How accurate are Dual-Energy X-Ray Absorptiometry measurements in clinical practice?: a retrospective single centre study.

Objective: The aim of this study was to identify common errors in DXA measurements at our hospital, establish standardization, and contribute to operator training.

Design: This descriptive study at a single centre analyzed 2375 spine and femur DXA images for errors. The reports were reviewed according to a checklist prepared considering The International Society for Clinical Densitometry(ISCD) official positions. Each image was reviewed by a Radiologist and Physical Medicine and Rehabilitation (PMR) specialist trained in the principles and standards of DXA scanning prior to the study. All the assessments were double-checked by other specialists and any incorrect images were reported. All the scans taken and available during the observation period were reviewed. The study included scan images taken in our centre, including lumbar spine and proximal femur measurements, for which all information was entered correctly and completely. Forearm measurements, whole-body measurements, or measurements performed in children were excluded.

Results: A total of 2375 DXA scan results from a single centre were analyzed. According to the evaluation criteria, 1023 (43.1%) lumbar spine and 1078 (45.4%) proximal femur DXA scans had at least one error. The most common error encountered was the presence of excessive degeneration (31.0%) for lumbar spine results, and inadequate hip internal rotation (23.9%) in the proximal femur results. There were found to be more errors in lumbar measurements according to BMI (p=0.04) and age (p≤0.001), and errors in the femur were higher in male gender (p<0.001).

Conclusions: Although the operator and auto-interpretation error rates in this study were low compared to the literature, the results showed that standardization in operator training is not at the desired level. It is hoped that this study will raise awareness of clinicians about DXA imaging errors, and that the training of clinicians who interpret DXA results should be considered as important as the training of technicians.