To estimate the sensitivity and specificity of computed tomography used for the detection of extranodal spread of metastatic head and neck squamous cell carcinoma, by experienced head and neck radiologists.

Participants had undergone a neck dissection for head and neck squamous cell carcinoma, together with computed tomography scanning prior to surgery (accessible for reporting). Computed tomography images were independently examined by two experienced head and neck radiologists. Nodal involvement by squamous cell carcinoma and the presence or absence of extranodal spread were recorded. Results were compared to the histological specimen. The sensitivity, specificity and positive predictive value of using computed tomography for the detection of nodal involvement and presence or absence of extranodal spread were estimated, and 95 per cent confidence intervals were calculated.

The study analysed 149 neck dissections. When using computed tomography to detect the extranodal spread of head and neck squamous cell carcinoma, radiologists A and B had sensitivities of 66 and 80 per cent, specificities of 91 and 90 per cent, and positive predictive values of 85 and 87 per cent, respectively.

The sensitivity and specificity of radiological detection of extranodal spread from head and neck squamous cell carcinoma is not well reported in the literature. Accuracy of reporting improves in the hands of experienced head and neck radiologists. This finding has clinical implications for surgical planning and adjuvant therapy requirements.