Construction of a prediction model for seroma after endoscopic thyroid-ectomy by breast approach
10.3969/j.issn.1009-9905.2025.05.001
- VernacularTitle:胸乳入路腔镜甲状腺手术后血清肿的预测模型构建
- Author:
Sheng-fei YANG
1
;
Yun-da ZHANG
;
Ming LIU
;
Shi-ran QIAN
;
Shu-xiong LI
;
Man ZHANG
;
Meng-ling WEI
;
Dong-wei LI
Author Information
1. 东莞东华医院 普外科(广东 东莞 523006)
- Publication Type:Journal Article
- Keywords:
Endoscopic thyroidectomy by breast approach;
Seroma;
Training cohort;
Prognostic factors;
Nomogram
- From:
Chinese Journal of Current Advances in General Surgery
2025;28(5):337-342
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To explore the prognostic factors of seroma after endoscopic thyroidectomy by breast ap-proach,and construct a nomogram to predict the possibility of cervical seroma.Methods:Data of patients undergoing endoscopic thyroid surgery in Dongguan Tungwah Hospital from January 2022 to May 2024 and Dongguan Songshan Lake Tungwah Hospital from May 2023 to August 2024 were retrospectively analyzed,and 1493 patients meeting the in-clusion criteria were selected.Among them,there were 1048 patients in Dongguan Tungwah Hospital as the training co-hort,1015 patients without seroma group and 33 patients with seroma group.There were 445 patients in Dongguan Songshan Lake Tungwah Hospital as the verification cohort,including 424 patients without seroma and 21 patients with seroma.Multivariate logistic regression analysis was used to obtain relevant independent prognostic factors,and R soft-ware established a nomogram model.Calibration curves,Hosmer-Lemeshow goodness of fit,ROC curves were used to evaluate the calibrability of the nomogram model,and clinical utility was assessed by clinical decision curves.Results:Multivariate logistic regression analysis showed that central lymph node dissection,diabetes,hyperthyroidism,and nod-ule size were independent prognostic factors related to seroma.Based on the prognostic factors,the nomogram of se-roma after ETBA was constructed.The calibration curves of the training and the verification group were in good agree-ment with the observed results,and the Hosmer-Lemeshow goodness of fit test was good,with the training cohort P=0.244 and the verification cohort P=0.803.The ROC curve of the training cohort showed that the area under the curve was 0.810(95%CI:0.740～0.879),and the ROC curve of the verification cohort showed that the area under the curve was 0.815(95%CI:0.722～0.909).Conclusion:The nomogram model based on the relevant prognostic factors ob-tained by multivariate logistic regression analysis has a good prediction effect on the seroma after ETBA,and can provide reasonable and individualized treatment plan for patients.
