Objective: This study aimed to determine which children are at risk of having residual cholesteatoma develop after initial surgery for either cholesteatoma or severe retraction pocket. This study aimed to identify the characteristics of each patient and assessing whether predictive factors exist to propose a well-guided therapeutic approach.
Design: This was a retrospective study. Mean follow-up period (after the last eradication intervention) was 42 months.
Patients: Two hundred thirty-one children (256 ears) with either cholesteatoma (n = 157) or severe retraction pockets (n = 99), both treated surgically, composed the patient group.
Statistical analysis: univariate analysis of event-free survivals (using log-rank test and Kaplan-Meier estimate) and multivariate analysis (using Cox regression model) of clinical history, surgical observations, residuals occurrence (using Kaplan-Meier's statistical method) were used.
Results: Posterior mesotympanum involvement, ossicular chain interruption after disease excision (and moreover combination of both), relative lack of experience of the surgeon, and presumed incomplete removal were identified as independent risk factors highly correlated with residuals' development (relative risks ranging from 2.02-4.84). Disease extension, type of process (cholesteatoma vs. retraction pocket), eventual surgical history, and surgical method initially used (closed or open) showed no such correlation.
Conclusion: The identification of any of these factors should instigate a mandatory second-look procedure regardless of other initial surgical findings and of technique used. The authors propose to advance this procedure from 12 to 9 months when all these factors are found in one child.