State of Minnesota, Respondent,

This opinion is nonprecedential except as provided by
Minn. R. Civ. App. P. 136.01, subd. 1(c).

STATE OF MINNESOTA
IN COURT OF APPEALS
A22-0426

State of Minnesota,
Respondent,

vs.

Brian Michael Garner,
Appellant.

Filed September 18, 2023
Affirmed
Worke, Judge
Dissenting, Ross, Judge

Itasca County District Court
File No. 31-CR-21-2347

Keith Ellison, Attorney General, St. Paul, Minnesota; and

Matti R. Adam, Itasca County Attorney, David S. Schmit, Assistant County Attorney,
Grand Rapids, Minnesota (for respondent)

Cathryn Middlebrook, Chief Appellate Public Defender, Benjamin J. Butler, Assistant
Public Defender, St. Paul, Minnesota (for appellant)

Considered and decided by Worke, Presiding Judge; Ross, Judge; and Smith,
John P., Judge.
*

* Retired judge of the Minnesota Court of Appeals, serving by appointment pursuant to
Minn. Const. art. VI, § 10.
2
NONPRECEDENTIAL OPINION
WORKE, Judge
Appellant argues that the district court erred by denying his petition for
postconviction relief seeking to withdraw his guilty plea because defense counsel was
ineffective for failing to investigate and advise him of a prescribed-use defense to his
first-degree driving-while-impaired (DWI) charge. We affirm.
FACTS
Respondent State of Minnesota charged appellant Brian Michael Garner with
first-degree DWI, criminal vehicular operation, fleeing a peace officer in a motor vehicle,
and driving after cancelation. See Minn. Stat. §§ 169A.20, subd. 1(7), .24(2), 609.2113,
subd. 1(6), 609.487, subd. 3, 171.24, subd. 5 (2020). Garner entered Norgaard pleas1 to
first-degree DWI and fleeing a peace officer in a motor vehicle, claiming that “drugs”
caused him to “blackout” at the time of the offenses. The state noted that blood testing
performed by the Minnesota Bureau of Criminal Apprehension (BCA) showed
amphetamine in Garner’s blood at the time of the offenses. Garner understood that, by
pleading guilty, he was “giving up any defenses.” The district court sentenced Garner
according to the plea agreement to concurrent sentences of 65 months in prison for the
DWI conviction and 22 months for the fleeing conviction.

1 A defendant enters a Norgaard plea when he “asserts an absence of memory on the
essential elements of the offense but pleads guilty because the record establishes, and the
defendant reasonably believes, that the state has sufficient evidence to obtain a conviction.”
Williams v. State, 760 N.W.2d 8, 12 (Minn. App. 2009), rev. denied (Minn. Apr. 21, 2009);
see also State ex rel. Norgaard v. Tahash, 110 N.W.2d 867, 871 (Minn. 1961).
3
Garner appealed, but this court stayed the appeal for Garner to seek postconviction
relief. In his postconviction petition, Garner requested to withdraw his guilty plea,
claiming that it was invalid because his attorney failed to investigate a prescribed-use
defense to the DWI charge.
At an evidentiary hearing, Garner testified that he had a prescription for Vyvanse
and was using it as prescribed during the offenses. Garner testified that he asked his trial
counsel to investigate the amphetamines that would show up in his blood because of his
use of Vyvanse. Garner recalled signing a release of information for trial counsel to
investigate the prescription. Garner testified that had trial counsel investigated a
prescribed-use defense, he would not have pleaded guilty.
But Garner also testified that he did not understand that the prescription was relevant
to his case. He also failed to provide a copy of his prescription at the hearing. And he
shifted his explanations regarding why he did not produce a copy, first claiming that the
pharmacy would not give him a copy and then claiming that he failed to obtain a copy
because he was in prison. He also admitted that he had not asked his postconviction
counsel to obtain a copy of his prescription.
Garner’s trial counsel testified that he and Garner demanded a speedy trial believing
that the state would not have the BCA test results in time for trial. Trial counsel noted that
Garner told him that the results would show “ Vyvanse . . . and marijuana” but Garner did
not state that Vyvanse would show in the report as an amphetamine. Trial counsel testified
that Garner never gave him “any information about a Vyvanse prescription” or that Garner
ever made him “aware” that “there existed a prescription” for Vyvanse.
4
Trial counsel testified that he discussed with Garner that he planned to review the
BCA test results to determine whether there was a viable defense of “intoxication by a
prescription drug.” Based on police reports and videos from the squad-car cameras
showing “multiple indicia of intoxication,” however, it appeared unlikely that Garner was
under only “therapeutic levels of . . . [a] controlled substance.”
Trial counsel testified that he previously investigated how Vyvanse would appear
on a drug test in a prior case with similar facts. Based on his experience, trial counsel
testified that it would “take a tremendous amount of time” to locate a n expert who would
conclude “that the amount of drugs in” Garner’s body was “within a therapeutic range” to
establish a prescribed- use defense. He explained that it was difficult to find an expert
because every prescription is unique and the affirmative defense requires a showing that
the defendant was taking the medication as prescribed. But trial counsel testified that he
never contacted the BCA scientist who performed Garner’s blood test. And in an email to
Garner’s postconviction counsel, the BCA scientist stated that the concentration of
amphetamine in Garner’s blood was “within the therapeutic range for . . . amphetamine.”
Trial counsel testified that he and Garner discussed the prescribed-use defense
immediately prior to tendering a plea of guilty. Trial counsel and Garner discussed the
pros and cons of using the blood-test results versus using a speedy-trial-demand tactic in
hopes of receiving a better plea offer. Trial counsel testified that the initial plea offers
required Garner to plead guilty to “at least” first-degree DWI with a guideline sentence or
a top-of-the-box sentence, because of the “egregious facts.” Trial counsel testified that
Garner leveraged the speedy-trial-demand tactic and received a better plea offer.
5
Trial counsel also testified that he warned Garner about the county attorney’s
practice of revoking offers if the defendant raised contested omnibus issues or went to trial.
Trial counsel advised Garner that if the offer was revoked, Garner could spend months
longer in a “condemned jail” while trial counsel “tried to come up with a different defense
that [he] didn’t think was viable in the first place.” Trial counsel testified further that before
pleading guilty at the plea hearing, he and Garner discussed whether to proceed as planned
or “do something different[]” such as investigate a prescribed-use defense.
The district court denied Garne r relief, concluding that Garner failed to show “a
reasonable probability that the outcome would have been different” had counsel not made
the alleged error because Garner provided “no evidence of a prescription.” This appeal
followed.
DECISION
Garner argues that his counsel was ineffective because he failed to investigate
Garner’s prescription and advise him of a prescribed-use defense to DWI. Because of the
ineffective representation, Garner claims that his guilty plea is invalid and he must be
allowed to withdraw it.
When, as here, a defendant files a direct appeal and then has it stayed to pursue
postconviction relief, we review the postconviction court’s decision using the same
standard as applied on direct appeal. State v. Beecroft, 813 N.W.2d 814, 836 (Minn. 2012).
We review legal questions on ineffective-assistance-of-counsel claims de novo. State v.
Nicks, 831 N.W.2d 493, 503 (Minn. 2013). We similarly review the validity of a guilty
plea de novo. State v. Raleigh, 778 N.W.2d 90, 94 (Minn. 2010). But we “give great
6
deference to a district court’s findings of fact.” State v. Andersen, 784 N.W.2d 320, 334
(Minn. 2010). We will not set aside factual findings unless they are clearly erroneous. Id.
Factual findings are not clearly erroneous if “reasonable evidence” supports them. State v.
Evans, 756 N.W.2d 854, 870 (Minn. 2008) (quotation omitted) . A postconviction
petitioner bears the bur den of proving the facts warranting relief by a fair preponderance
of the evidence. Minn. Stat. § 590.04, subd. 3 (2022).
A guilty plea may be invalid if the defendant received ineffective assistance of
counsel. State v. Ellis-Strong, 899 N.W.2d 531, 535-36 (Minn. App. 2017). To establish
an ineffective-assistance-of-counsel claim in the context of a guilty plea, the defendant
must show that “counsel’s representation fell below an objective standard of
reasonableness, and . . . a reasonable probability that, but for counsel’s errors, he would
not have pleaded guilty and would have insisted on going to trial.” Id. at 536 (quotation
omitted). To establish an ineffective-assistance claim, the defendant must meet both
prongs. See State v. Rhodes, 657 N.W.2d 823, 842 (Minn. 2003).
Garner claims that he pleaded guilty only because his trial counsel failed to
investigate and advise him of a prescribed-use defense to DWI.
Availability of defense
It is a crime for a person to drive a motor vehicle with “any amount of a controlled
substance listed in Schedule I or II.” Minn. Stat. § 169A.20, subd. 1(7). But it is an
affirmative defense when the defendant used the controlled substance according to the
terms of a prescription issued for the defendant. Minn. Stat. § 169A.46, subd. 2 (2022).
Because the DWI charge was based on the presence of amphetamine in Garner’s blood, the
7
prescribed use of Vyvanse—which contains amphetamine—may have established a
defense. See Minn. Stat. § 152.02, subd. 3(d)(1) (2020) (making “amphetamine” a
Schedule II controlled substance).
Trial counsel’s representation
Having determined that the prescribed-use defense may have been available to
Garner, we must next determine whether Garner showed that his attorney was ineffective
for failing to investigate and advise him of the defense.
There is a strong presumption that counsel’s performance was reasonable.
Andersen v. State, 830 N.W.2d 1, 10 (Minn. 2013). “An attorney’s representation meets
the objective standard of reasonableness if the attorney exercises the customary skills and
diligence that a reasonably competent attorney would exercise under the circumstances.”
Swaney v. State, 882 N.W.2d 207, 217 (Minn. 2016).
Counsel’s investigation into a defense is considered trial strategy. See Opsahl v.
State, 677 N.W.2d 414, 421 (Minn. 2004). While we generally decline to review
challenges to trial counsel’s strategy, id., the failure to investigate a defense may constitute
unreasonable representation if the failure was due to “inattention or neglect.” Swaney, 882
N.W.2d at 218. But even an incomplete investigation is objectively reasonable if supported
by the circumstances or if counsel made a reasoned decision that made further investigation
unnecessary. Id. In reviewing counsel’s investigation into a defense, we assess trial
counsel’s conduct “under prevailing professional norms, which includes a
context-dependent consideration of the challenged conduct as seen from counsel’s
perspective at the time .” Wiggins v. Smith, 539 U.S. 510, 523 (2003) (quotations and
8
citation omitted). Our review “must be highly deferential” to counsel’s performance.
Rhodes, 657 N.W.2d at 844 (quotation omitted).
Here, we cannot conclude that trial counsel was neglectful in his investigation or in
advising Garner about the prescribed-use defense. The record shows that trial counsel and
Garner demanded a speedy trial in a strategic move, believing that the state would not have
the test results in time for trial. Trial counsel had used this strategy in other cases.
After they received the blood-test results on the eve before trial, trial counsel
discussed with Garner the pros and cons of “doing something” with the test results, i.e.
investigating further in to a prescribed-use defense. But Garner never provided his trial
counsel with “any information about a Vyvanse prescription” or made his counsel “aware”
that “there existed a prescription” for Vyvanse. As the district court concluded: “[Garner]
has not produced verification of an existing prescription for Vyvanse, medical marijuana,
nor any other medication, in place at the time of the offenses or any time prior to the
offenses.” Without a prescription, there is no evidence of the prescribed use and there is
no affirmative defense. 2 See e.g., Grover-Tsimi v. State, A11-979, 2012 WL 34056, *5
(Minn. App. Jan. 9, 2012) (concluding that attorney was not ineffective for failing to
present affirmative defense that disorderly conduct was caused by epileptic seizure because
appellant does not have epilepsy).

2 Under the affirmative-defense statute, there must be evidence of (1) a current prescription
and (2) use in accordance with the terms of that prescription. See Minn. Stat. § 169A.46,
subd. 2 (stating that impairment-from-prescription-drug defense is available only if it is
“proven by a preponderance of the evidence . . . that the defendant used the controlled
substance according to the terms of a prescription issued for the defendant”).

9
Trial counsel had experience from a prior case with similar facts and knew that
locating an expert to testify regarding whether the substances in the sample were within a
therapeutic range would be difficult. Similarly difficult would be finding an expert to show
that Garner was taking the medication as prescribed. And counsel believed that, based on
his review of police reports and videos from the squad-car cameras showing “multiple
indicia of intoxication,” it was unlikely that Garner was under only “therapeutic levels of
. . . [a] controlled substance.” Trial counsel considered the defense even though Garner
never provided an actual prescription.
Trial counsel discussed the defense with Garner, but rather than proceed with trial
and raise the defense, they decided to use the leverage from demanding a speedy trial to
obtain a better plea offer. Garner received a “good” offer with a bottom-of-the-box
presumptive sentence for the felony-DWI offense.
The dissent suggests that we “overlook[ed] the district court’s unfavorable treatment
of [trial counsel’s] testimony,” and that we “credited the testimony as if the attorney in fact
discussed the defense with Garner, [while] the district court refused to do so.” The dissent
assumes that the district court refused to credit trial counsel’s testimony that he and Garner
discussed the prescribed-use defense. The district court found:
[Trial counsel] provided conflicting testimony
regarding whether he and [Garner] even thoroughly discussed
an affirmative prescription defense. He indicated that the
matter was discussed during an off-record conversation during
the plea hearing, but he immediately contradicted himself in
then asserting that he “was unaware of any prescription for
either Vyvanse or marijuana.”

10
The district court did not “refuse” to credit trial counsel’s testimony that he
discussed the possibility of the defense with Garner; rather, it found trial counsel’s
testimony conflicting only as to whether they discussed it thoroughly. Not having a
thorough discussion does not mean a discussion did not occur. As trial counsel testified,
and the district court credited, trial counsel did not believe the prescribed-use defense was
viable because of difficulties his office had experienced with the defense in an analogous
case and because of discovery showing Garner displaying “multiple indicia of intoxication”
that he did not believe would have been the result of prescribed usage.
Trial counsel was aware, based on his experience with the county attorney, that if
Garner continued with a contested omnibus hearing, the offer would have been revoked.
Trial counsel advised Garner that if the offer were revoked, Garner would likely spend
months “in [a] condemned jail” while they considered a defense that was not likely viable.
Additionally, trial counsel considered that if the offer were revoked, Garner could face
potential consecutive sentences. Trial counsel discussed with Garner that the facts of the
case—Garner fled for a significant distance, drove in excess of 100 miles per hour, and
crashed his vehicle carrying a passenger who suffered severe injuries —could expose him
to consecutive sentences. At the plea hearing, Garner confirmed that he wanted to plead
guilty rather than raise a defense.
Considering the deference afforded to counsel, and his perspective at the time,
including his experience with a prior case with analogous facts, his experience with the
county attorney, and his understanding of the facts and potential sentencing, we cannot
conclude that counsel provided objectively unreasonable representation. Counsel did not
11
neglect to investigate the defense; rather, after advising Garner of the relevant
circumstances, they reasonably decided to proceed with leveraging the speedy-trial
demand. Garner fails to show that his counsel’s performance fell below an objective
standard of reasonableness.
Prejudice
Although we have concluded that trial counsel’s representation was not
constitutionally ineffective and could conclude our analysis here, we also determine that
Garner fails to show prejudice.
There is no presumption of prejudice in an ineffective-assistance -of-counsel claim.
Gates v. State, 398 N.W.2d 558, 562 (Minn. 1987). Rather, Garner must show that “but
for [counsel’s] errors the result of the proceeding probably would have been different.”
Rhodes, 657 N.W.2d at 842 (quotation omitted). Specially, Garner must show that he
would not have pleaded guilty without trial counsel’s alleged error. See Ellis-Strong, 899
N.W.2d at 540.
Garner seems to claim that the district court legally erred by concluding that,
because Garner failed to present a prescription, he failed to show prejudice. The district
court stated that Garner produced no “verification” of a prescription, and that his “mere
assertion” that the prescription existed did not show prejudice. It is “implicit in these
findings” that the district court discredited Garner’s testimony regarding the alleged
Vyvanse prescription, and that Garner therefore failed to prove that he had such a
prescription without extrinsic evidence of it. See State v. Blom, 682 N.W.2d 578, 619
(Minn. 2004) (concluding that postconviction credibility findings were implicit in district
12
court’s determination that appellant did not show prejudice to support
ineffective-assistance claim). The district court’s credibility finding is entitled to
deference. See Evans, 756 N.W.2d at 870.
The BCA scientist’s statement that a potentially therapeutic amount of amphetamine
was in Garner’s blood does not undermine the district court’s credibility determination.
Having a therapeutic range of amphetamine in one’s system does not mean that a
prescription existed for the medication. The dissent notes that “the postconviction attorney
who did investigate learned from the same scientist that this amount was ‘within the
therapeutic range for amphetamine.’” Yet the postconviction attorney did not acquire any
prescription. Garner testified that his postconviction counsel spoke with the doctor who
prescribed him Vyvanse “about [his] prescription medication,” but Garner did not testify
that his postconviction counsel and the doctor discussed any Vyvanse prescription.
Garner also fails to explain why he would have insisted on raising a prescribed-use
defense instead of pleading guilty when he failed to prove that he had a prescription to
support the defense. Garner’s conclusory testimony that he would not have pleaded guilty
“[b]ecause the facts and [his] rights would have been proved” seems to change nothing on
this point. See Lee v. U.S., 582 U.S. 357, 369 (2017) (“Courts should not upset a plea
solely because of post hoc assertions from a defendant about how he would have pleaded
but for his attorney’s deficiencies. Judges should instead look to contemporaneous
evidence to substantiate a defendant’s expressed preferences.”). Without proving that he
had a Vyvanse prescription, Garner cannot show prejudice. See State v. Ecker, 524 N.W.2d
712, 718 (Minn. 1994) (holding that appellant failed to show prejudice to support
13
guilty-plea withdrawal based on alleged ineffective assistance in failing to investigate
possible exculpatory witnesses and defenses when appellant did not “show [that] witnesses
would have been found” or that “any defense . . . was realistic or appropriate under the
facts of his case”); see also Hill v. Lockhart, 474 U.S. 52, 59 (1985) (“[W]he[n] the alleged
error of counsel is a failure to advise the defendant of a potential affirmative defense . . . ,
the . . . ‘prejudice’ inquiry will depend largely on whether the affirmative defense likely
would have succeeded at trial.”).
Garner does not dispute that trial counsel advised him that the plea offer he accepted
was “objectively good.” He also does not dispute that trial counsel advised him that he
would go to prison if convicted on at least the first-degree DWI and would likely face
consecutive sentences. The presumptive sentence for first-degree DWI, calculated with
Garner’s 12 criminal -history points, was 72 months in prison (62 months-84 months
range). See Minn. Sent’g Guidelines 4.A (2021). The other top charges also carried
presumptive prison sentences and each charge was eligible for consecutive sentencing. If
the first-degree DWI were sentenced based on that criminal- history score and the other
charges were sentenced consecutively, Garner could have faced a significantly longer
sentence than he received with the plea agreement.
The plea offer accepted by Garner appears relatively favorable compared to prior
offers and other sentencing scenarios, which supports a lack of prejudice. See Lee , 582
U.S. at 367 (noting that “[t]he decision whether to plead guilty . . . involves assessing the
respective consequences of a conviction after trial and by plea”).
14
Given these considerations, even if Garner’s failure to prove that he had a Vyvanse
prescription does not alone defeat his claim of prejudice, the fact that he was offered a
favorable plea following counsel’s speedy-trial-demand tactic does. Based on this record,
Garner fails to show that his trial counsel was ineffective. The district court did not err by
denying Garner’s petition for postconviction relief.
Affirmed.

D-1
ROSS, Judge (dissenting)
The first time Brian Garner met his attorney, he gave the attorney all the information
necessary to reveal that the evidence might support a complete defense to the most serious
charge Garner faced. The attorney did nothing with that information and instead left Garner
to plead guilty to the charge even after he received the forensic evidence that supported the
defense. Because I am certain that failure to investigate or pursue the defense constitutes
ineffective assistance of counsel, I respectfully but strongly dissent.
Four simple circumstances present a basic legal question. First, a lawyer began
representing a defendant charged with impaired driving for having allegedly driven a car
after using methamphetamine. Second, the defendant repeatedly informed his lawyer that
he had not, in fact, used methamphetamine but had instead taken a named medication
regularly prescribed for a common attention disorder. Third, the law establishes that a
person who drives after having taken medication as medically prescribed is not, by virtue
of that conduct, guilty of impaired driving. And fourth, forensic laboratory testing
confirmed that the defendant indeed had not used methamphetamine but had ingested a
drug having the same chemical qualities as the medicine he told the lawyer he ingested, in
a dosage consistent wi th therapeutic use. The legal question is whether a lawyer armed
with this information provides adequate legal representation if he fails to do any of these
things: advise the defendant that the absolute defense of prescribed-use might be available
to avert an impaired-driving conviction altogether; investigate whether the defendant could
testify or provide other evidence that he had been prescribed the drug; or advise the
defendant not to hastily plead guilty to impaired driving and disclose to the state the

D-2
plausible affirmative defense to leverage a more favorable plea offer on remaining, lesser
charges. I believe that the only answer is no. I therefore disagree with the majority’s
holding that these failures were justified by the lawyer’s insightful “perspective at the time”
as part of some strategic plan to “use the leverage from demanding a speedy trial to obtain
a better plea offer.” I also dissent from the majority’s holding that, because the defendant
“was offered a favorable plea following counsel’s speedy-trial- demand tactic,” he
somehow was not prejudiced by his lawyer’s failure to adequately pursue the potential
complete defense.
The state charged Garner with four criminal offenses: first-degree driving while
impaired, criminal vehicular operation, fleeing a peace officer in a motor vehicle, and
driving with a canceled driver’s license. The impaired-driving charge was the most serious,
exposing Garner to the longest presumptive prison term. The state based that charge on its
allegation that Garner had driven a motor vehicle after using methamphetamine. But
Garner knew that he could not be guilty of impaired driving under that theory, because he
had not ingested methamphetamine. And that’s exactly what he told his attorney the first
time they met and repeatedly after that. Garner told his attorney that he had ingested an
amphetamine in the form of Vyvanse (which is a prescription drug for the treatment of
Attention-Deficit/Hyperactivity Disorder (ADHD)) and marijuana (which Garner told his
attorney he had been medically prescribed and which cannot form the basis of the impaired-
driving offense as charged). If what Garner told his attorney about having consumed
Vyvanse and not methamphetamine was true, then Garner had not engaged in impaired
driving by virtue of using methamphetamine. Garner was also certain that the laboratory

D-3
tests would confirm his assertion that he had ingested Vyvanse, not methamphetamine, and
he so informed his lawyer. And as the district court found, Garner “recalls signing releases
of information in order to facilitate [his attorney’s] investigation of the prescription.”
But the attorney never investigated whether Garner was prescribed Vyvanse or
whether his use was within a prescribed range. The district court explained away the
omission: “Given the difficulties his office had in an analogous case, [the attorney] saw
any prospective affirmative defense involving prescription drugs as laborious, if not
impossible.” The district court based this finding on the attorney’s testimony that, in a case
he co-chaired, “we had repeatedly struck out in trying to find someone in that case, an
expert [who] would testify . . . that this amount of . . . drugs in this person’s system was . . .
like within the therapeutic range . . . such that we could use that defense.” The district
court’s explanation, like the majority’s adoption of it, fails to incorporate the facts of this
case. In this case, the forensic scientist who completed the forensic testing expressly
reported the amount of amphetamine in Garner’s system; and the postconviction attorney
who did investigate learned from the same scientist that this amount was “within the
therapeutic range for . . . amphetamine.”
And unlike the majority’s description of events, the record also does not establish
that the attorney ever even discussed the prescribed-use defense with Garner. The majority
frequently references the attorney’s representation to the district court that he discussed the
defense with Garner. But these references overlook the district court’s unfavorable
treatment of this testimony. Although the majority credits the testimony as if the attorney
in fact discussed the defense with Garner, the district court refused to do so. The district

D-4
court instead found that the attorney “provided conflicting testimony regarding whether he
and [Garner] even thoroughly discussed an affirmative prescription defense. He indicated
that the matter was discussed during an off-record conversation during the plea hearing,
but he immediately contradicted himself . . . .” By basing its reasoning on facts that the
district court deliberately never found and indeed expressly doubted, the majority builds
its legal conclusions on marshy ground. My dissent by contrast rests on defense counsel’s
failure to investigate and adequately advise Garner on the prescription-use defense as it
bears on Garner’s guilty plea.
The state did not wait for the lab results to pitch Garner a deal: plead guilty to first-
degree driving while impaired and fleeing a peace officer in a motor vehicle and the state
would drop the other two charges. It seems to me that Garner’s attorney then had viable
strategic options. He could have advised Garner not to immediately accept the plea offer
while he promptly investigated whether the prescribed-use defense could thwart the state’s
impaired-driving charge altogether. A reasonable investigation would entail, at the very
least, informing Garner about the prescribed -use defense, verifying that Garner would
testify (with or without documentary evidence) that he was prescribed the Vyvanse he
claimed to have ingested, and waiting for the lab results to determine whether the forensic
evidence supported Garner’s assertion that he had ingested only Vyvanse, not
methamphetamine, and that he did so in a dosage that supported therapeutic rather than
recreational use.
Under this approach, if it turns out (as it in fact did turn out) that the lab results
confirmed what Garner had represented, the attorney could then move (or threaten to move)

D-5
the district court to dismiss the impaired-driving charge for lack of evidence or negotiate a
better plea deal on the strength of the potential complete defense. Or he could simply
recommend that Garner reject the plea offer and prepare for trial, knowing that the
impaired-driving charge stood on unsteady legs given Garner’s anticipated testimonial
evidence that he took only a prescription drug and the corroborating scientific evidence
supporting therapeutic use. A more aggressive strategic option would include disclosing
the potential defense to the state and counter- offering a more favorable plea deal without
awaiting the test results. Other viable options existed, and I need not outline them all. My
point is that the only clearly indefensible option is the one taken —ushering Garner to
quickly plead guilty while doing nothing with the prescribed-use evidence.
The majority says, “The record shows that t rial counsel and Garner demanded a
speedy trial in a strategic move, believing that the state would not have the test results in
time for trial. Trial counsel had used this strategy in other cases.” This reasoning is sound
only in one sense but clearly flawed as applied here. I say this because it simply doesn’t
matter that counsel “believ[ed] that the state would not have the test results in time for
trial.” Nor does it matter that he “had used this strategy in other cases.” In this case, the
state in fact did receive the test results before the time set for trial. No one can deny the
rationality of using an early speedy-trial demand to pressure the state into a choice of either
offering the defendant an attractive deal to elicit a guilty plea or possibly having to prove
a chemical-use charge with no scientific chemical-use evidence. But a defense attorney’s
theoretical strategy of making a speedy-trial demand is immaterial to our question, which
is whether Garner’s attorney provided effective assistance of counsel in light of the

D-6
potential prescribed-use defense given the information and evidence available before and
after the state made its plea offer. The majority does not— and I believe cannot—identify
any plausible strategic benefit from failing to take any steps to investigate the merit of the
defense, or at least raise the possibility of the defense with the prosecutor to renegotiate the
plea deal after the lab results corroborated Garner’s statements. The omission led Garner
to plead guilty to a crime for which the defense might have proven completely dispositive.
In sum, I am confident that there is no “strategy” justification for counsel’s failure
under the full, pre-plea picture. Garner had advised his attorney that the drugs in his system
were not methamphetamine, and the forensic evidence supported this claim. Garner had
advised his attorney that the drugs in his system were amphetamine, and the forensic
evidence supported this claim. The prescribed-use defense is available if the defendant
used the drug as prescribed, and the forensic evidence supported a claim of use in a
therapeutic quantity. If Garner’s attorney had simply asked Garner whether he had been
prescribed the drug, and Garner answered by telling his attorney what he later told the
district court in his postconviction testimony, the attorney could have used the information
to effectively represent Garner’s interests in advising him whether to plead guilty under the
terms the state offered. I believe that counsel’s performance fell below an objective
standard of effective representation.
I also believe that Garner has established that, but for his attorney’s failure to
investigate and apply the prescribed-use defense, he would not have pleaded guilty. “[T]o
satisfy the prejudice requirement, the defendant must show that there is a reasonable
probability that, but for counsel’s errors, he would not have pleaded guilty and would have

D-7
insisted on going to trial.” Hill v. Lockhart , 474 U.S. 52, 59 (1985) (quotation omitted).
Where, as here,
the alleged error of counsel is a failure to investigate or
discover potentially exculpatory evidence, the determination
whether the error “prejudiced” the defendant by causing him
to plead guilty rather than go to trial will depend on the
likelihood that discovery of the evidence would have led
counsel to change his recommendation as to the plea. This
assessment, in turn, will depend in large part on a prediction
whether the evidence likely would have changed the outcome
of a trial.

Id. Garner has made the required showing that he would not have pleaded guilty to impaired
driving had his attorney properly investigated and applied the defense. The only direct
evidence on this point is plain in the postconviction record:
Attorney: Do you believe you would have pleaded guilty if
your attorney had investigated the defense?

Garner: No.

. . . .

Attorney: And I should clarify; do you believe you would have
pleaded guilty [pursuant] to this plea agreement, to the felony
DWI?

Garner: No.

The district court concluded that Garner was not prejudiced by the failure to pursue
the possible prescribed-use defense, but it did so entirely on the following improper
reasoning:
If [the attorney] had obtained information from the prescriber
to confirm or deny the allegation of [Garner], then this
proceeding would not have been necessary. However, that is
dependent upon the provider actually providing the

D-8
information. . . . Ten months post sentencing, no evidence of a
prescription has been provided. . . . [Garner] has testified to an
alleged prescription, naming the facility, provider, and dosage,
and nothing more. The mere assertion of the existence of the
prescription does not show that the error would have made any
difference in the outcome. . . . [Garner] has not met his burden
to show that there is a reasonable probability that the outcome
would have been different.

Not only did the district court apply the wrong legal reasoning, its foundational premise
that “no evidence of a prescription has been provided” is also clearly erroneous. Testimony
is evidence. And Garner testified that Vyvanse is his “ADHD medication,” that “Rhea [del
Rosario] from Lakeview Behavioral Mental Health” prescribed the medication to him, that
he was prescribed the medication “in August . . . [of] 2021,” that his prescribed dosage was
“30 milligram[s]” initially and increased “to 50” milligrams “two weeks” later “because
the value . . . wasn’t high enough,” that he had the prescription “filled . . . [at] Walgreen’s,”
that he was on the medication “at the time of the offense,” and that “[I] wasn’t abusing
[the] medication” but “was taking it as prescribed . . . every day . . . at the same time.” The
district court treats this detailed, sworn testimony as a “mere assertion of the existence of
the prescription” rather than as specific evidence of a prescription, which it certainly is,
regardless of whether the district court believes it.
Most critical to our review on appeal, the district court made no factual finding that
Garner would have pleaded guilty to impaired driving had his attorney properly
investigated and advised him about the potential defense. The majority ignores the lack of
a fact-finding on this point and implicitly finds on its own that Garner certainly would have
pleaded guilty to impaired driving even if the defense had been properly investigated,

D-9
explained to him, and applied. I disagree with this approach for two reasons. First, it is not
our prerogative to find facts on appeal, as this is a duty left solely to the district court in
postconviction proceedings. See Scruggs v. State, 484 N.W.2d 21, 24 (Minn. 1992) (“The
postconviction court makes findings and conclusions to facilitate appellate review.”
(emphasis added)). Because the district court did not make the finding against Garner, we
should not resolve the appeal as if it had.
And second, I disagree with the reasoning in the majority’s fact- finding. The
majority reasons that “Garner . . . fails to explain why he would have insisted on raising a
prescribed-use defense instead of pleading guilty when he failed to prove that he had a
prescription to support the defense.” Here the majority seems to fall on the same mistake
that misled the district court. What matters here is not whether the district court or this
court believes that Garner had a prescription to support the defense, but whether Garner
believed he had evidence that could convince a jury at trial that he had a prescription to
support the defense. See State v. Ellis-Strong, 899 N.W.2d 531, 536 (Minn. App. 2017).
Again, his postconviction testimony on this issue was detailed and precise, and it was
sufficient to constitute evidence of a prescription with or without documentation. Put
differently, it is Garner’s informed understanding of what a jury might believe about the
defense at trial that bears on whether he would have pleaded guilty to impaired driving,
and whether the postconviction court believed he had a prescription is wholly irrelevant.
Garner testified that, had his understanding been informed by his attorney’s
investigation, he would not have pleaded guilty. Nothing in the record or logic undermines
that claim. I do not join the majority’s assertion that Garner “fails to explain why he would

D-10
have insisted on raising a prescribed-use defense instead of pleading guilty.” We do not
need Garner to explain why he would have raised the evidentially supported defense to
impaired driving instead of pleading guilty to impaired driving, because the explanation is
self-evident. The more obvious and relevant question is, why wouldn’t he raise the defense?
The majority is persuaded that Garner suffered no prejudice by his attorney’s failure
to investigate because “Garner received a ‘good’ offer with a bottom-of-the-box
presumptive sentence for the felony- DWI offense.” But we can call the sentencing offer
“good” only by assuming that Garner would have pleaded guilty to impaired driving in the
face of an informed decision about the defense against the impaired- driving charge. The
record lacks the necessary finding that he would have pleaded guilty, and I cannot join in
the assumption.
I would reverse the postconviction decision, allowing Garner to withdraw his guilty
plea and to seek a different deal or proceed to trial on the merits on all counts.