A generalized least significant change for individuals measured on different DXA systems

Area of Science:

• Medical Imaging and Instrumentation
• Biomedical Engineering
• Osteoporosis Research

Background:

• The least significant change (LSC) is crucial for monitoring bone mineral density (BMD) changes over time.
• Traditional LSC calculations assume measurements are taken on a single system.
• Comparing BMD measurements between different densitometry systems requires a new approach.

Purpose of the Study:

• To derive and validate a generalized least significant change (GLSC) for comparing BMD measurements between two different systems.
• To provide a method for accurate inter-machine comparisons in bone densitometry.
• To assess the impact of system upgrades and manufacturer changes on BMD measurement comparability.

Main Methods:

• Derived a generalized LSC (GLSC) formula accounting for precision errors and calibration relationships between two systems.
• Validated the GLSC using 10,000 simulated measurements between two dual-energy X-ray absorptiometry (DXA) systems.
• Analyzed the influence of correlation coefficients and sample sizes on GLSC accuracy.

Main Results:

• The GLSC was found to be significantly larger than the traditional LSC when comparing different systems (e.g., twice as large for Hologic QDR-2000 to QDR-4500 upgrade).
• GLSC values were 2.6 to 3.6 times larger when comparing scans between Hologic Delphi and GE Lunar Prodigy systems.
• A correlation coefficient below 0.95 increased %GLSC above 10%, and sample sizes over 30 minimally impacted GLSC accuracy.

Conclusions:

• The generalized least significant change (GLSC) is the most accurate method for comparing scans between dual-energy X-ray absorptiometry systems.
• GLSC accounts for inter-system calibration and precision errors, offering improved clinical utility for monitoring BMD.
• The GLSC provides a standardized approach for evaluating BMD changes across different densitometry equipment.