Reliability of MRI assessment of acute musculotendinous groin injuries in athletes

Andreas Serner, Frank W. Roemer, Per Hölmich, Kristian Thorborg, Jingbo Niu, Adam Weir, Johannes L. Tol, Ali Guermazi

To describe a multi-dimensional MRI assessment approach with a focus on acute musculotendinous groin lesions, and to evaluate scoring reproducibility. Male athletes who participated in competitive sports and presented within 7 days of an acute onset of sports-related groin pain were included. All athletes underwent MRI (1.5 T) according to a standardized groin-centred protocol. From several calibration sessions, a system was developed assessing grade, location and extent of muscle strains, peri-lesional haematoma, as well as other non-acute findings commonly associated with long-standing groin pain. Kappa (K) statistics and intraclass correlation coefficients (ICCs) were used to describe intra- and inter-rater reproducibility. Seventy-five athletes (mean age 26.6 ± 4.4 years) were included in the analyses, and 85 different acute lesions were observed. Adductor longus lesions were most common (42.7 %) followed by rectus femoris lesions (16.3 %). Kappa values ranged between 0.70 and 1.00 for almost all categorical features for acute lesions, with almost perfect intra- and inter-rater agreement (K = 0.89-1.00) for presence, number, location and grading of lesions. ICCs ranged between 0.77 and 1.00 for continuous measures of acute lesion extent. A standardized MRI assessment approach of acute groin injuries was described and showed good intra- and inter-rater reproducibility. • A multidimensional MRI assessment approach for acute groin injuries was described. • Standardized MRI assessment of acute musculotendinous groin injuries has high reproducibility. • Injury location and injury extent can be scored reliably using 1.5 T MRI.