In order to further reinforce the conclusion above， further subgroup analyses were performed based on the treatment dosages. Again an X2 test was completed to examine the heterogeneity； the results were P = 0.002(<0.10)， I2 =80% (>50%)； P = 0.38(>0.10)， I2 =5% (<50%) and P = 0.35 (>0.10)， I2 = 0%(< 50%)， respectively， which indicated that there was heterogeneity. Therefore a random-effect model was used. The results (WMD= 0.97； 95%CI= -0.38 to 2.32； P=0.16 for 0.2mg tamsulosin； WMD = 1.02； 95%CI=0.71 to 1.34； P<0.00001 for 0.4mg tamsulosin； and WMD = 1.07； 95% CI = 0.65 to 1.48； P<0.00001 for 0.8mg tamsulosin) indicated that there were significant differences between the tamsulosin groups and the placebo groups for BPH in the 0.4mg and 0.8mg dose subgroups (Figure 3). ......