Purpose

This test shows the parallelism to the reference system. Parallelism can be effectively analyzed through a "dilution-recovery" analysis in which different sera are analyzed at multiple dilutions. Dilution-recovery studies provide essential information that can validate the calibration system for the assay.

This analysis is specifically designed to demonstrate the ability of quantitative assays to generate the same interpolated result, irrespective of the dilution of the test sample analyzed or the region of the calibration curve used for interpolation.

 

Sample preparation: 10 samples were diluted in series of four dilutions (1:3, 1:9, 1:27 and 1:81). Samples were diluted using sample diluent serum. Dilutions were each prepared from undiluted sample.

 

Analysis and Computations

The obtained concentrations were then standardized with regard to undiluted concentration and expected concentrations were calculated. A comparison between the obtained versus expected concentration is shown in graphs 1-2, where it can be seen that the obtained concentrations for ImmunoCAP are close to the expected concentrations, but some of the results for samples processed on the Immulite 2000 are different from expected values.
 

A regression line was fitted to the dilution series for each sample. Since the concentrations are standardized, both the regression coefficient and R2 should be close to 1 if the dilution series produces expected results. Graph 1 shows the linear regression line performed for ImmunoCAP where the relationship between observed and expected concentration correlates well. The degree of determination and correlation drops for three samples tested with allergens Egg White (f1), Alternaria alternata (m6) on 3gAllergy test (Immulite 2000). The cause of the poor linearity for one of the Dust Mite (d1) dilution series is unknown.