"My client was arrested on July 4th weekend on a no refusal initiative," said Attorney [and former Forensic Science Commission chairman] Sam Bassett.
Bassett said his client was forced to surrender his blood. The sample was tested at the Austin Police Department's crime lab. The results, in a report 18 days later, revealed the blood alcohol content tested at 0.10, just above the legal limit.
"There's always a question when you are dealing with law enforcement crime labs, in my experience there are errors," said Bassett.
After fighting for 10 months, Bassett finally got a court order to have the blood sample retested. He sent it to the Southwestern Institute of Forensic Sciences in Dallas, an accredited and nationally recognized lab. And weeks later, he said the results were surprising.
"It was very surprising to me that there was such a difference," said Bassett.
The retest showed a problem, APD's analysis was 20 percent different.
The sample tested by the Southwestern Institute of Forensic Sciences now at a .08, right-on the legal limit.
One sample. Two different results. So what happened? Why did APD test higher?
"At least half of the ones that I have seen were performed incorrectly," said [attorney] Ben Florey.
Florey said he's seen this many times before. He says he's had to help a number of clients fight what he calls "bad science".
"They're back logged, and I would imagine that they make mistakes," said Florey.
"That starts to raise issues about the quality of the machinery, whether it's being properly calibrated. Whether the people doing the calibration are qualified to do it," said Attorney Bill Mange.
Wednesday, November 09, 2011
High error rates discovered for Austin PD blood tests in DWIs
When the defense had a blood sample in an Austin DWI case retested, a private lab came up with results 20% lower than the Austin PD blood test given on one of the city's "no-refusal" weekends. Reported My Fox Austin:
Subscribe to:
Post Comments (Atom)
.jpg)
47 comments:
I think it's fair to say that blood - as a source - yields an individualized and much more accurate result than does an arrestee's breath. However, if the testing process is flawed . . .
In any modern industrial endeavor if you can't distinguish good parts from bad parts you don't get to play.
Go to Toyota and tell you have invested in a state of the art hood ornament machine. Offer them a price that is 40% below their current supplier. Tell 'em you have retained the the Austin crime lab for metrology services. They'll laugh you out of the conference room.
In industry ISO and TL9000 standards establish creds for whatever company adopts them, Then you can have confidence in their products and processes. One way is the assurance that equipment is calibrated and maintained according to manufacturer's recommendations and companies must maintain those records for inspection. If anything goes wrong and those policies have been violated, they can be held accountable. Certainly we should ask our criminal justice departments to do no less. Standards should exist and they should be held accountable when they are not adhered to. People's lives and freedom are at stake. time for the legislature to change " recommended" to "mandated"; that "standard policies procedures" be set and adhered to.
Is this based on the same story where they interviewed 4 or 5 defense attorneys and not a single prosecutor? The same story where Faux News couldn't even get a return phone call from this "independent lab" from someone who was willing to discuss their procedures? The same story where the entire invesitgation was based upon one single blood test result (out of the hundreds or more done every year) and where Bassett's client STILL ENDED UP TAKING A PLEA? It looked like Chief Acevedo got ambushed with questions that were outside his area of expertise, and the reporters weren't willing to accept a completely reasonable response of the chief asking for time to get more information (he isn't a lab technician, after all). That Faux News piece was one sided, opinion-driven hack journalism at its worst. I was ashamed to see it coming from an Austin television station.
20 percent difference on a retest? I guarantee you that folks in both labs will say that a 20 percent (.02) drop 10 months later is within the range of normal and says nothing about the quality of the testing in APD. They really will testify to it. Really.
Twenty percent is nothing. I have a pending case where the .10 whole blood original test came back as .05 on a retest and the .14 serum came back as .10 on a retest, though the retest was two years later. The State plans to try to have some state employees testify that even a 50 percent decline is normal and has nothing to do with the quality of the original test. It matters not to the State that there is no scientific basis for such "opinions." They need the testimony so state crime lab people are going to give it to them. Those law enforcement crime lab folks do, after all, have to stick together.
Remember: this is government testing where nothing is ever bad and all is explainable. So long as they have evidence that they can say points to guilt, it is, in their view, always good evidence. They worry that if they ever say that anything they do is bad, then everything they do will get questioned. They also worry that if they blow the whistle on anything being done wrong, they will lose their job and career. Just look how well blowing the whistle worked out for HPD's Amanda Culbertson and Jorge Wong.
It is a myth that blood is a more reliable way to test for alcohol levels. But, that should come as little to no surprise to anyone: much of what the inherently biased "crime" labs do has serious reliability issues to everyone outside the "crime lab" world.
Troy
W. Troy McKinney
Schneider & McKinney, P.C.
440 Louisiana
Suite 800
Houston, TX. 77002
713-951-9994
713-224-6008 fax
http://www.texascriminaldefenselawyers.com
http://www.houstondwiattorneys.com/
wtmhousto2@aol.com
Board Certified Criminal Law - Texas Board of Legal Specialization
Board Certified in DWI Defense Law - National College for DUI Defense
NCDD certification program approved by the American Bar Association and
Accredited by the Texas Board of Legal Specialization.
E pur si muove, 9:40!
Lament the shortcomings of modern journalism all you want, and I join your frustration with minimalist local TV news. But they sat down with Acevedo, so in the bizarre worldview of modern professional journalism they "got both sides," and for good or ill if APD blood tests are coming up 25% higher than that reported by an independent lab, that's news.
Grits: It's not accurate to say "But a 20% error rate (on the high side, no less) means it's possible some borderline cases - where the blood test comes back close to the .08 mark - are being falsely charged," because we don't need either breath or blood to charge DWI, and can legally charge DWI for tests that come back lower than .08. .08 BAC is only ONE method of proving intoxication. The State can use the others regardless of how the BAC reads.
The lab that did the retesting (Southwestern Institute of Forensic Sciences is not a private laboratory as indicated in the original post. It is a non-law enforcement government lab operated by Dallas County. So this is not a private versus public lab issue. It may be a law enfocement versus non law enforcement lab issue.
Texas Maverick: Crime labs in Texas are accredited. Most are accredited under ISO 17025.
I'm confused. If blood with 0.1% alcohol is stored for ten months and then retested, would you expect to see 0.1% in the retest? Wouldn't it be expected that the alcohol concentration would be lower after 10 months? So the question that should have been asked is, how much lower would it be expected to be?
This is why Acevedo wanted to be able to be able to charge for having less than 0.08. "Hey, close enough, right? What's a 25% error? Now give us your $4000 and lower your prospects of becoming a productive member of society."
I notice that this guy isn't claiming that he was drinking. He just was't drinking enough to count. Well, unless you consider killing a child or two counting. But hey, why should a child or two matter.
Accreditation in and of itself does not mean a whole hell of allot in my experience.
Who audits them? Is it simply the auditing firm firm they pay to maintain their accreditation?
Are their customers provided access to audit them? And by customer I mean those effected by the services they provide.
As Paul Harvey use to say, And now for the rest of the story...
I reviewed the Fox story online and noticed the date of the blood draw was July 4, 2009, which adds a whole new dimension to this story.
From Oct. 1, 2007 through Oct. 1, 2009 APD contracted with the Southwestern Institute of Forensic Science to do Toxicology and other forensic work. Below is opening & paragraph 1 of the interlocal agreement.
"THIS CONTRACT is made and entered into by and between the ClTY OF AUSTIN, a municipal corporation, located in Austin, Texas, hereinafter called "City," and DALLAS COUNTY, acting through the SOUTHWESTERN INSTITUTE OF FORENSIC SCIENCES AT DALLAS, hereinafter called "County."
"1. DESCRIPTION OF WORK For the consideration hereinafter agreed to be paid to County by City, County shall provide toxicological, environmental and physical evidence analysis and other similar forensic analytical services for the Austin Police Department, in the City of Austin and County of Travis, Texas, hereinafter called the "Services," The Services are to be performed in a competent and professional manner, and performance shall conform to applicable professional standards for the Services. County shall also perform the Services in a timely manner, consistent with the needs Of the Austin Police Department."
This being the case, it is possible the Southwestern Institute of Forensic Sciences could have been the author of both test, but even if it wasn't, it had a duty to disclose to the defense attorney Bassett, that it was under contract with the Austin Police Department. This holds true for APD Chief Acevedo, as he is one of the Austin officials who executed this Interlocal Agreement.
It must also be noted the Southwestern Institute of Forensic Sciences is not part of the Dallas County Government, it is a division of the U.T. Southwestern Medical School, a State institution.
The Dallas County Crime Lab's name is the, "Institute of Forensic Sciences". Dr. Jeffery Barnard is county appointed director of the "Institute of Forensic Science" and also claims the title of Dallas Co. Chief Medical Examiner through an appointment from the U.T. Southwestern Medical School.
The toxicology expert cited in this story, "Dwain Fuller" use to be employed by the Southwestern Institute of Forensic Sciences, something he should have disclosed to the reporter.
I have already spoken to the Fox reporter and will be talking to him again this week, to learn the name of the Dallas Toxicologist who authored the report, which will likely be a problem as well, but I will wait before discussing that issue.
3:43, keep telling yourself that if it makes you sleep better at night, but IMO false DWI arrests likely account for the much higher dismissal rate for DWI cases in Austin. And you can ask Jeff Ward how reliable those other indicators are.
David, that's a fascinating Paul Harvey moment. Looking forward to the followup.
One of the things that accredtation requies is that a laboratory has an adequate plan for instrument maintenance and calibration, and that it exutes this plan to assre accuracy of test results.
Really, GFB appears to be leaping to judgment here on the basis of partial and suspect information - a fault he finds offensive in others, but which apparently is a perfectly fine thing to do when thats all you've got.
6:29, how great a leap did I really take? I wrote that "it's possible some borderline cases - where the blood test comes back close to the .08 mark - are being falsely charged." Notice the key word "possible." I leaped to no conclusions, only acknowledged the possibility.
Dfisher: I would expect that the original lab report that Bassett saw would have been printed on letterhead that would have identified the lab that did the testing. Is this incorrect? So would he have sent the samples for retesting to the same lab that did the original work? I don't think so.
And it's not clear why an expired interlocal agreement would constitute a conflict of interest in anyone's mind. The testing was done long after the agrrement ended.
Am I missing something here?
10:20:00 PM If the sample is stored properly you would expect the measurement to be the same. At least within the precision of the measurement system.
If stored improperly, the percentage of alcohol can go up or down.
So it might not be a testing issue at all? It might be a storage issue?? Gosh, that didn't seem like an option from the original post.
And can I ask, why is it being assumed that the problem was with the original lab work, and not with the lab that did the retesting?
3:43 here, Grits. I'm not "telling myself that." It's actually THE LAW. You gotta problem with State law, you know where they keep the Legislature. You gotta problem with prosecutors following and using the law the Legislature writes? Then there ain't no way to make you happy. The alternative is prosecutors just deciding they know better, and you don't seem to like that, either. Bad -- as in unsupported by sufficient evidence -- DWI arrests may be happening -- and getting dismissed in Austin, because the cops are poorly trained, stupid, or truly malicious, but that doesn't change the fact that a DWI arrest with a BAC below 0.08 is not, by law, a BAD DWI arrest. I sleep fine at night. I'm sure you're kept up at night by all these misinterpretations you insist on publishing.
I understand that's the law, 3:43/11:27, but I also understand that Austin PD has many more cases dismissed than other jurisdictions, apparently because they arrest on borderline cases that prosecutors frequently can tell won't result in sustainable convictions.
You can pretend all those cases aren't getting thrown out and everything is fine, if you like, but it doesn't change the fact that those other indicators aren't nearly as definitive as breath or blood tests. Nor does it reduce the likelihood that flawed forensics contributes to false DWI convictions. We're talking about a not-insignificant error rate, either by the Austin lab or the one in Dallas.
A couple of points: I was trained in medicine. It's a truism in medicine that labs can err. One almost always retests a lab report that appears out of whack.
2nd, I used to work at Travis Cty Jail. It was common to hear cops say, "Arrest 'em all and let the courts sort it out". That is likely to real source of the high dismissal rates. Indeed, back then it was reported that about a half of all those arrested had charges dismissed when a prosecutor had a look. But the arrest alone causes people their jobs, costs them a lot of money, and unfairly taints them with an accusation that will appear for years in background checks.
10:36:00 AM It would be an evidence preservation issue. A pretty damn important function of any lab assuming you think validation is important. If the BORs is not so important to you then push for amending them.
11:27:00 AM You do understand that the BAC level is a per se law. Precision is mighty damn important if the ledge sets a pass fail standard.
Yes we understand that one can be charged and convicted of DUI based on evidence other than BAC level. That's because the statutory language is "under the influence" not "while intoxicated". The per se "under the influence" BAC level was set by the legislature under threat of withholding federal highway dollars.
11/10/2011 07:22:00 AM,
You didn't miss something, I just didn't disclose all in my post, so here's something I withheld.
In my post I mentioned Dr. Barnard is the Dallas Co. appointed Director of the Crime Lab and also claims an Appointment from the U.T. Southwestern Medical School as Medical Examiner, (May 5, 2011 Federal Court Deposition) so everything he does must be held to the highest legal standard.
It here where both the Dallas Crime Lab and Medical Examiners office fail the smell test.
On August 18, 2009 the Travis Co. Medical Examiners Office received Accreditation by the National Association of Medical Examiners. (NAME) Based on the number of autopsies Travis Co. did at the time, they would need 5 ME's to meet the accrediting requirements, but Travis Co. had only two ME's, so what do they do? Since the Travis Co. Chief ME worked as a Dallas Co. Deputy Chief ME under Barnard until 1996, he went to Dr. Barnard and asked for his help in getting Travis Co. accredited. Now Dr. Barnard's ME's office is NAME Accreditation, so the scheme they came up with has implications to the Toxicology issues between Austin and Dallas, the subject in the original post.
The scheme between Travis Co. Chief ME David Dolinak and Dallas Co. Chief ME Barnard was, Travis Co. would submit the names of Jeffery Barnard, Keith Pinckard & Amy Gruszecki to NAME, as Travis Co Medical Examiners even though they were actually Dallas Co. ME's. This scheme worked and NAME believing Travis Co. had the 5 required medical examiners, accredited the Travis Co. Medical Examiners office. Below is the response to my open records request for the names of Dallas Co. ME's performing autopsies at the Travis Co. ME's Office in 2009.
**********
Dear Mr. Fisher:
Here are the names of the doctors that your requested:
Jeffrey Barnard
Keith Pinckard
Amy Gruszecki
IF you have any questions regarding your request and this response, please let me know.
Sincerely,
Daniel
Daniel C. Bradford
Assistant County Attorney
Travis County Attorney's Office
P.O. Box 1748
Austin, Texas 78767
Phone: 512.854.3718
Fax: 512.854.4808
email: daniel.bradford@co.travis.tx.us
*********
It is these cross agreements between the Travis Co. ME's office, the APD Crime Lab and the Dallas Co. ME's Office/ Crime Lab that the question of fraud arises. Actually, there is no question Dr. Barnard conspired with Travis Co. to fraudulently accredit the Travis Co. ME's Office. The only question here is, what type of service was Southwestern Institute of Forensic Sciences contracting with APD? Was it to simply run the test in Dallas, or were Dallas Co. toxicologist actually going to Austin the run APD's Blood Test? If the latter is true, then the same toxicologist could have run both test in Bassett's case.
It is also possible that Dallas Co. Toxicologist were running the test in Dallas and APD Toxicologist were putting their names on the results, claiming it to be their work. Having review the work product of both APD and Dallas, all these scenarios are within the realm of possibilities.
11/10/2011 07:22:00 AM,
You didn't miss something, I just didn't disclose all in my post, so here's something I withheld.
In my post I mentioned Dr. Barnard is the Dallas Co. appointed Director of the Crime Lab and also claims an Appointment from the U.T. Southwestern Medical School as Medical Examiner, (May 5, 2011 Federal Court Deposition) so everything he does must be held to the highest legal standard.
It here where both the Dallas Crime Lab and Medical Examiners office fail the smell test.
On August 18, 2009 the Travis Co. Medical Examiners Office received Accreditation by the National Association of Medical Examiners. (NAME) Based on the number of autopsies Travis Co. did at the time, they would need 5 ME's to meet the accrediting requirements, but Travis Co. had only two ME's, so what do they do? Since the Travis Co. Chief ME worked as a Dallas Co. Deputy Chief ME under Barnard until 1996, he went to Dr. Barnard and asked for his help in getting Travis Co. accredited. Now Dr. Barnard's ME's office is NAME Accreditation, so the scheme they came up with has implications to the Toxicology issues between Austin and Dallas, the subject in the original post.
The scheme between Travis Co. Chief ME David Dolinak and Dallas Co. Chief ME Barnard was, Travis Co. would submit the names of Jeffery Barnard, Keith Pinckard & Amy Gruszecki to NAME, as Travis Co Medical Examiners even though they were actually Dallas Co. ME's. This scheme worked and NAME believing Travis Co. had the 5 required medical examiners, accredited the Travis Co. Medical Examiners office. Below is the response to my open records request for the names of Dallas Co. ME's performing autopsies at the Travis Co. ME's Office in 2009.
**********
Dear Mr. Fisher:
Here are the names of the doctors that your requested:
Jeffrey Barnard
Keith Pinckard
Amy Gruszecki
IF you have any questions regarding your request and this response, please let me know.
Sincerely,
Daniel
Daniel C. Bradford
Assistant County Attorney
Travis County Attorney's Office
P.O. Box 1748
Austin, Texas 78767
Phone: 512.854.3718
Fax: 512.854.4808
email: daniel.bradford@co.travis.tx.us
*********
It is these cross agreements between the Travis Co. ME's office, the APD Crime Lab and the Dallas Co. ME's Office/ Crime Lab that the question of fraud arises. Actually, there is no question Dr. Barnard conspired with Travis Co. to fraudulently accredit the Travis Co. ME's Office. The only question here is, what type of service was Southwestern Institute of Forensic Sciences contracting with APD? Was it to simply run the test in Dallas, or were Dallas Co. toxicologist actually going to Austin the run APD's Blood Test? If the latter is true, then the same toxicologist could have run both test in Bassett's case.
It is also possible that Dallas Co. Toxicologist were running the test in Dallas and APD Toxicologist were putting their names on the results, claiming it to be their work. Having review the work product of both APD and Dallas, all these scenarios are within the realm of possibilities.
dfisher - Using the word "fraud" to describe the work arrangement between Dallas County and Travis County to help deal with a short term staffing shortage in Travis County seems a bit extreme. That would require the situation was misrepresened to the accrediting body. That seems pretty unlikely since the Dallas Office is accredited, and Barnard would be known to the accrediting organization as the director of the Dallas office. It seems more likely that this was a temporary fix implemented with the approval of the accrediting body.
So I guess my question is, when you use the word "fraud" are you doing that for dramatic effect (not a wise thing to do generally) or do you have information from the accrediting organization that this is a problem for them?
11/10/2011 06:31:00 PM,
I did not use the word fraud for dramatic effect, I used it for its legal definition.
The Texas Code of Criminal Procedure, Art. 49.25 limits one chief medical examiner per district, requires all deputy medical examiners to be employees of the chief medical examiner of the county and doesn't give the county, the chief medical examiner, or any division of the state the authority to circumvent the statute.
Article 49.25 doesn't require an autopsy in unattended deaths.
The National Association of Medical Examiner sets limits on how many autopsies a medical examiner can do per year to be accredited, but doesn't have a clause that allows its members to circumvent its rules to get other offices accredit.
So when the Travis County Commissioners Court, the Chief Medical Examiner Dolinak and Dr. Barnard and company conspired to violate the the Sections of Article 49.25, it is actually conspiracy to commit fraud, under the legal definition.
dfisher - So I take that to mean that the accrediting body did sign off on the arrangement. Kind of hard to see fraud/deception in that case.
10:17:00 Why do you think the accrediting body might set minimum staffing levels for an organization based on the amount of work load required of the personnel? Do you think these standards might written with regards to the time required to do the task properly and professionally? If the work load is 2.5 times greater than the maximum set by the standard do you believe the accrediting body can reasonably assume that the agency seeking accreditation is not committed to providing the necessary resources to maintain the work product required by the standard?
dfisher-
FYI...SWIFS is Dallas County. Barnard answers to County Commissioners.
http://www.ascld-lab.org/cert/cert262.pdf
The name on the lab report from SWIFS is Dr. Erin Spargo, Deputy Chief Forensic Chemistry (as seen in the news video).
It would be nice if measurement uncertainty didn't exist and every measurement was precise to the n-th degree. But that isn't the way things work. It's telling to me that people are quick to jump on this as a lab testing mistake issue without being at all concerned about the technical sources of interlaboratory variation, with what range is typically seen in interlab variation, with the reasons why a stored blood sample may change over time, or with what range of changes might be expected in this instance. If uncomplicated certainty is what you want, well you shouldn't be looking to science for that. Science isn't about certainty. It's about quantifiable uncertainty. That's unfortunate perhaps, but that's the way it is.
5:43, that's why I think it's something the Forensic Science Commission should examine.
GFB - The legislated investigative mission of the TFSC is limited in scope - allegations of serious negligence and misconduct. It's really about identifying the Fred Zains out there. And it is constituted for that mission - look at all the lawyers on it. For what you want you need a different commission - one comprised of scientists to deal with issues of science, measurement, statistics, etc. Even ME's bless their hearts would be useless to evaluate these types of issues. But good luck getting a commission like that.
Anon 1:07: "Yes we understand that one can be charged and convicted of DUI based on evidence other than BAC level. That's because the statutory language is "under the influence" not "while intoxicated". The per se "under the influence" BAC level was set by the legislature under threat of withholding federal highway dollars." Wrong. The only "under the influence" laws in texas pertain to underaged drinkers, because of 0 tolerance for minors. The other criteria they are talking about is not having physical and/or mental control of one's faculties. The statute uses language "driving while intoxicated." Read SS49.04.
5:43:00 Measurement uncertainty is inherent in all measurement systems. It's accounted for and dealt with successfully day in and day out in an infinite number of disciplines. It's a simple fact of life so there's no need to pretend like your the Lone Ranger.
I think a pertinent question is did the ledge take measurement uncertainty in to account when they adopted the standard? How much of the tolerance does the measurement system eat up, 5%, 40% or 110%?
Both labs pay fees to ASCLD/LAB for accreditation, which is required by law.
If the accreditation agency (ASCLD/LAB) was worth a flip, then it would step up to identify the differences in scientific results.
Then again, if ASCLD/LAB was worth a flip, then we wouldn't need a FSC.
Some Members on the FSC also pay ASCLD/LAB for their own lab accreditation.
So don't expect the FSC to throw stones from their glass houses, lest they want to jeopardize their own accreditation status.
Does anyone else see a problem with this arrangement?
"The other criteria they are talking about is not having physical and/or mental control of one's faculties."
If cops and prosecutors find themselves having to fall back on this provision frequently, I think we will end up seeing this provision invalidated because it is unconstitutionally vague. Just how do you define "not having physical and/or mental control of one's faculties?"
10:29 - It's hard to see how the legislature could have taken measurement uncertainty into account, since it is process specific and dependent on the particular instruments and processes used by a particular lab at a particular point in time. And from what I've heard, the legislature isn't typically that detail oriented.
11/11/2011 04:16:00 AM,
The accreditation certificate from your link uses the name of lab that was submitted to ASCLD. These accrediting organizations have no legal understanding of who they are accrediting. The below link will take you to the "Dallas County Government" site. Here you will you will find the county's lab is named "Institute of Forensic Sciences", not "Southwestern Institute of Forensic Sciences".
http://www.dallascounty.org/department/forensics/forensics_index.php
I know this is confusing, but that's what Dallas Co. and U.T. Southwestern Medical School intended when they gave these separate labs similar names.
It is this reason Dr. Barnard testified he holds an appointment from the Dallas Co. Commissioners Court as director of the Dallas Co. Crime Lab, (Institute of Forensic Sciences) and has a separate Co-Appointment from the U.T. Southwestern Medical School as Medical Examiner. This is a clever play on words between Dallas Co. and the State medical school, but in the end it will not protect them.
Anon 11:30. They don't use it by itself a lot, and when they do they have video to back them up. Sometimes BAC will be challenged on the technology, or operator error, or something like that, but if someone has a .20 it will likely be pretty obvious that he/she is not in control of their normal faculties.
Don
So the statute does not say under the influence.
What does it say?
One is a functional definition.
The second is a per se definition.
One can have the normal use of ones mental or physical faculties and still be convicted under per se condition (B)
So unless you're just an ignorant ass it's pretty damn obvious that innocent people will be convicted if the equipment is not accurate.
(2) "Intoxicated" means:
(A) not having the normal use of mental or physical faculties by reason of the introduction of alcohol, a controlled substance, a drug, a dangerous drug, a combination of two or more of those substances, or any other substance into the body; or
(B) having an alcohol concentration of 0.08 or more.
12:59 It's pretty obvious that at least some in the ledge have a clue about the scientific validity of breath testing because they specifically prohibited the FSC from investigating it in the enabling legislation.
5:43 - Sure measurement uncertainity is an issue in other disciplines and types of test. That doesn't mean people generally understand what it means or how it is significant in decision making.
For instance, let's say you are driving and your BAC is 0.80%. Your blood is drawn and sent to 10 different labs that all perform exactly the same test that has a standard deviation of +/- 2%. The average of all the lab's results is 0.80%. That makes it seem like a good test. Until someone points out that 50% of the labs would have gotten a result less that 0.80% (i.e., the test indicates you were innocent when you are really guily - a false negative result) and 50% of the labs would have gotten a result that was greater than 0.80%.
If your true BAC was 0.79% then almost 50% of labs would report a result equal to or greater than 0.80% - a false positve result. With a 2% standard deviation some labs might report a value as high as 0.84%, while another lab might report out a value as low as 0.74%.
Do those differences mean that it is a bad test, or that anyone did something wrong? Absolutely not. But many people would look at those differences and not understand what they mean. They would jump to the conclusion that someone screwed up, when what is really going on is a practical demonstration of measurement uncertainty in its most basic form.
seems to me if the tests are that far off +/- 2% seems it would not have taken a rocket scientist let alone a politician to have had enough brains so say for those tests on the mark at .8 will not be prosecuted since there is a large change a number of them will be INNOCENT! case closed!
For those who are interested in how industry qualifies a measurement system here's a pretty good article that discusses the various aspects in plain language other than statistics.
http://atmae.org/jit/Articles/hoffa101707.pdf
In spite of a previous comment, Dwain Fuller, the toxicologist contacted in the interview, has never been employed by Southwestern Institute of Forensic Sciences, but rather UT Southwestern Medical School at Dallas. I know Mr. Fuller to be objective and unbiased in his opinions.
Post a Comment