dissent">*546MEYERS, J.,

dissenting.

I agree with the majority that EMIT tests are reliable without a confirmation test, but do not believe that the analysis was necessary in this case. Here, there was no evidence that the victim, Michelle Briggs, died of a heart attack. The trial court was within its discretion to exclude the EMIT test results because the relevance of the results is questionable. What the Appellant is attempting to do is show that the cocaine test indicates that the victim died of a heart attack. At best, the test could only show that the cocaine may have caused a heart attack. But there is absolutely no evidence in the record reflecting that the victim suffered a heart attack before the collision. Allowing the evidence would be like admitting ballistic test results when the facts indicate that the victim was stabbed to death. Thus, the trial court did not abuse its discretion.

The majority should have conducted a harm analysis. As Presiding Judge Keller stated in Cain v. State: “Except for certain federal constitutional errors labeled by the United States Supreme Court as ‘structural,’ no error, whether it relates to jurisdiction, voluntariness of a plea, or any other mandatory requirement, is categorically immune to a harmless error analysis.” 947 S.W.2d 262, 264 (Tex.Crim.App.1997). The majority failed to consider the harm caused by the exclusion of the test results.

Even if the Court deems the EMIT test to be relevant here, the error was harmless. Although the results of EMIT tests may be admissible without a confirmation test, the evidence here did not indicate that the victim died of a heart attack. It is unlikely that a harm analysis would show that Appellant was harmed by the exclusion of the EMIT test. Therefore, I would affirm the holding of the court of appeals.

JOHNSON, J„

dissenting.

Appellant seeks to have the results of an unconfirmed EMIT test admitted as evidence that the complainant in this case was high on cocaine and amphetamines at the time of her death, thus she was responsible for the collision and her resulting death. There are a number of reasons to affirm the ruling of the court of appeals that, in the circumstances presented here, the results of an EMIT test, confirmed or unconfirmed, are not admissible.

This is what we know about the complainant’s blood sample: it was drawn at some time on or after October 21, 2007, and on or before October 25, 2007; it was not in the appropriate gray-topped test tube (which contains sodium fluoride) for proper preservation of the blood; the tube contained only the blood and an anti-coagulant; the sample was received by DPS-Austin on October 25, delivered by Matthew Ford; it was taken for testing by Megan Barton about one month later, on November 21, 2007, and subjected to an EMIT test on or after that date; it was subjected to gas chromatography (GC) by Renae Hawkins on October 23, 2008, about one year after it was drawn; the results of the two tests did not match.

This is what we do not know about the complainant’s blood sample: who drew the blood, where it was drawn, when it was drawn, and what conditions the blood was subject to after it was drawn and before it reached DPS. What version of the EMIT test was used? 1 Who is Matthew Ford? *547Who gave him the blood sample? Where did that person get it? Was it refrigerated? At what temperature? How hot did it get if not refrigerated? For how long? A gap of up to four days is a considerable period of time for unpreserved whole blood, and there is no established chain of custody. Just on the basis of what we don’t know about the sample and its peregrinations, the trial court correctly refused to admit the EMIT test.

But there are more reasons to exclude the EMIT results. EMIT is designed and sold as a screening test only. Each time that appellant tried to get either Barton or Hawkins to testify that an EMIT test is reliable for “determining the existence of drugs,” the answer was always stated in terms of reliable as a “presumptive test,” an “initial screening,” or a “screening test,” subject to confirmation by GC and mass spectroscopy (MS). Syva, the manufacturer of the EMIT test, notes that EMIT tests “provide either positive or negative results, indicating the presence of absence of a detectable drug.”2 EMIT d.a.u. and EMIT II “can give, in addition to positive/negative results, data that can be used to estimate the approximate concentrations of drug and drug metabolites in a sample.”3 But we don’t know what test was used.

Scientific evidence must not only be reliable, it must be relevant. Syva notes, “A positive result ... does not necessarily mean that the individual is intoxicated, since there is no established relationship between the amount of drug in the urine and intoxication.”4 Nor does there seem to be any correlation between the amount of drug in the blood and intoxication or an increased risk of a heart attack. Appellant has not produced evidence of a relationship between the amount of drug in blood and intoxication, as he must if he desires to use the result of a drug-detection screening test to prove that the complainant was high on cocaine or amphetamines or died from a heart attack instead of from the collision..

Scientific tests must also be appropriate to the circumstances. Generally, one would not demand a test for fingerprints in an investigation solely of a charge of failing to signal a left turn. Syva, in response to a FAQ — who uses the drug-of-abuse immunoassays and why — answers that “[m]ajor users of EMIT drugs-of-abuse assays include hospital laboratories and emergency decrements, drug and alcohol treatment programs, parole and probation agencies, prisons, work-release programs, the U.S. military, and medical or security departments of public institutions and private industry.”5 There is no mention of use in forensics. Given that EMIT tests are ' designed and manufactured as a screening device, that makes sense. Only forensics must satisfy the standard of beyond a reasonable doubt, while the listed “major users” are concerned with any indication of use. Even then, Syva, in response to a FAQ — is it necessary to confirm a positive result — responds, “The EMIT drugs-of-abuse test provides only a preliminary analytical test result. A more specific alternative chemical method must be used to obtain a confirmed analytical result. Gas chromatography/mass spectroscopy (GC/MS) is the preferred confirmatory method.” (Emphasis added). And the Syva pamphlet about EMIT test is *548clearly primarily concerned with urine testing; nowhere does it mention using an EMIT test on blood. Outside research is heavily weighted toward assays on urine.6

Written DPS protocol requires that presumptively positive EMIT results be confirmed by GC/MS, and if a positive EMIT test is not confirmed, the forensic scientist who performed the test may not testify in court about the results.

Q: Okay. Now the DPS has a protocol only to use the EMIT test as a screening device because that is the scientifically proven way to use EMIT; is it not?

A (Barton): Yes.

[[Image here]]

Q. Would you feel comfortable testifying that somebody had cocaine in their system based upon an unconfirmed EMIT test, actually where there’s a confirmation that says negative?

A. No, I would not.

Q: Would you say that it would be scientifically proven to draw any conclusions from this EMIT test at all?

A: I would not draw any scientific conclusion with just EMIT.

Q: And we can literally draw no conclusions to a scientific, to any acceptable scientific standard based on just the EMIT test and the result of this test alone.

A (Barton): No.

[[Image here]]

Q: Is it a pretty safe assumption for this Court that the EMIT test is considered reliable by the forensic toxicology community in Texas as only a presumptive screen?

A (Hawkins): Forensically, yes.

Q: Forensically; not for any other thing?

A: No.

Q: There is a difference in forensic and other uses of EMIT. Right?

A: Yes. The immunoassay test is very common in drug testing for employment testing.

[[Image here]]

Q: Also it is used — EMIT testing and that sort of thing-in things like operating rooms for quick answers when they’re treating emergency room patients and things like that in a clinical setting?

A: From what I’m told.

Q: But that is completely different from the forensic setting. Would you agree?

A: Yes.

Q: So for the forensic toxicology community in the state of Texas, the EMIT is not accepted as reliable as a test, unconfirmed, in and of itself, to show the presence of any drug.

A: Correct.

One of a trial judge’s roles is gatekeeper for admissibility of evidence. In this case, the trial judge performed admirably. He refused to admit testimony that, while “reliable” within the narrow confines of the manufacturer’s guidelines, was not considered by the manufacturer or DPS as sufficiently reliable forensically; it must be confirmed by a more accurate test, prefer*549able GC/MS. In effect, appellant wanted to use the EMIT test “off label.” And the chain of custody did not exist.

The trial court’s ruling may also have been influenced by the relevance requirement. Syva indicates in its EMIT brochure that benzoylecgonine, the metabolite of cocaine that is actually detected by the EMIT test, is retained in one’s system for 2-4 days. Both Barton and Hawkins testified that EMIT testing could not reveal when cocaine was ingested, how it was ingested, how much was ingested, how many times the user had used cocaine, whether the user was a frequent user, or whether the user had overdosed. Even considering the unconfirmed results of the EMIT test, the complainant could have ingested cocaine 10 minutes or four days before the collision.

The cause of death, blunt force trauma, does not correlate to use of cocaine, but it does correlate to a violent collision between two automobiles. The state’s collision reconstruction witness testified that the complainant’s car was struck from the rear while it was in gear and stopped on the shoulder of the road with the left tires on or close to the white fog line. With its tires pointing straight ahead and traveling at about 60 miles per hour, the right front of appellant’s truck struck the left rear of the complainant’s car with no indication of any avoidance maneuvering or hard braking before the point of impact. About half the width of each vehicle was within the zone of impact. Even if the complainant had ingested cocaine just before impact, it could have no relevance in a crash in which appellant literally ran up the complainant’s tailpipe, causing her car to violently rotate 180 and his car to use her left rear quarter panel as a spring board with enough force to launch his 5000-pound truck into a double layout somersault with a full twist. In the face of such evidence, cocaine use at some indeterminate time in some indeterminate amount had no relevance or probative value. The trial judge correctly ruled that the unconfirmed EMIT test would not be admitted, and that ruling should be affirmed.

I respectfully dissent.

. The manufacturer, Syva, makes at lest three versions — EMIT, EMIT d.a.u., and EMIT II. Syva is now part of Siemens Healthcare Diagnostics, Inc., part of the German pharmaceutical company. It describes EMIT as a “drug-of-abuse assay.” http://www.medical. siemens.com/siemens/en_GLOBAL/gg_diag_ *547FBAs/files/Drug_Testing/Education/0701518-UC l_EMIT_Abuse_FAQs_SJ_FINAL.pdf

. Id.

. Id.

. See e.g., http://labmed.yale.edu/Images/ Urine% 20DAU% 20by% 20EMIT_tcm45-9309.pdf Yale-New Haven Hospital (“The Emit D.A.U. Cocaine metabolite assay is a homogenous enzyme immunoassay intended for use in the qualitative analysis of benzoy-lecgonine (the metabolite of cocaine) in hu-irían urine.”). In its brochure on its EMIT tests, Syva lists 30 studies, at least 21 of which were done on urine. One is known to be on blood. Three are in-house studies by Siemens or an EMI-test competitor. I was unable to discover what body fluid was used in the remaining 5 studies.

Citation profile

Relationships

How this case has been treated — in progress