I’m a few more episodes in now, and as a scientist, I’m offended by that EDTA test. The defense witness they showed brought up many of the right points, but seemed to miss one important one.
The right points that she brought up are the limit of detection (LOD) and what a negative result actually means. Here’s how one would develop this assay. First, you would spike EDTA at increasing concentrations into an ideal buffer. You would then set up an experiment to reasonably detect that spiked in EDTA and show that in the absence of EDTA, you don’t get a signal, and at increasing concentrations you get an increasing signal. This would validate the assay (under those ideal conditions) and provide an LOD for that assay; in other words, what is the lowest concentration of EDTA that it is possible to detect (again, under those ideal circumstances). The important question is whether that LOD is actually higher than what would be in the blood sample. If the LOD is higher than the expected concentration, then obviously the assay won’t work to detect EDTA in your blood sample. In this case, they didn’t provide an LOD. Without an LOD, this assay is 100% meaningless. That the defense attorneys allowed them to get away with this is insane. Demand access to the raw data and determine the LOD.
Because it gets to point 2 that the defense witness brought up, which is that a negative result, particularly a negative result without an LOD provided, either means that their is no EDTA (which is the conclusion that the prosecution obviously wanted) or it means that your assay is not able to detect the low levels of EDTA. Without an LOD, either result is equally likely. Actually, I would tend toward believing it is more likely to be an assay problem than anything else.
Which brings me to point 3 that the defense expert didn’t bring up. The prosecution did not provide any details about the EDTA assay. As I mentioned, step 1 would be spiking EDTA into an ideal buffer and making sure that you could detect it. Step 2 would be trying to replicate the actual sample. In this case, the assay would have to show not that they are able to detect EDTA in ideal buffer, or even spiked into blood (which would be more challenging), but that they are able to detect EDTA when it is present in the same concentration that it is in Steven Avery’s blood sample AND when that sample has been smeared on to a surface, allowed to dry for a long time and then swabbed and tested.
There is no indication that they did this, and any scientist willing to sign off on the release of this data without an LOD probably didn’t do this. If you haven’t shown that your assay can detect EDTA that is present in low concentrations in dried blood, then your assay is unvalidated for this purpose. Without providing this data their data that there was no EDTA in that sample is meaningless. I find it very likely that their assay is completely incapable of detecting EDTA under the conditions that the blood in the car was held. In fact, I find it more likely than not that their assay would be completely unable to detect EDTA in those real world conditions, and it would be trivial for them to do it the right way to prove otherwise.