It’s an attempt to standardize a detection limit.
However, here is the issue. When calibrating an instrument, you establish a linear dynamic range.
Any quantitation that hits outside that range is considered estimated. This is true for both environmental work, and pharma.
So suppose you have a hit below your calibration range. The correct course of action is call that result estimated (J).
Now, let’s say you run an MDL study. The study itself is a farce because it doesn’t take into account the actual limits of signal to noise. What it does is it creates an approximated statistical value that is based on the recoveries of the extraction methods. Not what the instrument can actually detect. This is a problem because the detection of the instrument is what’s going to pull the quantization data. But let’s just say, that it does.
So, you have a kind of range established now.
ND → MDL-> LLOQ-> LDR-> UCP
ND= non detect
MDL = method detection limit
LLOQ = lower limit of quantitation
LDR= linear dynamic range
UCP = upper calibrarion point.
Hits between the LLOQ and the UCP are considered (mostly) to be accurate. Anything above the UCP and the result is flagged J for estimated result.
Anything below the LLOQ is also considered estimated as well, and the result is flagged J as well, for estimated result.
Now this is why MDL’s are ridiculous, any result that is less than the LLOQ but greater than ND (inside the MDL range) is, flagged J anyway, for estimated result. So reporting down to the MDL results in J flags anyway. But the issue that is NOT addressed, and should be, is: what is the signal to noise ratio with that “hit.” MDL’s don’t address that. And if the S/N is too small (typically >3) it isn’t considered a “hit” but according to the MDL it may be. It can be a source of a type I error. But it gets ignored because “well, MDL!”
Now I suppose when working at trace levels, type I is a bit better than type II, but often times the action level is above the MDL so it’s moot anyway.
Because any result below the LLOQ is flagged J anyway, it makes MDL’s redundant and that’s why the EPA regional labs typically don’t evaluate them themselves. But as “guidance criteria” they expect commercial labs to.
It’s a case of, what’s sauce for the goose, isn’t sauce for the ganger.
That’s why the running gag for hypocrisy in the environmental analytical chemistry industry is “is it ‘guidance criteria?’” Said tongue in cheek.
And who sees Elephants in space anyway and aren’t battleships or even frigates much larger?