Analyzing Predictors and Geospatial Trends of the Number of Visits and Timing of Antenatal Care in Zambia: A Generalized Linear Mixed Model (GLMM) Investigation from 1992 to 2018

Antenatal care (ANC) plays a crucial role in preventing and detecting pregnancy risks, facilitating prompt treatment, and disseminating essential information to expectant mothers. This role is particularly vital in developing countries, where a 4.65% rise in maternal mortality rate was observed in 2022, with 800 maternal and 7,700 perinatal deaths reported. The study aims to analyse the predictors and geospatial trends of the number of visits and timing of antenatal care in Zambia from 1992 to 2018, utilizing a generalised linear mixed model approach. The analysis utilized the Zambia Demographic Health Survey (ZDHS) database. The relationship between dependent and independent variables was examined using the Rao-Scott Chi-square test. Predictors of inadequate utilisation of ANC were identified through the multilevel generalised linear model. Spatial effects were modeled using Quantum Geographic Information System (QGIS) version 3.34.1 to develop univariate choropleth maps. A total of 45, 140 (31, 482 women had less than 8 ANC visits and 2, 004 had 8 or more ANC visits) women with a prior childbirth were included in the study. The findings in the study revealed that among women aged 15 to 19 years in 1992, 1996 and 2013/14, the rates of inadequate antenatal care utilization (less than eight ANC visits) was 90.87%, 90.99%, and 99.63%, respectively. Lack of formal education was associated with inadequate ANC from 1992 to 2018, with percentages ranging from 91.12% to 99.64%. They were notable geospatial variations in the distribution of ANC underutilization across provinces with Luapula, Muchinga, Northwestern, Northern and Eastern recording higher proportions. Furthermore, the study showed that age (25 to 29 and 30 to 34) was associated with increased risk of inadequacy ANC utilization (AOR, 2.94; 95% CI, 1.20 – 7.19 and AOR, 2.65; 95% CI, 1.05 – 6.65 respectively). Higher education and wealth index correlated with reduced odds of inadequate ANC utilization. In conclusion, this study highlights worrisome trends in ANC utilization in Zambia, with a significant rise in inadequacy, especially in adhering to the WHO's recommended eight ANC visits. Over the period from 1996 to 2018, there was a discernible decline in the prevalence of delayed initiation of ANC. The findings underscore a notable disparity between current ANC practices and established guidelines. Additionally, various factors predicting suboptimal ANC attendance have been identified. These insights call for targeted interventions to address the identified challenges and improve the overall quality and accessibility of ANC services in Zambia.


Background
About 140 million births take place every year and the proportion attended to by skilled health personnel has increased from 58% in 1990 to 81% in 2019 [1].Despite a 35% decrease in maternal deaths from 2000 to 2017, dropping from 451,000 to 295,000 annually, the grim reality persists: over 800 women continue to lose their lives daily due to pregnancy or childbirth complications.
The majority of these tragic occurrences unfold in sub-Saharan Africa, yet they are largely preventable with access to superior maternal health services [WHO, 2019; UN, 2021] [2,3].In 2022, Zambia experienced more than 800 maternal deaths and over 7,700 perinatal deaths.The primary reason of death of women was asphyxia, prematurity and new-born sepsis remain the leading causes of perinatal and new-born deaths, major causes of maternal deaths include obstetric hemorrhage, hypertensive disorders, pregnancy related infections, abortive outcomes, and other indirect causes [4].
The significance of antenatal visits during pregnancy cannot be overstated.Antenatal care ANC serves a critical role in preventing and identifying pregnancy risks, ensuring timely treatment, and delivering pertinent information to clients.Particularly crucial in the context of heightened maternal mortality rates, ANC aims to empower women, fostering a positive childbirth experience [5,6].It is essential for protecting the health of women and their unborn children, through preventive health care [4].
The World Health Organization (WHO) in 2016 advocated for a minimum of eight antenatal visits; one contact in the first trimester, two contacts in the second trimester and five contacts in the third trimester [7].Beyond the recommended frequency, ANC offers a spectrum of benefits, including health checks, risk detection, counselling, and an increased likelihood of skilled birth attendance, thereby playing a pivotal role in reducing both maternal and infant fatalities.As a comprehensive platform for essential services, ANC goes beyond routine check-ups to prevent complications, provide crucial birth counselling, and improve overall child health outcomes [8,9].Furthermore coverage of timely ANC initiation is low, globally, the ANC initiation rate is 58.6% but varies according to the continent.The estimated rate of early antenatal care visits is 48.1% in low income countries compared with 84.8% in high-income countries [10].In SSA, the initiation of ANC visits within the first trimester is 38% which ranges from 14.5% in Mozambique to 68.6% in Liberia [11].In Ethiopia, only 28% of women had their first ANC visit during the first trimester and varies across geographic regions [12].
In addition, the scarcity of skilled service providers and substandard quality of care pose significant barriers to maternal healthcare utilization in developing countries.Research indicates that older women are more likely to access healthcare services, possibly due to their experience with health services, confidence in household decision-making, or awareness raised by healthcare professionals about the risks associated with older age [13].Moreover, low levels of education and a lack of empowerment contribute to the reluctance of women to seek maternal care [14].The aim of this study was to investigate the geospatial distribution and factors associated with inadequacy of ANC utilization among women with prior childbirth in Zambia between 1992 and 2018.

Methods
This study involved secondary analysis of microdata, utilizing national-level data extracted from the 1992 to 2018 Zambia Demographic and Health Survey (ZDHS) program [20].The ZDHS is a comprehensive and nationally representative household survey conducted by the Zambia Statistics Agency in collaboration with global partners, including ICF International and the United States Agency for International Development (USAID).Employing a two-stage sampling process, the survey initially selects enumeration areas (EAs) and subsequently households.The nature of the DHS data facilitates the comparison of variables over time, enabling the monitoring of changes in indicators across diverse geographical regions [20].
Participation in the survey was limited to women aged 15-49 years from selected households who had consented to take part in the research.Detailed methods employed in the ZDHS are comprehensively documented [20].For this specific study, we extracted all pertinent variables from the women's data files (individual recode) 2018 ZDHS dataset.The data under examination pertains to the population of women in reproductive age group with at least one childbirth event.
The analysis involved the weighted sample from 2018 DHS.Data was accessed 13 th of January, 2024.The authors in this study did not have access to information that could identify individual participants during or after data collection.

Dependent and independent variables
The primary focus of this study centered on ANC attendance, categorized into two groups: the primary outcome variable, less than eight ANC visits (coded as "1"), and eight or more ANC visits (coded as "0").Additionally, a secondary variable, "Delay in first ANC initiation," was examined, with "early" defined as less than four months (coded as "0") and "delayed" as four months or more (coded as "1").Antenatal care holds paramount significance in averting pregnancy-related complications, providing essential counseling for the well-being of both the mother and the fetus, and ensuring preparedness for a health-facility delivery.Guided by the World Health Organization recommendations from 2018, the study underscored the ideal occurrence of the first ANC visit within the first trimester of gestation, with a minimum of 8 ANC visits throughout the pregnancy.
Furthermore, WHO recommended one contact in the first trimester, two contacts in the second trimester and five contacts in the third trimester [7].
Following an extensive literature review, we meticulously controlled for a wide array of demographic, socio-economic, behavioral, and community-level factors in our analysis.Specifically, we included variables that gauged the quality of antenatal care, such as discussions on "HIV transmission from mother to child," "preventive measures for HIV," "HIV testing discussions," "receipt of HIV test results during antenatal visits," "counseling post-HIV testing during antenatal care," and "HIV testing as part of antenatal visits," all coded as "yes" or "no." Additionally, the mother's education level was categorized into tiers: 0 for no education, 1 for primary education, 2 for secondary education, and 3 for higher education.Mother's age was segmented into five-year cohorts, spanning from 15-19 to 45-49.The household's Wealth Index was stratified into five quantiles: 1 for the poorest, 2 for the poor, 3 for the middle, 4 for the rich, and 5 for the richest.

Data Analysis
For descriptive purposes, percentages were computed for categorical variables using sample weighting for accurate representation.To determine association between the outcome variable (primary outcome = "less than eight ANC visits" and secondary outcome = "delayed first ANC utilization") and the categorical variables, the Uncorrelated Design Based Chi-square test (Rao -Scott Chi-square test) was used.This was selected to take into account clustering.To determine the factors associated with having less than eight ANC visits women with prior childbirth the study utilized the generalized linear mixed models (GLMMs) with the logit link and binomial family that adjusted for clustering and sampling weights were used to assess the association between the independent variables and the outcome variables (less than eight) [19].This approach accounted for the hierarchical structure of the data, with women aged 15 -49 nested within households and households nested within clusters.

Diagnostic test and model fitting
For model comparisons and assessment of goodness of fit, we used the probability F-test which showed that the adopted model explained fertility better than the null model (model without the explanatory variables), P<0.0001.The log likelihood ratio test and Akaike Information Criteria (AIC) were sufficiently explored to select the best fit model.Stata version 14.2 was used for the analysis.To assess multicolinearity among independent factors, the Variance inflation Factor (VIF) was used.There are no concerns with multicolinearity in any of the variables (all VIF<5) as seen in Table S1 below.The selection process identified the model with the lowest AIC, signifying optimal fit.
Additionally, the study calculated the intraclass correlation (ICC) to assess the homogeneity of the outcome variable within clusters.The ICC represents the proportion of between-cluster variation in the total variation, ranging from 0 to 1.An ICC of 0 indicates perfect independence of residuals.

Spatial Analysis
To assess the geographical distribution of antenatal care utilization among women aged 15-49 with prior childbirth.The study employed Quantum Geographic Information System (QGIS) version 3.34.1 to generate a univariate choropleth map.The spatial analysis was conducted at the provincial level, aligning each Women of Reproductive Age (WRA) with their respective provincial residence using geo-coordinate data collected during the DHS (1992 -2018).This geocoordinate data is based on pre-defined information that assigns each case to a specific province.
The unit of spatial analysis was defined as a cluster of sample households, as designated by ZDHS.
For consistency, a coordinate system of World Geographic System (WGS) 1984 Universal Transverse Mercator (UTM) Zone 36S was applied to facilitate accurate and standardized spatial representation.

Ethics statement
The methodologies employed in 1992 -2018 Zambia Demographic and Health Survey (ZDHS), including biomarker measurement protocols, received ethical approval from both the Inner City The user meticulously adhered to the given instructions, placing significant emphasis on the confidential nature of the data and the imperative of safeguarding the anonymity of households or individual respondents interviewed in the survey.

Results
Included in this study were data for 45,140 women, taken from a total sampled population of 59,979.Those without a history of prior child birth (n=14,839) were excluded.This analysis involved weighted sample from the 1992 to 2018 ZDHS (see Fig 1).

Legend: Figure 1: Description of sample derivation criteria
For this specific study, we extracted all pertinent variables from the women's data files (individual recode) 2018 ZDHS dataset.The data under examination pertains to women with a prior childbirth.

Characteristics of women who have previously given birth, linked with less than 8 ANC visit
The findings in the study revealed that majority of the women reported having less than eight ANC visits had no level of education throughout the survey years 1992 to 2018 (91.12%, 92.08%, 93.47%, 99.68% and 99.64% respectively).The relationship was significant (p<0.0001).Notably, a significant association (p<0.0001) was found between having less than 8 ANC visits and the absence of formal education, with 91.12%, 92.08%, 93.47%, 99.68%, and 99.64% of women reporting less than eight ANC visits in each respective year.
Additionally, the study revealed that the majority of participants reporting less than eight ANC visits belonged to the poorest socioeconomic stratum between 2007 and 2013/14 (98.68% and 99.62%, respectively), except for 2018, when a higher proportion (99.08%) was observed among the richer segment.Consistently, women not engaged in employment showed a higher proportion of less than 8 ANC visits from 1992 to 2018, with notable variations in 1996 and 2001/02.

Characteristics of women who have previously given birth, linked with delayed initiation of their first antenatal care (ANC) visit (four months or more).
The study investigated the characteristics of participants who delayed in the first antenatal care or missed their first trimester ANC visit.
The study examined factors influencing delayed or missed first ANC visits, with a focus on participant characteristics.The results revealed that a significant percentage of adolescents aged 15 to 19 (80.66%) delayed their first ANC visits compared to older women.
This difference was statistically significant (p=0.0002).Additionally, women with lower education levels (up to primary) exhibited a higher proportion of delayed ANC visits compared to those with higher education (secondary and above).Rural areas had a lower percentage of delayed ANC initiation (76.40%)compared to urban areas (79.86%).Regarding geographic distribution, the central province had the highest proportion of delayed first ANC initiation, followed by Northern (78.72%),Muchinga (78.38%),Copperbelt (78.07%), and Eastern province with the lowest percentage (70.89%).

Discussion
Antenatal care (ANC) remains a critical area of focus due to its profound impact on maternal and child health, including documented effects on intermediate variables such as birth weight [13].
This study investigated predictors of less than eight ANC visits, and among women aged 15 to 49 with prior childbirth in Zambia.Using a multilevel generalized mixed model with a binomial family and logit link function on data spanning from 1992 to 2018 from the Zambia Demographic and Health Surveys (ZDHS), the research aims to comprehensively understand predictors and the geographical distribution of less than eight ANC visits.
The study's findings revealed an upward trend, an increment of 14% from 1992 to 2018 of women reporting less than eight ANC visits.This indicates a significant gap, consistent between urban and rural settings, reflecting unsatisfactory compliance with WHO-recommended ANC visit levels in Zambia.These findings align with studies in Bangladesh [21], emphasizing the need for improved primary healthcare, especially in rural settings.Furthermore, the study explored the quality of ANC, demonstrating an improvement from 2007 to 2018.It stresses the importance of enhancing healthcare quality over time, in contrast to some countries, like Latin America, which struggle with low ANC coverage [22].Therefore, efforts to bridge the urban-rural gap healthcare policies should focus on improving accessibility in rural areas, recognizing the vital role of primary healthcare in promoting ANC in these settings.
The World Health Organization (WHO) initially advocated the Focused Antenatal Care (FANC) model in the 1990s, recommending a minimum of four ANC visits per pregnant woman [23].
However, in 2016, WHO revised its minimum recommended ANC visits from four to eight ANC visits.This change was reflected in the 2018 Demographic Health Survey.Given this transition, it's apparent that the 2016 WHO ANC model wasn't implemented before then, as the FANC model was in use.Consequently, the study is aware of the difference in required ANC visits between 2018 and preceding years [24,25].It is therefore, reasonable to assume that low attendance before 2018 adhered to the FANC model's minimum requirement of four visits, not eight.However, the study unveils a concerning trend: the proportion of ANC attendance fewer than eight visits has rather increased over time, indicating a decline in ANC attendance among pregnant women over the years.Proportion of delayed initiation of ANC visit in the study revealed an overall decline overtime this is similar to studies carried out in in Ethiopia [26][27][28] which showed that the trend for delayed ANC visits has been on the decline.
Furthermore, the study delved into the geospatial distribution of less than 8 ANC visits, revealing distinct patterns across provinces.The findings highlight that Western, Southern, Copperbelt, and Lusaka provinces consistently exhibited notably lower proportions of ANC visits less than 8 when compared to their counterparts.This gap could be attributed to varying levels of economic activities, employment rates, educational attainment, and the prevalence of urban landscapes over rural settings in these provinces [29].Conversely, provinces such as Luapula, Muchinga, Northwestern, Northern, and Eastern reported considerably higher proportions of less than 8 ANC visits, pointing towards region-specific challenges that necessitate targeted interventions for improved maternal healthcare access and utilization.
Analyzing factors influencing less than 8 ANC visits, the study identified education, wealth index, occupation, reading habits, television watching, and residing in the Copperbelt province as predictors of attending less than the recommended eight ANC sessions.These results align with those of previous studies, such as the work by Islam and Masud [21], indicating that these predictor variables were protective, correlating with a reduced risk of attending fewer than 8 ANC visits after accounting for other factors [30][31][32].Notably, the exception was the frequency of watching television, which was associated with an increased risk.These results align with prior research, suggesting protective effects, with the exception of television watching, associated with increased risk.Additionally, stratified analysis across survey years indicates age and education as predictors, emphasizing the dynamic nature of ANC utilization over time.Consequently, Public health campaigns, leveraging media channels, should be intensified to educate and empower women, fostering health literacy-related discussions and positively influencing their ANC behaviors.
Additionally, healthcare providers should emphasize comprehensive ANC education, covering not only the quantity of visits but also the content and quality of services.
In contrast, a study in Rwanda revealed contradictory outcomes, indicating that employed women had reduced odds of attending all ANC sessions [33].Conversely, our study aligns with research suggesting that regular reading of newspapers/magazines is associated with an increased likelihood of robust attendance at ANC visits.This association is attributed to higher literacy levels and education among women with access to newspapers, fostering health literacy-related discussions that positively influence their health-seeking behaviors [33][34][35][36].Furthermore, a study in Papua New Guinea reported similar findings of employment and education level being protective of the risk of less than 8 ANC visits [36].Additionally, a stratified analysis was conducted for each survey year from 1992 to 2018.The findings indicate that age predicted less than 8 ANC visits in 1992, 1996, and 2018.Education similarly served as a predictor, except in the year 2007.
Intriguingly, the results reveal that education acted as a protective factor.
Furthermore, the study revealed majority of adolescents aged 15 to 19 (80.66%) delayed their first ANC visits compared to older women.Additionally the study revealed that education, residence (rural and urban) and region were associated with delayed initiation of first ANC visit.These findings are similar to studies done in various countries that revealed that majority of women initiate first antenatal visit very late after the first trimester, which is against the WHO recommendation [37][38][39].A study done in Wollaita Soddo town revealed that education was associated with timing of first ANC visit.Women with secondary and higher education were more times likely to initiate than those with primary and who didn't have formal education [40], similar with a study conducted in Addis Ababa [41].
The stratified analysis revealed that only education consistently predicted the underutilization of ANC visits (less than 8 visits) from 1992 to 2018, except for 2007, where it showed no significant effect.Remarkably, education acted as a protective factor throughout these years.Previous studies have similarly highlighted the significant role of education in determining optimal ANC utilization [42][43][44][45].This association may stem from educated women's likely possession of adequate knowledge regarding ANC services and their understanding of the importance of early booking and attending the recommended ANC visits.Moreover, education may empower some women to overcome gender-specific discrimination and barriers, such as domestic challenges [46,47].

Limitations and strengths of the study
This study draws strength from the utilization of national data, providing a representative sample of the adolescent female population aged 15 to 49 in Zambia.Consequently, the study's findings are applicable and can be generalized to the specified target population of women with prior childbirth within this age range.However, it is essential to acknowledge the study's limitations.

Conclusion
In summary, this study sheds light on the critical challenges surrounding antenatal care (ANC) utilization in Zambia, exposing a concerning upward trend in less than 8 ANC visits, especially in meeting the prescribed eight ANC visits.The results underscore a substantial disparity between observed ANC practices and the established guidelines set by the World Health Organization Moreover, it notes a decreasing trend in the delay of initiating the first ANC visit from 1996 to 2018, even though it is still worrying.Persistent urban-rural discrepancies highlight the urgent need for tailored interventions aimed at rectifying significant deficiencies in ANC access and utilization across diverse settings.Notably, the study brings to light a noteworthy improvement in specific healthcare aspects, exemplified by a significant rise in HIV testing rates within ANC services.However, the observed decline in discussions about HIV transmission and preventive measures emphasizes the necessity of adopting a comprehensive approach to ANC, addressing both the quantity and quality of services.Additionally, the study identifies education, wealth index, occupation, reading habits, television watching, and residing in the Copperbelt as predictive factors for attending fewer than the recommended eight ANC sessions.

Fund
(ICF) institutional review boards (IRBs) and the Tropical Diseases Research Centre (TDRC) in Zambia.The consent process includes obtaining informed oral consent from each respondent, and for adolescents under 18 years, consent was obtained from a parent or guardian.Comprehensive information about the DHS consent process is available at https://www.dhsprogram.com/What-We-Do/Protecting-the-Privacy-of-DHS-Survey-Respondents.cfm.Authorization to use the ZDHS data was obtained from ICF Macro (see supporting information), and the dataset can be accessed at https://www.dhsprogram.com/data.
The reliance on the latest Zambia Demographic and Health Survey (ZDHS) dataset from 2018 follows a cross-sectional study design, implying that the results indicate correlation rather than causation between the outcome of interest and individual or contextual factors.Additionally, caution is advised when extending the findings to the broader adolescent age group of 10 to 19 years.Moreover, the contextual factors utilized in the study are derived from the ZDHS, potentially limiting their ability to fully capture the community experience.These considerations are crucial for a nuanced interpretation of the study's outcomes.

Table 2 : Key attributes of women who have previously given birth, linked with inadequate ANC visits (less than 8 ANC visit)
NS = Newspaper; MZ = Magazine; Rad = Radio; * = P-value <0.05 computed using the Rao-Scott Chi-square test.

Multivariable Generalized Linear Mixed Model from1992 to 2018 of less than 8 ANC visit antenatal care utilization (Stratified Analysis)
A comprehensive stratified analysis was conducted for each specific year from 1992 to 2018 using data from the Zambia Demographic and Health Surveys (ZDHS).The findings, detailed in

Table 4 ,
(25)(26)(27)(28)(29)age during selected survey years (1992 and 1996) was associated with reduced odds of less than 8 ANC utilization, and this association was statistically significant (p < 0.05).However, a significant shift occurred in 2018, revealing that older age was now linked to an increased risk of less than 8 ANC utilization.Specifically, the adjusted odds ratios for age groups AOR(25)(26)(27)(28)(29)and AOR(30-34) were 2.94 (95% CI, 1.20 -7.19) and 2.65 (95% CI, 1.05 -6.65), respectively.In 2018, older women had increased odds of having less than 8 antenatal care visits compared to women aged 15 to 19.Furthermore, women with a higher level of education consistently demonstrated a reduced risk of less than 8 ANC visits across the years, except for 2013, where this association was not statistically significant, holding all other factors constant.Moreover, women in rural settings were more at risk of less than 8 ANC visits, and statistically significant predictors were observed only in the period from 1992 to 2001/02.The adjusted odds ratios and corresponding confidence intervals for the