Developing Nurse‐Accessible Hypertension Prediction Tools for Low‐Income Populations: A Comparative Analysis of Machine Learning Algorithms With SHAP Interpretation

• Published in: International journal of nursing practice Vol. 31; no. 5; pp. e70060 - n/a
• Main Authors: Huang, Chuan, Xu, Jiaojiao, Qiu, Hai, Yue, Yuchuan
• Format: Journal Article
• Language: English
• Published: Australia Wiley Subscription Services, Inc 01.10.2025
• Subjects: Adult; Aged; Algorithms; Clinical nursing; Comparative analysis; Cross-Sectional Studies; Diabetes; Female; Humans; Hypertension; Hypertension - diagnosis; Hypertension - nursing; low‐income population; Machine Learning; Male; Medical screening; Middle Aged; Neural networks; Neural Networks, Computer; Nomograms; Nursing; Nutrition Surveys; Poverty; prediction; Predictions; Professional practice; Retrospective Studies; Risk assessment; Stratification; Transparency; low‐income population; prediction; nursing; machine learning; hypertension
• ISSN: 1322-7114; 1440-172X; 1440-172X
• DOI: 10.1111/ijn.70060

ABSTRACT Aim The aim of this study is to develop and compare machine learning algorithms for hypertension prediction in low‐income populations, with emphasis on model interpretability for nursing implementation in resource‐limited settings. Methods This retrospective cross‐sectional study analysed data from seven iterations of NHANES (2005–2018) focusing on low‐income populations. After LASSO regression identified eight key predictors, eight machine learning models were developed and evaluated using ROC curves, calibration plots and decision curve analysis, with SHAP methodology applied for interpretation. Results Among 12 506 participants, 39.96% had hypertension. Logistic regression and neural networks both achieved the highest discriminative ability (AUC = 0.853). SHAP analysis identified age as the most influential predictor, followed by waist circumference and diabetes status. A clinical nomogram with three‐tier risk stratification (< 30%, 30%–60% and > 60%) was developed for nursing assessment. Conclusion Neural network models with SHAP interpretation achieved optimal hypertension prediction (AUC = 0.853) while maintaining clinical transparency essential for nursing practice. The resulting nurse‐accessible nomogram with a visual scoring system supports evidence‐based screening in low‐income populations, pending external validation in clinical settings. Summary What is already known about this topic? Hypertension disproportionately affects low‐income populations with limited healthcare access, leading to delayed diagnosis and poor outcomes. Machine learning has shown promise in hypertension prediction but often lacks interpretability for clinical implementation. Nurses are positioned to implement early detection strategies but lack accessible prediction tools specifically designed for resource‐limited settings. What this paper adds? Development of machine learning models specifically calibrated for hypertension prediction in low‐income populations, with neural networks and logistic regression showing optimal performance (AUC = 0.853). Integration of SHAP methodology enhances model transparency, identifying age, waist circumference and diabetes status as the most influential predictors. Creation of a nurse‐accessible clinical nomogram with three‐tier risk stratification for enables direct implementation in community and primary care settings. The implications of this paper: It provides nurses with evidence‐based tools to efficiently identify at‐risk individuals in underserved communities, potentially improving early detection rates. It enhances nursing education by demonstrating the application of interpretable machine learning approaches in cardiovascular risk assessment. It supports policy development for task‐sharing in hypertension management, where nurses can effectively implement screening programmes with minimal technological requirements.