Comparison and cross-calibration of DXA systems: ODX-240 and Sophos L-XRA versus Hologic QDR-4500, for spinal bone mineral measurement. Translation of a reference database.

Replacement of dual-energy X-ray densitometry equipment may be necessary in time as a result of upgrading systems or new equipment. The lack of standardization in bone mineral density (BMD) measurements is known. Standardization efforts have been made for several years by the European Union under its organization COMAC-BME (Comité d'Actions Concertés-BioMedical Engineering) and by the International Committee for Standards in Bone Measurement. Cross-calibration is generally considered to be the result of linear regression between the measurements obtained with two densitometers. A major disadvantage of the regression method is the noncompatibility of the two formulae of calibration (Y versus X and X versus Y). In this study we considered cross-calibration in terms of a structural model that produced circular equations when, for example, three systems were cross-calibrated. Cross-calibration in this study was calculated from the measurement of the lumbar BMD of a population of 204 patients, with Hologic QDR-4500, ODX-240 and Sophos L-XRA systems. In vitro accuracy and short-term reproducibility of the three systems were studied. Using the structural calibration equation we transformed a reference database for L2-4 BMD obtained from a population of 983 French females, aged 11-47 years, on an ODX-240 to a reference database for a Hologic QDR-4500. A new young adult reference was obtained and consequently a new evaluation of the T-score for the Hologic QDR-4500.