目的：评价甲状腺结节的前后径与横径比值（A/T）≥1对其良恶性的诊断价值。方法：检索多个国内外数据库，收集2017年7月前发表的有关评价甲状腺结节A/T≥1对于恶性甲状腺结节诊断价值的诊断性研究。按照纳入标准筛选文献、提取资料和质量评价后，用Meta-Disc 1.4统计软件对数据进行分析。对纳入文献间的异质性进行评估，计算纳入研究中甲状腺结节A/T≥1诊断其为恶性的合并效应量，包括灵敏度、特异度、阳性似然比（PLR）、阴性似然比（NLR）、诊断比值比（DOR）、综合受试者工作特征曲线下面积（AUC）与Q*指数；采用敏感性分析评估研究结果的稳定性，并寻找研究间异质性来源。绘制Deek漏斗图评估研究结果所致的发表偏倚。结果：共纳入13篇研究，总计11 243枚结节，其中良性结节9 227枚，恶性结节2 016枚。13篇文献评估后质量高低不一，纳入研究间存在非阈值效应引起的异质性（r=0.352，P=0.239）。甲状腺结节A/T≥1诊断其为恶性的合并灵敏度为0.31（95% CI=0.29~0.33）、特异度为0.50（95% CI= 0.49~0.51）、PLR为1.32（95% CI=0.81~2.14）、NLR为0.85（95% CI=0.73~0.99）、DOR为1.51（95% CI= 0.77~2.90）、AUC为0.538 9、Q*值为0.529 2。纳入研究总体合并效应量结果稳定性好，异质性来源可能为研究方法，种族差异，是否采用盲法等，Deek漏斗图显示研究结果呈不对称分布，存在纳入研究结果所致的发表偏倚。结论：甲状腺结节A/T≥1对甲状腺结节的良恶性具有一定的诊断价值；但敏感度、特异度较低，需结合其他影像学征象综合判断。
Value of anteroposterior to transverse diameter ratio of the thyroid nodule in diagnosis of its benign and malignant nature: a Meta-analysis
Objective: To evaluate the value of anteroposterior to transverse diameter ratio (A/T) of the thyroid nodule greater than or equal to 1 in diagnosis of its benign and malignant nature. Methods: The studies published before July 2017 evaluating the value of thyroid nodule’s A/T≥1 in diagnosis of malignant lesion were collected through searching several national and international databases. After screening for inclusion criteria, data extraction and quality assessment, the data were analyzed with Meta-DiSc 1.4 statistical software. The heterogeneity among the included studies was assessed, and the pooled effect estimates of thyroid nodule’s A/T≥1 in diagnosing its malignant nature in the included studies were calculated, which included the sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR), diagnostic odds ratio (DOR), the area under the curve (AUC) of summary receive operating characteristic curve and Q* index. The stability of results of the overall pooled estimates of the included studies was validated by sensitivity analysis, and the sources of heterogeneity among the included studies were analyzed. The publication bias caused by the results of the included studies was evaluated by drawing the Deek’s funnel plot. Results: A total of 13 studies were included, involving 11 243 nodules that included 9 227 benign nodules and 2 016 malignant nodules. The quality of 13 studies was uneven, and there was a heterogeneity caused by non-threshold effect among the included studies (r=0.352, P=0.239). The overall pooled sensitivity, specificity, PLR, NLR, DOR, AUC and Q* index of the included studies of thyroid nodule’s A/T≥1 for diagnosing malignant nodule were 0.31 (95% CI=0.29–0.33), 0.50 (95% CI=0.49–0.51), 1.32 (95% CI=0.81–2.14), 0.85 (95% CI= 0.73–0.99), 1.51 (95% CI=0.77–2.90), 0.538 9 and 0.529 2, respectively. The results of the overall pooled estimates of included studies were stable, and the research methods, racial differences and whether by blind design might be sources of the heterogeneity of the studies. The Deek’s funnel plot showed an asymmetrical distribution of the results, indicating that there was a publication bias caused by the results of the included studies. Conclusion: Thyroid nodule’s A/T≥1 has certain value in differential diagnosis of benign and malignant thyroid nodules. However, its sensitivity and specificity are low, so combined judgment with other imaging signs is necessary.