| Longitudinal semicontinuous data, which usually contains a fair amount of zeros is common in biomedical and econometric research. Sometimes the existence of missing values in this kind of data creates additional difficulty in analysis. We have applied the two part random effect models to the longitudinal semicontinuous toenail data to compare the efficacy and the safety features of the two treatments. We have used the mixed effect logistic regression model and the mixed effect gamma regression model in the first and the second part of the models respectively. For the sake of simplicity in the analysis and in the interpretation, we have considered the independent random effects in the modeling purpose. We have applied the pattern mixture approach to each part of the two part models to identify the effective dropout patterns of the data. |
