目的 构建超声实时引导的动静脉内瘘穿刺流程，为提高护理操作安全、质量及教学提供指导依据。方法 于2024年4月—6月，通过文献回顾、专家会议形成超声实时引导的动静脉内瘘穿刺流程的初稿；采用德尔菲法，确定超声实时引导的动静脉内瘘穿刺流程终稿；护士进行相应培训后在临床中应用该流程进行动静脉内瘘穿刺。 结果 共开展2轮专家函询，第1轮函询问卷有效回收率为94%（15/16），第2轮函询问卷有效回收率为100%（15/15）；第2轮专家权威系数为0.930，肯德尔和谐系数为0.140，变异系数为0~0.166。最终形成9个一级步骤和35个二级步骤的超声实时引导的动静脉内瘘穿刺流程，各级操作步骤的重要性评分均数为4.47~5.00分。10名护理人员应用该流程进行超声引导穿刺98例次，一次穿刺成功率为95%。 结论 构建的超声实时引导的动静脉内瘘穿刺流程具有科学性，在临床的初步应用证实了其可行性及实用性，为血液透析患者动静脉内瘘超声引导穿刺操作同质化提供参考依据。

Objective To establish a real-time ultrasound-guided arteriovenous fistula puncture process and to provide guidance for improving the safety and quality of nursing operations and teaching. Methods From April 2024 to June 2024，the first draft of ultrasound-guided arteriovenous fistula puncture procedure was prepared through literature review and group decision making. Delphi expert letter consultation was used to determine the final draft of ultrasound-guided arteriovenous fistula puncture process. After receiving corresponding training，nurses applied this process in clinical practice for puncture of internal venous fistulas. Results Two rounds of expert consultation were carried out. The effective recovery rate of the first round was 94%（15/16），and that of the second round was 100%（15/15）. The expert authority coefficient of the second round was 0.930， the Kendall harmony coefficient was 0.140 and the coefficient of variation was 0~0.166. In the end，an ultrasound-guided arteriovenous fistula puncture process with nine primary steps and 35 secondary steps was formed. The average importance score of each step was 4.47~5.00. Ten nurses used this procedure to perform ultrasound-guided puncture 98 times，and the success rate of one puncture was 95%. Conclusion The established process of ultrasound-guided arteriovenous fistula puncture is scientific，and its feasibility and practicability have been confirmed in the preliminary clinical application，which provides a reference for the homogenization of ultrasound-guided puncture operation of arteriovenous fistula in hemodialysis patients.