Precision of dual X-ray absorptiometry and peripheral computed tomography using mobile densitometry units.

Irrespective of the method used for noninvasive bone mass determination, data comparison between different centers is a major problem as significant interunit variation may occur. We, therefore, have employed mobile densitometry units to reduce interunit variability in two large epidemiologic studies in Germany. Two cars were equipped with either two dual X-ray absorptiometry (DXA) instruments (QDR 1000 Hologic, USA) (car I) or a special purposed scanner for peripheral quantitative computed tomography (pQCT) (XCT 900, Stratec, FRG) (car II). The cars were moved across Germany 11,090 km and 1651 km during the studies over a period of 30 and 7 months, respectively. Precision in vitro was determined using hydroxyapatite phantoms. Forty-eight patients underwent duplicate measurements at the lumbar spine (n = 22) and hip (n = 26) for assessing reproducibility in vivo. Between the two series of scans, the car was moved 63 km. Long-term precision in vitro of the QDR 1000 instruments were 0.41% and 0.59% for BMD with no evidence of machine drift (rate of change per year 0.04% and 0.05%, respectively). Short-term reproducibility in vivo showed a coefficient of variation (cv) of 1.02% for spinal BMD (L2-L4) and 1.72% for femoral neck. Long-term precision in vitro of the pQCT scanner was 0.9%. Our study shows precision in vivo and in vitro and stability of the mobile densitometers similar to that achieved with stationary equipment. In conclusion, mobile densitometry may become a useful tool not only for epidemiologic surveys and clinical trials but also for routine evaluation in less densely populated areas.