# dilution linearity??? - how much is allowed and wht it means?? (May/31/2010 )

HI

just curious i was wondering..how much dilution recovery is allowed (like one of my collogue told that any recovery/linearity between 80-120% is allowed...so what if i am getting little less than that??) and what one mean by such as value???

what if myn spike recovery is good but nmy dilution linearity is bad..what does this indicate??

a kind of basic question..what is the ideal way to evaluate dilution linearity for an immunoassay??

thanks a lot

rick112 on May 31 2010, 05:39 PM said:

just curious i was wondering..how much dilution recovery is allowed (like one of my collogue told that any recovery/linearity between 80-120% is allowed...so what if i am getting little less than that??) and what one mean by such as value???

what if myn spike recovery is good but nmy dilution linearity is bad..what does this indicate??

a kind of basic question..what is the ideal way to evaluate dilution linearity for an immunoassay??

thanks a lot

hi rick... long time!!! well my thoughts on the above issue is that the limits of allowance is what we set.. it can be 80-120 or 70-140 or 95-105 depending on the stringency of your assay!!!

secondly if your spike recovery is fine and dilutional recovery is bad it means that at that particular dilutions tere is still sum kind of a matrix interference (again wen will u call it interference is upto the assay limits!!) while at those dilutions the matrix is not interfering with the spike sample as the amount of detectable alalyte has gone up by some factor now which is out of the interfering range of the dilution:analyte ratio may be!!! (i prefer the spike being the lowest point of the curve!!)

I hope i m not confusing you... essentially what i mean is that the limits of acceptance depends on the assay and tere is no set rules of recovery!!!

It can also be that your dilutional linearity is very good but spike does not recover at that dilutions!!!

for an immunoassay i wud say that you can try different dilutions and the dilutions from were u get the same amt of analyte concentration after multiplying with the dil factor (again inside the limits defined by u) can be adjudged a reliable range of estimation!!

If you are making a commercial product you would want your recovery to be within 95-105% and serial dilutions to be linear. There are no set rules...except how good do you want your test to be?

Recovery test is usually made by creating several pools low medium and high, preanalyzing them and spiking with a known amount of analyte usually at several concentrations.

Linearity serial dilutions of high sample in parallel with high standard both should lie on top of one another. If there is a matrix interference the dilution of the sample will deviate from the standard usually at the higher concentrations...higher matrix.

thanks a lot for the response....

well i would like to know..if all of you do a "dilution response curve" for samples???

also.how do you decide up on "minimum dilution point"?? what all parameters or criteria is look up on before u fix it??

Hello, I mostly do methods comparison plots between the assay under investigation and the 'reference' method(s). If the slope or intercept is incorrect...examine the standards or calibrators.

Also perform residual analysis by plotting the difference between the observed-expected v. expected values.

receiver operator curves.

If you believe matrix effects or heterophile antibodies these can be examined by dilution; with heterophile ab use animal sera.

If you believe you have a prozone effect then dilution will also assist.

I never heard of 'minimum dilution point' however, it is suggested to try to dilute actual samples as much as possible to minimize any matrix diferences. but if your assay requires a certain sensitivity you may lose it as the sample dilution increases.

hope this info helps you.