Semi-quantitative magnetic resonance imaging scoring of the knee detects previous injuries in professional soccer players

Abstract

Purpose: The medical examination ("medical") is an important procedure in professional soccer since it has high economic relevance. In addition to clinical tests, magnetic resonance imaging (MRI) is used to assess joint health. In the present study, the reliability of semiquantitative knee MRI during the "medical" in professional soccer was tested, and its relationship with clinical data and days missed due to knee injury was observed.

Methods: In this cross-sectional study, between 2012 and 2019, 69 newly assigned players (age 18-35 years) from a professional soccer club underwent MRI (3.0 T) of both knee joints during their "medical". Reported knee injuries and previously missed days due to injury were obtained from player anamnesis and the "transfermarkt.com" database. Based on the established "Whole-Organ Magnetic Resonance Imaging Score" (WORMS), two independent radiologists graded the MRI results. Further evaluation was based on the mean score of both knees.

Results: The mean WORMS for all subjects was 13.9 (median 10.5, range 0-61). Players with previous injuries had significantly higher scores than players without reported injuries (22.1 ± 17.7 vs. 8.9 ± 4.4, p < 0.002). Three outliers (previously undetected injuries) in the group of players without reported injuries were observed (6.7%). The WORMS was significantly correlated with a prior knee injury (r: 0.424, p < 0.0001) and days missed due to injury (r: 0.489, p < 0.001). Age was correlated with the WORMS (r: 0.386, p < 0.001). In a linear regression model, prior injury was the only significant predictor of a high WORMS (p = 0.001). The WORMS was a significant predictor of days missed due to injury (p < 0.0002) and prior injury (sensitivity: 78%, specificity: 91%, p = 0.006). The intraclass correlation coefficient was excellent (0.89).

Conclusion: Semiquantitative knee MRI for WORMS determination during the soccer "medical" is a robust and reliable method. Prior injury, even in players without documented trauma, was detected by the WORMS, and previously missed days due to injury were correlated with the semiquantitative MR knee score.

Level of evidence: Level III.

Keywords: Knee MRI; Medical; Professional soccer; Whole-Organ Magnetic Resonance Imaging Score.

Fig. 1

WORMS, correlation with age and days missed due to injury. a The median WORMS was 11 (9) in all players, 14 (26) in players with prior injuries and 8 (8) in players without prior injuries. Whiskers represent the IQR. Three outliers (> Q₃ + 1.5xIQR) were identified in the group of noninjured players (indicated by solid squares, absolute scores: 33, 35 and 40). b The mean WORMS was significantly higher in previously injured players than in age-matched players without reported injuries (22.1 ± 17.7 vs. 8.9 ± 4.4, p < 0.002). Whiskers represent the SD. c The WORMS (all players) increased with age (r: 0.386, p < 0.001). Dashed lines represent the 95% CI. d The WORMS (all players) correlated with days missed due to injury (r: 0.489, p < 0.0001). Dashed lines represent the 95% CI. *Levels of significance

Fig. 2

MRI of three soccer players (no prior injuries, prior injuries, no reported prior injuries). a Left knee of a 19-year-old striker with no prior injuries. The WORMS was 2 for both knees; hence, the total WORMS was 2. b Left knee of a 29-year-old midfielder with documented prior injuries. Note the severe cartilage damage, osteophytes and bone marrow oedema. The WORMS was 111 for the left knee and 10 for the right knee; hence, the total WORMS was 61. c Left knee of a 23-year-old defender with no documented prior injuries. The MRI demonstrates advanced joint damage and prior ACL repair. The WORMS was 53 on the left side and 27 on the right. The total WORMS was 40. Cartilage damage: arrows, bone marrow oedema: asterisk, osteophytes: arrowheads, cysts: star, ACL repair: plus sign

Fig. 3

WORMS as a predictor for days missed due to injury and prior injury. a Multiple regression analysis was performed to test whether the WORMS, age and playing position significantly predicted days missed due to injury. The results of the regression indicated that the WORMS and age explained 38% of the variance [R² = 0.38, F(5, 63) = 7.71, p < 0.0001]. The WORMS was a significant predictor of days missed due to injury (β = 3.40, p < 0.0002), as was age (β = 7.49, p = 0.013). The playing position was not a significant predictor (p = 0.794). Dashed lines represent the 95% CI. b Multiple logistic regression with prior injury as the outcome variable and the WORMS, age and playing position as predictors. The model had a positive predictive value for prior injury of 82%. The negative predictive value was 88% (sensitivity: 78%, specificity: 91%). The WORMS was a significant predictor of prior injury (β = 0.15 [0.06–0.28], OR = 0.86 [0.75–0.93], p = 0.006), as was age (β = 0.31 [0.11–0.54], OR = 0.73 [0.57–0.89], p = 0.004). The playing position had no significant influence (p > 0.35). Solid symbols indicate correctly predicted cases