 |
 |
William Nowatzke wrote: >Are there reviews or protocols out there for recommended method validation processes for putting new ELISA assays on-line? Specifically, what is considered acceptable for monitoring QC, especially with great variations between lot numbers or even from plate to plate? How about some sort of delta value between duplicates (within/between plates)? I have not seen any validation protocol for ELISA methods, however in any such protocol the exact position of the samples, controls and standards should be considered, especially when low concentrations are to be measured. I don't know why, but in my experience the outer positions often differ significantly both with respect to blank values and color intensity compared with the inner positions. If this phenomenon is present but not considered, serious systematic (if standards, controls and samples always are applied according to a fixed pattern) or quasi-random (if standards etc are applied randomly) errors can occur. This kind of effect is easy to disclose: Run the ELISA test with a plate containing only blanks and one standard in every second well. Compare with a Student's t-test the blank values from all outer positions with those of the inner positions, and do the same test for the standard values. If this test is significant, you must consider the difference between these positions. If not you can consider all positions to be equivalent. Mr Sten Öhman, PhD Elfin Lab & Milieuconsult - Part of the Wiltag group E-Mail address: [log in to unmask] Telephone int: +46 13 368941, Nat: 013-368940 Fax int: +46 13 368941, Nat: 013-368941 Mobile tel: 0709-526415 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%