Prediction model for hearing outcome in patients with idiopathic sudden sensorineural hearing loss

Abstract

The aim of this study is to develop a regression model for predicting hearing outcome in patients with idiopathic sudden sensorineural hearing loss (ISSNHL). A total of 174 consecutive patients with ISSNHL (average of the hearing levels at 250, 500, 1,000, 2,000, and 4,000 Hz was ≥40 dB; time from onset to treatment was ≤30 days) were retrospectively analyzed. They received steroid administration (400 mg/day of hydrocortisone sodium succinate followed by tapered doses) in combination with hyperbaric oxygen therapy. The hearing improvement rate compared to the unaffected contralateral ear was calculated. Correlations between the hearing improvement rate and four prognostic factors (patient's age, days from onset to treatment, initial hearing level, and the presence of vertigo) were examined by simple and multiple regression analyses. In the simple regression analysis, significant correlations were observed between the hearing improvement rate and all four prognostic factors. In the multiple regression analysis, the correlation was significant for patient's age, days from onset to treatment, and the presence of vertigo with partial correlation coefficients of -0.221, -0.324, and -0.329, respectively, but was not significant for the initial hearing level. We subsequently formulated a multiple regression equation for predicting the hearing improvement rate. The multiple correlation coefficient was 0.495 with a p value of 1.42 × 10(-9). Using this regression model, the hearing improvement rate is still difficult to predict with 95% probability, but is predictable with 70% probability.