Trend change in delayed first antenatal care visit among reproductive-aged women in Ethiopia: multivariate decomposition analysis

Reproductive Health, Volume 19, No. 1, Year 2022

Interpretation

Background: Early first antenatal care visit is a critical health care service for the well-being of women and newborn babies. However, many women in Ethiopia still start their first antenatal care visit late. We aimed to examine the trend in delayed first antenatal care visit and identify the contributing factors for the trend change in delayed first antenatal care visits in Ethiopia over the study period 2000–2016. Method: We analyzed the data on reproductive-aged women from the four consecutive Ethiopian Demographic and Health Surveys to determine the magnitude and trend of delayed first antenatal care visit. A weighted sample of 2146 in 2000, 2051 in 2005, 3368 in 2011, and 4740 women in 2016 EDHS were involved in this study. All statistical analysis was undertaken using STATA 14. Multivariate logistic decomposition analysis was used to analyze the trends of delayed first antenatal care visit over time and the contributing factors to the change in delayed first antenatal care visit. Results: The prevalence of delayed first antenatal care visit in Ethiopia decreased significantly from 76.8% (95% CI 75.1−78.6) in 2000 to 67.3% (95% CI 65.9−68.6) in 2016. Decomposition analysis revealed that 39% of the overall change in delayed first antenatal care visit overtime was due to differences in women’s composition, whereas 61% was due to women’s behavioral changes. In this study, residence, husband’s education, maternal occupation, ever told about pregnancy complications, cesarean delivery and family sizes were significantly contributing factors for the decline in delayed first antenatal care visit over the study periods. Conclusion: The prevalence of delayed first antenatal care visit in Ethiopia among women decreased significantly over time. More than halves (61%) decline in delayed first antenatal care visits was due to women’s behavioral changes. Public health interventions targeting rural residents, poor household economic status and improving awareness about pregnancy-related complications would help to reduce the prevalence of delayed first antenatal care visit. This study entailed 2000, 2005, 2011, and 2016 Ethiopian Demographic and Health Surveys (EDHSs) data on reproductive-aged women, and these EDHSs were nationally representative surveys conducted in nine regions and two administrative cities. The survey used a two-stage stratified cluster sampling design to select respondents by separating each region into rural and urban areas; a total of 21 sampling designs or strata were created. In the first stage, a total of 539 clusters (401 in rural areas and 138 in urban areas) for EDHS 2000, 540 clusters (395 in rural areas and 145 in urban areas) for EDHS 2005, 624 clusters (437 in rural areas and 187 in urban areas) for EDHS 2011 and 645 clusters (443 in rural areas and 220 in urban areas) for EDHS 2016 were selected with proportional allocation to cluster size. In the second stage, household listing operations were performed in all selected clusters. On average, 27 to 32 households per cluster were selected proportional to the cluster size. We extracted relevant factors for this study from the Kids Record (KR file) dataset. A total of weighted sample 2146 in 2000, 2051 in 2005, 3368 in 2011, and 4740 in 2016 under reproductive-aged women were included in the study (Fig. 1). Comprehensive sampling procedures were described in the EDHS report [4, 25–27]. The extracted sample sizes from four consecutive EDHSs All reproductive-aged women who had been given births or attended antenatal care during the last pregnancy within five years before the survey in all selected clusters in Ethiopia were the study population. Nonetheless, women who did not know the exact time of their first ANC visit were excluded from this study. Women asked to report the exact time of their first antenatal care visit (in months) in each survey. Our outcome variable was a delayed first ANC visit, which was determined based on the timing of the first ANC visit [4, 8]. The binary response variable for women is denoted by a random variable Yi, with two possible codes. The two possible values coded as Yi = 1 if ith women started their first ANC visit after four months and Yi = 0 if they started their first ANC visit before four months or at four months. We classified the independent variables in the study based on Andersen-Newman’s behavioral model for maternal health care utilization as predisposing, enabling, and need factors. In the first category, predisposing variables are socio-demographic and socio-cultural characteristics of the respondents that exist before their health condition. Some predisposing factors are maternal age, marital status, residence, religion, women’s occupational status, husband’s occupational status, women’s educational level, husband’s educational level, mass media exposure, parity, family sizes and living children in a household. In the second category, enabling (economic) factors reflect the means or facilitators to access health care services like household head and wealth index. In the last section, needing factors are the immediate causes to use health services and reflect the perceived health status of the respondents. Ever use contraceptive in a previous pregnancy, cesarean delivery of last births, wanted last pregnancy, told about pregnancy complications last births or a previous pregnancy, and ever had a terminated pregnancy (miscarriage, abortion, or stillbirths) are considered as needing factors. We extracted data on reproductive-aged women from the Kids Recode (KR file) data set. Before doing any statistical analysis, the data on women were weighted using sampling weights or pweights for probability sampling and non-response rate to restore the representativeness of the survey and get reliable statistical estimates. Data analysis in this study included descriptive and multivariate decomposition analysis of the change in delayed first ANC visit. After extracting relevant variables for the study, we appended data on reproductive-aged women obtained from the four 2000, 2005, 2011, and 2016 EDHSs together for trends and multivariate decomposition analysis. Besides, multicollinearity was checked using variance inflation factor (VIF) and a VIF less than 5 for each independent variable. Therefore, there was no multicollinearity between independent variables since its VIF value ranged from 1.01 to 2.15 with a mean VIF of 1.4. The trend period was divided into four phases such as; first phase (2000–2005), second phase (2005–2011), third phase (2011–2016), and fourth phase (2000–2016) to see the differences in the prevalence of delayed first ANC visit over time-based on different selected characteristics of women. The trend was assessed using descriptive analysis stratified by various selected predictor variables and examined separately for each phase. Multivariate decomposition analysis was used to identify the contributing factors to the trend change in the outcome variable between any two surveys over time. This analysis focused on how a delayed first ANC visit prevalence responds to differences in selected women’s characteristics and how these variables shape the differences across the surveys conducted at different times. Decomposition analysis aimed to identify the potential sources of variations in the prevalence of delayed first ANC visit in the last 16 years. Multivariate decomposition analysis for the non-linear response model used the output from a logistic regression model since it is a dichotomous variable to parcel out the observed change in delayed first ANC visit between surveys into components. The difference in the percentage of delayed first ANC visit over time is attributable to the compositional change between any two surveys and the difference in the effects of those selected independent variables. That means the change in delayed first ANC visit is divided into the differences in characteristics (endowment component) and the effect of the selected variables (coefficient component). The analysis focused on the decomposition of the trend change in delayed first ANC visit between the reference year (2000) and the recent year (2016). The recent EDHS 2016 and reference EDHS 2000 surveys are denoted by A and B, respectively. For logistic regression, log-odds or logit of delayed first ANC visit divided into two main parts as follow: where E represents endowments explained by characteristics, and C denotes coefficients (unexplained) [28]. We can rewrite the above equation as follow: where β0B is the intercept in the regression equation for EDHS 2000, β0A is the intercept in the regression equation for EDHS 2016, βijB is the coefficient of the jth category of the ith determinant in EDHS 2000, βijA is the coefficient of the jth category of the ith determinant in EDHS 2016, XijB is the proportion of the jth category of the ith determinant in EDHS 2000, and XijA is the proportion of the jth category of the ith determinant in EDHS 2016. Currently developed multivariate logistic decomposition analysis used for the decomposition analysis of delayed first ANC visit using mvdcmp STATA package [29].

AI Digest

Study Justification:

– Early first antenatal care (ANC) visits are crucial for the health of women and newborns.
– Many women in Ethiopia delay their first ANC visit.
– This study aimed to analyze the trend in delayed first ANC visits and identify contributing factors.

Highlights:

– The prevalence of delayed first ANC visits in Ethiopia decreased significantly from 2000 to 2016.
– Behavioral changes accounted for 61% of the decline, while differences in women’s composition accounted for 39%.
– Factors such as residence, husband’s education, maternal occupation, awareness of pregnancy complications, cesarean delivery, and family size influenced the decline.

Recommendations:

– Target public health interventions towards rural residents and those with poor household economic status.
– Improve awareness about pregnancy-related complications.
– Address barriers to early ANC visits, such as lack of education and access to healthcare services.

Key Role Players:

– Ministry of Health: Responsible for implementing public health interventions and policies.
– Healthcare Providers: Involved in delivering ANC services and educating women about the importance of early visits.
– Community Health Workers: Engage with rural residents and provide education on maternal health.
– Non-Governmental Organizations: Support initiatives to improve access to healthcare services and raise awareness.

Cost Items for Planning Recommendations:

– Training and Capacity Building: Budget for training healthcare providers and community health workers on ANC services and awareness campaigns.
– Infrastructure and Equipment: Allocate funds for improving healthcare facilities and ensuring they have necessary equipment for ANC services.
– Outreach Programs: Budget for community outreach programs to reach rural residents and provide education on maternal health.
– Information and Communication: Allocate funds for developing and disseminating informational materials on ANC and pregnancy-related complications.
Please note that the cost items provided are general categories and not actual cost estimates. The actual budget would depend on the specific context and implementation plan.

Strength of Evidence

N/A

Abstract

This study entailed 2000, 2005, 2011, and 2016 Ethiopian Demographic and Health Surveys (EDHSs) data on reproductive-aged women, and these EDHSs were nationally representative surveys conducted in nine regions and two administrative cities. The survey used a two-stage stratified cluster sampling design to select respondents by separating each region into rural and urban areas; a total of 21 sampling designs or strata were created. In the first stage, a total of 539 clusters (401 in rural areas and 138 in urban areas) for EDHS 2000, 540 clusters (395 in rural areas and 145 in urban areas) for EDHS 2005, 624 clusters (437 in rural areas and 187 in urban areas) for EDHS 2011 and 645 clusters (443 in rural areas and 220 in urban areas) for EDHS 2016 were selected with proportional allocation to cluster size. In the second stage, household listing operations were performed in all selected clusters. On average, 27 to 32 households per cluster were selected proportional to the cluster size. We extracted relevant factors for this study from the Kids Record (KR file) dataset. A total of weighted sample 2146 in 2000, 2051 in 2005, 3368 in 2011, and 4740 in 2016 under reproductive-aged women were included in the study (Fig. 1). Comprehensive sampling procedures were described in the EDHS report [4, 25–27]. The extracted sample sizes from four consecutive EDHSs All reproductive-aged women who had been given births or attended antenatal care during the last pregnancy within five years before the survey in all selected clusters in Ethiopia were the study population. Nonetheless, women who did not know the exact time of their first ANC visit were excluded from this study. Women asked to report the exact time of their first antenatal care visit (in months) in each survey. Our outcome variable was a delayed first ANC visit, which was determined based on the timing of the first ANC visit [4, 8]. The binary response variable for women is denoted by a random variable Yi, with two possible codes. The two possible values coded as Yi = 1 if ith women started their first ANC visit after four months and Yi = 0 if they started their first ANC visit before four months or at four months. We classified the independent variables in the study based on Andersen-Newman’s behavioral model for maternal health care utilization as predisposing, enabling, and need factors. In the first category, predisposing variables are socio-demographic and socio-cultural characteristics of the respondents that exist before their health condition. Some predisposing factors are maternal age, marital status, residence, religion, women’s occupational status, husband’s occupational status, women’s educational level, husband’s educational level, mass media exposure, parity, family sizes and living children in a household. In the second category, enabling (economic) factors reflect the means or facilitators to access health care services like household head and wealth index. In the last section, needing factors are the immediate causes to use health services and reflect the perceived health status of the respondents. Ever use contraceptive in a previous pregnancy, cesarean delivery of last births, wanted last pregnancy, told about pregnancy complications last births or a previous pregnancy, and ever had a terminated pregnancy (miscarriage, abortion, or stillbirths) are considered as needing factors. We extracted data on reproductive-aged women from the Kids Recode (KR file) data set. Before doing any statistical analysis, the data on women were weighted using sampling weights or pweights for probability sampling and non-response rate to restore the representativeness of the survey and get reliable statistical estimates. Data analysis in this study included descriptive and multivariate decomposition analysis of the change in delayed first ANC visit. After extracting relevant variables for the study, we appended data on reproductive-aged women obtained from the four 2000, 2005, 2011, and 2016 EDHSs together for trends and multivariate decomposition analysis. Besides, multicollinearity was checked using variance inflation factor (VIF) and a VIF less than 5 for each independent variable. Therefore, there was no multicollinearity between independent variables since its VIF value ranged from 1.01 to 2.15 with a mean VIF of 1.4. The trend period was divided into four phases such as; first phase (2000–2005), second phase (2005–2011), third phase (2011–2016), and fourth phase (2000–2016) to see the differences in the prevalence of delayed first ANC visit over time-based on different selected characteristics of women. The trend was assessed using descriptive analysis stratified by various selected predictor variables and examined separately for each phase. Multivariate decomposition analysis was used to identify the contributing factors to the trend change in the outcome variable between any two surveys over time. This analysis focused on how a delayed first ANC visit prevalence responds to differences in selected women’s characteristics and how these variables shape the differences across the surveys conducted at different times. Decomposition analysis aimed to identify the potential sources of variations in the prevalence of delayed first ANC visit in the last 16 years. Multivariate decomposition analysis for the non-linear response model used the output from a logistic regression model since it is a dichotomous variable to parcel out the observed change in delayed first ANC visit between surveys into components. The difference in the percentage of delayed first ANC visit over time is attributable to the compositional change between any two surveys and the difference in the effects of those selected independent variables. That means the change in delayed first ANC visit is divided into the differences in characteristics (endowment component) and the effect of the selected variables (coefficient component). The analysis focused on the decomposition of the trend change in delayed first ANC visit between the reference year (2000) and the recent year (2016). The recent EDHS 2016 and reference EDHS 2000 surveys are denoted by A and B, respectively. For logistic regression, log-odds or logit of delayed first ANC visit divided into two main parts as follow: where E represents endowments explained by characteristics, and C denotes coefficients (unexplained) [28]. We can rewrite the above equation as follow: where β0B is the intercept in the regression equation for EDHS 2000, β0A is the intercept in the regression equation for EDHS 2016, βijB is the coefficient of the jth category of the ith determinant in EDHS 2000, βijA is the coefficient of the jth category of the ith determinant in EDHS 2016, XijB is the proportion of the jth category of the ith determinant in EDHS 2000, and XijA is the proportion of the jth category of the ith determinant in EDHS 2016. Currently developed multivariate logistic decomposition analysis used for the decomposition analysis of delayed first ANC visit using mvdcmp STATA package [29].

Materials

N/A

Innovations

Based on the information provided, here are some potential innovations that could improve access to maternal health:

1. Mobile Health (mHealth) Solutions: Develop mobile applications or SMS-based systems that provide pregnant women with information and reminders about antenatal care visits, pregnancy complications, and other relevant health information. These solutions can help increase awareness and encourage timely healthcare-seeking behavior.

2. Community Health Workers: Train and deploy community health workers to provide education and support to pregnant women in rural areas. These workers can conduct home visits, provide antenatal care services, and refer women to healthcare facilities when necessary.

3. Telemedicine: Implement telemedicine programs that allow pregnant women in remote areas to consult with healthcare providers through video calls or phone calls. This can help overcome geographical barriers and ensure timely access to healthcare services.

4. Financial Incentives: Introduce financial incentives, such as conditional cash transfers or vouchers, to encourage pregnant women to seek early antenatal care. These incentives can help offset the costs associated with healthcare visits and motivate women to prioritize their maternal health.

5. Health Education Campaigns: Launch targeted health education campaigns to raise awareness about the importance of early antenatal care visits and address misconceptions or cultural barriers that may prevent women from seeking care. These campaigns can utilize various media channels, including radio, television, and community gatherings.

6. Improving Infrastructure: Invest in improving healthcare infrastructure, particularly in rural areas, by establishing or upgrading healthcare facilities that provide comprehensive antenatal care services. This includes ensuring the availability of skilled healthcare providers, necessary medical equipment, and essential medications.

7. Collaborations and Partnerships: Foster collaborations and partnerships between government agencies, non-governmental organizations, and private sector entities to pool resources and expertise in addressing maternal health challenges. This can lead to more coordinated and effective interventions.

It is important to note that the specific context and needs of the target population should be considered when implementing these innovations. Additionally, rigorous monitoring and evaluation should be conducted to assess their impact and make necessary adjustments for continuous improvement.

AI Innovations Description

The recommendation that can be developed into an innovation to improve access to maternal health based on the study mentioned is to implement targeted public health interventions. These interventions should focus on the following areas:

1. Targeting rural residents: The study found that residence was a significant contributing factor to delayed first antenatal care visits. Therefore, it is important to develop innovative strategies to reach and provide adequate maternal health services to women living in rural areas. This could include mobile clinics, telemedicine, or community health workers who can provide antenatal care services in remote areas.

2. Addressing poor household economic status: The study identified household economic status as an enabling factor for accessing maternal health services. To improve access, innovative approaches such as providing financial incentives or subsidies for antenatal care visits could be implemented. This could help alleviate the financial burden on families and encourage timely utilization of maternal health services.

3. Improving awareness about pregnancy-related complications: The study found that women who were informed about pregnancy complications were more likely to seek timely antenatal care. Therefore, innovative strategies should be developed to increase awareness among women about the importance of early antenatal care visits and the potential risks associated with delayed visits. This could include community education programs, targeted messaging through mobile phones, or interactive digital platforms.

Overall, the key recommendation is to develop and implement innovative approaches that specifically target rural residents, address economic barriers, and improve awareness about the importance of early antenatal care visits. By doing so, access to maternal health services can be improved, leading to better health outcomes for women and newborns.

AI Innovations Methodology

Based on the information provided, here are some potential recommendations to improve access to maternal health:

1. Strengthening Health Education: Implement comprehensive health education programs to increase awareness among women about the importance of early antenatal care visits. This can include educating women about the benefits of early care, common pregnancy complications, and the availability of maternal health services.

2. Community Outreach Programs: Establish community-based outreach programs to reach women in remote areas and provide them with information about antenatal care services. This can involve mobile clinics, community health workers, and partnerships with local organizations to ensure that women have access to the necessary care.

3. Improving Transportation: Address transportation barriers by improving infrastructure and transportation systems in rural areas. This can include building roads, providing transportation vouchers or subsidies, and establishing transportation networks specifically for pregnant women to facilitate their access to healthcare facilities.

4. Financial Support: Provide financial support to women who face economic barriers in accessing antenatal care. This can include implementing health insurance schemes, providing subsidies for antenatal care services, and offering financial incentives for early care visits.

To simulate the impact of these recommendations on improving access to maternal health, a methodology could be developed as follows:

1. Define the indicators: Identify specific indicators that measure access to maternal health, such as the percentage of women receiving early antenatal care visits, the distance traveled to access care, or the percentage of women who face financial barriers.

2. Collect baseline data: Gather data on the current status of the indicators in the target population. This can be done through surveys, interviews, or existing data sources.

3. Develop a simulation model: Create a simulation model that incorporates the potential impact of the recommendations on the identified indicators. This model should consider factors such as population size, geographical distribution, and socio-economic characteristics.

4. Input data and assumptions: Input the baseline data into the simulation model and make assumptions about the potential effects of the recommendations. For example, assume that the health education program will increase awareness by a certain percentage, or that the transportation improvements will reduce travel time by a certain amount.

5. Run the simulation: Use the simulation model to project the potential impact of the recommendations over a specified time period. This can be done by running multiple iterations of the model with different input values and assumptions.

6. Analyze the results: Analyze the simulation results to determine the potential changes in the identified indicators. This can include calculating the percentage increase in early antenatal care visits, the reduction in travel distance, or the decrease in financial barriers.

7. Validate the results: Validate the simulation results by comparing them with real-world data or expert opinions. This can help ensure the accuracy and reliability of the simulation model.

8. Refine and iterate: Based on the validation results, refine the simulation model and repeat the process to further improve the accuracy of the projections.

By following this methodology, policymakers and healthcare providers can gain insights into the potential impact of the recommendations on improving access to maternal health and make informed decisions about resource allocation and program implementation.

Author

Dima Health

Identifiers:

DOI: 10.1186/s12978-022-01373-2

Research Areas:

Community Interventions, Health System and Policy, Maternal Access, Maternal and Child Health, Quality of Care, Sexual and Reproductive Health, Social Determinants

Study Countries:

Ethiopia

Study Design:

Cross Sectional Study

Study Approach:

Quantitative

Participants Gender:

Female