Background: The quality of data concerning AE in clinical trials of nondrug interventions is often low. This may arise of missing definitions of terms describing AE within publications or the use of different terms for describing the same AE between trials. Often, a further major problem is an insufficient explanation of the given data (number and follow-up-time of patients included in the analysis, distinction between number of patients, and number of events). This results in high risk of bias and therefore in a loss of validity concerning the assessment of AE. An additional problem performing systematic reviews (SR) of non-drug interventions may arise from a large number of treatment comparisons. Objectives: To present an approach of assessing AE data of poor quality in trials to gain more certainty in the results of a SR. This approach is presented on the basis of a SR on non-drug local treatment options for benign prostate hyperplasia. Methods: The first task was to define generic terms for the plurality of different terms used in primary trials. However, the main task was to develop an algorithm in order to obtain robust assessments of AE. Therefore, relevant items should be identified to be able to operationalize the procedure for gaining signs for differences between groups. This should lead to a summary concerning reported AE for a high number of primary trials and of treatment comparisons. Results: Generic terms for AE were defined for the SR to represent all AE reported in the included trials. We identified four relevant items for the algorithm to be used: (i) number of studies, (ii) sample sizes, (iii) observed intergroup differences, (iv) homogeneity between study results. This algorithm made it possible to extract signs for differences between the groups on the basis of the given data in primary trials and present them clearly in a SR. Conclusions: This method to assess the given information about AE by low quality data in clinical trials seemed to be a reasonable approach for the evaluation of harms gaining the possibility of more certainty in the results of the SR.