Dual-Energy X-ray Absorptiometry Scan Autoanalysis vs Manual Analysis

The measurement of bone mineral density (BMD) with dual-energy x-ray absorptiometry (DXA) is valuable for the determination of 10-yr fracture risk and for antifracture treatment follow-up. Ensuring patient scans are performed with accuracy, and reliability is imperative, requiring both technician competence and regular machine calibration. With DXA, analysis of each scan can be performed either with the machine's default autoanalysis or can be optimized manually. For 1 yr, all patients sent for DXA measurements to the Saskatoon Osteoporosis Center had each lumbar spine and hip scan analyzed with both manual and autoanalysis methods and the 2 sets of scans compared. We compared the concordance between the 2 analysis methods by calculating a BMD percent error for all of the scans, with the manually adjusted scans acting as the reference standard. Mann-Whitney U tests were completed to test for statistically significance differences between analysis types. In this investigation, scans completed with manual analysis were more accurate with respect to BMD (up to 4.7% error) and T-scores (up to 0.38 difference). In addition, many errors were identified with autoanalysis. Consequently, technicians using DXA should not rely on autoanalysis but rather be trained in and use manual analysis.