Related Experiment Videos
Quality assurance for bone densitometry research studies: concept and impact
C C Glüer1, K G Faulkner, M J Estilo
1Osteoporosis Research Group, University of California at San Francisco 94143.
Summary
Implementing a quality assurance (QA) program in multicenter bone densitometry studies enhances data reliability and statistical power. This approach can significantly reduce study costs by improving precision and potentially decreasing required sample sizes.
Area of Science:
- Medical Imaging
- Radiology
- Clinical Trials
Background:
- Quality assurance (QA) is crucial for multicenter bone densitometry studies.
- Standardized QA protocols are needed to ensure data consistency and reliability across sites.
- Existing QA methods may have limitations, particularly phantom-based approaches.
Purpose of the Study:
- To present a comprehensive QA concept for multicenter bone densitometry clinical studies.
- To outline the key components and services provided by a central coordinating center for QA.
- To demonstrate the potential benefits of QA implementation on study power, precision, and cost.
Main Methods:
- Development of a QA framework involving consulting, certification, data review (scans, patient data, problem cases), and result correction.
- Utilizing QA standards for cross-calibration and longitudinal assessment of scanner stability.
- Acknowledging limitations of phantom-based QA data and presenting solutions for common research problems.
Main Results:
- Successful QA implementation can substantially enhance statistical power.
- Improved precision through QA can reduce the smallest detectable difference or sample size by over 50%.
- QA measures can lead to significant reductions in overall study costs.
Conclusions:
- A well-defined QA program is essential for high-quality bone densitometry results in multicenter trials.
- Balancing QA efforts with acceptable costs is key to maximizing net benefit.
- Basic QA measures are recommended even for individual clinical sites to improve data quality.