Related Experiment Videos
Time related bias in longitudinal studies using dual photon absorptiometry
S Rozenberg1, J Vandromme, A Peretz
1Interdisciplinary Group on Osteoporosis, Free Universities of Brussels (VUB-ULB), Hôpital Saint-Pierre, Belgium.
Journal of Endocrinological Investigation
|December 1, 1992
Summary
A study on bone mineral density (BMD) revealed a time-related measurement bias. This bias affected patient BMD slopes, indicating potential inaccuracies in longitudinal BMD assessments. Further investigation is needed to refine measurement protocols.
Area of Science:
- Osteoporosis Research
- Medical Measurement Science
- Biostatistics
Background:
- Longitudinal bone mineral density (BMD) assessments are crucial for patient management.
- Previous quality control measures did not detect subtle time-related measurement biases.
Purpose of the Study:
- To investigate a potential time-related bias in BMD measurements.
- To analyze BMD slopes and identify measurement inconsistencies over time.
Main Methods:
- Calculated regression equations for BMD over time in 95 women with at least 3 measurements over 3 years.
- Grouped patients based on the timing of their last BMD measurement (July-Dec 1989, Jan-June 1990, July-Dec 1990).
- Analyzed relative differences (RD) between predicted and actual BMD in relation to time.
Main Results:
- Significant differences (p < 0.05) in BMD slopes were observed between patient groups based on the timing of their last measurement.
- Significant temporal fluctuations (p < 0.001) in RD indicated a time-related error.
- Parallelism between RD variations and mean BMD values suggested a systematic bias.
Conclusions:
- A time-related measurement bias in BMD assessments was identified.
- This bias may impact the accuracy of longitudinal BMD monitoring.
- Refined quality control and measurement protocols are recommended to mitigate this error.