RESUMEN
OBJECTIVE: Determine predictive values of preoperative stages of pars tensa retractions, coexisting attic retraction and preoperative air conduction for ossicular destruction in isolated and combined pars tensa retraction, and predictors for successful tympanic grafts after surgery. STUDY DESIGN: Prospective case series study. SETTINGS: Tertiary referral center. PATIENTS: Forty-eight children ears without cholesteatoma were included in the study: 23 with isolated pars tensa retraction (median age, 11), 25 with combined pars tensa and attic retractions (median age, 13). INTERVENTION(S): Otomicroscopy, pure tone audiometry, and impedancmetry were carried out preoperatively. Three surgical procedures for isolated retractions were used: ventilation tube insertion alone or together with fascia graft or cartilage graft. Two surgical procedures for combined retractions were used: ventilation tube insertion alone and with cartilage graft. MAIN OUTCOME MEASURES: Incidence of ossicular destruction, postoperative retraction of the grafts. RESULTS: Stage of pars tensa retraction and preoperative air conduction thresholds do not predict long incus process defect in isolated group. Coexistence of an attic retraction predicts combined, long incus process and stapes superstructure defect (Chi = 3.943, p = 0.047, OR = 12.00). Retractions of grafts are predicted by mode of surgery, favoring cartilage graft (isolated group: Chi = 4.306, p = 0.0372,OR = 4.69; combined group Chi = 4.7836, p = 0.0287, OR = 0.1364). Stage of pars tensa retraction predicts poor outcome of fascia graft (Chi = 4.5347, p = 0.0332, OR = 12.00). CONCLUSION: Absence of correlation between stage of pars tensa retraction and air conduction thresholds with ossicular defects justifies surgical exploration of the auditory ossicles, even in lower stages of retraction. Combined ossicular defect is expected in combined retractions. The usage of cartilage graft proved to be more appropriate.