Development and validation for multifactor prediction model of sudden sensorineural hearing loss

Area of Science:

• Otolaryngology
• Medical Diagnostics
• Biostatistics

Background:

• Sudden sensorineural hearing loss (SSNHL) presents a significant global health challenge.
• Effective prediction models for SSNHL are currently lacking, hindering timely treatment.
• Early and rapid diagnosis is crucial for successful SSNHL management.

Purpose of the Study:

• To develop and validate a predictive model for SSNHL using readily available clinical data.
• To identify key indicators from conventional coagulation tests (CCT) and blood routine tests (BRT) associated with SSNHL.
• To create a nomogram for assisting physicians in early SSNHL diagnosis and treatment planning.

Main Methods:

• A retrospective study involving two cohorts (development and validation) of SSNHL patients.
• Evaluation of basic characteristics, CCT, and BRT results.
• Binary logistic regression analysis to build a prediction model and nomogram, validated using ROC curves and decision curve analysis.

Main Results:

• Thrombin time (TT), red blood cell (RBC), and granulocyte-lymphocyte ratio (GLR) were identified as significant predictors of SSNHL.
• The developed nomogram demonstrated good predictive performance with AUCs of 0.871 (development) and 0.759 (validation).
• No significant difference was found in ROC curves between the development and validation cohorts.

Conclusions:

• A validated multifactor prediction model for SSNHL has been established.
• The model incorporates easily and quickly accessible factors, facilitating early diagnosis.
• This tool can support physicians in making timely clinical treatment decisions for SSNHL.