Risk prediction models of radiodermatitis in breast cancer patients: a systematic review
10.3760/cma.j.cn211501-20240511-01177
- VernacularTitle:乳腺癌患者放射性皮炎风险预测模型的系统评价
- Author:
Xiaojie CHEN
1
;
Xiao ZHANG
;
Li WANG
;
Yunhong DU
Author Information
1. 山东第二医科大学护理学院，潍坊 261053
- Publication Type:Journal Article
- Keywords:
Radiodermatitis;
Forecasting;
Breast neoplasms;
Model;
Systematic review
- From:
Chinese Journal of Practical Nursing
2025;41(3):214-221
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To systematically review and evaluate risk prediction models for radiation dermatitis in breast cancer patients, providing a reference for developing higher-quality prediction models.Methods:A systematic search was conducted in PubMed, Embase, Web of Science, Cochrane Library, China National Knowledge Infrastructure, Wanfang Data, and China Biomedical Literature Database for studies related to risk prediction models for radiation dermatitis in breast cancer patients. The search timeframe was from database inception to February 28, 2024. Two researchers independently screened relevant literature based on inclusion and exclusion criteria, extracted the basic characteristics of the studies, and performed analysis. The included models were assessed for risk of bias and applicability.Results:A total of 8 articles involving 9 models were included, covering a total sample size of 1 291 cases, with 10 to 144 outcome events. The number of predictors ranged from 2 to 10, with common predictors including body mass index, smoking history, and breast volume, radiation dose, etc. The AUC values of the included models ranged from 0.76 to 0.98. The overall risk of bias for the models was relatively high, mainly due to issues related to study design, missing data handling, variable selection, and model validation.Conclusions:Existing risk prediction models for radiation dermatitis in breast cancer patients have certain limitations. It is recommended that future research further improve study designs, validate and optimize existing models, and develop high-performance risk prediction models.
