Surgeon's treatment outcome prediction: accuracy and communication

Achieving clear surgical margins is essential for treating soft tissue sarcomas (STS), as expected margins influence both prognosis and treatment decisions. When prognostic tools such as the Personalized Sarcoma Care (PERSARC) app are used for treatment planning, a preoperative estimate of expected surgical margins is required. The accuracy of these margin predictions and how margin uncertainty is communicated remain unclear.

We conducted a retrospective secondary analysis using PERSARC app data, electronic patient records, and audio-recorded treatment decision-making consultations. Preoperative margin predictions were compared with postoperative histopathology in 134 patients. Communication of uncertainty was analyzed in a subset of 37 consultations.

Predicted and observed margins differed for 42/134 patients, resulting in a sensitivity of 2.3% for predicting positive margins. Prediction accuracy was not associated with patient and tumor characteristics. Inaccurate margin predictions substantially altered PERSARC-based estimations for overall survival, local recurrence, and distant metastasis. During consultations, the possibility of positive margins was discussed with 22/37 patients. When discussed, surgeons frequently expressed high confidence in achieving negative margins. Communication of margin uncertainty did not differ between patients with negative versus positive postoperative margins.

Preoperative margin prediction in STS patients shows low accuracy for positive margins and is often accompanied by high expressed confidence and limited discussion of uncertainty. Treatment decisions were unlikely to be affected in this cohort, however inaccurate margin predictions may affect prognostic counseling. Increasing awareness of prediction limitations and transparent communication of uncertainty may support more realistic patient expectations and better informed decision making.