Evaluation of a trainer phantom in the learning phase of sentinel lymph node identification in breast cancer

The concept of a learning phase is difficult to implement in a university setting, as it is unacceptable to subject a patient who requires only lymphadenectomy to axillary dissection for the purpose of training surgeons. We therefore sought to evaluate intraoperative sentinel node detection using a phantom, the Senti-Trainer. Learning phases on the Senti-Trainer and detection rate were assessed in order to determine whether the proficiency of surgeons in training improved with the number of procedures.

Twenty residents each performed 30 detection procedures of a sentinel node on the silicon phantom. Each resident was evaluated at each procedure, and an observation report was made every five procedures. Evaluation was single-blind as the surgeons did not know the result of the previous detection and were aware of the results only after the thirtieth procedure.

The mean detection rate was 75% during the first procedure and reached 95% (or 5% detection errors) at the 30th procedure (p<0.0001; OR=6.33 with a 95% CI=[2.31; 17.33]). Proficiency in sentinel lymph node (SLN) identification also increased with the number of procedures performed. The ability to localize SLN improved during the learning phase with the increasing number of procedures performed. Mean detection time during the 30 procedures was 150 s (range: 115-210 s).

Training on a phantom showed that this is a valuable teaching tool that enables surgeons to become familiar with gamma probes. It cannot replace the clinical training phase, but is an important aid to proficiency in intraoperative detection.