Introduction

In 2020, there were over 800 preventable pregnancy and childbirth-related deaths every day, or about one death every two minutes1. In 2020, 287,000 women lost their lives both during and after pregnancy and childbirth worldwide. Low and middle-income countries accounted for about 95% of all maternal deaths, the majority of which were preventable2. Of the anticipated 253,000 maternal fatalities worldwide in 2020, 87% occurred in Sub-Saharan Africa and Southern Asia. Approximately 70% of maternal deaths (202, 000) occurred in Sub-Saharan Africa2,3 Yet maternal deaths only provide a portion of the picture. Twenty to thirty additional women will experience both short- and long-term problems, such as obstetric fistula, infections, a ruptured uterus, or pelvic inflammatory disease, for every woman who dies from pregnancy-related reasons4.

Ethiopia’s alarmingly high maternal mortality rate (MMR) persists despite all efforts and actions. The Demographic Health Survey of 2016 (DHS-2016) indicates that Ethiopia has an MMR of 412 per 100,000 live births5. In addition, every year over 500,000 Ethiopian women and girls experience disability as a result of complications during pregnancy and childbirth. Pregnancy problems, such as hemorrhage (29.9%), obstructed labor/ruptured uterus (22.3%), pregnancy-induced hypertension (16.9%), puerperal infection (14.68%), and unsafe abortion (8.6%), are the leading causes of maternal death in Ethiopia1,4.

Knowledge of pregnancy, labor, and delivery complications is crucial for safe motherhood6. Pregnancy danger signs are symptoms that should raise red flags for a pregnant woman and her fetus and necessitate prompt medical intervention7. The most typical warning signs of pregnancy include swollen hands and/or face, blurred vision, and heavy vaginal bleeding: Severe vaginal bleeding, protracted labor, and convulsions are dangerous indicators during labor and delivery; severe bleeding after delivery, unconsciousness after delivery, and fever are danger signs during the postpartum period8. Numerous problems that cause maternal deaths and contribute to prenatal deaths can have abrupt and severe onsets, and they are often unforeseen9. One of the most frequent reasons people fail to recognize a problem when it occurs and delay health care-seeking behavior is poor knowledge of pregnancy complications9,10. Ethiopia’s national reproductive plan places a strong focus on maternal and newborn health to lower the high rates of maternal and neonatal deaths10.

Achieving the 2030 Sustainable Development Goals, particularly target 16, requires extensive ongoing efforts11. The Ethiopian Government has initiated several reforms, such as launching the Community Based Health Insurance (CBHI) program in 201012, to enhance healthcare financing and meet SDG targets, particularly in reducing maternal mortality and improving access to primary healthcare13. Although maternal mortality rates in Ethiopia decreased between 2000 and 2016, they remain alarmingly high1,3. Our study data can assess whether maternal awareness of pregnancy-related issues has improved since the last Ethiopian Demographic and Health Survey (EDHS).

The prevalence of knowledge regarding pregnancy complications varies across different locations: 26.3% in Chiro town14, 74.4% in Debre Tabor Town15, 63.2% in Hosanna Town16, and 21.9% in Gedeo zone17. According to different literature, knowledge of pregnancy complications was significantly associated with age15,16,18, religion15 educational status16,19,20, occupation18,20, parity16,18, parity21, place of delivery9, and frequency of ANC visits15,16,18. Despite the availability of localized studies in Ethiopia, there remains a scarcity of nationally summarized data on women’s knowledge about pregnancy complications. Furthermore, these studies often do not assess the influence of factors such as mode of delivery, type of birth, birth spacing, regional variation, and household head on women’s knowledge of pregnancy complications. Therefore, the current study aimed to assess individual and community-level factors associated with women’s knowledge about pregnancy complications in Ethiopia based on the 2019 MEDHS data.

Methods

Study design, setting, and period

Cross-sectional data from the EMDHS 2019 were used in this study. The EMDHS is a five-year nationally representative survey. Since 2000, Ethiopia has carried out four EDHSs every five years. The country is divided into nine regions and two city administrations. Contextually, these regions are categorized as agrarian (Southern Nations, Nationalities, and People Region (SNNPR), Amhara, Benishangul-Gumuz, Gambela, Harari, Oromia, Tigray); pastoralist (Afar, and Somali); city administrations (Addis Abe and Dire-Dawa)22,23,24,25. We obtained the data for this study from the DHS website (www.dhsprogram.com) after requesting authorization online and providing justification for the investigation. From the woman record (IR file), we extracted the independent and dependent variables. EDHS is a nationally representative household survey that is carried out on a broad spectrum of demographics through face-to-face interviews. The Central Statistical Agency generated a sampling frame of all census enumeration areas for the 2019 Ethiopia Population and Housing Census (PHC), which was used by EMDHS 2019. An Enumeration Area (EA) is a geographic area covering an average of 131 households. A comprehensive list of 149,093 Enumeration areas was produced for the 2019 PHC. Enumeration areas were sampled separately in each of the 21 strata resulted from stratifying the nine regions and two city administrations into rural and urban categories. Before choosing the sample, a probability proportionate allocation was done based on the size of the Enumeration area. Prior to conducting the analysis, we verified the inclusion of the outcome variable in the EMDHS dataset and assessed all study variables for missing data. Cases with missing observations were omitted from the analysis. Furthermore, the dataset was weighted to correct for sample non-representativeness across regions in Ethiopia, thereby ensuring accurate estimates and standard errors. The study analyzed a weighted sample comprising 1,655women who gave birth in the last 12 months. The detailed sampling procedure is presented in a full EMDHS 2019 report26.

Study variables

Dependent variable

The outcome variable of this study is maternal knowledge about pregnancy complications.

Independent variables

Independent variables were classified into individual and community-level factors. Individual level factors were religion, maternal age, maternal education, head of household, wealth index, marital status, mode of delivery, type of birth, frequency of ANC visit, and sex of the child. Community-level factors were residence and administrative region.

Measurement of variables

The 2019 EMDHS data had eight dichotomized questionnaires that addressed maternal knowledge about pregnancy complications. The eight pregnancy complication questionnaires are S415A, S415B, S415C, S415D, S415E, S415F, S415G, and S415X. These eight questionnaires were used to collect data from mothers using No (0) and Yes (1) options.

Definition

Pregnancy complications

Pregnancy complications include vaginal bleeding, vaginal gush of fluid, severe headache, blurred vision, fever, abdominal pain, convulsion, and others27.

Knowledge about pregnancy complications

Assessment of what mothers described about pregnancy complications using the aforementioned eight questions.

Knowledge about pregnancy complications

A comprehensive assessment of participants’ knowledge regarding pregnancy complications was conducted using a set of eight items. Correct responses were assigned a value of 1, while incorrect responses were assigned a value of 0. The resulting scoring scale ranged from 0 to 8 points. Subsequently, participants’ level of knowledge was dichotomously categorized as either “good knowledge” (those surpassing the mean score) or “poor knowledge” (those falling below the mean score), based on the cumulative mean score of participants’ knowledge of pregnancy complications16,28.

Data management and statistical analysis

We weighted the data to account for the non-proportional distribution of samples among strata and regions before doing the descriptive data analysis. Next, using weighted frequencies, mean ± (standard deviations), and percentage, descriptive statistics were calculated and presented. STATA version 17 (STATA Corporation, IC., TX, USA) was used for all analyses.

Multilevel logistic regression was employed due to the hierarchical and nested nature of the EDHS data. To determine the random effect, intra-community correlation (ICC) was used, computed as, \(ICC = \frac{{\delta {\text{a}}^{2} }}{{\delta {\text{a}}^{2} + \delta {\text{b}}^{2} }}\) where \(\delta {\text{a}}^{2} \;{\text{ and}}\; \delta {\text{b}}^{{2}}\) are the community-level and individual-level variance respectively. Individual level variance (\(\delta {\text{b}}^{{^{2} }}\)) is equal to \(\pi^{2}\)/3 which is a fixed value. The median Odds Ratio (MOR) was calculated as MOR = e0.95*\(\sqrt {V{\text{a}}1}\), where Va1 is the variance in the empty model, and Proportional Change in Variance (PCV) was computed as PVC = \(\frac{Va1 - Va2}{{Va1}}\), where Va2 represents the neighborhood variance in the succeeding model and Va1 represents the variance of the empty model. Deviance (-2LL) was used to measure goodness of fit, and the Likelihood Ratio (LR) test was used to compare the models.

Patient and public involvement statement

The general public and patients (participants) were not consulted in the planning or design of the study. Because we used secondary data (DHS), patients were not consulted to interpret the results, nor were they asked to participate in the preparation or editing of this article to ensure its accuracy or readability.

Results

Background characteristics of study participants

A total of 1655 study participants were involved in this study. The majority, 998 (60.2%), of the study participants were in the age group of 15–29. The majority of the women, 218 (13.2%), were from the Amhara region, followed by Tigray with 203 (12.3%) and SNNPR with 200 (12.1%). Regarding educational status, 612 (37.0%) of the respondents had primary education. Concerning marital status, 1523 (92.0%) of the study participants were married (Table 1).

Table 1 Background characteristics of study participants in Ethiopia (Weighted, N = 1655).
Full size table

Obstetrics characteristics of the study population

The majority, 1463 (88.4%), of the participants gave birth by spontaneous vaginal delivery. Regarding the frequency of ANC visits, only 157 (9.5%) of the respondents had four or more ANC visits (Table 2).

Table 2 Obstetrics characteristics of the study population.
Full size table

Respondent’s knowledge of pregnancy complications

In the current study, the majority of the study subjects, 741 (44.8%, CI: 42.4%-47.2%), were found to be knowledgeable about pregnancy complications. Vaginal bleeding was the most frequently mentioned pregnancy complication by the study participants, 1105 (66.8%), followed by severe headache, 856 (51.7%) (Fig. 1).

Fig. 1
figure 1

Respondent’s knowledge about pregnancy complications in Ethiopia.

Full size image

Associated factors of knowledge about pregnancy complications

The random effect analysis result

According to the ICC in the null model, 17.1% of the total variability in the knowledge of pregnancy complications was attributed to changes between clusters, while the remaining variability was due to variation within clusters. However, the community-level variability decreased to 8.9% when considering both individual and community-level predictors (i.e., in the combined model). Our use of multilevel modeling proved more effective than the conventional single-level regression model, as indicated by the presence of a non-zero ICC in the null model. Furthermore, the variation in knowledge of pregnancy complications between clusters was demonstrated by a MOR of 2.12 in the null model. The proportional change in variance (PCV) of the multilevel model indicated that 53% of the variability in knowledge of pregnancy complications could be explained by the combined effects of individual and community-level factors. Deviance was used to compare models, and model III was identified as the best-fit model with the lowest deviance (Table 3).

Table 3 The random effect analysis result.
Full size table

The fixed effect analysis result

In the crude multilevel modeling, maternal education, frequency of ANC visits, and administrative region were significantly associated with knowledge of pregnancy complications. Women educated to secondary and higher levels were 1.54 (AOR = 1.54, 95% CI = 1.04–2.29) and 1.74 times (AOR = 1.74, 95% CI = 1.11–2.72) more likely to be knowledgeable about pregnancy complications than women with no education, respectively. The odds of good knowledge about pregnancy complications among women with four or more ANC visits (AOR = 0.74, 95% CI = 0.49–0.98) decreased by 26% compared to women with 1–3 ANC visits. Women living in Amhara (AOR = 2.10, 95% CI = 1.24–3.55) and SNNPR (AOR = 3.92, 95% CI = 2.10–7.31) were more likely to have good knowledge about pregnancy complications compared to women residing in Tigray. Conversely, women residing in Harari and Dire-Dawa were 80% (AOR = 0.20, 95% CI = 0.08–0.44) and 52% (AOR = 0.48, 95% CI = 0.24–0.95) less likely to be knowledgeable about pregnancy complications than women living in the Tigray region, respectively (Table 4).

Table 4 Factors associated with pregnancy complications in Ethiopia, EDHS 2019.
Full size table

Discussion

In this study, we investigated women’s knowledge about pregnancy complications in Ethiopia and identified several significant predictors. Our findings indicate that women’s knowledge levels were notably lower compared to those reported in a specific literature review focused on the country. Notably, maternal education, frequency of ANC visits, and administrative region emerged as significant predictors of knowledge about pregnancy complications.

In 2019, the prevalence of knowledge about pregnancy complications in the country was 741 (44.8%). Our study observed a lower prevalence rate compared to several studies conducted in different parts of Ethiopia. Specifically, higher prevalence rates have been reported in Debre Tabor (74.4%)15, Hossana (63.2%)16, Bahir Dar (59%)30, and Mechekel District (55.1%)28. These differences may be attributed to variations in sample sizes, study designs, and regional differences in antenatal care (ANC) utilization. Local socio-cultural factors may also influence knowledge about pregnancy complications, contributing to the observed discrepancies.

In contrast, our findings were higher than those reported in other Ethiopian studies. For example, lower prevalence rates were observed in Shashemane town (40%)9, Nekemte town (32%)20, Goba district (22.1%)19, Afar (7.9%)33, and Wolaita Sodo town (16.8%)18. The variations in these findings could be due to differences in local healthcare practices, levels of health education, and accessibility to ANC services.

When comparing our findings to studies conducted outside Ethiopia, we also observed higher prevalence rates compared to those reported in Tanzania (32%)31 and Rwanda (17%)32. The discrepancies between these international findings and our study may be attributed to differences in healthcare systems, health education levels, and accessibility to ANC services across these countries.

The multilevel analysis revealed that women who educated secondary and higher levels were 1.54 (AOR = 1.54, 95% CI = 1.04–2.29) and 1.74 (AOR = 1.74, 95% CI = 1.11–2.72) times more likely to be knowledgeable about pregnancy complications than women with no education respectively. This finding was supported by studies conducted in Debre Tabor15, Hossana Town19, Goba district19, and Afar33. The observed phenomenon may be attributed to the fact that older women, having experienced obstetric danger signs in previous pregnancies, may possess greater knowledge about these signs during subsequent pregnancies. This underscores the need for young women undergoing their first pregnancy to receive additional attention in counseling and health education. Studies indicate that health education delivered during antenatal care significantly improves mothers’ understanding of obstetric danger signs34.

In this study, the odds of good knowledge about pregnancy complications among women with four and above ANC Visits (AOR = 0.74, 95% CI = 0.49–0.98) was decreased by 26% to that of women with 1–3 ANC visits. This finding contrasts with studies conducted in Rwanda32 Goba district19, and Debre Tabor town15. The result indicating decreased knowledge about pregnancy complications among women with four or more ANC visits could stem from increased reliance on healthcare providers for managing their pregnancy, potentially reducing proactive information-seeking behaviors. Additionally, frequent visits might lead to information overload or a perception of comprehensive care, diminishing the perceived need for additional knowledge acquisition. Variations in the quality and content of ANC counseling could further influence the depth of understanding regarding pregnancy complications35.

Furthermore, knowledge of pregnancy complications varied significantly by region. Women residing in Amhara were 2.10 times more likely to have good knowledge (AOR = 2.10, 95% CI = 1.24–3.55), and those in SNNPR were 3.92 times more likely (AOR = 3.92, 95% CI = 2.10–7.31), compared to women residing in Tigray. This finding is consistent with a study conducted in Ethiopia36. Women residing in Harari and Dire-Dawa were 80% and 52% less likely, respectively, to be knowledgeable about pregnancy complications compared to women living in the Tigray region. This discrepancy might be attributed to differences in the socio-cultural characteristics of the regions. Another possible explanation could be variations in ANC utilization between less populated and pastoral regions (Harari and Afar) compared to densely populated regions33,37. Furthermore, the variation may be attributed to the stronger access to healthcare facilities in rural regions and city administrations. This suggests that access to healthcare services could vary across different regions of the country.

Strengths and limitations of the study

The study utilized extensive national survey data, providing sufficient statistical power and reliable estimates through sampling weights. By examining knowledge of pregnancy complications at individual/household and community levels, it explored hierarchical influences on outcomes. However, the cross-sectional nature of the data limits causal inferences between knowledge of fertility periods and independent variables. Additionally, reliance on self-reported information introduces the potential for recall bias.

Conclusion

This study found that nearly half (44.8%) of the study participants demonstrated knowledge about pregnancy complications. Maternal education, Frequency of ANC visits, and administrative region were significant predictors of knowledge about pregnancy complications. Such a low level of knowledge may make it more difficult for women who experience pregnancy complications to get access to obstetric care. Enhancing antenatal counseling services and disseminating health information regarding pregnancy complications in the community should be priorities.

Implications of the study

Our study underscores the need for targeted interventions to improve women’s knowledge about pregnancy complications in Ethiopia. Policies aimed at enhancing maternal education, increasing the frequency of antenatal care visits, and addressing regional disparities in healthcare access are crucial. Strengthening community-based health education programs and integrating comprehensive information on pregnancy complications into existing maternal health services can play a pivotal role in improving maternal health outcomes across the country.