Cross calibration of DXA as part of an equipment replacement program

The purpose of this study was to develop cross calibrations when replacing three dual-energy X-ray absorptiometers (GE Lunar DPXL [Madison, WI], DPXL, GE Lunar Expert [Madison, WI], Expert, Hologic QDR2000 [Waltham, MA]) with two new GE Lunar Prodigy instruments. Subjects previously scanned on the Expert or QDR2000 were transferred to Prodigy 1 and those previously scanned on the DPXL to Prodigy 2. A cohort of subjects was recalled for each old instrument, and approximately 20 subjects had lumbar spine and hip scans on each old instrument and the appropriate new instrument. An in vitro calibration was carried out using a Bona Fide Phantom (Bio-Imaging Technologies, Inc., Newtown, PA). Calibrations were fitted using a standardized principal components method. A Bland and Altman plot was used to calculate the mean difference and limits of agreement between instruments. Standardized bone mineral density (BMD) was also used to calibrate the Hologic to Prodigy 1. There was good agreement between instruments from the same manufacturer. As expected, BMD measured on the Prodigy was about 15% higher than the Hologic. Using standardized BMD to cross calibrate gave a mean difference of 3% at the lumbar spine. The limits of agreement following calibration are clinically significant, so it is not possible to apply a calibration to an individual subject for trending purposes, as the error is similar to the expected annual change in BMD, but can be used for cross calibration in clinical trials. The in vivo calibration gave better agreement than using standardized BMD. The phantom calibration was close to the in vivo calibrations at the spine, but not in some hip regions. When introducing a new instrument, a new baseline BMD has to be obtained for each subject.