The Hawkins Sign of the Talus: The Impact of Patient Factors on Prediction Accuracy.

BACKGROUND: Osteonecrosis is a complication of talar neck fractures associated with chronic pain and poor functional outcomes. The Hawkins sign, the radiographic presence of subchondral lucency seen in the talar dome 6 to 8 weeks after trauma, is a strong predictor of preserved talar vascularity. This study sought to assess the accuracy of the Hawkins sign in a contemporary cohort and assess factors associated with inaccuracy.

METHODS: A retrospective review of talar neck fractures at a level-I trauma center from 2008 to 2016 was conducted. Both the Hawkins sign and osteonecrosis were evaluated on radiographs. The Hawkins sign was determined on the basis of radiographs taken approximately 6 to 8 weeks after injury, whereas osteonecrosis was determined based on radiographs taken throughout follow-up. The Hawkins sign accuracy was assessed using proportions with 95% confidence intervals (CIs), and associations were examined with Fisher exact testing.

RESULTS: In total, 105 talar neck fractures were identified. The Hawkins sign was observed in 21 tali, 3 (14% [95% CI, 3% to 36%]) of which later developed osteonecrosis. In the remaining 84 tali without a Hawkins sign, 32 (38% [95% CI, 28% to 49%]) developed osteonecrosis. Of the 3 tali that developed osteonecrosis following observation of the Hawkins sign, all were in patients who smoked.

CONCLUSIONS: A positive Hawkins sign may not be a reliable predictor of preserved talar vascularity in all patients. We identified 3 patients with a positive Hawkins sign who developed osteonecrosis, all of whom were smokers. Factors impairing the restoration of microvascular blood supply to the talus may lead to osteonecrosis despite the presence of preserved macrovascular blood flow and an observed Hawkins sign. Further research is needed to understand the factors limiting Hawkins sign accuracy.

LEVEL OF EVIDENCE: Prognostic Level III. See Instructions for Authors for a complete description of levels of evidence.