The British Journal of Radiology, Vol 73, Issue 871 737-739, Copyright © 2000 by British Institute of Radiology

ARTICLES

The long-term performance of DXA bone densitometers

J Wells and PJ Ryan
Department of Nuclear Medicine, Medway Hospital, Gillingham, Kent, UK.

Long-term performance of a bone mass measuring device is an important criterion when considering the purchase of such equipment and has been regarded as an important feature of dual X-ray absorptiometry (DXA). The performance of a 6-year-old bone densitometer, the Lunar DPX alpha, which has undertaken 1500 scans annually over this period, was assessed. The short-term coefficient of variation calculated from 15 measurements with repositioning on a single day, using the Lunar aluminium phantom, was 0.242%. Long-term precision, also calculated by the coefficient of variation, was 0.548%. The manufacturer's quality control (QC) procedure was performed daily and allowed the machine to be used except on 15 occasions when bone density measurements could be acquired after rebooting. However, a 2.2% shift in phantom values occurred in July 1996 owing to a photomultiplier tube failure, but this did not produce a failure in the Lunar QC. The optical disc drive was replaced in July 1997. The machine failed to back up on six occasions over the last 2 years owing to software corruption and the acquired femur data were not saved on seven occasions owing to overloading of the memory buffer. In conclusion, expected hardware failure and minor software problems have occurred. We were concerned that the manufacturer's QC failed to detect a 2% shift in the phantom bone mineral density values and recommend regular measurements of the Lunar aluminum phantom in addition to the daily QC measurement of the tissue-equivalent block. We were nevertheless impressed by the long-term stability and reproducibility of the Lunar DPX alpha.