Estimation of Random Effects in a Logistic Regression

• Activity Number: 239
• Type: Contributed
• Date/Time: Monday, August 5, 2013 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): Jorgen Holm Petersen*+
• Companies: Kobenhavns Universitet
• Keywords: logistic regression model ; random effects model ; median odds ratio ; rater agreement

Abstract

A new approach to estimation in a logistic regression model with random effects is presented. An important example that will serve to motivate and illustrate the proposed model, is the study of Umbilical artery Doppler velocimetry used by obstetricians to assess the status of a fetus. The purpose of the assessment is to improve the fetus' chance of survival by choosing the optimal time of elective delivery. In the study, data related to 139 perinatal deaths were sent to 32 experts who were asked whether the use of Doppler velocimetry might have prevented each death. All raters have supplied dichotomous ratings of the same cases. The model includes two sets of parameters: one for the individuals (representing the severity of the 139 cases) being rated, and another for the 32 raters to have different propensities to score a given set of individuals positively or negatively - the rater bias. Estimating the rater variation without making distributional assumptions about the 139 cases is possible in this setting. The estimate is compared with standard approaches. Based on the model, the rater variation may be quantified as a suitable measure of the variation of the rater odds ratios.