Assessing the influence of a single study in the meta-anyalysis estimate