Radioisotope scanning is important to diagnose subacute thyroiditis (SAT), but it’s not always available. So we aim to establish a diagnostic scale for SAT without radioisotope scanning. The suspected SAT patients hospitalized in Yuebei people’s Hospital from January 2012 to December 2016 were selected and divided into study group and control group according to whether they were diagnosed as SAT. The clinical indexes of two groups were collected and the diagnostic scale of SAT was established by using binary logistic regression analysis. The effectiveness of the scale was evaluated by ROC curve. Of 309 patients, 58.25% of patients were confirmed with SAT and the remaining 41.75% of patients were not diagnosed with SAT. After univariate analysis, variables which were considered statistically different(P < 0. 05) between the two groups were selected as independent variables and the diagnosis of SAT was taken as dependent variable in the binary logistic regression model. The logistic regression model consisted of 4 variables, each was thyroid tenderness, firm on palpation, increased ESR and elevated thyroid hormone level. The P value of Omnibus tests was≤0. 001 and the Nagelkerke R Square was 0. 915. The diagnostic scoring scale was established with these four variables according to their regression coefficient. The area under the ROC curve for this diagnostic scale was 0. 991(95% confidence interval, 0. 982–0.999). The highest Youden index was 0. 912, the corresponding cut-off point was 7. Internally validation shows a sensitivity of 92. 78% and a specificity of 98.45% of our scale. We established and validated a diagnostic scale for SAT without the need for radioisotope scanning for the first time. It has good application in institutions that do not have radioisotope machines or among pregnant and lactating women.

中文翻译：

---

没有放射性同位素扫描即可建立亚急性甲状腺炎的诊断量表。

放射性同位素扫描对于诊断亚急性甲状腺炎（SAT）很重要，但并非总是可用。因此，我们旨在建立无需进行放射性同位素扫描的SAT诊断量表。根据2012年1月至2016年12月在岳北市人民医院住院的疑似SAT患者，根据是否被确诊为SAT，分为研究组和对照组。收集两组患者的临床指标，采用二元logistic回归分析法建立SAT的诊断量表。通过ROC曲线评估量表的有效性。在309名患者中，有58.25％的患者被确诊为SAT，其余41.75％的患者未被诊断为SAT。单变量分析后，被认为具有统计学差异的变量（P <0。05）选择两组之间的自变量，在二元logistic回归模型中将SAT的诊断作为因变量。Logistic回归模型由4个变量组成，每个变量为甲状腺压痛，触诊牢固，ESR增加和甲状腺激素水平升高。综合试验的P值≤0。001和Nagelkerke R Square为0.915。根据这四个变量的回归系数，建立了诊断评分量表。对于该诊断量表，ROC曲线下的面积为0. 991（95％置信区间为0. 982-0.999）。最高的尤登指数是0. 912，相应的临界点是7。内部验证显示我们的量表的敏感性为92. 78％，特异性为98.45％。我们首次建立并验证了SAT诊断量表，而无需进行放射性同位素扫描。它在没有放射性同位素机器的机构或孕妇和哺乳期妇女中有良好的应用。