Letters to the Editor: Mean values after treatment or mean difference?

Abstract

In a systematic review and meta-analysis Published in the Urology (1) entitled Low-Intensity Extracorporeal Shock Wave Therapy for Erectile Dysfunction: a Systematic Review and Meta-Analysis by Libo and Guizhong in 2017 , the authors explained that weighted mean difference (MD) was used for continuous variables in their study. However, the meta-analysis was performed on mean values after treatment in both intervention and control groups. In other words, the authors only used the endpoint scores as the mean difference in each group regardless of the baseline scores. Based on statistical literature, the mean difference is change average from baseline to endpoint. To calculate the mean difference, the difference between baseline and endpoint values should be considered (endpoint mean score minus baseline mean score). \xa0 This point was mentioned in the Cochrane Handbook for Systematic Reviews of Interventions. The mean difference (more correctly, difference in means ) is a standard statistic that measures the absolute difference between the mean value in two groups in a clinical trial and the mean of the differences is equal to the difference in means (2). \xa0Therefore, the result of Libo s meta-analysis (1) is incorrect due to the use of the endpoint scores (after treatment) and leads to wrong findings in assessment the efficacy of intervention. So, the meta-analysis should be performed on the mean difference.