Surgeons Cannot Predict Pilon Fracture Outcomes Based on Initial Radiographs.

ABSTRACT

This study sought to determine (1) whether surgeons can accurately predict functional outcomes of operative fixation of pilon fractures based on injury and initial postoperative radiographs, (2) whether the surgeon's level of experience is associated with the ability to successfully predict outcome, and (3) the association between patients' demographic and clinical characteristics and surgeons' prediction scores. A blinded, randomized provider survey was conducted at a level I trauma center. Seven fellowship-trained orthopedic traumatologists and 4 orthopedic trauma fellows who were blinded to outcome reviewed data regarding 95 pilon fractures in random order. Injury ankle radiographs, initial postoperative fixation radiographs, and brief patient histories were assessed. Midterm follow-up functional outcome scores obtained a mean 4.9 years after surgery were available for all patients. Main outcome measures were Pearson correlation coefficient-assessed functional outcomes and surgeon-predicted outcomes. A mixed-effect model determined the association between patients' characteristics and surgeons' prediction scores. Minimal positive correlation was observed between functional outcomes and prediction scores. No difference was noted between the attending and fellow groups in prediction ability. When surgeons' prediction confidence level was greater than 1 SD above the mean confidence level, correlation between functional outcome and prediction improved, although poor correlation was still observed. AO/OTA type 43C fractures, high-energy mechanisms, and older patient age were characteristics associated with lower prediction scores. Surgeons had poor ability to predict functional outcomes of patients with pilon fractures based on injury and initial postoperative radiographs, and level of experience was not associated with ability to predict outcome. [Orthopedics. 2020; 43(1) e43-e46.].