# LOD & LOQ by regression line

can anyone please help me to calculate LOD & LOQ from a linear curve.I need help in excel calculations.plzzzzz.

Thanks
Suri.

You cannot get LOD / LOQ from linear curve. You can only get bias of a test.

Hi

Thank you for your reply. But in ICH guidelines it is mentioned to calculate the LOD and LOQ by a linear regression curve.Then what is the best way to get LOD and LOQ’s for impurities in a HPLC method?Please suggest.

Hi Suri,

There are several ways for getting LOD & LOQ values like;

1. Go with signal to noise ratio.
LOD s/n ratio should be about 3.
LOQ s/n ratio should be about 10.
2. Go with Standard deviation of response & slope: formulae given in ICH Q2.

Hi Satish,

How to calculate the standard deviation of response and slope in excell.I never done this before.Can you please help me out??

Thanks

i am not sure if i am doing the correct way. Inject the analyte solution (at 100% concentration) at least 6 times. Then, you calculate the standard deviation. And slope (m) is obtained from linear curve from linearity (spiked spl)

LOD = 3 x stdev/slope

LOQ = 10 x stdev/slope

you may try doing it, and improve with advice from you authority.

regards

how to calculate 1)%of recovery 2)number of theoritical plates in column
3) PURPOSE OF LOD, LOQ

Dear Mr. Surinzp,

The easiest method of determining LOD and LOQ is by using the formulae mentioned by Mr. choy77. But the question still remains how to do it in Excel. Well assuming for linearity study you have two variables i.e. concentration (referred to as X variable) and response/assay (referred to as Y variable). First determine slope of the linearity curve by using formula (in Excel) S =SLOPE(Y values, X values), secondly determine standard deviation of line (also known as standard error of the predicted y-value for each x in the regression) by using formula R =STEYX(Y values, X values) and finally determine LOD and LOQ by using formulae LOD = 3R/S and LOQ = 10R/S.

Hope this helps.

Ovais

Dear Ovais,

I used your excel formulas and it worked. Thanks for your help.

Thanks
Suri.

Remember these values are estimates. You may still have to verify them. by preparing solutions at the LOQ and calculating the % RSD of the replicates.

Jay

Others are based on the dispersion properties of the linear regression, simple or weighted. the best way to deal with the detection and quantification was based on a method using the residual standard deviation of the limit of weighted regression for detecting and quantifying Eurachem approximation.

LOD = 3.3 SD of regression line/slope or 3.3 SD of y-intercept of regression line/slope

Do you know the link or free ebook to find the formula of SD of y-intercept of regression line?

In a source, i have read that LOD is from SD y-intercept of regresion line/slope
but i dont know to calculate the SD y-intercept of regresion line

maybe you can help me to give the valid source

sorry for mess english, Im Asian

thank you

SD is not standard deviation in the formula 3.3* SD/ Slope

This SD is called Residual standard deviation.

In EXCEL this formula is steyx(known y, known x).

inject about 5 lowest concentrations (ex: 0.1 to 0.5%) and get the area of peaks.

Now calculate SD by using excel formula “STEYX” and “SLOPE”

Best Regards,
Bujji Reddy Kanchi.