A Bayesian approach for detecting a disease that is not being modeled

Area of Science:

• Epidemiology
• Biostatistics
• Public Health

Background:

• Recent outbreaks of novel viruses like SARS, MERS, and Zika highlight the ongoing threat of emerging infectious diseases.
• These outbreaks have resulted in substantial loss of life and significant economic costs globally.
• The continued emergence and transformation of diseases necessitate improved detection and characterization methods.

Purpose of the Study:

• To develop and evaluate a Bayesian statistical model for the detection and characterization of previously unknown and unmodeled diseases.
• To assess the model's performance using historical patient-care report data.

Main Methods:

• Utilized a Bayesian statistical modeling approach.
• Applied the model to analyze historical patient-care reports.
• Evaluated the model's accuracy and effectiveness in identifying emergent diseases.

Main Results:

• The Bayesian model demonstrated capability in detecting and characterizing unknown diseases from patient data.
• Performance evaluation on historical data confirmed the model's utility.

Conclusions:

• The developed Bayesian model offers a promising tool for early detection and characterization of emergent diseases.
• This approach can enhance public health surveillance and response strategies for future outbreaks.