A few comments on meta-analysis of skewed continuous data