Section of Otolaryngology, Department of Surgery, University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin, USA.
Wisconsin Surgical Outcomes Research Program, Department of Surgery, University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin, USA.
Thyroid. 2023 Dec;33(12):1434-1440. doi: 10.1089/thy.2023.0234.
The use of thyroid ultrasound increases yearly, adding to costs and overdetection of clinically irrelevant nodules. We investigated which indications most commonly prompt referral for thyroid ultrasound and the diagnostic utility by indication. We performed a retrospective observational cohort study of adults (≥18 years) undergoing an initial dedicated thyroid ultrasound between 2017 and 2019 at a tertiary academic center. Indicated reasons for referral were categorized into suspected palpable nodule (SPN), compressive symptoms (CS), metabolic symptoms (MS), screening due to high-risk factors, follow-up of incidental finding on other imaging, and combination of factors. Percentage of ultrasounds with an identifiable nodule and with a nodule recommended for biopsy was compared by indication. Separate logistic regression models were used to identify factors associated with finding any nodule and a biopsy-recommended nodule. Among the 1739 patients included, the most common indication for thyroid ultrasound was SPN (40%), followed by incidental imaging (28%), CS (13%), combination (11%), MS (6%), and high-risk factors (2%). Overall, 62% of ultrasounds identified a nodule. Ultrasounds performed for incidental findings had the highest rate of nodule identification (94%), compared with 55%, 39%, and 43%, for SPN, CS, and MS, respectively ( < 0.05). Only 27% of ultrasounds identified a biopsy-recommended nodule. Nodules found incidentally had the highest rate of biopsy-recommended nodules at 55%. Rates of biopsy-recommended nodules for SPN, CS, and MS were 21%, 6%, and 10%, respectively. Logistic regression demonstrated that compared with patients referred for an SPN, those with incidental nodules were 10 times more likely to have a nodule found on ultrasound (odds ratio [OR] = 10.6 [CI 7.0-16.0]), while those referred for CS were half as likely to have a nodule (OR = 0.5 [CI 0.4-0.7]). Similar factors were associated with identification of biopsy-recommended nodules. Of all new dedicated thyroid ultrasounds, only a quarter find biopsy-recommended nodules, and nearly 40% do not identify a nodule at all. Notably, only 55% of ultrasounds done for SPN found a nodule. Ultrasound for CS and MS had the lowest rates of detecting nodules. Providing clear guidance on when to order thyroid ultrasounds can help reduce unnecessary health care utilization and potential overtreatment.
甲状腺超声的应用逐年增加,增加了成本,并过度检测到临床无关的结节。我们研究了哪些指征最常促使进行甲状腺超声检查,以及各指征的诊断效用。我们对 2017 年至 2019 年在一家三级学术中心接受初次甲状腺超声检查的成年人(≥18 岁)进行了回顾性观察性队列研究。指征性转诊原因分为可疑可触及结节(SPN)、压迫症状(CS)、代谢症状(MS)、高危险因素筛查、其他影像学检查偶然发现的随访以及多种因素。通过指征比较有可识别结节和推荐进行活检的结节的超声检查比例。使用单独的逻辑回归模型确定与发现任何结节和推荐活检的结节相关的因素。在纳入的 1739 例患者中,甲状腺超声检查最常见的指征是 SPN(40%),其次是偶然发现的影像学检查(28%)、CS(13%)、组合(11%)、MS(6%)和高危险因素(2%)。总体而言,62%的超声检查发现结节。与 SPN(55%)、CS(39%)和 MS(43%)相比,进行偶然发现的超声检查有最高的结节检出率(94%)。仅 27%的超声检查发现推荐活检的结节。偶然发现的结节有最高的推荐活检结节率,为 55%。SPN、CS 和 MS 的推荐活检结节率分别为 21%、6%和 10%。逻辑回归表明,与因 SPN 就诊的患者相比,偶然发现结节的患者在超声检查中发现结节的可能性高 10 倍(优势比[OR] = 10.6[CI 7.0-16.0]),而因 CS 就诊的患者发现结节的可能性低一半(OR = 0.5 [CI 0.4-0.7])。类似的因素与发现推荐活检的结节相关。在所有新的甲状腺专用超声检查中,只有四分之一发现推荐活检的结节,近 40%的超声检查根本没有发现结节。值得注意的是,只有 55%的 SPN 超声检查发现结节。CS 和 MS 的超声检查发现结节的比例最低。提供关于何时开具甲状腺超声检查的明确指导,可以帮助减少不必要的医疗保健利用和潜在的过度治疗。