Predicting Hypertension in the United States

This paper demonstrates the application of machine learning (ML) to predict patients with hypertension. The data was gathered from the New York City community health survey database for the 2018 survey year, which contains self-reported socio-demographic and health-related items. The study predicted individuals who were at risk of hypertensive conditions. Hypertensive respondents were identified using a battery of questions. The objective was to predict these individuals using social determinants of health (SDH) and clinical attributes. The analysis also shows the importance of clinical or pseudo-clinical measures to improve prediction accuracy. Our planet is under a severe pandemic, COVID-19. While this paper is on hypertension, a secondary conclusion was drawn. The world lacks a global database with clinical attributes for COVID-19 infected, recovered, and deceased patients. Machine learning with clinical data would immensely increase the potential for effective testing and a vaccine.