目的 构建并验证ICU成人患者中心静脉导管（central venous catheter，CVC）相关皮肤损伤（catheter-associated skin injury，CASI）风险预测模型。方法 采取便利抽样法，选取2023年1月—10月广东省某三级甲等综合医院符合纳入和排除标准的ICU患者为调查对象，通过单因素分析和二元Logistic回归分析确定危险因素并构建预测模型。采用受试者操作特征曲线下面积、校准曲线评价模型的区分度和校准度，采用临床决策曲线评价模型的临床有效性。结果 研究纳入434例ICU患者，发生CVC-CASI 68例，发生率为15.7%，共发生74例次CVC-CASI，其中非感染性渗出54例次，占73.0%。模型变量包括留置时间、置管部位、俯卧位通气、皮肤水肿、血清白蛋白、凝血酶原时间、活化部分凝血活酶时间及穿刺次数。模型验证效果显示，受试者操作特征曲线下面积为0.838（95%CI为0.784~0.893），灵敏度为77.9%，特异度为81.1%，最大约登指数为0.59。结论 ICU患者CVC-CASI的发生率较高，该研究构建的预测模型具有较好的预测效能，能够为临床医护人员早期识别高危人群和及时进行干预提供理论依据。

Objective This study developed and validated a risk prediction model of central venous catheter-associated skin injury（CVC-CASI） in patients in the intensive care unit. Methods Using convenience sampling，clinical data were obtained for ICU patients who were admitted to a tertiary grade A hospital in Guangdong Province between January 2023 and October 2023. Independent risk factors for CVC-CASI were identified by univariate and binary logistic regression analyses. And a risk prediction model was constructed. The discrimination and calibration of the model was evaluated by the area under the receiver-operating characteristic（ROC） curve and calibration curve，and the clinical effectiveness was assessed by the clinical decision curve. Results 68 of the 434 patients who were included in the study developed CVC-CASI，giving an incidence rate of 15.7%. 74 cases of CVC-CASI occurred，with non-infectious exudation accounting for 54 cases（73.0%）. Model variables included catheter dwelling time，catheterization site，prone ventilation，skin edema，serum albumin，prothrombin time，activated partial thromboplastin time，and puncture frequency. The model validation results showed that the area under the ROC curve was 0.838（95% confidence interval 0.784-0.893） with a sensitivity of 77.9% and a specificity of 81.1%. The Youden index was 0.59. Conclusion The incidence of CVC-CASI in ICU patients is relatively high. Our prediction model has good prediction efficiency and can provide theoretical basis for early identification of high-risk populations and timely intervention for clinical medical staff.