We greatly appreciate the interest Dr Haylen and colleagues have shown in our recent article evaluating the interrater reliability of the International Continence Society/International Urogynecological Association (ICS/IUGA) classification system for mesh-related complications. We are surprised by Haylen et al’s assertion that the classification system is not appropriate for retrospective use and was “designed for prospective ‘live’ use with full clinical information” because it’s clearly stated in the Preface of their most recent article they list “medical records and surgical audits” as among the possible applications of the system. If indeed the intent of the ICS/IUGA Standardization Committee is that the system only be used in a prospective fashion with optimal data collection, then: (1) the classification system should be amended to make this explicit and (2) the applicability of the system will be severely limited as it would preclude its use by government agencies and academic centers with large administrative databases and for the majority of surgical studies submitted for peer review publication.

Dr Haylen and colleagues attempt to reinterpret our reliability data by eliminating noncorrelation based on medical record issues suggesting that once these are eliminated, the agreement is “very acceptable.” Unfortunately, it is not clear what assumptions they made and we are not able to duplicate their calculations. As noted in Table 3 of our paper, disagreement between reviewers because of insufficient information in the medical record was noted in only 16 patients (12%) and eliminating them does not appreciably improve the reliability measures (kappa for vaginal category .34).

One of our criticisms that Haylen et al disagreed with was that the IUGA/ICS classification system does not allow the gradation of severity. To illustrate this point we offer an example. A patient experiencing urinary retention after a sling is categorized as 4B. However, it requires a minimal intervention of a simple sling lysis, with moderate burden to the patient. In contrast, another patient may be suffering from vaginal constriction and dyspareunia, and undergo numerous interventions in the effort to correct this, yet still may remain debilitated by chronic pain. Despite a much greater burden, this patient would be categorized as 1B.

We very much appreciate the significant efforts that these authors have made to develop a tool that has such clinical need. Unfortunately, our data suggest that the system has poor interrater reliability. We strongly encourage careful validity and reliability testing of this and any similar classification system before wide-spread adoption.