Radiofrequency Echographic Multi Spectrometry (REMS) for the measurement of bone density at the lumbar spine and hip - A "real-life" Australian validation study.

Abstract

To determine the diagnostic accuracy of Radiofrequency Echographic Multi Spectrometry (REMS) compared to dual energy X-ray absorptiometry (DXA) in a "real world" Australian tertiary hospital population, and to assess how much of the variability in REMS-bone mineral density (BMD) could be attributed to patient age, sex and weight. This single centre study recruited 69 patients in whom both REMS and DXA scans of the lumbar spine and proximal femur were performed according to manufacturer's specifications and standard clinical procedures, respectively. Overall, n = 69 patients (n = 27 male, n = 42 female) with the following parameters (mean ± standard deviation) were recruited: age (years) 61.6 ± 12.5; weight (kg) 73.5 ± 18.8; body mass index 27.3 ± 6.6. A simple multiple regression model for both REMS-femoral neck (FN) and REMS-lumbar spine (LS) BMD containing age (years) and weight (kg) explained 90% and 88%, of the total variability, respectively. In comparison, all factors (DXA machine, age, weight, sex) only explained 33-35% of the variability in DXA-FN and DXA-LS areal BMD measurements. Furthermore, REMS-BMD contributed <2% to DXA-BMD and DXA-BMD contributed <1% to REMS-BMD. Bland-Altman analysis was used to assess agreement between DXA-FN and REMS-FN (n = 68 patients) and DXA-LS and REMS-LS (n = 65 patients) measurements. Overall, there was poor interchangeability between DXA and REMS measurements. Most of the variability in REMS-FN and REMS-LS BMD was explained by a multivariable linear regression model which incorporated age and weight. In comparison, only 33-35% of the variability of DXA BMD at both these sites was explainable by a statistical model. Bland-Altman plots indicated that, in our hands, REMS-BMD and DXA-BMD could not be used interchangeably. Overall, our findings suggested that REMS-BMD may not reflect DXA-BMD to any clinically meaningful extent. These findings need to be confirmed in larger studies.