Zhu Rui, Shagjaa Tungalagtamir, Rossitto Giacomo, Caroccia Brasilina, Seccia Teresa Maria, Gregori Dario, Rossi Gian Paolo
Internal & Emergency Medicine Unit, Department of Medicine - DIMED, University of Padua, 35128 Padua, Italy.
Department of Endocrinology, Sichuan Academy of Medical Sciences & Sichuan Provincial People's Hospital, 610072 Chengdu, China.
J Clin Endocrinol Metab. 2023 Jan 17;108(2):496-506. doi: 10.1210/clinem/dgac654.
Determining the diagnostic accuracy of "exclusion" tests for primary aldosteronism (PA) compared to the aldosterone to renin ratio (ARR) is fundamental to avoid invasive subtyping in false-positive patients at screening.
To assess the accuracy of exclusion tests for PA using the diagnosis of unilateral PA as reference.
PubMed, EMBASE, Web of Science, and Cochrane Library databases were searched for studies published from January 1, 1970, to December 31, 2021, meeting tight quality criteria. Data were extracted following the PRISMA methodology. We performed a two-stage meta-analysis that entailed an exploratory and a validation phase based on a "golden" or "gold" diagnostic standard, respectively. Pooled specificity, negative likelihood ratio, diagnostic odds ratio, and summary area under the ROC curve (sAUROC) were calculated to analyze the accuracy of exclusion tests.
A meta-analysis of 31 datasets comprising a total of 4242 patients fulfilling the predefined inclusion criteria found that pooled accuracy estimates (sAUROC) did not differ between the ARR (0.95; 95% CI, 0.92-0.98), the captopril challenge test (CCT) (0.92; 95% CI, 0.88-0.97), and the saline infusion test (SIT) (0.96; 95% CI, 0.94-0.99). Solid information could not be obtained for the fludrocortisone suppression test and the furosemide upright test, which were assessed in only 1 study each.
The apparently high diagnostic accuracy of the CCT and the SIT was due to the selection of patients with an elevated ARR and thus a high pretest probability of unilateral PA; however, neither test furnished a diagnostic gain over the ARR. Therefore, the systematic use of these exclusion tests in clinical practice is not justified by available evidence.
与醛固酮肾素比值（ARR）相比，确定原发性醛固酮增多症（PA）“排除”试验的诊断准确性对于避免在筛查时对假阳性患者进行侵入性亚型分类至关重要。
以单侧PA的诊断为参考，评估PA排除试验的准确性。
检索PubMed、EMBASE、Web of Science和Cochrane图书馆数据库，查找1970年1月1日至2021年12月31日发表的符合严格质量标准的研究。按照PRISMA方法提取数据。我们进行了两阶段的荟萃分析，分别基于“金标准”或“黄金”诊断标准进行探索性和验证性阶段。计算合并特异性、阴性似然比、诊断比值比和ROC曲线下的汇总面积（sAUROC），以分析排除试验的准确性。
对31个数据集进行的荟萃分析，共纳入4242例符合预定义纳入标准的患者，发现ARR（0.95；95%CI，0.92 - 0.98）、卡托普利激发试验（CCT）（0.92；95%CI，0.88 - 0.97）和生理盐水输注试验（SIT）（0.96；95%CI，0.94 - 0.99）之间的合并准确性估计值（sAUROC）无差异。对于氟氢可的松抑制试验和速尿立位试验，由于每项仅在1项研究中进行评估，无法获得可靠信息。
CCT和SIT明显较高的诊断准确性是由于选择了ARR升高的患者，因此单侧PA的验前概率较高；然而，这两种试验均未比ARR提供诊断优势。因此，现有证据无法证明在临床实践中系统使用这些排除试验是合理的。
