Publications

Abstract

This study examined the concordance in BMD measurement and longitudinal change in BMD between the GE Lunar Prodigy and GE Lunar DPX. Even though a high concordance between the densitometers was observed on a single measurement occasion, a significant discordance in longitudinal changes in BMD was observed. INTRODUCTION: Measurement of bone mineral density (BMD) by dual-energy X-ray absorptiometry (DXA) technology plays an important role in the diagnosis and management of osteoporosis. The present study examined the concordance in BMD measurement and longitudinal change in BMD between GE Lunar Prodigy and DPX. METHODS: BMD at the lumbar spine and femoral neck was measured in 135 individuals (47 men and 88 women, mean age 73+/-9 years) using both GE Lunar DPX and Prodigy densitometers at baseline. In this group, 56 individuals (22 men and 34 women) had repeated BMD measurements using the DPX and Prodigy during a subsequent follow-up visit (average duration: 2.2 years). RESULTS: For a single BMD measurement, the coefficient of concordance between the Prodigy and DPX was greater than 0.98 at the lumbar spine and 0.96 at the femoral neck, with the slope of linear regression being approximately 1.0. During the period of follow-up, the lumbar spine BMD decreased by -0.5% (S.D. 1.8%) when measured by DPX, which was significantly different (p=0.002) from the change measured by Prodigy (mean change=0, S.D. 2.0%). However, there was no significant difference (p=0.95) in the rate of change in femoral neck BMD measured by DPX (mean=-1.6%, S.D.=2.9) and Prodigy (mean=-1%, S.D.=1.8%). The correlation in rates of BMD change between Prodigy and DPX was 0.63 at the lumbar spine and 0.52 at the femoral neck. Simulation analysis showed that the theoretical maximum correlation in rates of BMD change between Prodigy and DPX was 0.71. CONCLUSIONS: Despite both densitometers being highly concordant in a single BMD measurement, discordance in the assessment of BMD changes between the Prodigy and DPX densitometers was observed. These findings have implications regarding the assessment of response to therapy in a multi-centre setting when different densitometers are used.