State of New Hampshire v. Otto Keller

1. Whether the court erred in allowing the opinion testimony of the State’s witness on human performance, Colleen Scarneo.

Issue preserved by motion in limine, (A 5), defense objection, (T 112-13), and the court’s rulings. Supp. 33-45; T 113.

2. Whether the State introduced sufficient evidence to prove serious bodily injury.

Issue preserved by motion to dismiss at trial after State rested and the court’s ruling. T 158; 173-74.

 Citations to the record are as follows:

“M” refers to the transcript of the pretrial motion hearing on September 3 and December 6, 2021; “T” refers to the transcript of the trial on November 16 and 17, 2022; “S” refers to the transcript of the sentencing hearing on February 6, 2023; “Supp” refers to the supplement attached to this brief containing an order from which defendant appeals.

Otto Keller was charged with driving while under the influence of intoxicating drugs, causing an accident resulting in serious bodily injury to himself. A 3.

Keller moved in limine to exclude the opinion testimony of the State’s expert, Colleen Scarneo, that he showed signs and symptoms of impairment consistent with the use of certain drugs at the time of the accident. A 5. After an evidentiary hearing, the trial court (O’Neill III, J.) denied the motion by written order on January 28, 2022. Supp. 33-45. After a two day trial in November 2022, a jury convicted Keller as charged. T 219. The trial court (Attorri, J.) sentenced him to nine months in the house of corrections, suspended his driver’s license for eighteen months, and imposed a two year term of probation. S 23-24. The court granted Keller bail pending appeal. S 34-35.

On September 1, 2018, shortly after 1:00 a.m., Otto Keller crashed his Honda Civic into a parked car and the side of a building a few blocks from his home on Sheridan Street in Laconia. T 39, 57. The accident happened on the other side of the street from the direction Keller was traveling. T 40-41. Keller walked home and his girlfriend took him to the emergency room at Lakes Region General Hospital. T 39. His girlfriend also called the police and reported the accident. T 46-47.

In the emergency room, Keller was interviewed by Patrolman Sean McCormack at about 1:40 a.m., approximately thirty minutes after the incident. T 56, 65. The police had determined Keller owned the vehicle that crashed based on the registration. T 32. McCormack wanted to ask about what had happened and see if Keller showed any signs of alcohol or drug impairment. T 57-58. McCormack had been a patrolman for more than six years at the time, and had been trained to recognize the signs of drug impairment. T 54, 66. Keller’s eyes were not bloodshot or watery, his speech was not slurred and there was no odor of alcohol. T 58. In fact, McCormack saw no signs Keller was impaired. T 58, 66. McCormack did not administer any field sobriety tests to Keller. T 68.

McCormack asked Keller if he had consumed alcohol or drugs that evening, and Keller told him he had used heroin approximately five hours before the accident. T 59, 67. McCormack asked Keller to consent to a blood draw and Keller agreed. T 60. Keller’s blood was drawn a few minutes later. Id.

Several days later, on September 5, 2018, Patrol Officer Kira Goodheart contacted Keller about the accident. T 38. Goodheart had investigated the scene on the night of the accident, but had not seen or talked to Keller that night. T 28-29; 42. Keller told Goodheart he was home and agreed to meet with her there. T 38.

Keller was on his porch when Goodheart arrived, and she noticed his arm was in a sling. T 38. Goodheart did not specify in her testimony which arm was in a sling. Keller reportedly told her that as a result of the accident “he sustained a broken right humerus.” Id.

Keller told Goodheart he had used heroin, fentanyl, or a combination of the two, at least four hours or more before the accident and explained he was enrolled in a methadone clinic. He acknowledged he had struggled with substance misuse for about six years. T 39-40, 45, 47-48. He told Goodheart he had fallen asleep or passed out when he left his home at 1:00 a.m. to buy cigarettes, and that drugs did not contribute to the accident. T 39, 46, 49. Goodheart asked Keller to sign a medical release form to allow release of any records regarding his injuries and Keller agreed. T 41, 45. No medical records were offered in evidence at trial.

Consistent with Keller’s admission, the analyses of his blood showed the presence of methadone, fentanyl and amphetamines, but no alcohol. T 76, 77, 88, 97. Specifically, the blood analyses showed 198 nanograms per milliliter of methadone, with a range of uncertainty of plus or minus 30 nanograms (T 77-78), 53 nanograms per milliliter of fentanyl (T 88), and 290 nanograms of methamphetamines (T 97). 1 Keller was subsequently charged with a violation of RSA 265- A:3, I(b), knowingly driving a vehicle on a way while under the influence of intoxicating liquor and/or drugs, causing a motor vehicle collision resulting in serious bodily injury to himself. A 3.

At trial, the State called three experts to testify about the results of blood analyses. None of those experts could opine about whether Keller was impaired at the time of the accident based on the results. T 79; 92; 99-100. The State’s final witness was Colleen Scarneo. Scarneo was a consultant with a firm she started in 2020 called New England Forensic Toxicology. T 109. Until 2020, she worked for sixteen years for the New Hampshire Department of Safety, Division of State Police, Forensic Toxicology Laboratory. T 109-10. Scarneo was recognized as an expert in “human performance forensic toxicology.” T 112-13. Based on her review of three police reports regarding Keller’s accident, a statement by his girlfriend, blood sample collection information, medical records and blood testing lab reports, Scarneo gave the following opinion:

The jury found Keller guilty on November 22, 2022. T 219.

1. The court erred in denying Keller’s motion in limine and allowing the testimony of the State’s expert on human performance forensic toxicology, Colleen Scarneo. The methodology Scarneo used was not sufficiently reliable to meet the requirements for admission and her application of it to the facts of this case was so deficient as to skew even that methodology. Admission of Scarneo’s testimony prejudiced Keller because the jury heard unreliable evidence on the critical element of impairment.

2. The required element of serious bodily injury was not adequately supported by the evidence. There was no evidence supporting the conclusion that Keller’s broken arm was severe, permanent or resulted in protracted loss of function or impairment of health.

The opinion testimony of the State’s expert, Scarneo, improperly supported the conclusion that Keller’s accident was caused by drug use impairment. The court erred in allowing Scarneo’s testimony that Keller exhibited “signs and symptoms” “consistent” with drug impairment because the approach she used to reach her opinion was inherently unreliable. And even if her general methodology had met the minimum test for reliability, her application of it to the facts of this case was fundamentally flawed, providing an independent reason to exclude her testimony.

To sustain a guilty verdict on a charge of driving under the influence of drugs under RSA 265-A:3, I(b), the State must prove beyond a reasonable doubt that the defendant was impaired as a result of his use of alcohol or drugs. State v. Ducharme, 167 N.H. 606, 617 (2015). Impairment to any degree is sufficient. Id. But unlike the case with alcohol, there is no bright line or per se standard at which a concentration of drugs in a blood test proves the required impairment.

In this case, there was no percipient witness testimony that suggested or proved impairment. The only officer who saw Keller the night of the accident saw no signs of impairment. T 58, 66. There was no testimony from an eye- witness or accident reconstruction expert establishing that the manner of the accident proved impairment. To attempt to prove impairment, the State therefore offered the expert testimony of Scarneo.

Scarneo’s expert witness disclosure indicated she intended to give testimony “explaining the pharmacological and toxicological effects associated with the drugs found in [Keller’s] blood when used alone and concomitantly, as well as their effects on driving.” A 50. Beyond that, however, Scarneo promised to offer an opinion based on testimony and statements of witnesses and other discovery that Keller in fact demonstrated “signs and symptoms of impairment” “consistent with” the effects of drugs. Id. See also M 20. Defendant moved in limine to preclude Scarneo’s testimony on the grounds that it was inadmissible under New Hampshire Rules of Evidence 403 and 702, RSA 516:29-a, and Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993). A 5-18. The court held an evidentiary hearing and denied the motion. Supp. 33-45. Incorporating the reasoning of a sister court in State v. Kelley, Merrimack Cnty. Super. Ct. 09-CR-1051 (Jan. 4, 2011) (Smukler, J.) (A 32), the court held that Scarneo’s testimony satisfied the factors laid out in RSA 516:29-a and Daubert and was therefore reliable. Supp. 37-39. The court also concluded that Scarneo’s testimony would be helpful and would not usurp the jury’s role, because she would be putting evidence, including witness testimony, “into context, ” and her opinions involved “complexities beyond the ken of the average juror[.]” Supp. 42- 44. Keller renewed his objection at trial, but Scarneo’s testimony was allowed, and she was recognized as an expert in “human performance forensic toxicology.” T 112-13. Although the burden is not “especially onerous” because Rule 702 has been interpreted liberally in favor of the admission of expert testimony, the proponent must prove that expert testimony is admissible. Moscicki v. Leno, 173 N.H. 121, 125 (2020). This Court reviews a trial court’s determination of expert testimony admissibility for an unsustainable exercise of discretion. Baxter v. Temple, 157 N.H. 280, 286 (2008).

Scarneo was a consultant with a firm she started in 2020 called New England Forensic Toxicology. M 4. She has a bachelor’s degree in biology and a master’s degree in veterinary medical sciences from the College of Veterinary Medical Science at the University of Florida, with a concentration in forensic toxicology. M 6-7. Before she started her consulting firm, she was employed in the forensic toxicology laboratory of the New Hampshire Department of Safety, Division of State Police. M 5.

Scarneo concluded, and was allowed to testify, that Keller demonstrated “signs and symptoms of impairment” that were consistent with the effects of drugs found in his system. M 20. She did not identify the specific signs and symptoms she believed Keller exhibited or what the evidence of them was. She also did not explain what it means for a symptom or sign to be “consistent with” drug effects. Scarneo testified that, in general, her method to reach her human performance opinions consists of reviewing information about a defendant’s operation of a vehicle from police reports, witness statements and medical records. M 11-12. She looks at blood test toxicology results and reviews “the literature” for therapeutic and toxic and fatal concentrations to “come up with an understanding about just the numbers and what they may or may not mean in terms of their use.” M 12. Scarneo then draws on her general knowledge of pharmacokinetics and literature to determine “whether or not I feel that there was signs and symptoms of impairment that would be consistent with the drugs found in that individual’s body.” M 13. She conceded that it is a “challenging situation” to determine whether someone is impaired by drug use because there are no per se limits. M 20-21. She claimed she would not base her opinion on blood results alone, and did not do so in this case. M 29. Instead, she purported to call upon her experience, training and education to reach a conclusion about whether there were signs and symptoms consistent with impairment. M 21. The function of a trial court evaluating expert testimony is as a gatekeeper, “ensuring a methodology’s reliability before permitting the fact-finder to determine the weight and credibility to be afforded an expert’s testimony.” Baker Valley Lumber, Inc. v. Ingersoll-Rand Co., 148 N.H. 609, 616 (2002) citing Daubert, 509 U.S. at 592-95 (1993). “The proper focus for the trial court is the reliability of the expert’s methodology or technique.” Id. To evaluate reliability, the court must consider the four factors set forth in Daubert, applied in Baker Valley and codified in RSA 516:29 II:

(expert’s failure to satisfy any other set of reliability criteria when unable to satisfy Daubert factors justifies exclusion). Scarneo used no standardized rubric or defined assessment tool in making her assessment. M 39. As a result, no one else—including a court—can evaluate her methodology to ensure she didn’t miss a step, overlook an important part of the analysis or mis-apply part of the method. She readily conceded that “the interpretation and formulation of [her] opinion happens in [her] mind[.]” Id. Indeed, she couldn’t think of a way her approach could be tested other than for someone else to review the same material and come up with their independent opinion. M 40. That, of course, would test only her result in a particular case, not the reliability of her methodology.

With respect to peer review and general acceptance, Scarneo contended her approach has been accepted and published in peer review journals and reference books (M 42- 43), but she listed no such publications in her witness disclosure nor did she mention any in her testimony. When asked if she was testifying there are peer-reviewed studies approving techniques for determining whether someone’s driving was impaired based on blood results and police reports, she side-stepped the question:

M 43. But she didn’t identify any studies, peer-reviewed or otherwise, that approved her flexible, no check-list test for opining about signs and symptoms.

There is no known or potential error rate or confidence interval for Scarneo’s novel methodology, and no way to identify such an error rate. M 47. She identified no test of her methodology. M 49. When asked if her methodology could be tested by comparing her results to the results of a drug recognition expert evaluating the same person in person, she said she didn’t see how that would work. Id. In the end, Scarneo failed to connect her opinion that Keller showed “signs and symptoms” of impairment to any reliable scientific methodology. Although Scarneo claimed her method is standard and accepted, the State did not sustain its burden to prove that to be the case. Even more fundamentally, the State didn’t explain how an expert opinion about “signs and symptoms consistent with the use of drugs” could be reliable when the signs and symptoms are not even identified.

The court’s reliance on the Superior Court’s reasoning in State v. Kelley was misplaced. Supp. 38-39. The Kelley court found Scarneo qualified to offer an opinion on a driver’s symptoms and possible impairment, and concluded that her methodology was reliable enough to meet the threshold for admissibility. A 37-39. But the analysis in that case does not support admissibility of Scarneo’s opinion here.

As a threshold matter, the record in Kelley appears to have been quite different than the record here. For example, the Kelley court found that the record supported a finding that Scarneo’s methods have been subject to peer review. A 38. Even if accurate as a description of the record in that case, the point is not relevant here, because the State did not offer evidence of peer review. More fundamentally, the Kelley court appears to have conflated Scarneo’s expertise with the need for peer review of the scientific principles underlying her inquiry. It supported its peer review finding with evidence that Scarneo had attended conferences and classes “on the effects of alcohol and drugs on human performance and behavior.” Id. Keller does not dispute there is such a relationship, but rather challenges the method Scarneo used to identify its significance in this case. The Kelley court did not analyze that issue. In any case, Scarneo’s education does not constitute peer review of her methods.

Similarly, the Kelley court relied on the existence of error rates for the laboratory tests measuring drug blood levels for the third Daubert factor of whether such error rates are known for the challenged technique. A 38. But Scarneo conceded here that her methodology has no determinate or determinable error rate, confidence interval or range of uncertainty. M 47. The Kelley court also misstepped on the general acceptance factor, concluding it was satisfied for Scarneo’s methodology because the known side effects of drugs “have such widespread and general acceptance” in the scientific community. A 39. The general acceptance of that concept says nothing about the reliability of Scarneo’s approach to identifying signs and symptoms of impairment in an individual case.

Finally, the Kelley court concluded that Scarneo’s methodology was reliable because it has been tested. Its support for this conclusion was that Scarneo has attended seminars and written articles on toxicology as it applies to drivers under the influence, and applied her extensive knowledge and experience to the case record. A 37-38. This analysis falls short of the rigorous analysis required by Daubert, and begs the question of how her approach has been tested, and by whom. Scarneo concedes that tests have not been done of her methodology. M 49.

The court here did not articulate its own analysis of the Daubert factors beyond incorporating “the reasoning contained in Kelly [sic].” Supp. 38. For the reasons described above, Kelley neither supports the conclusion here nor can withstand scrutiny on its own terms. Scarneo’s testimony was not reliable and should not have been admitted. Because Scarneo’s opinion testimony was critical to support the allegation that Keller was impaired at the time of the accident, this abuse of discretion was prejudicial, warranting reversal of Keller’s conviction.

Even if Scarneo’s methodology was sufficiently reliable in the abstract, her application of it in this case was so flawed as to render her testimony inadmissible for that independent ground. See State v. Langill, 157 N.H. 77, 87-88 (2008) (RSA 516:29-a, I(c) requires a court to examine whether a witness reliably applied the methodology to the facts of the case.) Though exclusion of expert testimony on the ground that application of a methodology was unreliable “is warranted only if the methodology was so altered by a deficient application as to skew the methodology itself, ” id. at 88, citing United States v. Gipson, 383 F.3d 689, 697 (8th Cir. 2004), that is the case here.

Scarneo claimed to have followed her general methodology in Keller’s case: she stated that she reviewed the police reports, a statement by Keller’s girlfriend, blood sample collection forms and lab results, and Keller’s hospital records from the accident. A 47; M 16-18. She claimed to be looking in these materials for “signs and symptoms of impairment that would be consistent with the use of those drugs that were found in the person’s blood.” M 27-28. Other than the blood laboratory results and Keller’s own statements to police about his drug use, however, Scarneo did not rely on anything in the materials she reviewed to support her conclusion that Keller demonstrated “signs and symptoms of impairment.” M 20. She did not point to a single piece of information in—or missing from—the police reports, witness statements or medical records that informed her opinion in any way other than the evidence of drug use. Even though she claimed to have reviewed his report (A 47), she did not mention in either her disclosure or testimony Officer McCormack’s conclusion, based on his personal observation of Keller minutes after the accident, that Keller showed no signs of impairment. In fact, other than including Keller’s medical records and his girlfriend’s statement in the section of her disclosure listing “data or other information I considered in forming my opinions” (A 47), she seemed not to have made use of them at all. Although police reports, witness statements and medical records are key items in the list of things Scarneo reviews as part of her methodology, she made no mention in her testimony of how these items in Keller’s case supported her opinion. Disregard of, or failure to mention, some of these items might arguably go to weight, rather than admissibility, but Scarneo failed to analyze the significance to her opinion of any of them other than the evidence of drug use or measurement. Thus, taking Scarneo at her word and assuming these inputs are in fact critical to her methodology, this Court must conclude that her failure to explain how they applied here rendered her conclusory opinion inadmissible.

Scarneo never identified the “signs and symptoms” of Keller’s alleged impairment either. Her opinion was little more than conclusion:

M 20.

In reality, despite her acknowledgement that the level of drugs in a person’s blood alone is not sufficient to prove impairment, Scarneo did not use a reliable “methodology” to evaluate Keller’s impairment, but improperly reached her opinion that he showed “signs and symptoms” based on his blood levels. She admitted that fact in her deposition, but changed her testimony at the hearing on the motion in limine:

M 38.

She confirmed that testimony at trial:

And that’s your opinion? Yes it is.”

You were asked those questions and gave those answers?

T 137. Scarneo provided no explanation for why she changed her testimony.

Scarneo’s failure to apply her own methodology, or to explain how she did, means that even if her approach could be reliable if properly applied, her opinion is not admissible here. The application of her purported methodology to information beyond blood test results is missing, thus skewing the methodology itself. Langill, 157 N.H. at 88. Scarneo’s opinion is based on virtually nothing other than her own ipse dixit, and should have been excluded.

Keller was convicted of one felony count of Aggravated Driving While Intoxicated – Serious Bodily Injury. RSA 265- A:3, I(b). In addition to the other elements of DUI, this aggravated offense requires a collision that results in serious bodily injury to the driver or another. State v. Fogg, 170 N.H. 234, 236 (2017). The injury underlying the charge against Keller was “a right proximal humerus fracture” sustained by Keller himself. A 3. Whether an injury constitutes, a “serious bodily injury” is generally a question of fact for a jury. See State v. Scognamiglio, 150 N.H. 534, 536 (2004); State v. Plaut, 124 N.H. 813, 813 (1984). But as with any other required element of a crime, the State bears the burden of proving a serious bodily injury beyond a reasonable doubt. To prevail on a challenge to the sufficiency of the evidence, “the defendant must prove that no rational trier of fact, viewing all of the evidence and all reasonable inferences from it in the light most favorable to the State, could have found guilt beyond a reasonable doubt.” State v. Wilmot, 163 N.H. 148, 154 (2012). Even under this standard, the evidence presented at trial did not support the conclusion that Keller suffered a “serious bodily injury.” A challenge to the sufficiency of the evidence is reviewed de novo. State v. Cable, 168 N.H. 673, 677 (2016).

The element of serious bodily injury in the charge against Keller was based on his own statement that he broke his arm in the accident. Officer Goodheart, who saw Keller five days after the incident, testified she “noticed that he was in an arm sling, ” and that Keller told her he had broken his arm in the collision. T 38. Goodheart subsequently got Keller’s consent to release his medical records, (T 41), but no records were introduced into evidence. Goodheart’s testimony was the entirety of the evidence presented on the injury at trial, and was insufficient to prove serious bodily injury.

RSA 265-A:3, I(b) incorporates RSA 625:11, which defines serious bodily injury as “any harm to the body which causes severe, permanent or protracted loss of or impairment to the health or of the function of any part of the body.” This Court has held that the definition is in the disjunctive, so that an injury must be severe, permanent or protracted. State v. MacArthur, 138 N.H. 597, 600 (1994). But a broken arm need not by definition be any of those. Even though Keller had released his medical records, the State presented no evidence that the injury to Keller’s arm was severe or permanent. Likewise, there was no evidence that the function of his arm was impaired for a protracted period. The bare fact that Keller’s arm was in a sling at a single point in time after the accident is not sufficient to give rise to a reasonable inference that his injury was severe or protracted, let alone permanent.

The State’s apparent strategic decision not to call any doctor or offer medical records, required the jury to speculate about the significance of Keller’s injury and whether it was severe, permanent or protracted. This effectively allowed the State to turn a broken arm into serious bodily injury, as a de facto matter of law. The State should not be permitted to forgo actual proof of facts demonstrating serious bodily injury and invite the jury to infer that a broken arm is always a severe, permanent or protracted injury regardless of any other facts.

To be sure, this Court has found cases of broken bones severe enough to constitute serious bodily injury. See Scognamiglio, 150 N.H. at 537 (broken nose, swollen discolored eyes, clogged breathing passages, sinus infection); State v. Candello, 170 N.H. 220, 223-24 (2017) (rib fractures, spleen laceration, three day hospitalization). But in those cases, the juries heard from medical professionals and other witnesses supporting the extent and severity of the injuries. As in those cases, the State may have been able to prove beyond a reasonable doubt that Keller’s broken arm was a serious bodily injury as well. But it made no effort to do so. Keller’s conviction should accordingly be reversed.

WHEREFORE, Mr. Keller respectfully requests that this Court reverse his conviction.

Undersigned counsel requests fifteen minutes of oral argument before a full panel.

The written decision of the court denying the motion in limine is appended to the brief. The other appealed decision was not in writing and therefore is not appended to the brief. This brief complies with the applicable word limitation and contains approximately 4768 words.

I hereby certify that a copy of this brief is being timely provided to the Criminal Bureau of the New Hampshire Attorney General’s office through the electronic filing system’s electronic service.