In their retrospective investigation, Roberts and colleagues analyzed surgical and nonsurgical outcomes on the basis of whether magnetic resonance imaging (MRI) of the knee joint done prior to treatment had been ordered by an orthopaedic surgeon or by a primary care practitioner (PCP). Of 1592 patients included in this study, 747 had been originally evaluated and referred to the authors’ practice by a PCP and 845 had been initially evaluated by an orthopaedic surgeon in their practice. These two subgroups were equally matched demographically. The MRI ordering trends were tracked for both groups. The results revealed more frequent arthroscopic interventions on the basis of the MRI results ordered by the orthopaedic surgeons. Specifically, the orthopaedic surgeons were more likely to order MRI studies for younger patients, patients with acute symptoms, and those with a history of trauma. This subgroup of patients more often proceeded to arthroscopic intervention. In contrast, patients who underwent MRI ordered by their PCP were more likely to receive a total knee arthroplasty because of severe osteoarthrosis, a procedure and diagnosis not requiring MRI.

A salient point from this study is that the orthopaedic surgeons were less likely than PCPs to order MRI of the knee for patients with severe osteoarthrosis who subsequently underwent total knee arthroplasty (4.3% versus 9.2%, respectively). This is a particularly relevant point in light of the current health-care climate, as the authors state in their article. Overutilization of sophisticated imaging, such as MRI, burdens the health-care system and adds to the national debt. While there is no doubt that sophisticated imaging has improved the health of many, its overuse certainly offers no advantage to the health of our nation. It has been suggested that as many as 20% to 50% of high-tech imaging tests do not provide information that improves patient outcome and thus are unnecessary. The American College of Radiology appropriateness criteria were established in an effort to curb this overutilization. Roberts et al. address these criteria as they relate to knee MRI in their study.