Objective  To investigate the current status and diagnostic efficacy of preoperative ultrasonography (US) in detecting central lymph node metastasis (CLNM) from papillary thyroid carcinoma (PTC).

  Methods  Data of preoperative US findings and surgical pathology of PTC patients were collected from 13 Grade-A tertiary hospitals, and a retrospective analysis of preoperative US assessment for CLNM was conducted. The patients were divided into the metastasis and non-metastasis groups according to postoperative pathological diagnosis of CLNM. Two criteria were used to evaluate the diagnostic efficacy. One was based on the hypothesis that the US findings of the central lymph node matched the diagnostic criteria for metastatic lymph nodes; the other was based on the hypothesis that the US findings detected typical CLNM characteristics.

  Results  The data of a total of 1, 016 patients were collected. Those who did not fulfill the criteria were excluded, resulting in 925 cases with preoperative US data of the central lymph nodes. Of the 925 patients, 825 underwent central lymph node dissection. Of the included cases, 384 were metastatic and 441 were non-metastatic. When the detection of the central lymph node was used as the diagnostic criteria, the detection rate of CLNM was 18.9% (3.0%-50.0%). The sensitivity, specificity, and accuracy for CLNM detection were 33.59%, 89.11%, and 63.27%, respectively. When the detection of the typical CLNM characteristics was used as the diagnostic criteria, the detection rate was 8.2% (0-33.0%). The sensitivity, specificity, and accuracy for CLNM detection were 17.97%, 97.96%, and 60.73%, respectively.

  Conclusions  The detection rate and sensitivity of US for the diagnosis of CLNM are low, indicating that the 2012 domestic guideline is in line with the current evaluation condition, which recommends prophylactic central neck dissection under the premise of function preservation.