Predicting Changes in Systolic and Diastolic Blood Pressure of Hypertensive Patients in Indonesia Using Machine Learning.

Abstract

Purpose of review: This retrospective study investigated factors that influence the occurrence of decreased systolic and diastolic blood pressure including sociodemographic and economic factors, hypertension duration, cigarette consumption, alcohol consumption, duration of smoking, type of cigarettes, exercise, salt consumption, sleeping pills consumption, insomnia, and diabetes. These factors were applied to predict the reality of systolic and diastolic decrease using the machine learning algorithm Naïve Bayes, artificial neural network, logistic regression, and decision tree.

Recent findings: The increase in blood pressure, both systolic and diastolic, is very harmful to the health because uncontrolled high systolic and diastolic blood pressure can cause various diseases such as congestive heart failure, kidney failure, and cardiovascular disease. There have been many studies examining the factors that influence the occurrence of hypertension, but few studies have used machine learning to predict hypertension. The machine learning models performed well and can be used for predicting whether a person with hypertension with certain characteristics will experience a decrease in their systolic or diastolic blood pressure after treatment with antihypertensive drugs.