[Skip to Navigation]
Letters to the Editor
February 2008

The Missing Data Problem in Meta-analyses

Arch Gen Psychiatry. 2008;65(2):238. doi:10.1001/archgenpsychiatry.2007.31

The article by Leichsenring and colleagues1 presented an interesting meta-analysis of 17 studies to examine the efficacy of short-term psychodynamic psychotherapy. As in many other meta-analyses, they do not report any analyses of the treatment dropout rates and attrition rates during the follow-up phase of the studies. However, every meta-analysis can be only as good as the studies it is based on. Factors that influence effect size estimations include study design issues and sample characteristics.2 Missing data are particularly problematic because they impose a serious threat to the internal and external validity of a clinical trial.

Add or change institution