Purpose of the study: Early and correct diagnosis of meniscal injuries in children is essential to preserve the meniscus and knee function. Often isolated, these lesions may be difficult to identify. The diagnostic procedure involves an excess of diagnostic arthroscopies and multiple series of magnetic resonance imaging (MRI) procedures performed by radiologists not specialised in paediatric diagnosis. The purpose of this study was to determine the concordance between MRI and arthroscopic findings in children with meniscal lesions.

Material and methods: This was a retrospective analysis of MRI series then arthroscopy performed on 96 knees in a paediatric population (age 9 – 17 years) explored between 1995 and 2008 for suspected traumatic meniscal injury. Discoid mensci were excluded. All arthroscopic procedures were reviewed by one operator and MRI by one radiologist familiar with paediatric pathology. All files were complete. Agreement with the arthroscopic findings (gold standard) was determined by the presence of a meniscal lesion, its localisation, and its type.

Results: Agreement reached 72% for presence or not of a lesion. It was 55% for localisation and 45% for type. In this context, the sensitivity of MRI was 85% with 42% specificity. There were 11 false positives and 7 false negatives.

Discussion: These figures are disappointing compared with those reported in the adult population (sensitivity 92% and specificity 87%) and express a decline in sensitivity, specificity and positive predictive value for MRI in children. The experience of a paediatric radiologist could improve the sensitivity of MRI in these situations.

Conclusion: This study shows that the agreement is not perfect for a suspected meniscal injury in children. MRI does not provide adequate diagnostic information and cannot be used for precise preoperative planning due to a lack of descriptive precision.