Truace
Subscribe

AI model flags children at risk of getting no vaccines with 80% accuracy, in 27 African countries

Study title: Predicting zero-dose vaccination status in 27 sub-Saharan African countries: a machine learning approach

To develop a machine learning (ML)-based predictive model for assessing the risk of zero-dose children using Demographic Health Survey data in sub-Saharan Africa. This study analysed pooled Demographic and Health Survey data from 27 countries in sub-Saharan Africa collected between 2016 and 2024.

Read the study abstract

To develop a machine learning (ML)-based predictive model for assessing the risk of zero-dose children using Demographic Health Survey data in sub-Saharan Africa. This study analysed pooled Demographic and Health Survey data from 27 countries in sub-Saharan Africa collected between 2016 and 2024. Data preprocessing included imputation, balancing of unequal classes and systematic feature selection. Seven ML models were trained and evaluated using performance metrics such as accuracy, recall and F1-score. Feature…

By the Truvace desk1 min read1 sourceEvidence: Moderate · 60/100 — Observational study · 1 independent sourcePeer-reviewedTRV-2026-1153· cite & verify
Predicting zero-dose vaccination status in 27 sub-Saharan African countries: a machine learning approach
Sensitivity of endemic behaviour of COVID-19 under a multi-dose vaccination regime, to various biological parameters and control variables by Dagpunar, John; Wu, Chenchen. CC BY 4.0 · https://creativecommons.org/licenses/by/4.0

In brief

Among the seven models evaluated, LightGBM demonstrated the best overall performance, achieving the highest accuracy (80%), sensitivity (91%), and area under the receiver operating characteristic curve (AUROC = 0.78). Our findings suggest that machine learning, particularly LightGBM and XGBoost, is moderately effective in predicting the risk of zero-dose vaccination among children.

Main points

  1. To develop a machine learning (ML)-based predictive model for assessing the risk of zero-dose children using Demographic Health Survey data in sub-Saharan Africa.
  2. This study analysed pooled Demographic and Health Survey data from 27 countries in sub-Saharan Africa collected between 2016 and 2024.
  3. Data preprocessing included imputation, balancing of unequal classes and systematic feature selection.

The gain

Tested on survey data from 27 sub-Saharan African countries, the best AI model (LightGBM) predicted which children were at risk of getting no vaccines with 80% accuracy and 91% sensitivity.

The rundown

To develop a machine learning (ML)-based predictive model for assessing the risk of zero-dose children using Demographic Health Survey data in sub-Saharan Africa. This study analysed pooled Demographic and Health Survey data from 27 countries in sub-Saharan Africa collected between 2016 and 2024.

Data preprocessing included imputation, balancing of unequal classes and systematic feature selection. Seven ML models were trained and evaluated using performance metrics such as accuracy, recall and F1-score.

Sources

  1. Peer-reviewedBMJ Health & Care Informatics2026-09-18

The debate