Menu
Menu
Menu
Menu
Background About 31 million children in sub-Saharan Africa (SSA) suffer from immunisation preventable diseases yearly and more than half a million children die because of lack of access to immunisation. Immunisation coverage has stagnated at 72% in SSA over the past 6 years. Due to evidence that full immunisation of children may be determined by place of residence, this study aimed at investigating the rural-urban differential in full childhood immunisation in SSA. Methods The data used for this study consisted of 26 241 children pooled from 23 Demographic and Health Surveys conducted between 2010 and 2018 in SSA. We performed a Poisson regression analysis with robust Standard Errors (SEs) to determine the factors associated with full immunisation status for rural and urban children. Likewise, a multivariate decomposition analysis for non-linear response model was used to examine the contribution of the covariates to the observed rural and urban differential in full childhood immunisation. All analyses were performed using Stata software V.15.0 and associations with a p<0.05 were considered statistically significant. Results More than half of children in urban settings were fully immunised (52.8%) while 59.3% of rural residents were not fully immunised. In all, 76.5% of rural-urban variation in full immunisation was attributable to differences in child and maternal characteristics. Household wealth was an important component contributing to the rural-urban gap. Specifically, richest wealth status substantially accounted for immunisation disparity (35.7%). First and sixth birth orders contributed 7.3% and 14.9%, respectively, towards the disparity while 7.9% of the disparity was attributable to distance to health facility. Conclusion This study has emphasised the rural-urban disparity in childhood immunisation, with children in the urban settings more likely to complete immunisation. Subregional, national and community-level interventions to obviate this disparity should target children in rural settings, those from poor households and women who have difficulties in accessing healthcare facilities due to distance.
We sourced data from the most recent Demographic and Health Surveys (DHS) of 23 countries conducted between 2010 and 2018. These 23 countries were included based on two principal reasons: countries having DHS data within the study period (ie, from 2010 to 2018) and comparability of data due to availability of required variables within the datasets. These surveys were executed by the DHS Programme. DHS is executed in low-income and middle-income countries in partnership with local organisations in the respective countries. It is usually conducted at 5-year interval and follows a common execution procedure. The survey focuses on crucial maternal and child health factors such as immunisation, maternal healthcare utilisation, malaria and other essential indicators.16 A two-stage stratified sampling procedure is adopted in selecting research participants. In the first phase, clusters/enumeration areas (EAs) are selected guided by a sample frame developed during the preceding census of the respective countries, while in the second stage, a sample of households is drawn from each selected EAs. The full sampling procedure has been documented elsewhere.16 A total of 26 241 children who had complete information on all the variables considered were eligible for this study. We relied on the Strengthening the Reporting of Observational Studies in Epidemiology statement in conducting this study and writing the manuscript (online supplemental file 1). bmjgh-2020-003773supp001.pdf Full immunisation coverage of children 12–23 months was the outcome of interest. Children within 12–23 months are expected to receive the rudimentary immunisation dosages. Following the WHO recommendations, a child was considered to be fully immunised if the child has received Bacille Calmette-Guerin against tuberculosis; at least three doses of polio vaccine; three doses of diptheria, tetanus toxoids and pertussis vaccine and one dose measles vaccine17 as illustrated in table 1. Subsequently, a child between 12 and 23 months who had received all nine doses was classified as fully immunised (coded 1), while any child without all the dosages was categorised as non-fully immunised (coded 0). Recommended immunisation for SSA children Source: modified from WHO.17 BCG, Bacille Calmette-Guerin; DPT, Diphtheria-Pertussis-Tetanus; SSA, sub-Saharan Africa. The key independent variable was residential status, that is, rural or urban area. Based on some immunisation literature,9 11 13 we selected and controlled for nine (9) covariates grouped into child factors: sex of child, and birth order as well as maternal factors: maternal age, education, wealth quintile, occupation, sex of household head, health insurance subscription status and distance to health facilities. Sex of child was either male or female while birth order ranged from one (1) to six (6) or more. Maternal age was measured in completed years categorised into 5-year interval (15–19, 20–24, 25–29, 30–34, 35–39, 40–44, 45–49); education was captured as no education, primary, secondary and higher; wealth quintile was grouped as poorest, poorer, poor, richer and richest; occupation was measured as not working or working; distance to health facility was either not a big problem/no problem or a big problem; and finally health insurance subscription status was either yes (ie, subscribed) or no (ie, not subscribed). The descriptive summaries of countries by year of survey, sample size and the prevalence of full immunisation by place of residence were presented in table 2. Similarly, a pooled analysis of the proportion of children who were fully immunised by residential status with respect to each of the covariates were presented in table 3. A χ2 test of association was conducted to investigate if there exist a difference between the maternal and child characteristics by place of residence (table 4). The proportion of children aged 12–23 months with full immunisation status across the 23 countries was presented in figure 1. For the inferential analysis, we fitted a Poisson regression with robust standard errors to explore the predictors and direction of full immunisation status with respect to the covariates (table 5). In the final analysis, a multivariate non-linear decomposition model,18 which is similar to the Fairlie and Blinder-Oaxaca was employed to decompose the disparity in immunisation status due to residential status (table 6). This technique was used to evaluate the variation in full immunisation status between rural and urban children and to identify the contribution of each covariates to the explained immunisation variation between rural and urban children. The multivariate decomposition analysis was weighted, and other analyses were conducted using Stata software V.15.0 and adjusted for the complex survey design. The variance inflation factor (VIF) was used to check for the presence of multicollinearity which showed no evidence of multicollinearity (mean VIF=2.98, maximum=3.40, minimum=1.04). Full immunisation coverage per country. Details and summaries of countries’ immunisation status by residence n/a, not applicable. Weighted full immunisation coverage by explanatory variables (n=26 241) * p < 0.05. Weighted frequencies of explanatory variables by residence (n=26 241) * p < 0.05. Multivariable poisson regression of full immunisation on the exposure variables *P<0.05. †P<0.01. ‡P<0.001. Ref, reference category; RR, relative risk. Multivariate decomposition of child and maternal factors associated with full immunisation inequality between rural and urban residence * p < 0.05, ** p < 0.01, *** p < 0.001. The Inner-City Fund International further ensures that the procedures of DHS are consistent with the regulations for the respect of human subjects as recommended by the US Department of Health and Human Services. Comprehensive information about the ethical protocols is accessible through http://goo.gl/ny8T6X. Authors sought and obtained permission to use the data from the Measure DHS Programme after our intent for the data was accessed. Patients and the public were not involved in the design and conduct of this research.
N/A
