采用荟萃分析方法评价超声(US)对卵巢交界性肿瘤(BOT)的诊断价值。

检索万方、中国知网、维普、PubMed、Highwire、Springer、EBSCO等中英文数据库，检索1990年1月至2023年3月发表的文献。选择符合标准的4篇文献采用诊断准确性研究质量评估(QUADAS2)标准评价。采用Meta-Disc 1.4软件进行荟萃分析，评价US对BOT的诊断价值。

共纳入4篇文献和244例患者。合并敏感度、合并阴性似然比不存在异质性，采用固定效应模型，分别为0.65(95%CI：0.59～0.72)和0.43(95%CI：0.36～0.51)。合并特异度、合并阳性似然比、合并诊断比值比存在异质性，采用随机效应模型，分别为0.85(95%CI：0.83～0.87)、4.58(95%CI：2.35～8.93)和10.85(95%CI：4.85～24.31)。拟合受试者工作特征曲线(ROC)曲线下面积为0.687。

US单独应用对BOT具有一定诊断价值，但合并敏感度较低，单独用于BOT的诊断具有一定局限性，临床应结合其他指标联合诊断以提高诊断价值。

To evaluate the diagnostic value of ultrasound (US) in borderline ovarian tumor (BOT) by meta-analysis.

Wanfang, CNKI, CQVIP, PubMed, HighWire, Springer, EBSCO and other Chinese and English databases were searched for literatures published from January 1990 to March 2023. Four articles meeting the criteria were evaluated by quality assessment of diagnostic accuracy studies-2 (QUADAS2) standard, and Meta-disc 1.4 software was used for meta-analysis to evaluate the diagnostic value of US for BOT.

Four articles and 244 patients were analyzed. There was no heterogeneity in the combined sensitivity and the combined negative likelihood ratio. The fixed effect models were used for analysis, and the values were 0.66 (95%CI: 0.59-0.72) and 0.43 (95%CI: 0.36-0.51), respectively. There was heterogeneity in the combined specificity, positive likelihood ratio and diagnostic odds ratio. The random effect models were applied, and the values were 0.85 (95%CI: 0.83-0.87), 4.58 (95%CI: 2.35-8.93) and 10.85 (95%CI: 4.85-24.31), respectively. The area under the curve of fitting receiver operator characteristic (ROC) was 0.687.

The application of US alone has certain diagnostic value for BOT, but the combined sensitivity is low. Utilizing US alone for BOT diagnosis has certain constraints, suggesting that clinical diagnosis should incorporate other indicators to enhance diagnostic accuracy.