Analysis of longitudinal binary outcomes in clinical trials with low percentage of missing values