Machine learning evaluation of a hypertension screening program in a university workforce over five years.

Area of Science:

• Public Health
• Health Informatics
• Machine Learning Applications

Background:

• Global hypertension prevalence is high, particularly in low- and middle-income countries.
• Workplaces offer a practical setting for early hypertension diagnosis and treatment.
• Machine learning approaches are underutilized for workplace health assessments.

Purpose of the Study:

• To evaluate a workplace screening strategy for hypertension using machine learning.
• To analyze employee health checkup data for hypertension trends.
• To assess the feasibility of using k-means clustering for hypertension detection in the workforce.

Main Methods:

• Utilized an anonymized dataset of 1,723 employees from a university.
• Analyzed data including demographics, blood pressure, year group, department, and gender.
• Applied machine learning, specifically k-means clustering, to determine blood pressure status (low, normal, high).

Main Results:

• The average workforce age is 42.
• Hypertension was prevalent in employees over 40, irrespective of gender or job type.
• A consistent decrease in hypertension prevalence was observed from 2018 to 2022.

Conclusions:

• Machine learning is a feasible and sustainable tool for periodic workplace health monitoring.
• This approach can aid in diagnosing and controlling hypertension among the workforce.
• Workplace screening strategies integrated with machine learning show promise for public health interventions.