Zou Guoling, Zeng Chuandao, Li Chenyang, Hu Wei
Department of Otolaryngology, Ankang Central Hospital Ankang 725000, Shanxi, China.
Otolaryngology Head and Neck Surgery Center, Xi'an People's Hospital (Xi'an Fourth Hospital) Xi'an 710004, Shanxi, China.
Am J Transl Res. 2025 May 15;17(5):3413-3423. doi: 10.62347/UDCK5613. eCollection 2025.
To investigate the factors influencing postoperative recurrence of auricular pseudocysts and to develop recurrence risk prediction models using logistic regression and Cox regression analyses.
This retrospective study analyzed clinical data from 215 patients who underwent surgical treatment for auricular pseudocysts between January 2015 and December 2022. Univariate analysis identified factors associated with recurrence, which were further assessed using multivariate logistic regression and Cox regression. Recurrence prediction models were constructed, and their predictive performance was evaluated using receiver operating characteristic (ROC) curves and area under the curve (AUC) values.
Univariate analysis identified age, cyst size, surgical approach, and postoperative adjuvant therapy as significant factors associated with postoperative recurrence (P<0.05). Multivariate logistic regression and Cox regression identified age <53.5 years, cyst size <2.5 cm, fenestration surgery, and absence of postoperative adjuvant therapy as protective factors against recurrence (P<0.05). The constructed models showed stable AUC values for 90-day and 120-day predictions (AUC = 0.718). No significant difference in predictive performance was observed between logistic regression and Cox regression models for 6-month recurrence risk (P = 0.934).
Age, cyst size, surgical approach, and postoperative adjuvant therapy are critical factors influencing postoperative recurrence of auricular pseudocysts. The recurrence prediction models based on logistic regression and Cox regression demonstrate high efficiency in predicting short-term recurrence and can guide postoperative management strategies.
探讨影响耳廓假性囊肿术后复发的因素，并使用逻辑回归和Cox回归分析建立复发风险预测模型。
本回顾性研究分析了2015年1月至2022年12月期间接受耳廓假性囊肿手术治疗的215例患者的临床资料。单因素分析确定与复发相关的因素，进一步采用多因素逻辑回归和Cox回归进行评估。构建复发预测模型，并使用受试者工作特征（ROC）曲线和曲线下面积（AUC）值评估其预测性能。
单因素分析确定年龄、囊肿大小、手术方式和术后辅助治疗是与术后复发相关的重要因素（P<0.05）。多因素逻辑回归和Cox回归确定年龄<53.5岁、囊肿大小<2.5 cm、开窗手术和无术后辅助治疗是预防复发的保护因素（P<0.05）。构建的模型在90天和120天预测中显示出稳定的AUC值（AUC = 0.718）。逻辑回归和Cox回归模型在6个月复发风险的预测性能上无显著差异（P = 0.934）。
年龄、囊肿大小、手术方式和术后辅助治疗是影响耳廓假性囊肿术后复发的关键因素。基于逻辑回归和Cox回归的复发预测模型在预测短期复发方面显示出高效率，并可指导术后管理策略。
