There are significant differences in the DXA reference ranges currently used in Australia for both the proximal femur and the lumbar spine. These differences lead to differences in the diagnosis of osteoporosis, and hence differences in therapeutic decisions, depending on which machine a patient is scanned. To overcome these problems the ANZBMS recommends that the Geelong Osteoporosis Study (GOS) reference range is used in Australia for interpretation of DXA scans in Caucasians. At present the GOS data is available only for females in the lumbar spine and proximal femur. Acquisition of a male data set is currently under way.

• Lunar (and Hologic) T scores will become slightly less negative but the difference will be small. In the range of T between -1.5 and -3.0 the difference will be 0.1 T score units or less. At very low BMDs the differences will be larger but at that level of BMD it is unlikely that the T score differences would impact on a decision to treat.
  
• Norland T scores will become slightly more negative. The change in T scores may approach 0.5 T score units as the T score approaches -3.0.

Effect of Implementation of the GOS Reference Range on Femoral Neck T scores

• Lunar T scores will become slightly more negative but the difference will be small. In the range of T between -1.5 and -2.5 the difference will be -0.2 T score units or less and 0.16 T score units or less below -2.5.
  
• Norland T scores will become less negative. The magnitude of the change will depend on which of the reference ranges was being used previously but the change may exceed 0.5 T score units as the T score approaches -3.0.
  
• Hologic T scores will generally become less negative. The magnitude of the change will depend on which of the reference ranges was being used previously but the change will usually be £ 0.25 T score units below T of -2.0.

Logistics of Implementation of the GOS Reference Range
In Norland and Hologic instruments the software design lends itself to a relatively easy implementation of the GOS reference ranges; this is also true for more recent Lunar machines using "enCORE" operational software. The respective Australian agents should be contacted for assistance.

In Lunar instruments using older DOS operating systems, the software design does not lend itself easily to adding the GOS reference ranges. While the newest version of software for the Lunar Prodigy contains the GOS reference range, for older machines it may not be available for some time. However in the interim, due to the relatively small changes in converting from the Lunar reference range to the GOS reference range, continued use of the Lunar reference range is considered acceptable and would have minimal effect on the diagnosis.

Policies & Pos. Statements INDEX