Unstimulated interferon gamma may be a useful pleural fluid biomarker in the diagnosis of tuberculous pleural effusion (TPE). However, the exact threshold of pleural fluid interferon gamma and its accuracy during routine clinical decision-making is not clear. We assessed the performance of pleural fluid interferon gamma in diagnosing TPE and tried to identify a useful assay threshold. We queried the PubMed and Embase databases for publications indexed until May 2020 that provided both sensitivity and specificity data on unstimulated pleural fluid interferon gamma for diagnosis of TPE. A bivariate random effects model was employed to compute summary estimates for diagnostic accuracy parameters, both overall as well as at threshold ranges of <2, 2 to 5, and >5 IU/ml. We retrieved 2,048 citations, of which 67 publications (7,153 patients) were assessed in our review. The summary estimates for sensitivity, specificity, and diagnostic odds ratio were 0.93 (95% confidence interval [CI], 0.91 to 0.95), 0.96 (95% CI, 0.94 to 0.97), and 310.72 (95% CI, 185.24 to 521.18), respectively. Increasing interferon gamma thresholds did not translate into any substantial change in diagnostic performance; however, eight studies using thresholds of >5 IU/ml showed poorer diagnostic accuracy estimates than other studies with lower thresholds. None of the prespecified subgroup variables significantly influenced relative diagnostic odds ratios in a multivariate meta-regression model. All publications demonstrated a high risk of bias. Unstimulated pleural fluid interferon gamma level provides excellent accuracy for diagnosing TPE and has the potential of becoming a first-line test for this purpose.

中文翻译：

---

用于诊断结核性胸腔积液的非刺激性胸腔积液干扰素伽马：系统评价和荟萃分析

未受刺激的干扰素 γ 可能是诊断结核性胸腔积液 (TPE) 的有用胸水生物标志物。然而，胸水干扰素γ的确切阈值及其在常规临床决策中的准确性尚不清楚。我们评估了胸腔积液干扰素 γ 在诊断 TPE 方面的性能，并试图确定一个有用的检测阈值。我们在 PubMed 和 Embase 数据库中查询了索引到 2020 年 5 月的出版物，这些出版物提供了用于诊断 TPE 的未受刺激的胸水干扰素 γ 的敏感性和特异性数据。采用双变量随机效应模型计算诊断准确性参数的汇总估计值，包括总体以及 <2、2 至 5 和 >5 IU/ml 的阈值范围。我们检索了 2,048 次引文，其中 67 篇出版物（7, 153 名患者）在我们的审查中进行了评估。敏感性、特异性和诊断优势比的汇总估计值为 0.93（95% CI，0.91 至 0.95）、0.96（95% CI，0.94 至 0.97）和 310.72（95% CI，185.24 至 521.18） ， 分别。增加干扰素伽马阈值并没有转化为诊断性能的任何实质性变化；然而，八项阈值大于 5 IU/ml 的研究显示出比其他阈值较低的研究更差的诊断准确性估计。在多元元回归模型中，没有一个预先指定的亚组变量显着影响相对诊断优势比。所有出版物都显示出高偏倚风险。