Dual X-ray absorptiometry quality control: comparison of visual examination and process-control charts

Area of Science:

• Medical Imaging
• Quality Control
• Bone Densitometry

Background:

• Dual X-ray absorptiometry (DXA) is crucial for monitoring bone mineral loss during treatment.
• Ensuring DXA scanner accuracy through rigorous quality control is essential for valid measurements.
• Several statistical methods exist for DXA scanner quality control, each with varying effectiveness.

Purpose of the Study:

• To compare the efficacy of five different quality control procedures for DXA scanners.
• To evaluate methods based on their ability to detect scanner malfunctions using phantom and simulated data.
• To identify the optimal quality control strategy for DXA scanner performance monitoring.

Main Methods:

• Analysis of five longitudinal phantom datasets from a clinical trial.
• Analysis of simulated datasets to assess method performance.
• Comparison of visual inspection, Shewhart charts (with and without filters), moving average/standard deviation charts, and CUSUM charts.

Main Results:

• Visual inspection is subjective; regular Shewhart charts have high false alarm rates.
• Shewhart charts with filters reduce false alarms but lower sensitivity.
• The CUSUM method demonstrated high sensitivity and low false alarm rates, estimating malfunction dates.
• Moving average/standard deviation charts approached CUSUM performance.

Conclusions:

• The CUSUM method is recommended for monitoring DXA scanner performance due to its superior sensitivity and low false alarm rate.
• Shewhart charts may be acceptable at individual clinical trial sites if centrally validated by CUSUM analysis.