Poisson Regression Models for Cohort Studies

Poisson Regression Models for Cohort Studies

1. Prospective

   • Subjects are classified based on their exposure status at the beginning of the study

2. Retrospective

   • Determination of the exposure status is made for the past

3. In a cohort study, a disease-free study population is selected on the basis of their exposure to a factor and followed over a period of time to determine the occurrence of disease

   • Objectives
     • Describe the different measures of incidence in a cohort study
       • Incidence measures
         • measure that quantifies the number of new cases of a disease that develop in a population at risk during a specific time period
           • Incidence density
             • the number of events per person time
           • Cumulative Incidence
             • risk or incidence proportion
     • Define the concept of crude and adjusted relative risk (RR) using a Poisson regression model.
       • Stratified Analysis
         • The confounding variables: exposure, disease and strata
         • RR: estimated for the association of interest.
         • Mantel–Haenszel method: test the homogeneity of the RRs
       • Poisson Regression Model
         • estimates the incidence of a disease or healthrelated event under different conditions
       • Definition of Adjusted Relative Risk
         • Relative Risk Estimation
           • When the exposure is dichotomous (1 denotes presence and 0 denotes absence), the confidence interval is given by the following expression:
         • it is used to obtain a comparison between the exposed group and the unexposed group under the assumption that the confounding variables have the same value
       • Interaction Assessment
         • there are no interactions between the exposure and potential confounding variable
     • Assess confounding and interaction effects using a Poisson regression model.
       • Confounding Variable
         • It is a distortion in the estimated magnitude of association that is produced when an extraneous factor
           • it is recommended to determine if the RR crude overestimates or underestimates the association of interest using as crude a reference the RR adjusted.

Images and files in the map