Observer error in the assessment of nodal disease in head and neck cancer.

Abstract

There is no previously published information on clinicians' abilities to accurately differentiate between different stages of node positive disease in head and neck cancer. Forty-two surgeons examined standardized nodes in a model neck and estimated nodal size. Each recorded their confidence in their ability to perform the task using a visual analogue scale. Reference nodes of known size were provided for comparison during a second examination of each node. The study was repeated after 1 month. Accuracy was poor and was not dependent on experience or confidence. There was a tendency to underestimate the size of smaller nodes. Estimates were strongly influenced by volume independent of largest diameter (p <.001). Reference nodes aided accuracy (p =.031). Subjects were not consistent on repeated testing (p <.001). Both trainees and specialists are poor at accurately staging nodal disease using palpation alone.