This method is used for detecting some potentially unpublished studies, and then adjusting publication bias in the summarized effect size.
This methods is regarded as sensitivity analysis, and the results can not be regarded as determinative conclusions.