Does subspecialty referral affect the pretest probability for an abnormal finding on knee MRI of pediatric patients?

Pediatricians and orthopedists comprise the largest referral basis for knee MR examinations at our institution. In an era of cost optimization, the purpose of this study was to compare differences in pretest probability for an abnormal finding on knee MRI based on referral subspecialty.

A retrospective review of 501 consecutive knee MR examinations of pediatric patients (56% male; mean age, 14 years; age range, 1-18 years) referred by nononcology orthopedic surgeons and 93 consecutive knee MR examinations of patients (47% male; mean age, 14 years; age range, 2-18 years) referred by general pediatricians from 2005 to 2009 were reviewed. Two patient groups based on the MR report were established: patients with entirely normal MRI findings and those who had a clinically significant MRI finding or findings. The latter group included children with any internal derangement, a discoid meniscus, a neoplasm, or evidence of infection or inflammation.

The incidence of an entirely normal knee MRI from pediatrician referral and orthopedic referral was 24% (22/93) and 17% (87/501) (p = 0.15), respectively. The respective incidence of selected specific injuries identified from pediatrician and orthopedic referral included any internal derangement, 39.8% and 48.3% (p = 0.13); neoplasm, 2.2% and 1.4% (p = 0.64); and inflammation, 6.5% and 2.0% (p = 0.03).

Despite differences in subspecialty training, we found no significant differences in the proportion of normal knee MRI examinations and no statistical difference in the proportion of patients with internal derangement. Further investigation is necessary to determine whether these findings translate into overall cost-savings or differences in patient outcome.