Artificial intelligence-assisted decision support for postpartum family planning: a machine learning framework for personalized contraceptive recommendations in resource-limited settings
Main Article Content
Abstract
Background: For postpartum women, access to proper family planning methods in underserved regions is limited, which has contributed to the high rate of unwanted pregnancies and their associated sequelae. Despite efforts to curtail these challenges through the institution of the World Health Organization (WHO) medical eligibility criteria (MEC) provided guidance, their application in postpartum care is impeded by factors such as breastfeeding status, contraceptive availability, prior side effects, and privacy concerns.
Methods: This study instituted a comparative analysis featuring machine learning frameworks like logistic regression, Random Forest and XGBoost trained using a synthetic dataset of 8,000 anonymized postpartum records derived from a publicly available contraceptive method choice dataset and augmented with postpartum-related variables to achieve a prediction goal of seven WHO- approved contraceptive methods with the best performing model integrated with a Telegram bot for accessibility.
Results: The XGBoost model achieved the best performance with a test accuracy of 88.5% and a macro-averaged F1-score of 0.734, demonstrating balanced predictive performance across the seven contraceptive classes.
Conclusion: This study, as a proof of concept shows the urgency in using AI to combat lack of access to healthcare. Despite the excellent outcome, the need for further validation with real world data is imperative after which clinical deployment can be mooted. This model, if successfully deployed in clinical settings may support postpartum women in making informed decisions about their health, thereby contributing positively to maternal health especially in underserved regions. This study, though promising, is however limited by the use of synthetic data which can affect model performance on unseen dataset.
Article Details
Issue
Section
This work is licensed under a Creative Commons Attribution 4.0 International License.
How to Cite
References
1. Cleland J, Conde-Agudelo A, Peterson H, Ross J, Tsui A. Contraception and health. Lancet. 2012;380(9837):149-156. https://doi.org/10.1016/S0140-6736(12)60609-6
2. United Nations Department of Economic and Social Affairs, Population Division. World family planning 2022: Meeting the changing needs for family planning—Contraceptive use by age and method (UN DESA/POP/2022/TR/NO. 4). United Nations; 2022.
3. World Health Organization, UNICEF, UNFPA, World Bank Group, United Nations Department of Economic and Social Affairs. Trends in maternal mortality 2000–2020. WHO; 2023.
4. Sedgh G, Bearak J, Singh S, Bankole A, Popinchalk A, Ganatra B, et al. Abortion incidence between 1990 and 2014: Global, regional, and subregional levels and trends. Lancet. 2016;388(10041):258-267. https://doi.org/10.1016/S0140-6736(16)30380-4
5. United Nations. Transforming our world: The 2030 agenda for sustainable development. United Nations; 2015.
6. Borda MR, Winfrey W, McKaig C. Return to sexual activity and modern family planning use in the extended postpartum period: An analysis of findings from seventeen countries. Stud Fam Plann. 2020;51(2):125-141. https://doi.org/10.1111/sifp.12116
7. Topol E. High-performance medicine: The convergence of human and artificial intelligence. Nat Med. 2019;25(1):44-56. https://doi.org/10.1038/s41591-018-0300-7
8. Jiang F, Jiang Y, Zhi H, Dong Y, Li H, Ma S, et al. Artificial intelligence in healthcare: Past, present and future. Stroke Vasc Neurol. 2017;2(4):230-243. https://doi.org/10.1136/svn-2017-000101
9. World Health Organization. Medical eligibility criteria for contraceptive use. 5th ed. WHO; 2015.
10. Lim T. Contraceptive Method Choice [dataset]. UCI Machine Learning Repository; 1999. https://doi.org/10.24432/C59W2D
11. World Health Organization. Family planning/contraception methods. WHO; 2022.
12. Kuhn M, Johnson K. Applied predictive modeling. Springer; 2013.
13. Chen T, Guestrin C. XGBoost: A scalable tree boosting system. In: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining; 2016. p. 785-794. https://doi.org/10.1145/2939672.2939785
14. Breiman L. Random forests. Mach Learn. 2001;45(1):5-32. https://doi.org/10.1023/A:1010933404324
15. He H, Garcia EA. Learning from imbalanced data. IEEE Trans Knowl Data Eng. 2009;21(9):1263-1284. https://doi.org/10.1109/TKDE.2008.239
16. Hastie T, Tibshirani R, Friedman J. The elements of statistical learning. 2nd ed. Springer; 2017.
17. Ribeiro MT, Singh S, Guestrin C. “Why should I trust you?”: Explaining the predictions of any classifier. In: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining; 2016. p. 1135-1144. https://doi.org/10.1145/2939672.2939778
18. Python Telegram Bot Developers. python-telegram-bot documentation. 2024. Available from: https://python-telegram-bot.org
19. Melaku MS, Yohannes L, Sharew B, Derseh MH, Taye EA. Application of machine learning algorithms to model predictors of informed contraceptive choice among reproductive age women in six high fertility rate sub-Saharan Africa countries. BMC Public Health. 2025;25(1):1986.
20. Dey AK, Dehingia N, Bhan N, Thomas EE, McDougal L, Averbach S, et al. Using machine learning to understand determinants of IUD use in India: Analyses of the National Family Health Surveys (NFHS-4). SSM Popul Health. 2022;19:101234.
21. Kebede SD, Mamo DN, Adem JB, Semagn BE, Walle AD. Machine learning modeling for identifying predictors of unmet need for family planning among married/in-union women in Ethiopia: Evidence from performance monitoring and accountability (PMA) survey 2019 dataset. PLOS Digit Health. 2023;2(10):e0000345.