A prognostic nomogram for all frequencies sudden sensorineural hearing loss based on the commixed index of inflammatory–immune–hemostasis–nutrition
摘要
All-frequencies sudden sensorineural hearing loss (AF-SSNHL) is a common otology emergency that greatly impacts patient’s quality of life. Given that inflammatory-immune-hemostasis-nutrition-related parameters are commonly reported prognostic factors, this study aimed to compare and evaluate the clinical value of such parameters to develop a prognostic nomogram. The analysis prospectively enrolled 245 patients with AF-SSNHL patients from August 2018 to May 2022. Patients were divided into training and validation cohorts randomly. The predictive values of the hematological indexes were calculated by receiver operating characteristic analysis, and predictive abilities were evaluated using logistic regression. A predictive prognostic model was developed using logistic regression and internally/external validated using calibration curves, decision curve analysis, and ROC analysis; a nomogram was constructed as its graphical representation. The prognostic nutritional index (PNI) and neutrophil lymphocyte D-dimer albumin score (NLDA) were superior to other hematological indexes in predicting the outcome of AF-SSNHL. The predictive prognostic model revealed that a high value of NLDA [odds ratio (OR), 4.363; 95% confidence interval (CI), 0.828–10.416; P < 0.05] and a low level of PNI (OR, 2.439; 95% CI, 1.188–5.006; P < 0.05) were independent risk factors for ineffective AF-SSNHL treatment. The concordance indexes for internal and external validation of the predictive model were 0.815 (95% CI, 0.754–0.875, P = 0.770) and 0.788 (95% CI, 0.663–0.913, P = 0.348), respectively. The areas under the curves of the nomogram-based predicted probabilities (nomogram scores) were 0.815 (95%CI, 0.754–0.875; P < 0.05) and 0.849 (95%CI, 0.737–0.961; P < 0.05), respectively. PNI and NLDA were superior hematological parameters and independent predictors of AF-SSNHL prognosis. Our predictive model, visualized as a nomogram based on vertigo, hearing level onset, PNI, and NLDA, may facilitate individualized prediction of treatment outcome of AF-SSNHL.
Trial registration: ChiCTR1800017072, https://www.chictr.org.cn/showprojen.aspx?proj=29023, Registration Date: 07/10/2018.