Patient variables impact lumbar spine dual energy X-ray absorptiometry precision

Abstract

Changes in bone mineral density are used to monitor osteoporosis therapy. To determine whether a change in bone mass is clinically significant, the precision of bone mineral density measurements must be known. We therefore measured the impact of vertebral body exclusion on dual energy X-ray absorptiometry (DXA) precision. At one university and one Veterans Affairs DXA center, three radiology technologists each scanned 30 participants twice, with repositioning between scans, to estimate DXA precision. Three International Society for Clinical Densitometry-certified physicians reviewed all lumbar spinal scans to note the presence of focal structural defects. We calculated precision for subsets of vertebrae, and for virtual samples of patients with and without physician-identified vertebral focal structural defects. We graphed the reciprocal of least significant change versus bone area to determine the dependence of precision on interpreted scan area. Within each sample, greater interpretable bone area improved precision. The contribution of interpreted bone area to precision differed among the samples, ranging from 57 to 94%. Greater population bone mineral density heterogeneity and presence of focal structural defects each decreased precision. All bone densitometry centers must determine precision using a sample representative of their served populations. Failure to do so may lead to incorrect determination of least significant change. Population heterogeneity, vertebral body exclusion and presence of focal structural defects each decreases precision.