Analysis of breath alcohol measurements using compartmental and generalized linear models

Abstract

Pharmacokinetic parameters are important for clinical application. Traditionally, compartmental model together with non-compartmental approach are commonly adopted to deal with the analysis of pharmacokinetic data. In this study, we apply the alternative generalized linear model as proposed by Wake eld (2004) to the breath alcohol measurements of Chinese subjects in Hong Kong. Both compartmental and generalized linear models are tted to each subject involved. Four core pharmacokinetic parameters including the time to maximum blood alcohol concentration (BAC) level, the BAC level attained at peak, the rate of clearance and elimination half-life are examined. The parameter estimates under the two models are then compared. Finally, we also extend the concept from the individual to the population level.