3). The husband's education factor had a signi cant effect on the observed positive change in the proportion of delayed rst antenatal care visit. Women whose husband's education was secondary and above showed a signi cant positive contribution to the observed percentage decrease in the delayed rst antenatal care visit over the study period which contributed about 27.1% (Table 4).

Discussion
Delayed antenatal care is a major cause of pregnancy-related complications and the death of mothers and babies [22,23]. This study explored the magnitude, trends, and identi ed the signi cant contributing factors of delayed rst ANC visit in Ethiopia using the EDHS 2000-2016. The proportion of delayed rst antenatal care visits declined from 76.8% in 2000 to 67.3% in 2016. This nding was supported by the study done in Nigeria and Zimbabwe [24,25]. The possible reason for the decline in the trend of delayed rst ANC visits could be due to the role of health extension workers in enhancing the maternal health care services and by providing adequate ANC services at the households and community level in Ethiopia [26]. Besides, the establishment of the Health Development Army(DHA) aimed to enhance the achievement of the Health Extension Program [24] and improve the health care accessibility to meet the primary target of the MDG agenda [27] might help in decreasing the trend of delayed rst ANC visits.
In the decomposition analysis, the proportion of women who delayed the rst antenatal care visit have been signi cantly declined over time. Therefore, understanding the potential sources of changes has public health advantage for assessing which are the contributing factors for the change in delayed rst antenatal care visit and dealing with what is making progressively in the reducing delayed rst antenatal care visit to evaluate the implemented strategies and work on the identi ed contributing factors for delayed rst antenatal care visit.
This study showed that the contribution of behavior changes was more important than that of the composition changes which contributed by about 61% to the decline in the proportion of delayed rst antenatal care visits in the past 15 years during the study period. About 39% of the change in the delayed rst antenatal care visit over the study period 2000-2016 was observed due to the difference in the composition of women's characteristics over time. An increase in the composition of rural residing women showed a signi cant negative effect on delayed rst antenatal care visits and contributed about 8.2% of the change in the delayed rst antenatal care visit which implies that women residing in rural areas had delayed rst antenatal care visit. This nding was supported by the study done in Vietnam, and Zambia [28,29], in which women who lived in rural areas were initiated later and fewer antenatal care visits. The possible reason might be there is a lack of access to health care services in rural areas due to limited nances, the inaccessibility of health facilities and health care providers, lack of transportation services, and had no better opportunities for health information in the rural areas [30].
An increase in the composition of women who were not working showed a signi cant negative impact on the decline in the proportion of delayed rst antenatal care visit. This nding was supported by the study done in public health centers in Addis Ababa, Ethiopia in which women who had no occupation were more likely to delay the rst antenatal care visit [31]. The possible reason might be women who had no occupation were more likely to be economically dependent and consequently unable to access maternal health services recommended by health workers [32]. This implies that women who had no occupation or were not employed may not get important information about antenatal care visits from others in the workplace.
Decreasing in the composition of women who did not tell about pregnancy complications showed a signi cant positive contribution to the decline in the delayed rst antenatal care visit. This nding was supported by the study done in Malaysia and South Gondar, Ethiopia [33,34]. The possible reason might be the women may be aware of pregnancy complications and understanding the importance of antenatal care visits for pregnant women and fetuses. Besides, it suggests that health education had paramount importance in the improvement of the timing of ANC initiation and get intended to bene t early [35,36]. This implies that most women were not aware of the risks of pregnancy, pregnancy danger signs, or when to begin in ANC women may prefer to wait four months before con rming their pregnancy and sharing the conditions with others.
The composition change by categories of the number of family size associated with both negative and positive contribution to the decline of delayed rst antenatal care visit. An increase in the proportion of women who had three and four family size in the house negatively affected the proportion of delayed rst antenatal care visit, while decreasing in the proportion of women who had ve and above family size in the house positively affected the decline in the delayed rst antenatal care visit. This nding was supported by studies done in Cameroon, Ethiopia [37,38]. The possible explanation might be experienced a challenge to attend antenatal services due to time constraints considering their responsibilities for their children and health-related care services as well as resource constraints could affect health care utilization for treating pregnancy complications [39]. This result indicates that having a higher number of family sizes in the household may enhance the con dence of women to get delivered without attending early antenatal care visits and also may not give greater attention and care about the early antenatal care visits during the pregnancy period.
Decreasing in the composition populations in women who have not watched television and no listened to a radio program among women over the study period had a signi cant effect on the decline in the proportion of delayed the rst antenatal care visit. This nding was supported by studies done in North India, South India, and Northern Ethiopia [40][41][42]. The possible reason might be women might access information about antenatal care and could understand the timing of the rst antenatal care visit as well as get the awareness of where to go when they feel any pregnancy complications.
The composition change in the women who experienced a caesarean delivery among pregnant women over the study period had a signi cant impact on the decline in the proportion of delayed rst antenatal care visit. This nding was supported by the study done in Australia [43]. This could be women who have had a previous caesarean delivery are aware of the risk of pregnancy complications and also have increased risks of uterine rupture, placenta previa, and ectopic pregnancy [44]. This implies that caesarean delivery has occurred in the advanced facility setting and had got information for the next pregnancy without happening of high illness or any pregnancy-related complications during antenatal care visits. The husband's educational level was the only signi cant contributing factor for the decline in the delayed rst antenatal care visit due to the effect of characteristics. The change in behavioral characteristics of the husband who were secondary and above educational level had a positive contribution to the decline in the delayed rst antenatal care visit. This nding was supported by the other studies done in Ghana and Jamaica [45,46], in which women's husbands had secondary and above educational levels were less likely to delay the rst antenatal care visit. The possible reason could be the husband's education had a great role in improving awareness of health matters and the importance of ANC in particular. Advice from the husband may enable women to seek ANC and utilize the service early in pregnancy [47]. This result indicates that the husband's education had a signi cant effect on initiating women to understanding the importance of antenatal care visits for their pregnancy and educated husband accompany his wives for ANC.

Conclusions
Delayed rst antenatal care visit had shown a signi cant decrease among pregnant women over the last fteen years, but it was still high. The most important factors for the change in delayed rst antenatal care visit were the place of residence, husband's education, told about pregnancy complications, cesarean delivery, family size, and the watching of television, and listening of radio to change in the proportion of delayed rst antenatal care visit. Most of the change in delayed rst antenatal care visit has been occurred due to the improvement of health care behaviors through health care providers on maternal health care services and the timing of the rst antenatal care visit. Enhancing awareness about pregnancy complications in households and give information about the cause of poor pregnancy outcomes. From these ndings, the interventions of public health such as health education and community communication education should primarily be given attention to the contributing factors to reduces delaying to the rst ANC visit.

Consent of publication
Not applicable.