目的 研究子宫动脉频谱形态异常联合血清 CC 趋化因子配体 19（CCL19）水平预测系统性 红斑狼疮（SLE）患者不良妊娠结局（APO）的价值。方法 选取 2015 年 1 月至 2022 年 1 月于青岛市妇女儿童医院治疗的妊娠合并 SLE 患者 80 例，根据妊娠结局将其分为 APO 组（34 例）和 Non-APO 组（46 例）。比较两组子宫动脉频谱形态、血清 CCL19 水平，分析两者与 SLE 患者 APO 的相关性及预测效能。结果 与 Non-APO 组相比，APO 组子宫动脉频谱形态异常构成比、血清 CCL19 水平均升高（P < 0.05）。Logistic 风险回归分析显示，子宫动脉频谱形态、血清 CCL19 是 SLE 患者APO的危险因素（P < 0.05）。ROC曲线显示，血清CCL19、预测SLE患者APO的最佳截断点为335.89 ng/L，AUC 为 0.816；子宫动脉频谱形态异常预测 SLE 患者 APO 的 AUC 为 0.694；两者联合预测的 AUC 为 0.858。结论 子宫动脉频谱形态异常、血清 CCL19 水平 是SLE 患者APO 的危险因素，两者联合检测对其具有较高的预测效能。 

Objective To investigate the value of morphological abnormalities of uterine artery on Doppler ultrasound combined with serum CC chemokine ligand 19（CCL19） level in predicting adverse pregnancy outcomes （APO） in patients with systemic lupus erythematosus （SLE）. Methods Eighty pregnant women with SLE were selected from January 2015 to January 2022 in Qingdao Women and Children′s Hospital and divided into APO group （34 women） and non-APO group （46 women） according to their pregnancy outcomes. The morphology of uterine artery on Doppler ultrasound and serum CCL19 level were compared between the two groups. The correlation and predictive efficacy of the combined detection with APO in SLE patients were analyzed. Results As compared with non-APO group， APO group had an increase in morphological abnormalities of uterine artery and serum CCL19 level （P < 0.05）. Logistic risk regression showed that the morphology of uterine artery and serum CCL19 were risk factors for APO in patients with SLE （P < 0.05）. The ROC curve showed that the optimal cut-off point for serum CCL19 to predict APO in patients with SLE was 335.89 ng/L， and the area under curve （AUC） for predicting APO in patients with SLE was 0.816. The AUC for morphological abnormalities of uterine artery on Doppler ultrasound to predict APO in patients with SLE was 0.694. The AUC for the combined prediction was 0.858. Conclusions Morphological abnormalities of uterine artery on Doppler ultrasound and serum CCL19 level are risk factors for APO in patients with SLE， and their combined detection has a higher predictive efficacy.