Application of confidence interval in the assessment of dose proportionality

Aim:To introduce an statistically relevant approach for analysis of dose proportionality. Methods:First,linear regression of pharmacokinetic parameters vs doses was carried out and (1-α)% confidence interval of the estimated slope was calculated,which permited to define the which was relevant to the range of dose and knowledge of experiment.At last,conclusion was drawn according to the relationship between the confidence interval and the critical region.Results:The assessment of dose proportionality to satraplatin in phase I clinical trial indicated that the confidence interval model,which included consideration of variations in data collection,was more rigorous than traditional method.Conclusion:Coupled with statistical and expirical knowledges,the confidence interval model is suitable for pharmacokinetics study.Moreover,this model offers clear criteria which is helpful for the standardization of the assessment of dose proportionality.