Development and validation of the nomogram for predicting preoperative vocal cord palsy in thyroid cancer patients.

Abstract

Low incidence of preoperative vocal cord palsy (VCP) promotes a diagnosis model to eliminate patients without the necessity of preoperative laryngoscopy assessments, avoiding medical costs and discomfort. However, previous studies lacked a comprehensive strategy and external validation data to effectively detect VCP in thyroid cancer patients. This study aimed to develop a VCP scoring system that could calculate cumulative VCP risks and determine preoperative laryngeal examinations based on the clinical characteristics of VCP patients from the Union Hospital, Tongji Medical College of Huazhong University of Science and Technology. A retrospective study recruited 5,354 thyroid cancer patients was performed. Preoperative VCP incidence was recorded, and a prediction table was constructed using independent, significant risk factors for preoperative VCP. The visualized nomogram, including five parameters, was proportionally assigned 0 to 100 points. Finally, the diagnostic performance was confirmed by verifying the nomogram in the internal and external cohort. The incidence of preoperative VCP by preoperative laryngoscopy assessment was 1.57%. Age at diagnosis (OR: 1.04; P=0.006), history of neck surgery (OR: 11.57; P<0.001), voice symptoms (OR: 32.75; P<0.001), large nodule diameter (OR: 1.04; P<0.001) and suspicious neck lymph nodes (OR: 3.25; P<0.001) were identified as independent risk factors. The nomogram was proven to be acceptable discrimination in internal and external sets, and the cut-off value was 94.7. We identified clinical risk factors related to preoperative VCP and established a nomogram for VCP clinical discrimination with an excellent performance in the external cohort.