Presentation of knee pain in a middle-aged or older patient often leads to an x-ray of the knee. We know from previous research that knee symptoms correlate very poorly with plain x-ray findings. Magnetic resonance imaging (MRI) produces excellent pictures of articular cartilage, synovium and bone marrow that might be expected to produce information that correlates better with the patient experience of symptoms and guides treatment planning. Well perhaps not. In a study from Framingham published in the
BMJ, 710 patients over the age of 50 with normal x-rays of the tibiofemoral joint underwent an MRI scan of the knee. Of the 710, 206 (29%) reported knee pain in the preceding month. However, 631 (89%) had some abnormality characteristic of OA. Osteophytes were the most common finding, in 524 (74%) patients. 492 (69%) patients had articular cartilage defects. The prevalence of abnormalities increased with age, as might be expected. Abnormalities were almost universal in patients reporting knee pain (90–97%, depending on definition of pain) but still remarkably high in patients 1 not reporting knee pain (86–88%), calling into question the clinical significance of MRI findings. So the pictures may be better and certainly more interesting, but whether imaging will lead to more effective treatment seems doubtful based on the findings of this study.