Agreement between lunar densitometers in the capital region of Denmark-a phantom study

In clinical practice, patient follow-up requires scans to be performed on the same densitometer to ensure comparability. This study evaluated the agreement and equivalence between nine Lunar iDXA and four Lunar Prodigy scanners, assessing their interchangeability for longitudinal spine BMD monitoring. The manufacturer's encapsulated spine phantom was scanned 15 times on each of nine iDXA and four Prodigy densitometers under routine quality-control conditions by a single experienced operator. For each scanner, L2-L4 BMD repeatability was assessed using mean, SD, and coefficient of variation. Between-scanner equivalence was tested pairwise using two one-sided tests with ±1% equivalence bounds with Holm-Bonferroni correction. Exploratory method-comparison plots were used to assess agreement across vertebral regions. All iDXA scanners demonstrated low within-scanner variability (CV: 0.07-0.23%) and closely clustered mean BMD values, with a small negative bias relative to the phantom nominal value (mean -0.63%). All pairwise iDXA comparisons met the predefined ±1% equivalence criterion. Graphical analyses confirmed high agreement across all vertebral levels. In contrast, Prodigy scanners showed higher variability (CV: 0.17-0.33%), a positive bias (mean +1.5%), and failed to meet equivalence criteria in most pairwise comparisons. Under the conditions of this phantom study, the nine iDXA scanners in the region showed high agreement and may be sufficiently equivalent for longitudinal spine BMD monitoring across different sites. However, the Prodigy scanners did not demonstrate sufficient agreement for interchangeable use. These findings support regional harmonization of iDXA-based BMD measurements, although validation in vivo, including femoral and full-body measurements, is warranted.