I was honored to speak on September 10, 2015, at the ICJE on the impact of the Georgia Supreme Court decision Williams v. State, 296 Ga. 817, 771 S.E.2d 
373 (2015).  The Williams decision has a far reaching impact on the admissibility of blood, breath and urine tests in Georgia DUI cases.  The following is an excerpt of my presentation that addressed why this analysis is important and how to apply it to DUI investigations.

The Williams decision held that the U.S. Supreme Court’s opinion in Missouri v. McNeely, 133 S. Ct. 1552 (2013) applies to Georgia DUI investigations when it stated that “[f]or the first time in Georgia, the trial court now must determine ‘the voluntariness of the consent’ at a suppression hearing challenging the chemical test.”

The old, arguably clear-cut statutory rules of implied consent, were primarily focused on whether: (1) the warning was read immediately after arrest; (2) the proper section was read (over-21, under-21 or CDL); and (3) all reasonable independent test requests were honored. After McNeely and Williams, the trial court must now engage in a more complex and careful analysis of whether the driver’s consent was truly free and voluntary as opposed to merely an acquiescence to the authority asserted by a police officer.

The 4th Amendment protections guarantee that a police officer may not search an individual unless one of three things exist: (1) search warrant; (2) exigency or other exception; or (3) true consent.

It is fair to wonder how it is that the legal community has ignored the implications of the 4th Amendment for so long, especially since DUI cases are some of the most fiercely litigated criminal cases – not only in Georgia, but across the country as well. Now, with the benefit of hindsight we have the clarity that we lacked for the last 46 years starting with a universal misinterpretation of Schmerber v. California, 384 U.S. 757 (1966).

In Schmerber, a driver who had been arrested for DUI was taken to the hospital and asked to submit to a blood test. He refused, and the officer ordered the hospital to draw the blood any way, without a warrant. In that case, the U.S. Supreme Court found that the dissipation of evidence (alcohol in his blood stream) supported a finding of exigent circumstances that would justify a nonconsensual warrantless blood draw.

For the next 46 years, the legal community behaved as if Schmerber had created a per se rule that exigency was always present in a DUI case. After all, isn’t the evidence always dissipating?   And if exigency was omnipresent in a DUI case, then the 4th Amendment was always inapplicable. As a result this issue was rarely revisited until McNeely was decided in 2013.

McNeely clarified the Court’s holding in Schmerber, stating that the 4th Amendment does apply, always had applied, and that Schmerber  did not create a per se exigency rule for DUI investigations. This decision took the legal community by surprise, especially since the facts of McNeely are nearly identical to those in Schmerber – DUI arrest, driver refused, and officer stated it would take 90 minutes or more to get a warrant.

McNeely made a lot of waves amongst DUI practitioners across the country. We began to look at our implied consent statutes through a 4th Amendment lens and realized there may be significant issues with the onerous language read to drivers upon being arrested.  In most states, we tell the driver that the law requires the driver to submit to testing, and if he does not submit, his license would be suspended.

Georgia was only the fifth state in the nation to examine McNeely’s applicability to our own state’s practices. However, Williams did not get to the issue of whether our implied consent language is problematic under the 4th Amendment. Rather, it was presented with the simple question of whether the 4th Amendment applies to DUI cases, and it answered that question in the affirmative.

In the weeks and months immediately following the Williams decision, there was much confusion in Georgia about how Williams would affect the admissibility of blood and breath tests. One particularly problematic argument surfaced regarding whether the decision impacted breath test cases. Essentially, the argument was that since Williams was a blood test case, it perhaps didn’t apply to breath test cases. Many practitioners, judges, prosecutors and defense attorneys commented that a breath test is obviously much less intrusive than a blood test, so surely there must be fewer protections associated with breath testing.

This argument sounds familiar from law school and is quite emotionally compelling.  However, it has no basis in the law. And the reason for that is quite simple. We don’t ever have to wonder whether a breath test is a search that implicates 4th Amendment protections.   We already know that it does, because the U.S. Supreme Court told us that it does back in 1989, in Skinner v. Ry. Labor Executives’ Association, 489 U.S. 602, 616-617 (1989).

The US Supreme Court already engaged in the lengthy intrusiveness analysis. Although a lay person might find this to be an interesting debate, there can be no debate about it amongst legal practitioners. A breath test is a search, and now, under McNeely, that means the 4th Amendment applies to breath testing just as it applies to blood testing. This was clearly recognized on June 29, 2015, by a Fulton County State Court Judge:

This Court finds that the holding in Williams applies to all requests for consent to test blood, breath, or other bodily substances pursuant to the implied consent law.

(In the interest of protecting the identity of the driver, the citation has been intentionally omitted).

ln State v. Medicine, 2015 S.D. 45 (June 10, 2015), South Dakota became the first state to reject their current implied consent language, holding that the language was inherently coercive and “where there is coercion, there cannot be consent.”

Georgia’s current language is quite similar to the problematic language found in Medicine.  The first sentence tells the driver that “Georgia law requires you to submit.” Our implied consent warning arguably creates the coercive situation Medicine disapproved of:

Whether fabricated or an honest recitation of purported statutory authority, “[w]hen a law enforcement officer claims authority to search [an individual], he announces in effect that the [individual] has no right to resist the search. The situation is instinct with coercion—albeit colorably lawful coercion. Where there is coercion there cannot be consent.

 State v. Med., 2015 S.D. 45, ¶ 13, 865 N.W.2d 492, 498.

And since the Williams decision did not address whether the Georgia implied consent warning is inherently coercive, our trial courts are left to navigate this issue without much guidance from our higher courts. Trial courts must decide how to weigh this language on a case-by-case basis.

For example, some trial courts may agree with South Dakota and find that the language itself creates coercion that cannot survive a 4th Amendment analysis. On the other hand, other trial courts may weigh that factor differently depending on the circumstances. One recent opinion from DeKalb County State Court addressed this issue:

Given the language of the balance of the implied consent warning concerning the consequences of refusal and consent, however, the Court finds that the language of the implied consent warning does not ordinarily invalidate consent, although it is a factor to be considered under the totality of the circumstances. Despite the language in the first sentence as to a ‘requirement’ to submit to a test, a defendant in a DUI case is given more advice as to his or her choices than many Defendants in appellate and trial cases where consent has found to be valid.

[This Court] finds that, in the absence of other evidence, the implied consent gives a fairly reasonable, though not pellucid, explanation that the Defendant has a real, though difficult choice, and does not subtly coerce a positive response or overbear the Defendant’s will.

This language, however, must be weighed in the totality of the circumstances; it potentially causes more confusion about the Defendant’s choice than a simple explanation of the options and a question of whether the Defendant consents.

(In the interest of protecting the identity of the driver, the citation has been intentionally omitted).

A Fulton County State Court Judge, agreed in part with the DeKalb County State Court finding that the language standing alone did not automatically invalidate the search.

Given the entirety of the language contained within the implied consent warning concerning the consequences of refusal and consent, the Court finds that the language of the implied consent notice does not, standing alone, invalidate [driver’s] consent to the state-administered test of his breath. The language concerning adverse consequences of refusal in the notice adequately describes how the implied consent law operates so that [the driver] could make an informed choice. The implied consent notice gives a fair explanation of the real, though difficult choices available to an individual arrested for driving under the influence and did not coerce a positive response or overbear [the driver’s] will.

(In the interest of protecting the identity of the driver, the citation has been intentionally omitted).

Interestingly, a different Fulton County State Court Judge took a similar position as that of South Dakota in Medicine.

Here, the evidence adduced at the hearing demonstrates only the reading of the implied consent warnings and a response to the question posed. While the officer testified that he didn’t pose the question in any particular tone of voice, he provided no testimony that would indicate that the Defendant freely assented. Instead, the circumstances reflected that the Defendant was in handcuffs at the time, in the back of a police car, and Defendant was not informed that the test was not mandatory. Moreover, the leading language of the implied consent warnings begins “Georgia law requires that you submit…

(In the interest of protecting the identity of the driver, the citation has been intentionally omitted).

Keep in mind, this factor (“Georgia law requires you to submit”) will almost always be on the right side of our scale, in other words, tending to toward involuntary. However, it is up to the trial court to determine the evidentiary weight of this factor.  Does it weigh as much as a bowling ball, or as little as a feather?

The specific facts of the case will be crucial to this analysis.  For instance, what happens when the driver asks questions about the warning and the officer replies with the standard response that he can only re-read the card?  This issue was addressed in early September by a Gwinnett County State Court Judge.

The primary question for this Court though is whether or not the Defendant actually consented to the state administered test. After Defendant was placed under arrest, the statutory implied consent warnings were read to him as he sat in the back of the police car. Collins said, “do I have to take this, then yes.” The Defendant then asked, “is this mandatory?” And, the officer said he could not tell him but he could re-read the implied consent warnings. Defendant then said ”yes.”

Based on the totality of the circumstances in this case, the State merely showed that Defendant acquiesced to the officer’s request that he submit to a breath test but was unable to show actual consent.

(In the interest of protecting the identity of the driver, the citation has been intentionally omitted).

Almost every driver arrested for DUI will be under arrest when implied consent is read to him. But this formal arrest may not always be weighted heavily, depending on each individual trial court’s opinion on the meaning and significance of a formal arrest.

Some trial courts will look for unusual use of force or other facts of the case when evaluating whether the arrest itself created impermissible coercion. For example, one Gwinnett State Court opinion recently found that the facts supported the suppression of the test where the arrest was more forceful than normal, and the driver had previously refused the PBT.

The Defendant stated that she would submit to the state-administered test only after she was and pushed upon the hood of a patrol car, forcibly placed in handcuffs, and read the Implied Consent advisement.
The Court finds that the most probative evidence of whether the Defendant actually consented to the state-administered blood test was the Defendant’s refusal to submit to a preliminary breath test. The preliminary breath test, which is less invasive that a blood test, was requested prior to arrest, and the Defendant was told by [the officer] that the preliminary breath test was voluntary.
Because the Defendant did not voluntarily submit to a pre-arrest preliminary breath test, it does not logically follow that after being forcibly arrested and read the Implied Consent advisement, the Defendant would actually consent to a more invasive blood test.

(In the interest of protecting the identity of the driver, the citation has been intentionally omitted).

Again, the formal arrest factor will almost always be on the side of the scale tending toward involuntary, but it is up the trial court to determine how heavy this factor is, and this is normally done on a case-by-case basis.

In two of the out-of-state decisions, approving of their individual state’s implied consent language, the driver had also been advised of his constitutional rights. In State v. Brooks, 838 N.W.2d 563 (2013), the driver was permitted to speak with an attorney twice before consenting to the test. And in State v. Moore, 354 Or. 493, 318 P.3d 1133, (2013), the driver was Mirandized before submitting to the test.

In Georgia, Miranda is almost never read during a DUI investigation. Likewise, it is rare that the driver has any opportunity to speak to a lawyer before consenting. The trial court must evaluate the absence of constitutional warnings on a case-by-case basis.

 

 

 

Not surprisingly, the issue of being too intoxicated to consent has been raised quite a few times since Williams. If the driver does not understand what he may or may not being agreeing to, how can this be true consent? In almost any other context, we would all agree that intoxication affects voluntary consent. So how does that factor affect consent during DUI investigations? In grappling with that decision, most trial courts have looked for excessive impairment when evaluating consent. For example, one State Court Judge in DeKalb County held that:

[u]nder the totality of the circumstances, including the actual language of the implied consent warning, it appears that the Defendant was generally attempting to acquiesce in officer’s instructions, but was too mentally impaired to do so, so that his capacity for self-determination [was] critically impaired, and his answer to the implied consent advisement was a mere acquiescence to authority and not an actual consent.

(In the interest of protecting the identity of the driver, the citation has been intentionally omitted).

That same State Court Judge came to the opposite conclusion in a different case:

In the present case, while the Defendant’s judgment was likely somewhat impaired by alcohol, he was engaged with the decision as to whether he should consent and the consequences of not consenting. There was no evidence he was not coherent, he was not rendered mentally deficient, and his capacity for self-determination was not critically impaired. He was apparently able to read the implied consent notice card and was prepared to answer the question after he had done so. He had a lot of time to think about his options. His will was not overborne; nor was he subtly coerced – the officers attempted to respond appropriately, in a non-coercive fashion to his questions.

(In the interest of protecting the identity of the driver, the citation has been intentionally omitted).

A Fulton County Judge similarly found the driver’s consent valid, despite intoxication.

In the present case, even if [the driver’s] judgment was somewhat impaired by alcohol, he was coherent, was not rendered mentally deficient, and his capacity for self-determination was not critically impaired. His will was not overborne; nor was he coerced; instead, the officer acted at all times in a professional and courteous manner.

(In the interest of protecting the identity of the driver, the citation has been intentionally omitted).

Intoxication is not the only kind of mental impairment that may be relevant to the Williams analysis. In Gwinnett, one State Court Judge granted the Williams motion where the driver had suffered a recent head injury.

Defendant had a visible head injury and a nose bleed when the officer approached him and spoke with him. According to the officer’s testimony, [the driver] stated that he had hit his head, but was very confused about the accident and disoriented and did not know where he was. As a result of the head injury, field sobriety evaluations were not requested.

After Defendant was placed under arrest, the statutory implied consent warnings were read. Defendant did not ask the officers any questions about the test but did consent to a blood test as requested. However, this Court finds that [the driver] lacked the capacity to consent based upon his head injury, his confusion and disorientation and his statement that he did not know where he was.

(In the interest of protecting the identity of the driver, the citation has been intentionally omitted).

Garrity involved police officers who were under investigation. They were told that they had the right to refuse questioning, but if they did, they would be subject to immediate termination. As you might expect, they cooperated by making statements to the investigators. Those statements were used as admissions in a subsequent criminal prosecution. The U.S. Supreme Court found that those statements were made under unconstitutional duress and coercion, and as a result should have been excluded at trial.

This very compelling language from the Supreme Court in Garrity stands in contrast with the “difficult choices” language found in out-of-state decisions interpreting McNeely.  

Despite this apparent contradiction, some Georgia trial judges have adopted the same approach as State v. Padley, 849 N.W.2d 867, 872-873 (Wis.App. 2014), and have rejected the Defendant’s Garrity argument. One Fulton County Court Judge addressed this point.

The implied consent notice gives a fair explanation of the real, though difficult choices available to an individual arrested for driving under the influence and did not coerce a positive response or overbear [the driver’s] will.

(In the interest of protecting the identity of the driver, the citation has been intentionally omitted).

 

Again, this factor almost always will be tending toward non-voluntary, however the trial court must determine its weight.

 

According to one DeKalb County State Court Judge, if the driver does not understand English, then it seems that under Williams, submission to a breath test is likely mere acquiescence to the officer’s claim of authority.

Defendant is not a native English speaker, and understands little English.
If the age, education, and intelligence of the Defendant is relevant to voluntary consent, his understanding of English is even more relevant. The State did not show by a preponderance of the evidence that the Defendant understood what he was told and actually and voluntarily consented to a sampling of his breath. Instead, it appears more probable that the Defendant simply acquiesced to authority. Thus, there is no exception to the requirement of a warrant in this case.

Referencing Schneckloth v. Bustamonte, 412 U.S. 218 (1973).(In the interest of protecting the identity of the driver, the citation has been intentionally omitted)..

It is worth mentioning that in Brooks, the driver was permitted to speak with an attorney twice before deciding to submit to testing.

Moore might normally be applicable to Georgia, except in that case, Moore was Mirandized, which is very unusual in Georgia.

Wisconsin’s implied consent warning does not inform the driver that the law requires submission, as does the Georgia warning. Rather, the driver is told that “this law enforcement agency now wants to test one or more samples…”

Because California does not have standard implied consent warning, each officer simply explains the law to each driver. As a result, we do not know what the officer told the driver in Harris. 

As discussed earlier, South Dakota was the first state to declare it’s implied consent statutory language to be coercive under the 4th Amendment analysis.

There is plenty of debate both in Georgia, and across the country about the impact of McNeely on warrantless chemical tests in DUI cases. What is clear is that, absent a ruling similar to that in South Dakota, Georgia practitioners will have to continue to examine the facts on a case-by-case basis until we have more guidance from our higher courts.

 

On April 24, 2015, we (D. Benjamin Sessions, Robert Chestney and I) were excited to meet with Georgia Public Defenders at the State Bar to discuss the new Georgia Supreme Court decision Williams v. State, ___ Ga. ___ (Case No. S14A1625, decided 3/27/15) and how it affects any state-administered test submitted to pursuant to implied consent.

In March of 2015, the Georgia Supreme Court held that a trial court must determine whether a police officer’s request for a blood sample from a driver who has been arrested for DUI complies with the protections afforded by the Fourth Amendment.

In Williams, the Supreme Court distinguished between “implied consent” under OCGA § 40-5-55 and “actual consent” which is required to permit a warrantless search. Under Williams, the trial court must determine whether actual consent was given, “which would require the determination of the voluntariness of the consent under the totality of the circumstances.” Williams v. State, ___ Ga. ___ (Case No. S14A1625, decided 3/27/15).

For 49 years prior to Missouri v. McNeely, 133 S. Ct. 1552 (2013), the legal community was under the impression that Schmerber v. California, 384 U.S. 757 (1966) had established a “DUI Exception” to the Fourth Amendment.  In 1966, the Schmerber decision held that because there was evidence of a crime in the defendant’s blood (alcohol), and that evidence was rapidly dissipating, this type of warrantless search would be permitted under the exigency exception to the Fourth Amendment.  In 2013 however, the McNeely opinion explicitly corrected that misinterpretation of Schmerber, advising that there is no such per se rule regarding blood testing and exigency, but that rather each case must be decided on a case-by-case basis.

At the outset, the first question to be addressed is whether the Williams and McNeely holdings apply to breath testing, since both of those cases involved a blood test.  Happily the US Supreme Court resolved this issue for us more than 25 years ago in Skinner v. Ry. Labor Executives’ Ass’n, 489 U.S. 602, 616-617 (1989).  Skinner held that a breath test “implicates similar concerns about bodily integrity and . . . should also be deemed a search.” Skinner v. Ry. Labor Executives’ Ass’n, 489 U.S. 602, 616-617 (1989).

Under Williams the trial court must determine voluntariness of the consent under the Fourth Amendment.  But before we can get to that question, first we have to get past the threshold inquiry of whether the driver even consented to the breath/blood test.  Consent is almost never requested under the Georgia implied consent regime.  Drivers are not asked if they will consent to a search and test of their bodily substances – they are asked, “will you submit [to the testing] under the implied consent law?” Webster’s defines the intransitive verb “submit” as follows: “to yield oneself to the power or authority of another.” Legally speaking, “submission” means “[a] yielding to the authority or will of another.

The statement, “Georgia law requires you to submit to [testing]” is false. No law in Georgia requires a person to submit to the test. Georgia has no criminal statute making it illegal to refuse testing. Compare the Minnesota Supreme Court decision in State v. Brooks, 838 N.W.2d 563 (Minn. 2013). There, the advisement given “informs drivers that Minnesota law requires them to take a chemical test . . .” But in Minnesota, that statement is true – “It is a crime for a person to refuse to take a test required under the implied consent law.

The recent out-of-state cases addressing McNeely are not appropriately applied for the purposes of resolving the Fourth Amendment concerns here in Georgia.  In Minnesota, drivers are advised that they have the right to speak with an attorney prior to deciding whether to consent to a test, whereas in Georgia, drivers have no such right. State v. Brooks, 838 N.W.2d 563 (2013).

In Wisconsin the implied consent advisement makes no mention of drivers being required to submit to chemical testing.  Rather, Wisconsin drivers are simply notified that the officer would like to test their breath or blood and what the consequences would be should he/she refuse. State v. Padley, 2014 WI App 65,1, 354 Wis. 2d 545.

The Williams decision did not adopt California’s perfunctory analysis of actual consent in People v. Harris, 234 Cal. App. 4th 671, review filed (Apr. 2, 2015).  But rather, Williams remanded the case to the trial court for a more meaningful Fourth Amendment analysis.  By contrast, the Harris opinion seems to suggest a shifting of the burden to the defendant to provide evidence that the consent was involuntary when it held that the lack of evidence that defendant was reluctant to submit was sufficient to support a finding that the defendant actually consented to testing.

[T]he record demonstrates that defendant verbally agreed to a blood test after being admonished by Deputy Robinson under the implied consent law, and that he did not verbally refuse to give a blood sample or demonstrate a desire to withdraw his consent either verbally or by physically resisting the phlebotomist’s attempt to draw his blood

People v. Harris, 234 Cal. App. 4th 671, 690, 184 Cal. Rptr. 3d 198, 214 (2015), review filed (Apr. 2, 2015).

Unlike the discussion in Harris, Williams reminds us that mere acquiescence to the authority asserted by a police officer cannot substitute for free consent.

In Garrity v. New Jersey, the US Supreme Court made clear that, just as Odysseus did not have a “free choice” when faced with the option of steering his ship towards a treacherous whirlpool or risk being tossed against the rocks by the sea,  a decision made under duress is not a free choice under the Fourth Amendment.  The Garrity opinion is tremendously instructive here, particularly when the government attempts to rely on the convenient, but not necessarily relevant passage from Harris: “the choice to submit or refuse to take a chemical test ‘will not be an easy or pleasant one for a suspect to make,’the criminal process’ often requires suspects and defendants to make difficult choices.” People v. Harris, 234 Cal. App. 4th 671, 687 (2015), review filed (Apr. 2, 2015).

I was honored to speak for the Office of the State Public Defender in Casper, Wyoming last week (February 6, 2015). One of the topics I spoke about was blood testing in DUI (DWUI) cases. Below is an excerpt of that talk.

The blood testing process can be broken down into three critical stages: the blood draw, the storage and transportation of the sample, and the testing at the lab.

One of the most important things to look for is whether an iodine swab or isopropyl swab was used during the blood draw. The forensic blood test kits that most police agencies use include an alcohol-free swab. However, periodically the nurse, or hospital employee will inadvertently use a hospital (alcohol-based) swab instead.

It is imperative that the puncture site is carefully cleaned to avoid the introduction of candida albicans, which is a yeast that lives on human skin and creates ethanol (alcohol) as a waste product.

This is the same type of yeast that is problematic for those suffering from the rare auto-brewery syndrome.  Those afflicted with this syndrome have an excess amount of yeast in their intestines.  When the person consumes starchy foods, or foods high in carbohydrates, the yeast ferment – creating ethyl alcohol.  This can lead to BAC results that have been documented as as high as .37 in the absence of any alcohol being consumed from an outside source.

Although the blood kit comes with it’s own needle, often medical professionals have a preference for the needles they normally use at the hospital, often they prefer a butterfly needle.

Using a butterfly needle may cause unnecessary breakage of the red blood cells.  Although  this won’t have an effect on a whole blood test, it could affect the results of a hospital blood test.

The gray-stop tube is the only acceptable tube for this type of forensic blood draw.  This is because it contains the two necessary compounds, a preservative and an anti-coagulant.

The preservative does not directly affect the yeast, but rather it cuts off its food supply.  Yeast only creates alcohol if it has sugar (glucose) to consume.  In the absence of a preservative, the blood will break down into glucose, thus feeding the yeast.  The preservative is designed to stop the blood from decomposing, thus starving any yeast that may have been introduced into the blood sample.

In order to ensure that these tubes are sterile, they are bombarded with radiation by the manufacturer.  The gray-stop tubes are vacuum-sealed.  But be sure to check the expiration date of the gray-stop tube.  Although the powders (preservative and anticoagulant) will not expire, the vacuum will expire.  If the vacuum seal is compromised, there is the possibility of yeast contaminating the tube.

The powders must be mixed homogeneously throughout the sample.  The manufacturer provides very specific instructions regarding the proper mixing of the blood sample.  The tubes must be inverted gently 5-8 times (depending on the brand).  Otherwise the powder will clump at the bottom of the tube where it will be ineffective at preventing the blood from turning into glucose or from clotting.

Our blood is made up of many different parts.  In order to measure any different substance in the blood (in our case, alcohol) we first have to separate the alcohol parts (ethanol molecules) from the other parts of our blood.

The molecules *race* through the headspace column, pushed through by an inert gas (e.g., helium, or nitrogen) until hitting the flame ionization detector.  The headspace column is designed to attract the molecules at different rates of *stickiness* which will have the effect of slowing down the molecules at different speeds.  Thus, as the different types of molecules race through 30 meters of the headspace column they start to spread out from one another. Once properly separated the quantity of molecules can now be measured.  The actual measurement depends upon, in layman’s terms, the intensity of the flame as the molecules rush through it.

This is a chromatogram, which is a graphical representation of the BAC results.  A good chromatogram will have tall, skinny and symmetrical peaks.  You can see that there is good separation on the graph because the two peaks have no overlap.

When the molecules do not properly separate in the headspace column, then we have bad separation, which means the flame ionization detector’s measurement may be the result of two different types of molecules coming out at the same time.

Here you can see  where the blue molecules did not properly separate from the green molecules.  When they hit the flame ionization detector the results of both molecules may be improperly attributed — artificially inflating the ethanol result.

Here you can see what co-elution looks like on the chromatogram.  You can see that the acetone is combining with the ethanol.  Since the BAC is calculated by measuring the area under the peak, where the peaks overlap, this number may be reported as higher than the true number.  This is particularly a problem if you client is a diabetic.  Depending on the individual, a diabetic may naturally have acetone in his blood.