Construction and validation of a risk prediction model for perineal radiation dermatitis in cervical cancer patients undergoing radiotherapy
10.3760/cma.j.cn211501-20240621-01629
- VernacularTitle:宫颈癌放疗患者会阴部放射性皮炎风险预测模型的构建及验证
- Author:
Fang ZHANG
1
;
Zhao WANG
1
;
Xueyu LI
1
;
Xiaocen CHEN
1
Author Information
1. 天津医科大学肿瘤医院放射治疗科 国家恶性肿瘤临床医学研究中心 天津市肿瘤防治重点实验室 天津市恶性肿瘤临床医学研究中心，天津 300060
- Publication Type:Journal Article
- Keywords:
Uterine cervical neoplasms;
Intensity modulated radiotherapy;
Radiation dermatitis of the perineum;
Risk factors;
Prediction model
- From:
Chinese Journal of Practical Nursing
2025;41(11):858-865
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To construct a risk prediction model for radiation dermatitis in the perineum of cervical cancer patients undergoing radiotherapy, and verify the predictive effect of the model, so as to provide theoretical basis for reducing the risk of perineal radiation dermatitis in clinical practice.Methods:A prospective study design was used to conveniently select 247 cervical cancer patients who received radiotherapy at Tianjin Medical University Cancer Institute & Hospital from January 2022 to March 2023 as the modeling group. They were divided into a group with perineal radiodermatitis ( n = 130) and a group without perineal radiodermatitis ( n = 117). The data of the two groups were compared, and a risk prediction model was constructed based on the results of logistic regression analysis. Used the area under the receiver operating characteristic (ROC) curve to test the predictive performance of the model. From May to November 2023, 109 patients were selected as the validation group to conduct external validation of the model. Results:Among the 247 patients in the modeling group, the age of the group with perineal radiodermatitis was (45.31 ± 7.25) years old, and the age of the group without perineal radiodermatitis was (46.70 ± 6.43) years old. The incidence of perineal radiation dermatitis during radiotherapy in the modeling group was 52.6% (130/247), while in the validation group it was 44.0% (48/109). The final predictive variables included in the model were high blood sugar or not ( OR = 0.304, 95% CI 0.174-0.530), vaginal bleeding/discharge or not ( OR = 4.338, 95% CI 2.397-7.854), involvement of 1/3 of the vagina or not ( OR = 0.497, 95% CI 0.281-0.879), whether topical skin protectants could be used( OR = 3.479, 95% CI 1.596-7.583), whether skin protectants could be used in a standardized manner ( OR = 2.399, 95% CI 1.201-4.789), whether the skin could be cleaned with the correct cleaning agent ( OR = 3.155, 95% CI 1.603-6.209), and whether the skin could be wiped with the correct method ( OR = 3.325, 95% CI 1.806-6.121), totaling 7 predictive factors (all P<0.05). The area under the ROC curve of the prediction model was 0.866, the sensitivity was 0.735, and the specificity was 0.852, the Youden index was 0.587. Model validation results showed the area under the ROC curve was 0.762, the sensitivity was 0.815, the specificity was 0.717, and the accuracy was 80.73%, indicating good predictive performance. Conclusions:The predictive model constructed in this study had good predictive performance and could provide reference for clinical nursing staff to screen high-risk patients with perineal radiation dermatitis.
