Related Experiment Videos
Effect of the reporting-interval size on critical difference estimation: beyond "2.77"
1Department of Chemical Pathology, St. Vincent's Hospital, Sydney, and Faculty of Medicine, University of New South Wales, Sydney, Australia. gjones@stvincents.com.au
Clinical Chemistry
|March 4, 2006
Summary
The reporting interval size impacts critical difference (CD) calculations for pathology results. Larger reporting intervals increase CD, necessitating careful selection based on clinical needs and assay variation.
Area of Science:
- Clinical Chemistry
- Laboratory Medicine
- Analytical Science
Background:
- The reporting interval (RI) is crucial for numerical pathology results, defining the bin size for analyte reporting.
- The impact of RI size on critical difference (CD) between serial patient results was previously unexamined.
Purpose of the Study:
- To investigate and quantify the influence of reporting interval size (RIS) on critical difference (CD).
- To develop a model for predicting CD based on RIS and biological variation.
Main Methods:
- Modeled the effect of varying RIS on CDs using spreadsheet software.
- Applied the model to laboratory data, incorporating analytical precision and within-person biological variation.
Main Results:
- An approximately linear increase in CD was observed as RIS grew relative to combined analytical and within-person variation.
- A revised formula for CD was proposed: CD = 2(1/2) x z x (SDa(2) + SDi(2))(1/2) + 1.5 x RIS.
- This relationship held true across different probabilities and for both uni- and bidirectional changes.
Conclusions:
- Reporting interval selection should align with specific assay requirements and clinical needs.
- For detecting subtle analyte concentration changes, a smaller RIS is recommended, relative to assay precision and biological variation.