Predicting Respiratory Diseases Attributed to PM2.5 Air Pollution in Nairobi County Using Random Forest Model

Area of Science:

• Environmental Health
• Data Science
• Public Health

Background:

• Urban air pollution, particularly fine particulate matter (PM2.5), poses significant risks to respiratory and cardiovascular health.
• Nairobi County faces challenges with air quality, necessitating effective tools for disease prediction and prevention.

Purpose of the Study:

• To evaluate the predictive performance of a Random Forest model for identifying respiratory diseases associated with PM2.5 exposure.
• To identify key demographic and environmental factors influencing respiratory and cardiovascular health outcomes in Nairobi.

Main Methods:

• Utilized a comprehensive dataset including demographic information and air quality metrics.
• Developed and validated a Random Forest machine learning model to predict disease outcomes.
• Performed feature importance analysis to determine key predictive variables.

Main Results:

• The Random Forest model achieved high accuracy (79.97%) and Area Under the Curve (AUC) of 0.872.
• Sensitivity and specificity were 81.88% and 73.27%, respectively, demonstrating effective classification.
• Age and PM2.5 concentrations were identified as the most significant predictors of health outcomes.

Conclusions:

• The study confirms the utility of machine learning for predicting PM2.5-related respiratory diseases.
• Highlights the critical roles of age and air pollution levels in disease etiology.
• Emphasizes the urgent need for air quality interventions in urban environments like Nairobi.