Reply

Sir

We appreciate the opportunity to provide additional information about our article [1]. The questions you raise point out some limitations of meta-analysis in general. In this study, meta-analysis involves comparing studies with heterogeneous definitions of erectile dysfunction (ED) and variation in follow-up duration across studies.

The definition for ED differed across studies and ranged from new reports of ED postoperatively to a decline in International Index of Erectile Function (IIEF) score. Whether or not ED was self-reported, if all patients were questioned about ED, or if a standardised questionnaire to evaluate ED was used, is described for each study in Table 1. In retrospective studies where patients were not asked about erectile function in advance, new onset ED after urethroplasty is self-reported and is subject to recall bias. In meta-analysis, the most weight was given to studies that prospectively questioned all patients about ED with a standardised questionnaire.

The time frame by which ED was determined was not provided in the methods for all studies included in our meta-analysis. The prospective studies that evaluated ED typically used a time frame of 6–18 months to evaluate postoperative erectile function. The study follow-up time frame is listed in Table 1.

Random effects meta-analysis was used to calculate the incidence of ED rather than simply dividing the number of cases of ED by the total number of patients [2].