Related Experiment Videos
A computer program for estimation of reference percentile values in laboratory data
Computer Programs in Biomedicine
|January 1, 1979
Summary
Physicians can better interpret clinical laboratory results using percentile rankings. A new FORTRAN program estimates these reference percentiles using regression, offering more stable and less biased results than traditional interpolation methods.
Area of Science:
- Clinical laboratory science
- Biostatistics
- Medical informatics
Background:
- Clinical laboratory results are conventionally reported in standard units.
- Interpreting these results can be challenging without context of a reference population.
- Percentile rankings offer a valuable supplementary method for result interpretation.
Purpose of the Study:
- To describe a FORTRAN program for estimating reference percentile values.
- To provide physicians with enhanced data interpretation tools for clinical laboratory results.
- To improve the accuracy and stability of percentile estimations.
Main Methods:
- Development of a FORTRAN program for percentile estimation.
- Utilizing regression analysis over 20% of the distribution for estimation.
- Comparison of regression-based estimation with interpolation methods.
Main Results:
- The FORTRAN program effectively estimates reference percentile values.
- Regression-based estimates demonstrated negligible bias.
- Percentile estimates derived from regression were more stable than those from interpolation.
Conclusions:
- The described FORTRAN program provides a robust method for estimating clinical laboratory result percentiles.
- Regression-based percentile estimation offers advantages in stability and bias over interpolation.
- This approach can aid physicians in more accurate interpretation of laboratory data.