In response to Laser versus conventional fenestration in stapedotomy for otosclerosis: a systematic review.

We would like to thank the authors for taking the time to address several important issues regarding our article on laser versus conventional fenestration in stapedotomy. We agree with the authors that some considerations need to be taken into account when interpreting the findings of our systematic review. The authors once more stress that no uniform definition of sensorineural hearing loss is used in the included studies, and the pure-tone audiometric frequencies that were used differ. The difference in pure-tone audiometric frequencies is one of several heterogeneities discussed in our article, along with differences in approach to type and settings of the laser, the conventional fenestration technique, and follow-up duration. With regard to another remark made in the letter: The article written by the author’s group was neither included nor shortlisted. Quite a few years ago, Dr. Matthew Yung had the great idea to start an international otological database in order to gather important surgical data quicker and facilitate research. He contacted many esteemed surgeons to join this Common Otology Database (COD). In the initial idea, he proposed that apart from entering the data into the COD, independent auditors were needed for site visits to go through the records for verification—or “regular external validation” as he calls it. The article written by van Rompaey et al. was excluded during the screening process for several reasons. First, no mention was made of the type of laser that was used for laser fenestration, nor was there any mention of the laser settings that were used. Standardization of treatment was not achieved: Different types of lasers and surgical techniques were used in the 20 different tertiary-referral centers from which data were collected. Second, the results of 608 surgically treated cases of otosclerosis, with at least 3 months of follow-up and performed in 20 tertiary-referral otologic centers (the benchmark group) over a 5-year period, were included. This means that on average six patients were included per center per year for analysis (608 patients/20 centers/5 years56 patients/center/year). Because far more than six otosclerosis patients usually are surgically treated in tertiary referral centers, selection bias is most likely an issue. Third, it is possible to enter patient data after the date on which the actual surgical procedure was performed in the online Common Otology Database (COD) (tested on January 22, 2014). This being the case, the preoperative data could be submitted in the postoperative phase by the surgeon, resulting in the inherent risk of becoming a retrospective database; again, selective reporting cannot be excluded. The need for external data validation during site visits becomes important at this point, as proposed by Yung in 2005. A full disclosure of the participating sites and numbers per site and audit validation dates in the publication would have resulted in inclusion in our rapid systematic review. We agree that there is no evidence supporting the superiority of one of both techniques, which is also the conclusion of our review. Lacking high quality and consistent evidence, decisions regarding which method to use are instead based on expert opinion. We prefer the use of laser in stapedotomy when available, taking into account a theoretical risk of both footplate fracture and mechanical trauma.