医学
抽吸
内镜超声
活检
多中心研究
随机对照试验
放射科
超声波
外科
机械工程
工程类
作者
Shenglin Xu,Jianian Guo,Mengbin Qin,Yiteng Meng,F. Xie,Weiguang Qiao,Haiyan Hu,Peng Peng,Jahan Rownoak,S. H. Heng,Finang Ung,Yaping Ye,Jing Wang,Weixin Li,Yingying Zou,Li Zou,Shao Hui Huang,Side Liu,Junfen Wang,Jun Yao
标识
DOI:10.14309/ajg.0000000000003389
摘要
Conclusions regarding the suction techniques of EUS-FNB remain controversial. This study aimed to compare the diagnostic accuracy of the dry suction versus wet suction technique in solid pancreatic lesions (SPLs) and determine the optimal number of passes for EUS-FNB. This investigation was conducted as a multicenter, randomized, controlled, non-inferiority trial. Patients with SPLs were randomly allocated to receive either the dry or wet suction technique. The primary outcome was diagnostic accuracy. The secondary outcomes included sensitivity, specificity, optimal number of needle passes, specimen quality, procedure time, and adverse events. Of the 200 patients, 193 were included in the final analysis, with 96 in the dry suction group and 97 in the wet suction group. The diagnostic accuracies were 97.92% and 96.91% in the dry and wet groups, respectively, with a 1.01% difference between the study groups (two-sided 95% CI, -3.47% to 5.48%, P=0.659). The overall adverse event rate was 2.6%. No significant differences were observed in sample adequacy (98.9% vs. 98.9%, P = 1) or blood contamination (P = 0.796). Regarding procedure time, there was no statistical difference (18.68±8.03 min vs. 19.36±8.89 min, P=0.626); however, more procedural steps were required in the wet suction technique. No significant difference was found between the cumulative diagnostic accuracy of each needle (1st pass 93.78% vs. 2nd pass 95.34% vs. 3rd pass 97.41%, P = 0.225). The dry suction technique is non-inferior to the wet suction technique for EUS-FNB in SPLs. In the absence of rapid on-site evaluation (ROSE), only one pass was required to achieve more than 90% diagnostic accuracy. (ClinicalTrial.gov number NCT05549856.).
科研通智能强力驱动
Strongly Powered by AbleSci AI