Objective: Assessment of validity of the Japanese Otological Society and the European Academy of Otology and Neurotology (EAONO-JOS) cholesteatoma staging system (EJS) through international collaboration to investigate: (a) feasibility of retrospectively staging previously acquired data, (b) strengths and weaknesses of the staging system.
Method: Nine centers with prospective cholesteatoma databases were recruited. Cases were allocated to EJS Stage at each center using details from time of initial surgery. An independent rater also staged the cases and noted any discrepancies. At one center results from database staging were compared with staging based on contemporaneous operative records. Inter and intrarater reliabilities were calculated, and recidivism rates calculated according to Stage.
Results: Of 1482 cases of cholesteatoma, 320 (22%) were Stage 1, 977 (67%) Stage 2, 153 (11%) Stage 3 and 4 (0.3%) Stage 4. No database contained details of all parameters required for accurate staging. Staging discrepancies occurred in >10% cases but inter and intrarater reliability of staging were high (Kappa 0.8; 95% confidence interval between 0.7-0.9). At 5 years, 11% had residual and 8% had recurrent cholesteatoma: rates increased with Stage but generally not significantly (Kaplan-Meier Log Rank analysis).
Conclusion: The EJS Staging system provides an internationally agreed standard for collecting data to classify cholesteatoma severity. Although data can be applied retrospectively with reasonable reliability, prospective data collection would prevent inaccuracies that arise from missing data fields. To enhance validity of the EJS system, we propose clearer definitions of some categories. Refinements to definitions of stage may improve prognostic utility of the EJS but should be made using evidence powered by large-scale collaboration.