The utility of MRI scans for a painful knee in the elderly patient

We assess the requesting patterns and overall utility of MRI scans in our institution over a 6-month period for knee pain in an elderly population.

We performed a retrospective cohort study analysing patients undergoing a knee MRI between June 2018 and January 2019. Clinical notes were reviewed to determine the nature of any surgical intervention that was performed as a result of the MRI findings. A simple economic analysis was then performed using local pricing information.

One hundred and nine MRI scans were performed. The majority of scans were ordered by primary care physicians with no preceding plain radiographs (55%, n = 60). Meniscal tears were reported in 68% of patients. Degenerative changes were noted in 82% of patients. One patient was noted to have a flipped meniscal fragment. This was the only patient to undergo surgical intervention of any kind. Univariate analysis demonstrated no correlation between the presence of a meniscal tear (p = 0.61) or degeneration (p = 0.56) and the need for intervention. The only significant predictor of intervention was the presence of locking (p < 0.001). For the 101 patients that underwent unnecessary MRI scans, our institution would have saved €17,170 in unnecessary costs over 6 months.

We suggest that all elderly patients with knee pain undergo plain radiographs before MRI scan and only those with locking should be considered for an MRI scan at all. Primary care physicians are overusing MRI as an imaging modality in this cohort at a cost of over €30,000 per year to our institution.