Can you spot a bad DXA report?

I make no apologies for harping on a familiar theme. As reimbursement for DXA is falling, fewer colleagues are purchasing new DXA instruments and most reports are coming from centers that have been providing this service for some time. Most produce excellent, reliable data, but there are still too many unacceptable reports that accompany troubled patients who cannot understand why their osteoporosis is getting worse when they are taking their therapy properly.

Before embarking on an extensive and expensive work-up to see where the problem might lie, first check that the report is correct.

Problem number 1: The follow-up DXA has been performed at a different center than the earlier one, sometimes on a scanner from a different manufacturer. This may be unavoidable for many reasons and there is little to be done about this after the fact. Check that the patient is taking therapy regularly and correctly. Next go through your check list of possible secondary causes of bone loss — history and physical exam first and then appropriate laboratory studies. I can’t resist the temptation to again mention the value of bone resorption markers when nothing pans out from the above work-up.

Problem number 2: Neither the reporting physician nor the technician performing the scan has undergone formal DXA training. To my knowledge the only group offering certification after a training course and examination is the International Society for Clinical Densitometry. I think that the society will be able to tell you which facilities in your area are staffed by certified personnel. If not, politely check with the center directly. Spotting reports from centers not up to par is not difficult. Most DXA scanners have a prepared print out that can be spewed from the instrument without careful review. You should insist that each report forwarded to your office contains the images of the analyzed skeletal sites. It is easy to spot an analysis performed on a prosthetic hip — yes that does happen, particularly with the increasing trend to report “dual femur” data. Likewise on the image of the lumbar spine vertebral fractures, scoliosis, osteophytes and artifacts (zippers, vascular clips, even navel rings) can’t be missed. These may not come across well in faxed reports so ask that an original be mailed to you as well. The images on newer scanners are near X-ray quality.

Problem number 3: The report does not follow established guidelines. Reporting osteopenia at L-1 and L-3, osteoporosis at L-2, and normal at L-4 is a no-no! The correct interpretation should be made on the combined data of all evaluable vertebrae L-1 to L-4. Be wary of “cookie cutter” reports or reports that state “mild osteopenia” as I have seen twice in the last week.

Problem number 4: Make sure that follow-up scans are reporting the same skeletal site each time. If L-1 to L-4 is reported at baseline and L-2 to L-4 at follow-up, insist that the report be re-analyzed and that comparable skeletal sites are used. The recommended site for serial measurement of the proximal femur is the total proximal femur, not the femoral neck. The reason for this is the better precision at the total femur site which contains 20 g to 30 g of bone mineral compared with the femoral neck with only 3 g to 5 g. Positioning to optimize the femoral neck site is tricky and extremely difficult to reproduce accurately after an interval of two years, particularly if the technician performing the scan has also changed.

Taking care of all of the above is time consuming and is not really your job. Unfortunately unless you have full confidence in the center performing the study I cannot think of any alternative.