What were the results of the EDTA test?
The FBI released a 637 page report detailing the process of their EDTA test. Some very helpful Reddit users have broken down the information here and here. Dr. Marc Lebeau, chief of the Chemistry Unit at the FBI, gave extensive testimony regarding the test on March 5 and March 6, 2007. Janine Arvizu, an independent Laboratory Quality Auditor, gave testimony on March 9, 2007. There are a number of common questions related to the test:
What is LC/MS/MS anyway?
LC/MS/MS is tandem mass spec- essentially more powerful than regular ol' LC/MS. In very simplified layman's terms:
LC is liquid chromatography, which is added to MS for another layer of assurance in mixed samples like blood. It is used to separate, identify, and quantify different components in a mixture. The liquid sample is passed through a pump containing an absorbent material, and different components of the mixture will get "stuck" to the material slightly differently, therefore flowing out of the pump at different rates. Imagine, for example, pumping different objects through a magnetized tube. A non-magnetic material like paper will pump through very easily, while objects like paperclips will become "stuck" to the magnetized tube and take longer to pump through.
MS is mass spectrometry. It is a process that bombards a compound (blood, in this case) with energy, breaking it up into separate ions. The ions are then accelerated and subjected to an electronic or magnetic field, which "sorts" them based on their mass-to-charge ratio. Results display the relative abundance of detected ions, which is referred to as a "chemical fingerprint." That "fingerprint" (or fragmentation pattern) is searched against a database of known chemicals, and identified.
Was there a Limit of Detection (LOD)?
One LOD was established to determine how small of a concentration of straight EDTA the test is practically capable of finding in a clean sample- that was 13 ug/mL (13 PPM). The tube maker's website states the tube sample would have 1800mg/L of EDTA in it. The test could detect amounts less than 1% of what would've originally been in blood from the vial.
They also tested blood in various sizes dried onto solid surfaces, swabbed up, and put through their protocol to determine the smallest detectable size of blood spot they could reliably find EDTA from. They tested 1 uL, 2uL, and 5uL samples. EDTA was readily detected in the 1uL and 5uL samples. It was indicated in the 2uL sample, but based on a problem with the ion levels, the analyst marked it as "not detected." An LOD is set at a level where you can detect your substance reliably 50% of the time you know it's present. That means at levels just above and below you may or may not detect all of the time, but the further away you get, the more reliable the test becomes. It's not a surprise the 2uL sample didn't work perfectly, as it was so close to the LOD. By the time they got up to 5uL, they were 100% accurate in detecting EDTA in known positive samples.
To give an idea of the sizes being used: 1uL of blood would create a drop about 1/50 the size of a penny.
Wasn't this just some dusty old procedure that had only been used in OJ's case and never again since? And didn't it fail in OJ's case?
The EDTA test was developed in 1995 and peer-reviewed in 1997. In the OJ Simpson case, a sample was initially found to contain EDTA, but it was later discovered that was a carryover from EDTA samples previously run through the machine. In the test during the Avery trial, new equipment and methods were used to ensure no carry-over. The new instrumentation and additional checks went through the FBI's QA division, which separately signed off on the protocol. LeBeau and another lab member were tested for blinded control samples to determine their ability to detect EDTA presence/absense based on pre-set FBI criteria for making that call. Both passed 100%.
The EDTA test was brought up in a separate case, Cooper vs. Gaughnour. The blood in question in that case was a tiny (about 1uL) drop found on a T-shirt that contained both the victim's blood and the defendant's blood. The T-shirt had been washed with a detergent containing EDTA. Therefore, all spots on the T-shirt would have EDTA present, possibly tainting the results, and it was not used in trial.
It should be noted in both prior cases involving the EDTA test, the test erred in favor of the defendant- finding EDTA in non-preserved blood. It did not fail to detect EDTA in planted blood. Problems from the OJ trial were resolved with updated testing, and problems from the Cooper case were resolved by testing control samples from other areas in the car to ensure it had not been cleaned with an EDTA-containing product.
Didn't the FBI Chemistry Unit make up their own procedure and not use something peer reviewed?
No. After the OJ case, research scientists (separate division from LeBeau) at the FBI peer-viewed the method and published their findings in the Journal of Analytical Toxicology.
Some minor updates were made to that procedure as noted above. In addition, negative control swabs were taken from areas next to the stains to determine whether a positive in a bloodstain might be attributable to environmental EDTA- all negative controls were negative when tested.
Didn't the defense not have an opportunity to do their own testing?
The FBI saved portions of each sample in case the defense wanted to have independent testing done. See this page regarding the timeline of the blood vial in evidence. The defense had ample opportunity to test the blood, if they so desired.
Doesn't EDTA break down over time?
The FBI actually performed a stability test with 33 month old EDTA dried blood spot cards. Scientific literature indicates that EDTA does not break down readily absent harsh intervention- and that makes sense because it's a preservative. The 33 month old DBS cards showed the presence of free EDTA (EDTA is in excess in vials) in 100% of cards tested, but only showed the Fe-EDTA complex in 60% of cards. Another study shows Fe-EDTA can break down faster with exposure to UV light, which may explain it's absence in 40% of those samples. There was a lawyer versus scientist argument over degradation versus dissociation, but EDTA was ultimately still detectable in all of the nearly 3 year old cards.
EDTA wasn't ruled detected in the 2uL sample of preserved blood. Couldn't the same thing have happened in the swabs from the Rav 4?
EDTA was indicated in the 2uL sample of preserved blood, but because the ions were "off" it was ruled not detected. There was no indication or evidence of EDTA in the swabs from the Rav 4.
Could Manitowoc County officers dilute the blood before planting it, making EDTA undetectable?
According to Bruce McCord, who leads a forensic research team at Florida International University, it would be almost impossible. Because the testing equipment could detect EDTA in amounts less than 1% of what was used in the vial, the blood would have to be so diluted it would appear to be water. Blood that diluted would probably not show up as a blood stain in Teresa's vehicle.
So does this conclusively prove the blood wasn't planted?
No. In a test like this, it is virtually impossible to prove a negative. However, this test could detect EDTA in proportions much smaller than what would've been present in the vial, and do so very reliably. An EDTA degradation study done over more than two years showed EDTA does not easily break down in dried blood stains, and these blood stains had only been drying for a few days before they were swabbed. The test was so sensitive that diluting the mixture before planting the blood to the point that EDTA was undetectable would result in a stain that no longer resembled blood. While it is still technically possible the blood was planted from the vial, the results of this test make that extremely improbable.