Atieh [11] found no significant difference in perforation risk. In these studies, occurred in the two groups of the RCTs are almost identical perforations. Maybe due to the fact that they included only one RS, while our study included 22, they see no deviation.

The review of Stacchi [12] also described a lower incidence of membrane perforation during piezosurgery (10.9%) than during conventional surgery (20.1%). These results are comparable with ours.

Geminiani [13] found a significantly lower incidence of membrane perforations by the meta-analysis of the retrospective studies. Such a difference was not detected during meta-analysis of the data collected from the randomised clinical trials. They describe that the differences are most likely due to the inherent limitations of retrospective studies that include biases in selection of control and exposure to risk variables. They say that RCTs should be considered the main, because these trials have the preferred design for assessing differences in the outcome of a systematic review.

While incorporating selected non-randomised and non-controlled prospective and retrospective studies, our current data show that there is less perforation of the Schneiderian membrane when using piezosurgery. This might be a weakness resulting from the inclusion of the non-controlled studies.