Bone bruise vs. non-displaced fracture on MRI: a novel grading system for predicting return-to-play.

Objective: To devise an MRI grading scheme for osseous contusion patterns in elite hockey players for predicting return-to-play (RTP).

Methods: A retrospective review was performed to identify traumatic lower extremity osseous injuries in professional hockey players. A total of 28 injuries (17 players) were identified over a 10-year period. All had MRIs acquired at ≥ 1.5 T within a mean interval of 2 days from initial injury. MRIs were retrospectively reviewed by 3 musculoskeletal radiologists for osseous contusion pattern, classified as grade 1 (mild), 2 (moderate), or 3 (severe). Grade 3 contusions were further subdivided by the presence or absence of fracture, defined as discrete cortical disruption on MRI or follow-up CT. RTP was calculated from date of injury to next game played based on game log data. Statistical analysis was performed using ANOVA and post hoc unpaired t test.

Results: Mean RTP for grade 1, 2, and 3 injuries was 2.8, 4.5, and 20.3 days, respectively. Grade 3 injuries without and with cortical fractures had mean RTP of 18.3 and 21.4 days, respectively. ANOVA analysis between groups achieved statistical significance (p < 0.001). Post hoc t test demonstrated statistically significant differences between grade 3 and grades 1 (p < 0.001) and 2 (p < 0.001) injuries. There was no statistical difference in RTP between grade 3 subgroups without and with fracture (p = 0.327).

Conclusion: We propose a novel MRI grading system for assessing severity of osseous contusions and predicting RTP. Clinically, there was no statistically significant difference in RTP between severe osseous contusions and nondisplaced fractures in elite hockey players.