¶1 A police officer (Officer) testified at trial, based on information he received in his police training, about the human body's average rate of elimination of alcohol-the "burn-off rate." Randy Lynn Harvey, who was standing trial for driving under the influence (DUI), objected for lack of foundation. The district court overruled the objection, and a jury found Harvey guilty. Harvey subsequently filed a motion for a new trial, asserting that Officer gave expert testimony as a lay witness. The district court denied the motion. Harvey appeals, and we reverse.

BACKGROUND

The Arrest

¶2 Around one o'clock in the morning in late November 2015, Officer, a fourteen-year police veteran with advanced roadside impairment training and about 200 DUI arrests under his belt, spotted a pickup truck traveling without its taillights illuminated and initiated a traffic stop by turning on his overhead lights. Harvey, who was driving the truck, continued for about one block and stopped at a red light as Officer, with his patrol car's lights still flashing, followed him. Harvey proceeded through the light after it had turned green, travelled another half block, and then pulled over, bringing the truck to a stop. Officer noted that the street was "pretty dead" and that there was "quite a bit of parking [at] that time of night on the side of the road."

¶3 When Officer approached the truck, Harvey rolled down the window about two inches, but he did not look Officer in the eye. Officer observed that Harvey's left eye "appeared to be glassy and bloodshot." Officer also detected "a strong odor of an alcoholic beverage emanating from the vehicle" and saw a "passenger attempting to hide a ... half-consumed [bottle of] vodka ... in a bag behind her seat." When asked, Harvey denied that he had been drinking.

¶4 Officer requested that Harvey perform several field sobriety tests (FST). Officer recalled that Harvey appeared to have good balance when he exited his vehicle to take the FSTs. But with regard to his ability to perform the FSTs, Harvey advised Officer that (1) he had a titanium ankle replacement because he had been shot with a .40 caliber bullet in one leg and (2) he had been bitten in the calf on the other leg by a shark when he was a child.

¶5 Officer had Harvey perform three standardized FSTs. The first was the horizontal gaze nystagmus test (HGN test). In administering the HGN test, Officer noticed six out of a possible six indicators that Harvey had alcohol in his system. Officer next administered the vertical gaze nystagmus test (VGN test), but Harvey's performance during this test did not show any indicators of the presence of alcohol. At trial, Officer testified that the VGN test detects when someone has a high level of alcohol in his blood. The third FST was the walk-and-turn test, which Harvey had indicated he could probably complete despite his injuries, but he advised Officer that "numbness" in his calf would "prevent him from feeling when his feet touched heel to toe." Officer recalled telling Harvey that he "would take that into consideration." Harvey's performance during this test showed six out of eight possible indicators of the presence of alcohol in his system. Because Harvey's prior leg injuries prevented him from performing the last standardized FST-the one-leg-stand test-Officer opted to have Harvey recite the letters of the alphabet and count backward. Harvey performed both tests without error.

¶6 Based on Harvey's performance on the walk-and-turn and HGN tests and the odor of the alcoholic beverage, Officer arrested him for DUI. After transporting Harvey to the police station, Officer requested that Harvey take a breath test to determine his blood alcohol content (BAC). Even after being warned that refusal to take the test may result in the loss of his driver license, Harvey refused to submit to it. Officer then obtained a search warrant to collect Harvey's blood for testing, and a sample was drawn at about three o'clock in the morning, two hours after Harvey was pulled over. Harvey's blood was subjected to two identical tests-a screening test and a confirmation test-with each test producing two results. The screening test measured Harvey's BAC at .075 and .076. The confirmation test yielded two BAC results of .081.

¶7 The State charged Harvey with DUI. Under Utah law, Harvey could be found guilty of DUI if (a) a subsequent chemical test showed that he had a BAC of .08 or greater at the time of the test, (b) he was under the influence of alcohol to a degree that rendered him incapable of safely operating a vehicle, or (c) he had a BAC of .08 or greater while operating a vehicle. Utah Code Ann. § 41 -6a-502(1) (LexisNexis 2014). In Harvey's case, the DUI was charged as a third-degree felony based on two prior DUI convictions within the last ten years. See id. § 41-6a-503(2)(b).

The Trial

¶8 The State's forensic scientist testified that he could not "rule out the possibility" that Harvey's BAC was below .08 at the time it was tested. He stated that the tests he ran indicated that Harvey's BAC would have been "around [.07 or .08] and probably within those two numbers."

¶9 In addition to recounting the details of the arrest, Officer testified that "the average burn-off rate [for alcohol] is approximately .015" per hour. Harvey's counsel immediately objected to this testimony for lack of foundation. The court asked the State to "lay some foundation with respect to how [Officer] would know" about alcohol burn-off rates. Officer testified that he learned about how the body burns off alcohol in his training at the police academy. Harvey's counsel again objected, but the court overruled the objection, stating, "[I]f this is something he learned in [the police academy], he can testify to it. It goes to the weight." On cross-examination, Officer stated that he could not calculate Harvey's precise alcohol burn-off rate and could provide only the average because rates depend on weight, gender, and what someone has eaten.

¶10 At the close of the State's case, Harvey moved for a directed verdict because the State's forensic scientist testified that Harvey's BAC may have been below .08 and because Officer did not notice a pattern of unsafe driving. Based on the forensic scientist's testimony, the district court determined that there was "a real probability" that Harvey's BAC was under .08. Thus, the court granted the motion in part, finding that a reasonable jury could not find Harvey's BAC was .08 or higher at the time of the blood test. But the district court allowed the case to proceed to the jury on the other two prongs of the statute, namely (1) whether Harvey was under the influence of alcohol to a degree that rendered him incapable of safely operating a vehicle or (2) whether Harvey had a BAC of .08 or greater while operating a vehicle. See Utah Code Ann. § 41 -6a-502(1)(b)-(c). The jury found Harvey guilty.

¶11 Subsequently, Harvey moved for a new trial, arguing that Officer's burn-off-rate testimony was improper expert testimony from a lay witness. The district court denied the motion, explaining that Officer had "sufficient training and expertise to relay the information" that he had learned in the police academy "about the average blood alcohol dissipation rate." The district court imposed a suspended prison sentence and placed Harvey on supervised probation. Harvey appeals.

ISSUE AND STANDARD OF REVIEW

¶12 The issue presented on appeal is whether the district court erred in admitting Officer's burn-off-rate testimony. "A decision to admit or exclude expert testimony is left to the discretion of the [district] court, and that decision will not be reversed unless it constitutes an abuse of discretion." Belnap v. Graham , 2016 UT App 14 , ¶ 8, 366 P.3d 852 (cleaned up). And "[e]ven if we determine the testimony was erroneously admitted, the defendant must show that the error was prejudicial." State v. Peraza , 2018 UT App 68 , ¶ 23, 427 P.3d 276 , cert. granted , 429 P.3d 460 (Utah 2018). 1

ANALYSIS

¶13 The State makes no effort to defend the district court's ruling on the admissibility of Officer's burn-off-rate testimony. Rather, the State's theory is that Officer's burn-off-rate testimony-even if erroneously admitted-was harmless because there was overwhelming evidence to convict Harvey under the alternative impairment theory of DUI, namely, that Harvey was under the influence of alcohol to a degree to render him incapable of safely operating a vehicle. See Utah Code Ann. § 41 -6a-502(1)(b) (LexisNexis 2014). In other words, for the burn-off-rate testimony to be harmless, we must conclude that the other evidence presented at trial demonstrating that Harvey was incapable of safely operating a vehicle was so strong that there is no reasonable likelihood that the admission of the burn-off-rate testimony affected the jury's verdict. Thus, the State's argument is a delicate house of cards built on the foundation of showing there was ample evidence of Harvey's incapacity.

¶14 We first consider the admissibility of Officer's burn-off-rate testimony. Then, having determined that his testimony was improperly admitted, we consider whether its admission was harmless based on the strength of the evidence showing that Harvey was under the influence of alcohol to a degree rendering him incapable of safely operating a vehicle.

I. Erroneous Admission of Officer's Burn-Off-Rate Testimony

¶15 We agree with Harvey that Officer was not a qualified expert on the subject of alcohol burn-off rates and therefore should not have been allowed to testify about this subject. The Utah Rules of Evidence state that "a witness who is qualified as an expert by knowledge, skill, experience, training, or education may testify in the form of an opinion or otherwise if the expert's scientific, technical, or other specialized knowledge will help the trier of fact to understand the evidence or to determine a fact in issue." Utah R. Evid. 702(a). The district court reasoned that Officer could testify as an expert about the burn-off rate of alcohol because the metabolism of alcohol was "something he learned in [the police academy], [so] he can testify to it. It goes to the weight." Thus, the district court determined that the alcohol burn-off rate constituted special knowledge that must be provided by an expert.

¶16 Testifying about alcohol burn-off rates is generally viewed by courts as the province of experts. But learning about something in cursory fashion during training does not make a person an expert. Indeed, in Negrini v. State , where an officer attempted to testify about the alcohol burn-off rate based on information he "heard during one lecture at the police academy," a Texas court held that the officer was not qualified as an expert on the subject. 853 S.W.2d 128 , 133 (Tex. Ct. App. 1993). And when burn-off-rate testimony has been admitted in Utah, such testimony has been offered by an expert with a scientific background. See State v. Folsom , 2019 UT App 17 , ¶ 25, 438 P.3d 992 (noting the admission of burn-off-rate testimony from a forensic toxicologist); State v. Bradley , 578 P.2d 1267 , 1268-69 (Utah 1978) (admitting burn-off-rate testimony from a medical examiner).

¶17 We acknowledge that law enforcement officers may offer expert testimony on a range of subjects provided that sufficient foundation is laid for the specific testimony. See United States v. Figueroa-Lopez , 125 F.3d 1241 , 1244-45 (9th Cir. 1997) (admitting expert testimony of a special agent opining on the behavior of drug traffickers); State v. Rothlisberger , 2006 UT 49 , ¶ 31, 147 P.3d 1176 (admitting expert testimony of a police officer on determining whether a quantity of drugs is for distribution purposes); Salt Lake City v. San Juan , 2015 UT App 157 , ¶ 2, 353 P.3d 623 (admitting expert testimony of a police officer on detecting signs of impairment). The question in this case, however, is whether Officer was qualified to testify as an expert on the topic of alcohol burn-off rates based on the foundation provided at trial.

¶18 Unlike signs of impairment or the behavior of drug traffickers, the burn-off rate of alcohol from the human system is not something that an officer can observe and form an expert opinion about based on training and experience. See State v. Ohotto , 261 Or.App. 70 , 323 P.3d 306 , 310-11 (2014) (stating that testimony regarding "alcohol absorption and elimination rates" is not derived from an individual's "training and prior experience as a police officer"). Instead, it is necessarily scientific testimony derived from lab testing, based on technical training, and presented by an expert qualified in that area. Burn-off-rate testimony requires "a formulaic calculation derived from scientific understandings of physiological processes that cannot be achieved through reading a training manual, conducting routine [DUI] investigations in the course of law enforcement, and attending [a training] course." Id. at 311.

¶19 The error in this case was admitting Officer's burn-off-rate testimony where the State failed to establish a sufficient foundation that Officer had the necessary scientific expertise, based on education, training, or otherwise, to offer burn-off-rate testimony. It is not that officers, as a class, can never be qualified to provide such testimony. See State v. Claybrook , 193 Ariz. 588 , 975 P.2d 1101 , 1103 (App. 1998) ("The expert witness who presents BAC retroactive extrapolation evidence can be a police officer or the operator of the machine if properly certified and in addition possesses superior knowledge, experience and expertise on the question." (cleaned up)); Commonwealth v. Gonzalez , 519 Pa. 116 , 546 A.2d 26 , 31-32 (1988) (holding that a police officer was qualified as an expert on the elimination of alcohol from the bloodstream where he worked in the police laboratory and had specific education in the pharmacology and toxicology of alcohol). But here, no evidence was presented to show that Officer "understood the process of alcohol elimination or relevant factors for consideration." See People v. Beck , 418 Ill.Dec. 594 , 90 N.E.3d 1083 , 1111 (2017). Nor was any testimony offered to explain the nature, extent, or depth of his training and education on burn-off rates of alcohol.

¶20 Participating in a course at the police academy where burn-off rates were generally mentioned shows merely that Officer was exposed to information-not that he is an expert in any of the subjects touched on in the course of study. 2 Thus we agree with Harvey that the testimony at trial failed to establish that Officer had sufficient training to qualify as an expert on the subject of alcohol burn-off rates. Therefore, Officer's testimony about the alcohol burn-off rate was improperly admitted.

II. The Harm From the Admission of Officer's Burn-Off-Rate Testimony

¶21 We must next determine whether the erroneous admission of Officer's burn-off-rate testimony was harmless. See State v. Reece , 2015 UT 45 , ¶ 33, 349 P.3d 712 ("Errors are often harmless where there is overwhelming evidence in the record of the defendant's guilt."); see also State v. Dunlap , 2010 UT App 122U, para. 3, 2010 WL 1920043 (stating that improperly admitted evidence is harmless if other evidence provides "overwhelming proof that [a defendant is] guilty of DUI"). In DUI cases, this court has oft stated that a jury may properly consider the entirety of the evidence in determining whether the level of impairment due to consumption of alcohol makes it unsafe for an individual to drive. For example, we have determined that even if a defendant's statements about the number of alcoholic beverages he had consumed was improperly admitted, such error was harmless beyond a reasonable doubt because the prosecution produced other "strong evidence" of the defendant's guilt, including the odor of alcohol on the defendant, blood shot eyes, failure of two out of three field sobriety tests, inability to count to thirty, admission to drinking, slow and slurred speech, failure to stop or slow down at a red light, inability to remember details about the night, and refusal to submit to a breathalyzer test. Salt Lake City v. Kane , 2007 UT App 44U, paras. 4-5, 2007 WL 479633 ; see also Dunlap , 2010 UT App 122U, paras. 5-6 (noting that even if the defendant's intoxilyzer results had been excluded, the jury could have considered all the other evidence, including the defendant's driving in the wrong direction, urinating himself, slurred speech, and lying on the ground next to his vehicle, to determine whether defendant was incapable of safely operating a vehicle).

¶22 Here, the State argues that other evidence-apart from the BAC evidence-conclusively establishes that Harvey was incapable of safely operating a vehicle. The State makes three specific points. First, the State argues that Harvey demonstrated an unusual driving pattern. For example, Officer tried to stop Harvey around one o'clock in the morning, but Harvey did not immediately pull over, even though there was ample room and the streets were empty at that hour. And instead he continued for a block and a half and even stopped at a red light, all while the patrol car's overhead lights were flashing. Second, when Officer approached Harvey's truck, Harvey displayed suspicious behavior. He rolled down his window only a few inches. He maintained his gaze straight ahead and refused to make eye contact with Officer. Officer noticed signs that Harvey had likely been drinking an alcoholic beverage. Not only was the odor of alcohol emanating from Harvey's truck, but Harvey's passenger was trying to hide a half-consumed bottle of vodka as Officer approached. Harvey's left eye-the only one Officer could see because Harvey declined to make eye contact-appeared glassy and bloodshot. Third, even though Harvey denied that he had been drinking alcohol, he refused to take a breath test. 3

¶23 Unlike this case, all the cases we have identified where a driver was convicted of DUI for being under the influence of alcohol to a degree that rendered a person incapable of safely operating a vehicle, see Utah Code Ann. § 41 -6a-502(1)(b) (LexisNexis 2014), as opposed to having a BAC over the legal limit, have in common the presence of some evidence of the driver's incapacity to operate the vehicle safely, see State v. Salgado , 2018 UT App 139 , ¶¶ 33-35, 37, 427 P.3d 1228 (noting that a defendant's creating a "traffic hazard" by driving considerably below the speed limit, non-responsiveness to several attempts made by an officer to pull over, poor performance on field sobriety tests, maintaining an "intent focus straight ahead," very droopy eyelids, and confusion about being stopped were enough for the State to meet its "burden of producing believable evidence on each element of DUI"); State v. Van Dyke , 2009 UT App 369 , ¶¶ 12-13, 15, 36-37, 223 P.3d 465 (holding "[e]ven in the absence of direct evidence that the defendant drove recklessly or violated traffic rules, the jury was free to consider all of the evidence presented," including the defendant's slurred speech, glazed eyes, incoherence, staggering, unsteadiness, "slow and lethargic" movements, "lack of control of his actions," "impaired judgment," and boisterous behavior before driving "to determine whether his level of impairment was such that it was unsafe for him to drive" and that "[u]nder these circumstances, there was sufficient evidence from which the jury could find, beyond a reasonable doubt that [the defendant's] level of intoxication made it unsafe for him to be driving"); Rosengreen v. State , Dep't of Public Safety , 2003 UT App 183U, para. 5, 2003 WL 21373157 (concluding that even assuming an officer failed to comply strictly with the standards for the administration of FSTs, the officer's testimony that the defendant was incapable of safely operating a vehicle based on the odor of alcohol on the defendant's breath and the defendant's performance on three FSTs, admission that he had consumed two or three beers, swaying during and between FSTs, difficulty in maintaining his balance, and chattiness constituted sufficient and reliable evidence of defendant's incapacity to operate a vehicle safely).

¶24 Such evidence of incapacity is lacking in the testimony offered by the State's witnesses against Harvey. And this lack of evidence of Harvey's incapacity is the fundamental problem with all the other non-BAC evidence offered by the State. None of it shows that Harvey was impaired such that he could not safely operate his truck. 4 Officer did not point to any abnormalities in Harvey's driving patterns. Now the State argues that Harvey displayed an unusual driving pattern because he did not immediately pull over after Officer activated the lights on his patrol car. But taking a block and an half to pull over hardly qualifies as an indication of impaired driving. Just as likely is that Harvey was thinking clearly enough to buy some time for his passenger to try to hide the open bottle of vodka before pulling over. And Officer did not note anything unusual about the manner in which Harvey pulled over. Officer offered no indication that Harvey slurred his speech during their interaction; he did not mention the odor of alcohol on Harvey's breath. On the contrary, Officer stated that Harvey kept his balance and did not lean on anything when he got out of his truck in preparation to complete the FSTs. And no mention is made of Harvey struggling to comply with Officer's orders during the FSTs.

¶25 It is beyond question that Harvey had been drinking. In fact, Harvey's counsel admitted, "There was alcohol in [Harvey's] blood. We all acknowledge that. If that were the issue here today[,] ... I couldn't ask you not to find him guilty." Apart from his counsel's concession, there is abundant evidence that Harvey was drinking on the night in question. There was an open container of vodka, the truck reeked of alcohol, Harvey avoided making eye contact, Harvey rolled down the window only two inches, and Harvey's eye appeared to be bloodshot. But at most, these circumstances show that Harvey had been drinking and that he was trying to hide that he had been drinking. We note that Harvey has a rather lengthy criminal history, including two recent DUIs. Thus, he had good reason-at least from his perspective-to be circumspect as he tried to avoid additional interaction with the legal system. But none of these circumstances-even when considered in aggregate-indicate that Harvey was impaired to such an extent that he could not safely drive. And we do not regard Harvey's refusal to submit to a breath test as evidence of impaired driving. It certainly indicates Harvey's consciousness that he had been drinking and driving, but there is no indication that his refusal equates to the necessary level of impairment.

¶26 Thus, the only indication that Harvey was impaired necessarily comes by way of the two failed FSTs. But these two failed tests, standing alone, are insufficient to support a finding beyond a reasonable doubt that Harvey was so impaired that he was incapable of safely operating his vehicle. Although Harvey failed the HGN test, such failure did not show incapacity. Rather, it merely indicated that Harvey had alcohol in his system, a fact that Harvey readily concedes. But Harvey also passed the VGN test. So although there was admittedly alcohol in Harvey's blood, the combination of the two tests could potentially show-at the very most-that there was not a high amount present.

¶27 Harvey also failed the walk-and-turn test. But we deem Harvey's failure of this FST to be an unreliable indication of his impairment because Harvey had suffered some serious previous injuries (i.e., a shark bite and a bullet wound ) to both legs-injuries that likely compromised his ability to complete the walk-and-turn test successfully.

As Harvey points out, studies show that fifty percent of completely sober drivers fail the walk-and-turn test. Steven J. Rubenzer, The Standardized Field Sobriety Tests: A Review of Scientific and Legal Issues , 32 Law & Human Behavior 293, 297 (2008). Thus, it is little surprise that Harvey, given his previous injuries, also failed this test.

¶28 Further, in our survey of cases, courts generally hold that FSTs provide a means of estimating only BAC, not the degree of incapacity. See Coffey v. Shiomoto , 60 Cal.4th 1198 , 185 Cal.Rptr.3d 538 , 345 P.3d 896 , 905 (2015) (recognizing that the HGN test, the walk-and-turn test, and the one-leg-stand test are "extremely accurate in discriminating between BACs above and below 0.08 percent" (cleaned up)); State v. Ortez-Olivia , 2018 WL 5879517 , at *3 (Del. Super. Ct. Nov. 8, 2018) ("The standardized field sobriety tests ... are highly accurate and reliable tests for detecting blood alcohol concentrations at or above 0.10%, above the legal limit in this State of 0.08%."); State v. Lasworth , 131 N.M. 739 , 42 P.3d 844 , 845 (Ct. App. 2001) (recognizing that standardized field sobriety tests "enabled officers in the field to accurately estimate whether a motorist's BAC was at or above 0.08 percent ninety-one percent of the time"). 5

¶29 Thus, in light of the weak evidence of impairment, there is a reasonable likelihood that the jury would have reached a different verdict if Officer's burn-off-rate testimony had been excluded.

CONCLUSION

¶30 Because the evidence of Harvey's incapacity to operate a vehicle safely was not so overwhelming as to render the erroneous admission of Officer's burn-off-rate testimony harmless, we determine that Harvey was prejudiced by the admission of the testimony in question. Therefore, we must vacate his DUI conviction and remand for a new trial.