摘要
Abstract
Objective:To analyze the risk factors of post-mastectomy pain syndrome(PMPS)and construct a risk prediction model.Methods:A total of 275 patients with breast cancer surgery admitted to the hospital from February 2021 to February 2024 were retrospectively selected as the research subjects.According to whether PMPS occurred after surgery,they were divided into PMPS group and non-PMPS group.The clinical data of the two groups of patients were collected.Univariate and multivariate Logistic regression were used to analyze the risk factors of PMPS.The decision tree method and the Logistic regression algorithm were used to construct its risk prediction model.The predictive value of the two models for PMPS was compared.Results:Among 275 breast cancer surgery patients,58 cases developed PMPS during the follow-up,with an incidence of 21.09%.Logistic regression analysis showed that diabetes,tumor location,preoperative anxiety and depression,axillary lymph node dissection,and postoperative chemoradiotherapy were risk factors for PMPS in breast cancer patients(P<0.05).The decision tree model was generated based on risk factors and selected four explanatory variables:tumor location,diabetes,preoperative anxiety and depression,and postoperative chemoradiotherapy,among which tumor location was the most important influencing factor for PMPS.The area under the curve(AUC)of the decision tree model was 0.773[95%CI(0.719,0.821)],which was greater than 0.705 of the Logistic regression model[95%CI(0.647,0.758)],P<0.05.Conclusions:Diabetes,tumor location(upper quadrant),preoperative anxiety and depression,axillary lymph node dissection,and postoperative chemoradiotherapy are risk factors for PMPS in breast cancer patients.The decision tree model constructed in this study has a significant advantage in predictive efficiency compared with the Logistic regression model.关键词
Logistic回归/决策树法/乳腺癌/术后疼痛综合征/风险预测模型Key words
Logistic regression/decision tree method/breast cancer/postoperative pain syndrome/risk prediction model
