When combining continuous data using meta-anlyses most often we use mean, standard deviation and number of sample in each group. Is it possible to combine data reported as median and interquartile range? Is there anyone who has experience in this regard?