Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2024 Nov 29;7(4):ooae132.
doi: 10.1093/jamiaopen/ooae132. eCollection 2024 Dec.

A retrospective cohort study on predicting infants at a risk of defaulting routine immunization in Uganda using machine learning models

Bartha Alexandra Nantongo  1   2 , Josephine Nabukenya  1 , Peter Nabende  1 , John Kamulegeya  3
Affiliations

A retrospective cohort study on predicting infants at a risk of defaulting routine immunization in Uganda using machine learning models

Bartha Alexandra Nantongo et al. JAMIA Open. .

Abstract

Objectives: Using machine learning models to predict infants at risk of defaulting routine immunization (RI) and identify significant features for Uganda.

Materials and methods: Principal component analysis reduced dimensionality. Datasets were balanced using synthetic minority over-sampling technique. k-Nearest Neighbors, Decision Trees, Random Forests (RFs), Support Vector Machine (SVM), Naïve-Bayes, Logistic Regression (LR), XGBoost, Adoptive-Boosting, and Gradient-Boosting were used on Uganda's 2016 Demographic and health survey data with social-economic and demographic factors as predictors. Experiments with and without K-fold cross-validation were performed. Models were evaluated for accuracy, recall, precision, and area under a curve (AUC).

Results and discussion: Experimental results revealed that the rate of defaulting increases as an infant's age increases at 5.3% Bacille Calmette-Guérin (BCG), 7.3% pentavalentI, 22.9% pentavalentIII, and 22.1% for measles. Significant predictors for BCG were immunization card, polio0, cluster altitude. Reception of pneumococcal1, BCG, and district for pentavalentI; polio3, pentavalentII for pentavalentIII; polio active and pentavalentIII for measles.RF had the best performance at predicting vaccine defaulting with 96%, 95%, 94%, 84% accuracy for BCG, PentavalentI, pentavalentIII, measles, respectively. Similarly, RF had the same precision, recall, AUC at 1.0. However, XGBoost, SVM, LR displayed the worst discriminatory power among infants who received the vaccine from defaulters with AUC ≤0.57.

Conclusion: Immunization card, preceding vaccines reception, and district were the most influential predictors. RF was the best classifier among the 9 models to predict defaulting RI. The study recommends regular outreaches, daily vaccination, provision of immunization cards, and accessible water sources to reduce defaulting.

Keywords: defaulting; infants; machine learning models; prediction; routine immunization.

PubMed Disclaimer

Conflict of interest statement

The authors have no competing interests to declare.

Figures

Figure 1.
Figure 1.
Vaccination coverage by region. Regions in Uganda as per demographic health survey 2016.
Figure 2.
Figure 2.
Study methodology.
Figure 3.
Figure 3.
Dataset shape and distribution.
Figure 4.
Figure 4.
ROC AUC for experiment 2.
See this image and copyright information in PMC

References

    1. Chandir S, Siddiqi DA, Hussain OA, et al.Using predictive analytics to identify children at high risk of defaulting from a routine immunization program: feasibility study. JMIR Public Health Surveill. 2018;4:e63. - PMC - PubMed
    1. Galadima AN, Zulkefli NAM, Said SM, Ahmad N.. Factors influencing childhood immunisation uptake in Africa: a systematic review. BMC Public Health 2021;21:1475. 10.1186/s12889-021-11466-5 - DOI - PMC - PubMed
    1. UNICEF. Uganda CO Humanitarian Annual Situation Report. UNICEF; 2018. Accessed June 20, 2024. https://www.unicef.org/media/76021/file/Uganda-SitRep-Dec-2018.pdf
    1. Khan ZA, Feng Z, Uddin M, et al.Optimal policy learning for disease prevention using reinforcement learning. Hindawi Sci Program. 2020;2020:1.
    1. Mosser JF, Gagne-Maynard W, Rao PC, et al. Mapping diphtheria-pertussis-tetanus vaccine coverage in Africa, 2000–2016: a spatial and temporal modelling study. Lancet. 2019;393:1843-1855. Accessed June 2024. https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(19)30226... - PMC - PubMed

LinkOut - more resources