Spatial-temporal patterns and predictors of timing and inadequate antenatal care utilization in Zambia: A Generalized Linear Mixed Model (GLMM) investigation from 1992 to 2018

Abstract

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 over 800 maternal and 7,700 perinatal deaths reported. The study aimed at investigating the spatial temporal patterns and associated factors of timing and inadequacy of antenatal care utilization in Zambia, from 1992 to 2018 using a Generalized Linear Mixed Effect Model (GLMM) approach. The study 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 higher education (AOR, 0.30; 95% Confidence Interval, CI, 0.14-0.63; p = 0.001), and wealth index (AOR, 0.34; 95% CI, 0.17-0.70; p = 0.003) 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.

Fig 1. Description of sample derivation criteria.

For this specific study, we extracted all pertinent variables from the women’s data files (individual recode) 1992 to 2018 ZDHS dataset. The data under examination pertains to women with a prior childbirth.

Fig 2. Trends of antenatal care attendance (less than 8 ANC sessions) and delayed ANC initiation from 1992 to 2018.

The figure was generated in excel. The figure shows the proportion of ANC visits of less than 8 from 1992 to 2018.

Fig 3. Content of antenatal care offered between 2007 and 2018.

The variables captured to measure content of antenatal care were (1) talked about HIV transmission of mother to child (2) talked about things to do to prevent getting HIV (3) talked about getting tested for HIV (4) got results of HIV test as part of antenatal visit (5) received counseling after tested for HIV during antenatal care (6) tested for HIV virus as part of antenatal visit.

Fig 4. Shows the spatial-temporal patterns of mothers with ANC visits less than 8 by provinces in Zambia in 1992.

The maps were generated using QGIS version 3.4.1. (Source: author generated).

Fig 5. Shows the spatial-temporal patterns of mothers with ANC visits less than 8 by provinces in Zambia in 1996.

The maps were generated using QGIS version 3.4.1. (Source: author generated).

Fig 6. Shows the spatial-temporal patterns of mothers with ANC visits less than 8 by provinces in Zambia in 2001/02.

The maps were generated using QGIS version 3.4.1. (Source: author generated).

Fig 7. Shows the spatial-temporal patterns of mothers with ANC visits less than 8 by provinces in Zambia in 2007.

The maps were generated using QGIS version 3.4.1. (Source: author generated).

Fig 8. Shows the spatial-temporal patterns of mothers with ANC visits less than 8 by provinces in Zambia in 2013/14.

The maps were generated using QGIS version 3.4.1. (Source: author generated).

Fig 9. Shows the spatial-temporal patterns of mothers with ANC visits less than 8 by provinces in Zambia in 2018.

The maps were generated using QGIS version 3.4.1. (Source: author generated).