Predicting Contraceptive Discontinuation in Nigeria: A Machine Learning Analysis of Family-Planning Method Adherence Using NDHS 2023-24 Calendar Data
- 1Marwadi University
Abstract
BackgroundContraceptive discontinuation -- stopping a method while still at risk of an unintended pregnancy -- is a major but under-examined driver of unmet need in Nigeria, where modern contraceptive prevalence remains far below the countrys FP2030 target. Distinguishing discontinuation that a health system could plausibly prevent from discontinuation that reflects a womans achieved fertility intentions is essential for targeting intervention, but this distinction has not been studied systematically, or compared algorithmically, in the Nigerian literature.
MethodsWe used the reproductive calendar from the 2023-24 Nigeria Demographic and Health Survey to create 14,178 contraceptive use episodes for 7,723 women and identified two outcomes: Gap A, twelve-month discontinuation (n = 11,624), and Gap B, among women who discontinued within twelve months, method-related versus intentional discontinuation (n = 3,701). We evaluated four algorithms (survey-weighted logistic regression, elastic net, Random Forest, and XGBoost) on the same held-out test set (80/20) and reported the same area under the receiver operating characteristic curve (AUC).
ResultsDiscontinuation after 12 months was 32.3 percent (survey-weighted), a value that is very similar to the Nigeria-published life-table value of 37.2 percent. The best-performing algorithm in both cases, Gap A (AUC = 0.699) and Gap B (AUC = 0.664), was Random Forest, while the worst was XGBoost (AUC = 0.621) in Gap B, which struggled even with the basic logistic baseline. The most common predictor across all four algorithms and both outcomes was contraceptive method type, with long-acting reversible contraceptive (LARC) much less likely to be discontinued than short-acting methods. Among individuals who discontinued within 12 months, reasons were almost evenly split between method-related (49.7 percent) and intentional (49.2 percent).
ConclusionsFactors (side effects, cost, access, and efficacy dissatisfaction) account for approximately half of Nigerias contraceptive discontinuation problem and could be reduced with counseling, task shifting to pharmacists, and ongoing investments to increase the availability of longer-acting methods (LARCs) in Nigeria. Algorithm performance is outcome-dependent, not fixed, and it is better to compare algorithms by outcome than to default to a particular algorithm in DHS-based machine learning research.
Related
- Patient-Reported Tolerability, Treatment Initiation, and Discontinuation of Postoperative Progestin Therapy After Cesarean Scar Repair in Women Planning Pregnancy
- Association between interspousal communication on family planning and contraceptive use: a cross-sectional study in Temeke municipality, Tanzania.
- Introduction of contraceptive implants as a family planning strategy in Honduras and Chiapas: Results from the Salud Mesoamerica Initiative (SMI)
- Impact of family planning messages on modern contraceptive use: an observational study of women of reproductive age in Zambia 2007–2018
- Predicting One-Year CPAP Adherence and Use Trajectories Utilizing Baseline and Early Use Data
BibTeX
@misc{nosaihaza2026predicting,
title = {{Predicting Contraceptive Discontinuation in Nigeria: A Machine Learning Analysis of Family-Planning Method Adherence Using NDHS 2023-24 Calendar Data}},
author = {E. Nosa-Ihaza and A. Ssembuusi and L. P. Luponya and U. N. Nosa-Ihaza},
howpublished = {medRxiv},
year = {2026},
month = sep,
doi = {10.64898/2026.09.12.26362911},
url = {https://www.medrxiv.org/content/10.64898/2026.09.12.26362911v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Predicting Contraceptive Discontinuation in Nigeria: A Machine Learning Analysis of Family-Planning Method Adherence Using NDHS 2023-24 Calendar Data
AU - E. Nosa-Ihaza
AU - A. Ssembuusi
AU - L. P. Luponya
AU - U. N. Nosa-Ihaza
PY - 2026
DA - 2026/09/12
PB - medRxiv
DO - 10.64898/2026.09.12.26362911
UR - https://www.medrxiv.org/content/10.64898/2026.09.12.26362911v1
AB - BackgroundContraceptive discontinuation -- stopping a method while still at risk of an unintended pregnancy -- is a major but under-examined driver of unmet need in Nigeria, where modern contraceptive prevalence remains far below the countrys FP2030 target. Distinguishing discontinuation that a health system could plausibly prevent from discontinuation that reflects a womans achieved fertility intentions is essential for targeting intervention, but this distinction has not been studied systematically, or compared algorithmically, in the Nigerian literature. MethodsWe used the reproductive calendar from the 2023-24 Nigeria Demographic and Health Survey to create 14,178 contraceptive use episodes for 7,723 women and identified two outcomes: Gap A, twelve-month discontinuation (n = 11,624), and Gap B, among women who discontinued within twelve months, method-related versus intentional discontinuation (n = 3,701). We evaluated four algorithms (survey-weighted logistic regression, elastic net, Random Forest, and XGBoost) on the same held-out test set (80/20) and reported the same area under the receiver operating characteristic curve (AUC). ResultsDiscontinuation after 12 months was 32.3 percent (survey-weighted), a value that is very similar to the Nigeria-published life-table value of 37.2 percent. The best-performing algorithm in both cases, Gap A (AUC = 0.699) and Gap B (AUC = 0.664), was Random Forest, while the worst was XGBoost (AUC = 0.621) in Gap B, which struggled even with the basic logistic baseline. The most common predictor across all four algorithms and both outcomes was contraceptive method type, with long-acting reversible contraceptive (LARC) much less likely to be discontinued than short-acting methods. Among individuals who discontinued within 12 months, reasons were almost evenly split between method-related (49.7 percent) and intentional (49.2 percent). ConclusionsFactors (side effects, cost, access, and efficacy dissatisfaction) account for approximately half of Nigerias contraceptive discontinuation problem and could be reduced with counseling, task shifting to pharmacists, and ongoing investments to increase the availability of longer-acting methods (LARCs) in Nigeria. Algorithm performance is outcome-dependent, not fixed, and it is better to compare algorithms by outcome than to default to a particular algorithm in DHS-based machine learning research.
ER -