Hypertension prediction using machine learning algorithm among Indonesian adults

Abstract

Early risk prediction and appropriate treatment are believed to be able to delay the occurrence of hypertension and attendant conditions. Many hypertension prediction models have been developed across the world, but they cannot be generalized directly to all populations, including for Indonesian population. This study aimed to develop and validate a hypertension risk-prediction model using machine learning (ML). The modifiable risk factors are used as the predictor, while the target variable on the algorithm is hypertension status. This study compared several machine-learning algorithms such as decision tree, random forest, gradient boosting, and logistic regression to develop a hypertension prediction model. Several parameters, including the area under the receiver operator characteristic curve (AUC), classification accuracy (CA), F1 score, precision, and recall were used to evaluate the models. Most of the predictors used in this study were significantly correlated with hypertension. Logistic regression algorithm showed better parameter values, with AUC 0.829, CA 89.6%, recall 0.896, precision 0.878, and F1 score 0.877. ML offers the ability to develop a quick prediction model for hypertension screening using non-invasive factors. From this study, we estimate that 89.6% of people with elevated blood pressure obtained on home blood pressure measurement will show clinical hypertension.