Computed Radiographic Measurement of Limb-Length Discrepancy<sbt aid="1049352">Full-Length Standing Anteroposterior Radiograph Compared with Scanogram</sbt>

Abstract

Although a scanogram is commonly used to measure limb-length discrepancy, there are several potential pitfalls associated with this imaging technique. The purpose of the present study was to evaluate the results obtained with use of a full-length standing anteroposterior radiograph of the lower extremities and to compare them with those obtained with use of a scanogram. Both imaging studies were performed using computed radiography. One hundred and eleven patients with limb-length discrepancy had a full-length standing anteroposterior radiograph and a scanogram made on the same day. The patients included seventy-nine children and thirty-two adults in whom the discrepancy was secondary to trauma (55%), congenital shortening (18%), Blount disease (14%), or another cause (13%). Limb length and limb-length discrepancy were measured utilizing both imaging studies. The agreement between the standing anteroposterior radiograph and the scanogram was assessed with use of the correlation coefficient r, and the limits of agreement between the two imaging studies were assessed. An average magnification of 4.6% (3.3 cm) was observed in association with the measurement of lower extremity length with use of the full-length standing anteroposterior radiograph. The mean difference in limb-length-discrepancy measurements between the two techniques was 0.5 cm, and the limits of agreement (that is, the mean plus or minus two standard deviations) were 0.5 to 1.5 cm. When the limb-length discrepancy on the standing anteroposterior radiograph was compared with that on the scanogram, the correlation coefficient r was 0.96. A difference of >0.5 cm between the limb-length discrepancy measured on the standing radiograph and that measured on the scanogram was associated with a mechanical axis deviation of >2 cm. Remaining variables, including age, gender, etiology, and scanogram ruler inclination, did not correlate with a difference in the measurement of limb-length discrepancy with use of these two imaging studies. The measurement of limb-length discrepancy on a standing anteroposterior radiograph was very similar to that on a scanogram, especially in the absence of substantial mechanical axis deviation. These findings support the use of a standing anteroposterior radiograph of the lower extremities as the initial imaging study for patients presenting with unequal limb lengths. This approach allows for a more comprehensive radiographic evaluation of the lower extremity, including deformity analysis, while reducing the expense and radiation exposure as compared with the use of additional imaging studies for the assessment of limb-length discrepancy.