Online Program Home
My Program

Abstract Details

Activity Number: 672 - Methods for Infectious Disease Epidemiology
Type: Contributed
Date/Time: Thursday, August 2, 2018 : 10:30 AM to 12:20 PM
Sponsor: Section on Statistics in Epidemiology
Abstract #328561 Presentation
Title: A Unified Probability Model for Pathogenic Etiology of Infectious Diseases
Author(s): Nong Shang*
Companies: CDC
Keywords: Etiology; population attributable risk; latent class model
Abstract:

A traditional epidemiology approach to estimate etiology fraction of a pathogen to an infectious disease is through applying population attributable risk (PAR) idea comparing carriage rates of the pathogen between disease cases and healthy controls. However, there are conceptual and technical difficulties to extend the method to situations with many pathogens. Recently, a new approach called partially latent class model (pLCM) was introduced to address the issues of PAR. It allows imperfect laboratory test results and is able to estimate multiple pathogens' etiology fractions simultaneously. However, the definitions of pLCM's model parameters as well as its major model assumptions lack clear epidemiological interpretations. In this paper, we propose a probability model that considers carriage of a pathogen elevating risk of developing disease. Under this model structure, both PAR and pLCM would produce the same etiology fraction estimation if the elevated risks from different pathogens are additive. The model provides epidemiological interpretations over pLCM's parameters and model assumptions. Consequently, a procedure can be developed to examine pLCM's model assumptions.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2018 program