# Construction and efficacy evaluation of a prognostic assessment model for corticosteroid treatment in patients with sudden sensorineural hearing loss based on lipid metabolism profile

## TL;DR

This study builds a model to predict how well corticosteroid treatment will work for sudden hearing loss based on patients' lipid metabolism profiles.

## Contribution

A novel prognostic nomogram based on lipid metabolism factors for predicting corticosteroid treatment outcomes in SSNHL patients.

## Key findings

- Older age, diabetes, and higher ApoB/ApoA ratio are risk factors for poor prognosis.
- Higher HDL-C acts as a protective factor for better treatment outcomes.
- The nomogram showed good predictive accuracy (AUC of 0.785 in training and 0.744 in validation).

## Abstract

To explore the correlation between lipid metabolism profile, clinical indicators and prognosis of corticosteroid treatment in sudden sensorineural hearing loss (SSNHL) patients, and construct/verify a prognostic assessment model based on lipid metabolism profile for clinical individualized treatment.

A retrospective study enrolled 446 SSNHL patients (divided into training set, n = 312; validation set, n = 134) who received corticosteroid treatment. Patients were grouped into good/poor prognosis groups by 4-week post-treatment pure tone average (PTA) improvement. Differences in indicators were compared; multivariate logistic regression identified independent factors to build a nomogram. Calibration curves, decision curve analysis (DCA) and ROC curves evaluated the model.

Poor prognosis group had higher age, diabetes/hypertension rates, ApoB/ApoB/ApoA ratio, non-HDL-C, disease duration, total deafness rate, and lower HDL-C/ApoA (all p < 0.05). Independent factors: older age, diabetes, higher ApoB/ApoA ratio, longer disease duration (risk factors), higher HDL-C (protective factor, all p < 0.05). The nomogram showed good consistency (Hosmer-Lemeshow p = 0.960/0.472), predictive efficacy (AUC = 0.785/0.744) and clinical net benefits in both sets.

Age, diabetes, HDL-C, ApoB/ApoA ratio and disease duration are key factors for SSNHL corticosteroid treatment prognosis. The nomogram based on these indicators has reliable predictive efficacy, serving as an effective tool for clinical prognosis assessment and individualized treatment.