Magnetic resonance imaging of the hip: poor cost utility for treatment of adult patients with hip pain.

BACKGROUND: Although MRI is frequently used to diagnose conditions affecting the hip, its cost-effectiveness has not been defined. QUESTIONS/PURPOSES: We performed this retrospective study to determine for patients 40 to 80 years old: (1) the differences in hip MRI indications between orthopaedic and nonorthopaedic practitioners; (2) the clinical indications that most commonly influence treatment decisions; (3) the likelihood that hip MRI influences treatment decisions separate from plain radiographs; and (4) the cost of obtaining hip MRI studies that influence treatment decisions (impact studies).
METHODS: We retrospectively assessed 218 consecutive hip MRI studies (213 patients) at one institution over a 5-year interval. Medical records, plain radiographs, and MRI studies were reviewed to determine how frequently individual MRI findings determined treatment recommendations (impact study). The cost estimate of an impact study was calculated from the product of institutional MRI unit cost (USD 436) and the proportion of impact studies relative to all studies obtained either for a specific indication or by an orthopaedic/nonorthopaedic clinician.
RESULTS: Nonorthopaedic clinicians more frequently ordered hip MRI without a clinical diagnosis (72% versus 30%, p < 0.01), before plain radiographs (29% versus 3%, p < 0.001), and with less frequent impact on treatment (6% versus 15%, p < 0.05). Hip MRI most frequently influenced treatment when assessing for a tumor (58%, p < 0.001) or infection (40%, p < 0.001) and least frequently when assessing for pain (1%, p < 0.002). Hip MRI impacted a treatment decision independent of plain radiographic findings in only 7% of studies (3% surgical, 4% nonsurgical). Hip MRI cost was least when assessing for a neoplasm (USD 750) and greatest when assessing undefined hip pain (USD 59,000). The cost of obtaining an impact study was also less when the ordering clinician was an orthopaedic clinician (USD 2800) than a nonorthopaedic clinician (USD 7800).
CONCLUSIONS: Although MRI can be valuable for diagnosing or staging specific conditions, it is not cost-effective as a screening tool for hip pain that is not supported by history, clinical examination, and plain radiographic findings in patients between 40 and 80 years of age. LEVEL OF EVIDENCE: Level IV, economic and decision analysis study. See Instructions for Authors for a complete description of levels of evidence.