Quality control and evaluation of bone densitometry by DXA

Objective To analyze the quality control and evaluate different types dual X-ray absorptionmetry( DXA) from the same manufacturer,and the same DXA in the different time of measuring the bone densitometry. Methods Conduct the accuracy and precision tests with DPX-L and Prodigy from GE-Lunar in 2008 and 2012 respectively. Lumbar spine phantom designed by the authors was used in the accuracy tests. In the precision tests,the lumbar spine phantom,female lumbar spine and quality control phantom of the Prodigy were used. All the subjects were tested by the same technician under standard the technical requirements. The accuracy tests scanned the phantom for 10 times consecutively on the same day and calculate the percentage of the measured value and true value of bone mineral density( BMD). The precision tests scanned the lumbar spine phantom,female lumbar spine and quality control phantom once a day for consecutively 25 days,and calculate the coefficient of variation( CV). Results The BMD was used as the description indicator,and the accuracy errors of DPX-L and Prodigy were 0. 80% and 3. 80%( P 0. 001) in 2008. The accuracy errors of Prodigy before and after the application were 3. 80% and 3. 70%( P 0. 05). The precisions errors of DPX-L( after the application of 15 years until 2008) and Prodigy( started to use in 2008) were 1. 88% and 0. 33%( P 0. 001). The precision error of DPX-L was 5. 6 times bigger than that of Prodigy. Female lumbar spine scanning showed that the precision error of DPX-L was 4. 9 times bigger than that of Prodigy( 9. 17% /1. 89%). Conclusions The accuracy and precision errors varies with the different types of DXA and with the different application time of the same DXA. The uncorrectable precision errors should be more concerned. Compared with the long-term precision,the short-term precision had less confounders and stronger comparability.