Protocol Online logo
Top : Forum Archives: : ELISA

ELISA. REPORTING TEST REPRODUCIBILTIY - can i run my batch again to get coefficients of variation (Jan/05/2009 )

hello everyone.
i ran these samples i had, in singles and got interesting results. as i had run the samples in singles, obviously i cannot calculate coefficients of variation. i need to order new kits. should i order enough to run the batches once more or double the number of kits so they can be run in duplicate. funds are limited so i want to know if you guys think that it is scientifically acceptable to describe test reproducibility and precision by running the two tests on the same samples at diffErent times. ( p.s samples were stored safely in alliquotes so refreezing and thawing is not an issue here)
waiting anxiously for suggestions


I believe you may be referring to 2 different methods: intraassay variation (replicates within a run) and interassay variation (single points tested in different runs). Usual parameters for validating an assay are:

1. Methods comparision; new assay vs. reference
2. Linearity (serial dilutions of high sample)
3. Spike and Recovery (know amount/concentrations spiked into samples at low, mid and high levels)
4. Interassay precision
5. Intraassay precision
6. Residual plot
7. Receiver operator curve
8. Assay Range
9. Sensitivity
10. Specificity
11. Interference

HOWEVER, you mentioned you purchased the test kits. The manufacturer should have precision information in addition to the information I cited above. You should not have to do any tests.