A machine learning analysis of perceived healthcare access barriers among women of reproductive age in Sub-Saharan Africa: using demographic and health surveys

Abstract

BACKGROUND

In Sub-Saharan Africa (SSA), women of reproductive age face significant barriers to accessing healthcare, undermining progress towards Universal Health Coverage. Understanding the complex determinants of this multifaceted problem requires advanced analytical approaches.

OBJECTIVE

This study aimed to identify and rank the key predictors of perceived healthcare access barriers among women in SSA using a multi-country dataset and an interpretable machine learning (ML) framework.

Study Design

We analyzed recent Demographic and Health Survey data from 28 SSA countries (n=467,965 women). A composite measure of healthcare access barriers was used as the outcome. After preprocessing and feature selection, eight ML models were trained and evaluated. The best-performing model was interpreted using SHapley Additive exPlanations (SHAP).

RESULTS

The pooled prevalence of perceived healthcare access barriers ranges from 17.8% (95% CI, 17.0–18.6%) in South Africa to 72.1% (95% CI, 71.6–72.7%) in Malawi. The Light Gradient Boosting Machine (LightGBM) was the top-performing model (Accuracy:65.2%, F1-score: 72.8%, ROC-AUC:68.2%, and PR-AUC: 72.9%). SHAP analysis identified the wealth index as the most powerful predictor, followed by use of internet, media exposure, mobile phone ownership, and educational level. Lower socioeconomic status and limited access to information and communication technologies were strongly associated with greater access barriers

CONCLUSION

The findings underscore that socioeconomic and digital inclusion factors are paramount in determining healthcare access. This highlights the need for integrated strategies that go beyond health-sector solutions to address underlying social and economic determinants. This study provides an evidence-based, actionable ranking of the determinants of perceived healthcare access problems. The results can inform targeted policies on poverty alleviation, digital inclusion, and education to reduce disparities and advance equitable healthcare access for women in SSA.