MINIMUM DETECTION LEVEL

If I am going to write a blog entitled The Thinking Chemist, I should at least write on topics relating to chemistry once in a while, so here goes.

The minimum detection level, or MDL, of an analyte, is statistically defined as the lowest concentration of that analyte that can be detected reliably 95% of the time under a given set of conditions, equipment, and analyst.  This is determined empirically and cannot be determined in advance by any sort of calculation, and must be determined by actually running tests on a standard concentration of a known analyte. It can change if major changes are made in the test equipment, testing done by a different technician, or even being done in  different room in the same building.

No test, no matter how well and how accurately performed will give you exactly the same number in 100% of the repetitive samples run.  There will always be a slight deviation that should be around a very low number, which will obviously be different for each analyte.

After one runs the repetitive samples  their standard deviation is determined and then multiplied by the Student #, which is different for different numbers of samples.  The resultant number is the MDL for that analyte under those conditions.