BACKGROUND

For bone health assessment, dual-energy X-ray absorptiometry (DEXA) is recommended to measure bone mineral content and areal bone mineral density (aBMD) in the lumbar spine. However, intermachine differences were not taken into account when developing these recommendations. According to the International Society of Clinical Densitometry (ISCD), phantom-based cross-calibration is adequate after replacing the DEXA system from a different manufacturer. For different DEXA equipment, individual calibration equations were found to be necessary to fit the observed values with the given densities.

METHODS

The BMD European Spine Phantom (ESP) measurements (L1, L2, and L3) were assessed on 3 machines. We used the Welch test in the one-way analysis of variance (ANOVA) with a Tamhane T2 test, linear regressions, and Bland-Altman analysis to assess the consistency of measurements and establish cross-calibration equations.

RESULTS

The coefficients of variation (CV)% of the phantom BMD values measured using the 3 systems were less than 3.0%. The 3 DEXA systems were highly correlated with BMD in the lumbar spine, with correlation values ranging from 0.933 to 0.984 (P<0.0001). The cross-calibration regression models of the ESP measurements yielded the highest prediction accuracies with the lowest prediction errors (the standard error of the estimate ranged from 0.004 to 0.008 g/cm; P<0.0001). After the regression equations were applied, the differences in BMD values among the 3 systems were negligible. In addition, the Bland-Altman plot showed that almost all data points were within the 95% limits of agreement.

CONCLUSIONS

A strong agreement for BMD measurement was established between the 3 DEXA systems. Cross-calibration equations for the lumbar spine BMD values need to be applied to transform the Hologic Discovery A or GE Lunar iDXA measurements into SONIALVISION SMIT measurements to comply with the ISCD standards for patient continuity of care in assessment during clinical diagnosis.