UNITED STATES v. PARASMO (2026)

Frank Parasmo, a medical doctor licensed to prescribe opioids in New York state, appeals from a judgment of conviction following a jury trial at which he was found guilty of thirty-two counts of unlawfully distributing a controlled substance in violation of 21 U.S.C. § 841. On appeal, Parasmo contends that (1) the district court incorrectly instructed the jury to apply an objective – rather than a subjective – standard of intent; (2) the district court improperly admitted expert testimony and evidence regarding New York State medical standards for the prescription of opioids; and (3) he received ineffective assistance of counsel.

These challenges fail. While we agree that the district court issued a defective charge – albeit one consistent with our precedent at the time – that error was harmless because it is “clear beyond a reasonable doubt that a rational jury would have found the defendant guilty absent the error.” United States v. Ng Lap Seng, 934 F.3d 110, 129 (2d Cir. 2019) (internal quotation marks omitted). Parasmo's evidentiary challenges similarly fall short because the expert testimony and evidence of New York State medical standards assisted the jury without usurping its role. Finally, we decline to address Parasmo's ineffective-assistance-of-counsel claim, which is better left for a motion pursuant to 28 U.S.C. § 2255. Accordingly, we affirm the judgment of the district court.

I. Background

Between 2014 and 2015, Parasmo issued prescriptions for large quantities of oxycodone and hydrocodone to at least twenty patients. Many of these patients displayed numerous red flags indicating that they should not have been prescribed either drug. Evidence introduced at trial included testimony that Parasmo continued prescribing opioids even after (1) he became aware that patients were addicted to opioids; (2) he learned that patients were diverting or selling their prescribed medications; (3) urine screens were negative for the prescribed opioids, indicating that the patients were not using the drugs as directed but instead diverting them to third parties; (4) patients were simultaneously abusing unprescribed narcotics; and (5) insurers, pharmacies, and the New York State Medical Society repeatedly issued written warnings that he was overprescribing opioids.

On January 2, 2019, a grand jury in the Eastern District of New York indicted Parasmo on thirty-five-counts of distributing oxycodone and hydrocodone without authorization, in violation of 21 U.S.C. § 841. App'x at 35. Parasmo proceeded to trial and was convicted on thirty-two counts on October 7, 2021. Parasmo filed a motion for acquittal under Federal Rule of Criminal Procedure 29 on December 15, 2021. On July 4, 2022, following the Supreme Court's decision in Ruan v. United States, 597 U.S. 450, 142 S.Ct. 2370, 213 L.Ed.2d 706 (2022), Parasmo filed a supplemental motion for acquittal, or in the alternative, for a new trial, pursuant to Federal Rule of Criminal Procedure 33. These motions were denied by the district court on January 30, 2023. Parasmo timely appealed that decision.

II. Discussion

Parasmo now challenges his conviction, arguing that (1) the district court's jury instructions were erroneous in light of Ruan; (2) several evidentiary rulings by the district court were erroneous; and (3) his trial counsel provided ineffective assistance of counsel. For the reasons set forth below, we conclude that the first two arguments lack merit, while the ineffective-assistance claims are best left for another day.

A. The Erroneous Jury Instruction Was Harmless.

Parasmo first contends that the district court improperly instructed the jury that it should apply an objective – rather than a subjective – standard of intent for each of the counts in the indictment. That instruction, Parasmo argues, ran afoul of the Supreme Court's holding in Ruan, which was decided after he was found guilty at trial.

“A jury instruction is erroneous if it misleads the jury as to the correct legal standard or does not adequately inform the jury on the law.” United States v. Wilkerson, 361 F.3d 717, 732 (2d Cir. 2004) (internal quotation marks omitted). “We review a claim of error in jury instructions de novo, reversing only where [the] appellant can show that, viewing the charge as a whole, there was a prejudicial error.” United States v. Moses, 109 F.4th 107, 114 (2d Cir. 2024) (internal quotation marks omitted). “Even where charging error is identified, ․ we will not reverse a conviction if the government can show harmlessness, i.e., show that it is clear beyond a reasonable doubt that a rational jury would have found the defendant guilty absent the error.” Ng Lap Seng, 934 F.3d at 129 (internal quotation marks omitted).

1. The Jury Instruction Was Erroneous in Light of Ruan.

Parasmo's statute of conviction, section 841(a)(1), makes it unlawful for any person to “knowingly or intentionally” “distribute[ ] or dispense” a controlled substance “[e]xcept as authorized.” 21 U.S.C. § 841(a)(1). As provided by federal regulation, registered doctors – like Parasmo – are “authorized” to issue prescriptions for controlled substances, but only if those prescriptions are issued “for a legitimate medical purpose” and “in the usual course of ․ professional practice.” 21 C.F.R. § 1306.04(a).

In Ruan, the Supreme Court considered the “state of mind that the [g]overnment must prove to convict ․ doctors of violating [section 841(a)(1)].” 597 U.S. at 454, 142 S.Ct. 2370. The Court ultimately adopted a “subjective” standard, explaining that when a defendant produces evidence that he is “authorized to dispense controlled substances,” the government must then “prove beyond a reasonable doubt that the defendant knew that ․ [ ]he was acting in an unauthorized manner.” Id. (emphasis added). In reaching that conclusion, the Court rejected the government's contention that the statute “implicitly contain[s] an ‘objectively reasonable good-faith effort’ or ‘objective honest-effort standard,’ ” observing that section 841(a)(1), like other criminal statutes, omits any mention of “words such as ‘good faith,’ ‘objectively,’ ‘reasonable,’ or ‘honest effort.’ ” Id. at 465, 142 S.Ct. 2370.

At the same time, Ruan did not “bar all consideration of objective criteria.” United States v. Bauer, 82 F.4th 522, 528 (6th Cir. 2023). To the contrary, the Supreme Court explained that the government can refer to “objective criteria such as ‘legitimate medical purpose’ and ‘usual course’ of ‘professional practice’ ” as circumstantial evidence of a defendant's subjective intent. Ruan, 597 U.S. at 467, 142 S.Ct. 2370 (quoting 21 C.F.R. § 1306.04(a)). “[T]he more unreasonable a defendant's asserted beliefs or misunderstandings are, especially as measured against objective criteria, the more likely the jury will find that the [g]overnment has carried its burden of proving knowledge.” Id. (alteration adopted and internal quotation marks omitted).

Here, the district court's jury instruction on mens rea – which predated the Supreme Court's decision in Ruan – conveyed that the government had the burden to prove that Parasmo knowingly and intentionally acted in an unauthorized manner. Specifically, the district court instructed that the government had to prove the following elements beyond a reasonable doubt: (1) that Parasmo “knowingly and intentionally distributed the controlled substance alleged in the indictment,” and (2) that he “knowingly and intentionally prescribed the controlled substances outside the bounds of professional medical practice and not for a legitimate medical purpose.” App'x at 2405. As to the second element, the district court elaborated that “when [a doctor] knowingly and intentionally acts outside the bounds of professional medical practice, and without a legitimate medical purpose in prescribing controlled substances, he is doing so in an unlawful manner.” Id. at 2406–07. And in setting out the meaning of “knowledge,” the district court made clear that knowledge “[could not] be established merely by demonstrating that the defendant was negligent, careless[,] or foolish,” id. at 2407, and that it was “not enough for the government to prove negligence, malpractice, carelessness[,] or sloppiness” on Parasmo's part, id. at 2410.

Parasmo primarily takes issue with the next portion of the jury instruction, which directed the jury to “determine whether the defendant acted in good faith.” Id. at 2411. Specifically, the district court instructed the jury that:

A doctor prescribes a drug in good faith in medically treating a patient when he prescribes the drug for a legitimate medical purpose in the usual course of practice; that is, the doctor has prescribed the drug lawfully. Good faith in this context means acting reasonably and with the honest exercise of best professional judgment as to a patient's needs; that is, the defendant acted in accordance with what he reasonably believed to be the standard of medical practice generally recognized and accepted in the State of New York. If you find that the defendant acted in good faith in prescribing the drugs, then you must find him not guilty. The government bears the burden of proving beyond a reasonable doubt that the defendant acted without a good faith belief that his distribution of the controlled substances ․ was for a legitimate medical purpose in the usual course of medical practice.1

Id. at 2411–12 (emphases added). According to Parasmo, this instruction is no longer good law after Ruan because it impermissibly transformed the subjective-intent standard into a lesser, objective one when it defined good faith to mean “acting reasonably” and requiring “reasonabl[e] belie[f].” Considering this charge in light of the instructions as a whole, we agree. See Moses, 109 F.4th at 114.

Of course, as Parasmo acknowledges, the district court did not have the benefit of the Supreme Court's holding in Ruan, and the good-faith instruction given in his case was consistent with our precedent at the time. Indeed, in United States v. Wexler, we approved of a similarly formulated charge that “[g]ood faith in this context means the honest exercise of best professional judgment as to a patient's medical needs,” and “that the doctor acted in accord with what he should have reasonably believed to be proper medical practice.” 522 F.3d 194, 205–06 (2d Cir. 2008). But as Ruan made clear, the government cannot meet its burden merely by proving that the physician lacked “objective good faith” or failed to act reasonably in issuing the prescriptions. 597 U.S. at 465, 142 S.Ct. 2370.

The government, for its part, asserts that “Ruan is wholly consistent with the jury instructions” issued by the district court and that “[n]othing about the good-faith charge modified th[e] correct articulation of the mens rea” standard. Gov't Br. at 43. As support, the government points to Ruan’s continued approval of objective criteria, including evidence of accepted professional standards of care, to evaluate the credibility of a doctor's asserted beliefs. But that argument misses the mark. The district court's good-faith instruction did not clearly inform the jury that it could use the gap between objective standards and Parasmo's conduct as circumstantial evidence to determine whether he acted with the requisite (subjective) intent. Instead, the charge incorrectly suggested that good faith, in and of itself, required that Parasmo was (objectively) “acting reasonably.” App'x at 2411.

To be sure, the instructions that the district court gave during Parasmo's trial at times hinted at the correct subjective standard of intent. As noted above, the district court's initial instructions as to the authorization element incorporated a “knowing[ ] and intentional[ ]” mens rea requirement and reiterated that the government had to prove more than that Parasmo was simply “negligent, careless[,] or foolish” in writing the prescriptions at issue. App'x at 2405–07, 2410; see also Ruan, 597 U.S. at 465–66, 142 S.Ct. 2370 (explaining that having a “defendant's criminal liability [turn] on the mental state of a hypothetical ‘reasonable’ doctor ․ reduces culpability on the all-important element of the crime to negligence” (internal quotation marks omitted)). And the good-faith instruction incorporated some subjective terms by referring to “good faith” as involving the “honest exercise of best professional judgment as to a patient's needs,” while further emphasizing that the government “b[ore] the burden of proving ․ that [Parasmo] acted without a good faith belief,” id. at 2411.

But other aspects of the good-faith instruction – either explicitly or implicitly – endorsed an objective standard of reasonableness. For example, the district court first explained that a doctor acts “in good faith ․ when he prescribes the drug for a legitimate medical purpose in the usual course of practice.” Id. It also defined “good faith” to require “acting reasonably,” and directed the jury to consider what Parasmo “reasonably believed to be the standard of medical practice.” Id. (emphases added). And at times, the district court referred to the second element without any clear reference to the mens rea requirement. See, e.g., id. at 2410 (“What the government must prove beyond a reasonable doubt is that ․ he was not writing those prescriptions for a legitimate medical purpose, but was instead writing them outside the usual course of professional practice.”).

Ultimately, because the jury charge in Parasmo's case blended pre- and post-Ruan instructions, we conclude that it was legally deficient. See United States v. Sabhnani, 599 F.3d 215, 237 (2d Cir. 2010) (finding instructions erroneous on de novo review where the “charge either fails to adequately inform the jury of the law, or misleads the jury as to the correct legal standard” (internal quotation marks omitted)).

2. The Erroneous Jury Instruction Was Harmless Because Overwhelming Evidence Supported the Guilty Verdict.

Nonetheless, as noted above, “[e]ven where charging error is identified, ․ we will not reverse a conviction if the government can show harmlessness, i.e., show that it is clear beyond a reasonable doubt that a rational jury would have found the defendant guilty absent the error.” Ng Lap Seng, 934 F.3d at 129 (internal quotation marks omitted). Based on the evidence presented at trial, we conclude that a jury would have reached the same result if properly instructed, and therefore, the instructional error was harmless.

To begin, the government's expert witness, Dr. Seth Waldman, testified regarding the generally accepted standards of medical practice in New York from 2009 through 2015. Dr. Waldman explained, among other general principles known to New York practitioners during the relevant period, that prescribing opioids like oxycodone presents risks of addiction and overdose, and that those risks increase when patients concurrently use controlled substances or illegal drugs like cocaine, heroin, and amphetamines. Due to these risks, Dr. Waldman testified that a practitioner acting in the usual course of medical practice would monitor a patient's use of prescribed medications, including by drug testing the patient. Such testing is “important,” he explained, because monitoring for the presence of other substances – or the absence of the prescribed medication – can reveal misuse. App'x at 665–66. In particular, taking oxycodone in combination with other illegal substances “becomes extremely dangerous.” Id. at 666. Dr. Waldman also testified that a negative oxycodone test could indicate a very different problem – namely, that the patient is “diverting that medication” by “selling it or giving it to someone else.” Id. at 667. Dr. Waldman further detailed the signs of “drug[-]seeking behavior,” which could include patients “repeatedly using their prescriptions faster than [scheduled],” “[l]osing prescriptions, having prescriptions stolen on a repetitive basis,” “receiving prescriptions from multiple doctors for the same condition,” and exhibiting physical signs “that the patient is using drugs.” Id. at 671–72.

Dr. Waldman's testimony is particularly compelling given the evidence the jury heard related to Parasmo's prescription of opioids to: (1) patients whose toxicology reports repeatedly reflected the use of illegal substances; (2) patients whose toxicology reports repeatedly reflected the absence of prescribed opioids, indicating that patients were not actually taking (and perhaps diverting) their prescribed medications; and (3) patients who had long histories of substance abuse, overdoses, and unsuccessful attempts at rehabilitation. The record also reflects that Parasmo himself recognized that he should cease prescribing opioids with respect to certain patients, and that third parties – including other physicians, pharmacists, insurance companies, and law-enforcement officers – repeatedly raised concerns to Parasmo about his prescription practices with patients who were doctor-shopping, taking other illegal substances, and/or diverting their prescriptions.

The district court's order denying Parasmo's post-trial motions for a new trial – albeit applying the pre-Ruan standard for intent – thoroughly documented the evidence introduced at trial, which included the following examples:

• Leonard Marino repeatedly reported to Parasmo, over the course of several years, that his medication had been lost, stolen, or run out early, which as Dr. Waldman explained, was a “sign that the patient is overusing or selling the medication.” Id. at 803. Marino's medical records also showed “multiple prescriptions from different doctors ․ overlapping in a very short period of time,” indicating that he was actively engaging in “doctor shopping” – a pattern that Parasmo was alerted to in a November 28, 2011 phone call from a pain management specialist. Id. at 797–98, 801. In 2014 and 2015, Marino also returned multiple abnormal toxicology reports that either showed no oxycodone (indicating that he was not taking his prescribed medication) or indicated the presence of other substances (in addition to his prescribed oxycodone). Despite indicating in Marino's chart in July 2015 that he would reduce the number of pills given to Marino to “wean” him off the drug, Parasmo increased his dosage two months later and continued to prescribe him oxycodone into December 2015 – even after he continued to return abnormal toxicology reports.

• John Lettenberger had a documented history of a substance-abuse disorder. His toxicology reports in 2009 revealed positive tests for cocaine, and Parasmo's notes in 2011 reflected his attendance at Narcotics Anonymous meetings. In April 2014, Parasmo's notes indicated that Lettenberger asked for fifty additional oxycodone pills, which he claimed to “owe[ ]” someone. Id. at 790–91. Although Parasmo refused to provide the additional pills, Dr. Waldman testified that Lettenberger's statements were “concerning” because they showed “that the patient is already and has been in the process of diverting at least some of his medications.” Id. at 791. Two months later, on June 2, 2014, Lettenberger tested positive for cocaine, but negative for oxycodone, his prescribed medication. According to Parasmo's records of an appointment the following day, Lettenberger told him he “had some coke at a party” and that after running out of oxycodone pills, he bought Vicodin “on the street.” Id. at 792–93. Parasmo's records from that appointment also reflect that Lettenberger was “mov[ing] about quite well” and was “very animated,” which, as Dr. Waldman explained, was inconsistent with Lettenberger's claim of having “more pain than oxy can handle.” Id. at 792. Nevertheless, on June 26, 2014, Parasmo issued a prescription for 240 30-mg oxycodone pills.

• Rocco Oliveri had a history of substance abuse and abnormal toxicology test results, which indicated the presence of buprenorphine, a drug administered for the treatment of substance abuse. In December 2012, Oliveri asked Parasmo to renew his prescription one week early; Parasmo obliged and prescribed 120 30-mg oxycodone pills, even though Oliveri continuously tested negative for the prescribed medications. After noting in August 2013 that he intended to “wean [Oliveri] from oxycodone,” id. at 869, Parasmo nevertheless continued to write prescriptions for 150 30-mg oxycodone pills. A week later, Oliveri told Parasmo that “his son's friends took his oxys from him and beat him up,” and that “he ha[d] never taken any himself” because “he gives all his [ ]oxys to his son.” Id. at 870–71. Oliveri's addiction specialist “specifically requested that the doctor not write any more oxycodone” for Oliveri. Id. at 871. And yet, on November 12, 2013, Parasmo prescribed more oxycodone for Oliveri.

• Leslie Finnegan-Andrews also had a history of abnormal toxicology test results, which indicated the presence of numerous unprescribed drugs. On January 23, 2014, Parasmo noted “oxycodone addiction – wean” on Finnegan-Andrews's medical record. Id. at 1690. In October 2015, Parasmo noted that Finnegan-Andrews was “continually asking for more pain meds,” “her brother who needs pain meds lives with her,” and “[m]aybe that is why she keeps asking me for more pain meds.” Id. at 715–16. An October 26, 2015 drug test indicated the presence of unprescribed Dilaudid in her system, while a November 13, 2015 test returned negative for any opioids. On December 17, 2015, despite noting that Finnegan-Andrews had “r[un] out of meds less than two weeks after [he] gave her the meds,” and noting that she had received a seven-day supply of Oxycodone through a walk-in clinic, id. at 1643, Parasmo issued her two prescriptions: one for sixty-three 20-mg oxycodone pills, and another for sixty 40-mg oxycontin pills. Id. at 604–05.

• Maria Scalcione had a history of substance abuse and had been discharged by a prior doctor for drug-seeking behavior. Id. at 896. During the time she was treated by Parasmo, Scalcione consistently returned abnormal test results, which indicated the presence of numerous unprescribed substances and the absence of her prescribed medications. Between June 2014 and December 2015, Scalcione consistently tested negative for Oxycodone. In October 2015, Scalcione's sister contacted Parasmo and reported that Scalcione was not taking the oxycodone pills and was instead “selling them.” App'x at 914–18, 922–23. Notwithstanding this clear evidence, Parasmo prescribed Scalcione ninety 30-mg Oxycodone pills on December 29, 2015.

The record included additional evidence that would compel a reasonable jury to conclude that Parasmo subjectively understood that his prescription practices fell outside the usual course of professional practice. See Ruan, 597 U.S. at 454, 457, 142 S.Ct. 2370; 21 C.F.R. § 1306.04(a). Former Drug Enforcement Administration Task Force Officer Dean Steinmann testified about an interview he conducted with Parasmo on July 1, 2015. According to Steinmann, Parasmo mentioned that he had received a letter “a couple years back” from the New York State Medical Society informing him that he was in the “top 10 percent of physicians” in New York state for “[o]xycodone writing or prescriptions.” App'x at 299–300. Parasmo told Steinmann that he “felt nervous about that letter” and that “he was getting a lot, more and more pain patients,” even though he and his staff “kept trying to weed out or get rid of” them. Id. at 300. Parasmo also “informed” Steinmann that the physician with whom he shared an office “d[idn't] prescribe any pain management medication,” as was the case with a “number of other physicians” with whom Parasmo “had spoken.” Id. at 301. Parasmo told Steinmann that, in light of his “big influx” of pain patients, he decided to put up a sign in his office in March 2015 indicating “that he wasn't going to be writing any[ ]more controlled substance or opioid prescriptions.” Id. at 302. But as the record demonstrates, Parasmo continued to issue oxycodone prescriptions well after March 2015 and through at least December 2015.

3. The Dissent Misconstrues the Evidence Introduced at Trial.

Resisting this overwhelming evidence, the dissent insists that the error is not harmless because Parasmo's office had none of the hallmarks of a traditional “pill[ ]mill,” and because some of his patients reported genuine medical concerns. Dissent at –––– – ––––. But nothing in section 841 or our caselaw suggests that the government is required to prove that a doctor was operating a so-called “pill mill” – a medical practice where controlled substances (typically opioids) are prescribed at a high volume, for cash payment, and with little or no genuine patient evaluation. The statute requires only proof that the defendant “knowingly or intentionally acted in an unauthorized manner.” Ruan, 597 U.S. at 457, 142 S.Ct. 2370; see also 21 C.F.R. § 1306.04(a). And knowingly issuing prescriptions without a “legitimate medical purpose” clearly meets that standard. 21 C.F.R. § 1306.04(a).

Furthermore, while Parasmo's office was not run like a typical “pill mill,” he certainly profited financially from these prescriptions. The record reflects that patients returned to his office with striking regularity – sometimes seeing him multiple times a month. See App'x at 976–77. Parasmo himself noted that “he had a big influx” of “pain patients” and that he was “prescribing more and more pills,” id. at 302, to the point where he was in the top 10% of physicians by amount of oxycodone prescriptions issued in New York, id. at 299. And he was compensated for each visit, either directly by the patient or through insurance reimbursements, enabling him to profit from prescriptions written without legitimate medical justification.

The dissent nonetheless contends that a jury could plausibly conclude that Parasmo acted in good faith, and that he believed that his prescriptions were appropriate to manage his patients’ conditions because he ordered numerous tests for his patients, referred them to specialists to “treat their pain-inducing conditions,” “tried to wean them off the controlled substances,” and “looked for alternative treatments ․ including prescribing medications other than a controlled substance.” Dissent at ––––. But that portrayal reflects a highly selective view of the record. As explained above, Parasmo routinely ignored the results of drug screens and continued prescribing large quantities of opioids to patients even after they tested positive for other drugs or tested negative for their prescribed medications. He ignored requests from addiction specialists when they requested that he stop writing oxycodone prescriptions for their shared patients. See App'x at 871. And even after noting in patients’ files that they needed to be weaned off controlled substances, he frequently increased their prescriptions at their next appointments. See id. at 811; 1280–89.

Take, for example, Frankie Campanelli, whom the dissent offers as an example of Parasmo's good faith in prescribing opioids. Campanelli had a history of back pain and had previously had surgery for back issues. At her first visit in October 2011, Parasmo prescribed her thirty 10-mg Percocet pills. Id. at 975. After that, her prescriptions dramatically increased at each visit. On January 17, 2012, Parasmo increased her dose to ninety 15-mg Oxycodone pills. Id. By May 2012, Campanelli's drug test indicated the presence of unprescribed Alprazolam, Hydrocodone, and Oxymorphone, but not Oxycodone. Id. Nevertheless, in July 2012, Parasmo again increased her Oxycodone prescription, this time to ninety 30-mg pills. Id. By October 2012, Campanelli was up to 150 30-mg pills. And in November, Parasmo increased her Oxycodone prescription to 240 30-mg pills, notwithstanding Campanelli reporting that she had seen a different doctor who injected her with steroids. On May 17, 2013, Parasmo increased Campanelli's prescription yet again, this time to 240 30-mg Oxycodone pills and forty 15-mg Oxycodone pills. Eleven days later, Parasmo wrote another prescription for 240 more 30-mg Oxycodone pills, with no explanation noted in her chart. Id. at 976–77. On August 8, 2013, Parasmo's notes indicated that Campanelli's urine tests came back positive for heroin, that he intended to “wean the 15s,” and that she would get “no more meds” if she continued to test positive. Id at 977. Notwithstanding that notation, Parasmo increased Campanelli's prescription the following month to 240 30-mg Oxycodone pills, 240 15-mg Oxycodone pills, and 180 Xanax pills – even after she had another abnormal toxicology screen. In November 2013, Parasmo noted that Campanelli was taking sixteen tablets of 30-mg Oxycodone a day and referred her to pain management. Id. at 979. A month later, he again noted in her chart the need to “wean meds.” Id. And yet in January 2014, Parasmo wrote Campanelli a prescription for 240 30-mg Oxycodone pills, 240 15-mg Oxycodone pills, and 120 2-mg Xanax pills. Id. at 979–80. So much for “weaning” her from these powerful and addictive pain killers.

Consider also Summer Ferro, whom the dissent cites as evidence of Parasmo's good faith. When Ferro first consulted with Parasmo on January 6, 2011, she complained of “panic/anxiety attacks, asthma, and stomach problems,” and had a history of substance abuse. Id. at 693–94. Another physician who shared Parasmo's office warned him that Ferro was “abusing her pain meds.” Id. at 694. In December 2012, Parasmo communicated to Ferro that he could not treat her “gynecological, gastroenterological[,] or pain issues” because they were “not [his] area of expertise.” Id. at 699. He also recorded in her chart that she “only comes to me for pain meds” and that her problems “will not be solved with this kind of treatment.” Id. at 1216. Yet, this acknowledgement did not change his prescribing practices. Despite Ferro's repeated positive drug tests for cocaine and marijuana – and despite acknowledging at one point that he “[c]annot give controlled substances” to her – Parasmo continued to prescribe Ferro large quantities of opioids, including multiple prescriptions for 120 Percocet pills between 2013 and 2015 – sometimes on the very same day he documented these concerns. Id. at 700–04, 1222, 1224.

Parasmo's knowledge is perhaps most starkly illustrated by his treatment of Maria Scalcione. Scalcione began seeing Parasmo in 2011, complaining of “severe pain in [her] neck” and “tingling” in her “arm and hand.” Id. at 895. There was substantial evidence that she had substance-abuse issues – among other indicators, she had been discharged from a prior doctor for drug-seeking behavior. Id. at 896. In December 2011, Parasmo wrote her a prescription for 120 2-mg Xanax pills and 180 5-mg Percocet pills. See id. at 897. On February 3, 2012 and February 29, 2012, Parasmo wrote Scalcione prescriptions for Xanax, Percocet, and Hycodan cough syrup (a cough suppressant containing hydrocodone, which is an opioid). On February 29, 2012, Scalcione tested positive for marijuana and cocaine, and negative for Xanax, Percocet, or hydrocodone. See id. A drug test in March 2012 yielded the same results. Nevertheless, on May 18, 2012, Parasmo wrote Scalcione a prescription for 180 5-mg Percocet pills and 120 2-mg Xanax pills. Id. at 898. On June 29, 2012, Scalcione tested positive for marijuana, but not for the prescribed Xanax or Oxycodone, and Parasmo noted that she had not taken Percocet in ten days. Id. at 899. On the same day, he received a note from the New York State Monitoring Program stating that Scalcione had received an additional sixty Oxycodone pills from another prescriber. Id. at 898. Despite these warning signs, he refilled her prescription that day. Id. at 899.

By June 2013, Scalcione's prescription had increased to 150 30-mg Oxycodone pills. In December 2013, United Healthcare alerted Parasmo that Scalcione was “on a high dose of opioids.” Id. at 900. And yet, later that month he prescribed her an additional 150 30-mg Oxycodone pills. In June 2014, May 2015, June 2015, October 2015, November 2015, and December 2015, Scalcione consistently tested negative for Oxycodone. Id. at 900–01. In October 2015, Scalcione's sister contacted Parasmo and stated that she would “report” him if he continued writing her prescriptions, stating that Scalcione was “not taking” the oxycodone pills but was instead “selling them.” Id. at 922–23. Despite unmistakable indications that Scalcione was not taking the prescribed medication – including her testing negative for the drug over an eighteen-month period – and was instead diverting it, Parasmo nonetheless prescribed her ninety 30-mg oxycodone pills on December 29, 2015, id. at 902. This behavior – continuing to prescribe opioids despite clear evidence that the patient was not taking them – made clear that the prescriptions were not for a legitimate medical purpose and that Parasmo knew it.

It is true that many of Parasmo's patients had complicated medical and personal histories. But the complexity of their medical conditions does not erase the fact that Parasmo ignored glaring red flags and blatant warnings that his patients were diverting the pills he prescribed. On this record we cannot agree with the dissent that there was a “substantial likelihood that a properly instructed jury would have concluded that the government failed to prove that Dr. Parasmo acted with a guilty intent.” Dissent at ––––.

4. Pabisz and Tureseo Support a Finding of Harmless Error.

Parasmo contends that our holdings in United States v. Pabisz, 936 F.2d 80 (2d Cir. 1991), and United States v. Tureseo, 566 F.3d 77 (2d Cir. 2009), compel the conclusion that the instructional error here was not harmless. We are not persuaded.

In Pabisz, we reversed the defendant's conviction for willfully attempting to evade his federal income taxes because, in light of the Supreme Court's intervening decision in Cheek v. United States, 498 U.S. 192, 111 S.Ct. 604, 112 L.Ed.2d 617 (1991), “the jury was erroneously instructed that they could not credit his good faith defense unless they found his beliefs reasonable,” Pabisz, 936 F.2d at 81. But in that case, Pabisz testified at trial to the research he had conducted and actions he had taken that led him to conclude that he was not required to file individual income-tax returns – that is, evidence of his subjective beliefs. That testimony created a risk that the jury had convicted him because they found his honestly held beliefs to be objectively unreasonable. See id. at 83. And in Tureseo, we concluded that an erroneous jury instruction as to the mens rea requirement for aggravated identity theft was not harmless because at least some testimony presented at trial suggested that the defendant did not know “that he was using the identity of an actual person,” as was required for a conviction. 566 F.3d at 86.

Here, unlike the defendant in Pabisz, Parasmo did not testify at trial (as was his right) and thus did not introduce any direct evidence of his subjective beliefs. And unlike in Tureseo, where the only evidence of the defendant's subjective intent referred to by the court was thin, the proof at Parasmo's trial overwhelmingly established that Parasmo possessed the requisite mens rea for the offenses charged. As discussed above, ample evidence shows that he prescribed large quantities of opioids to patients who displayed numerous red flags, even after receiving warnings from third parties that those patients were diverting their medication, doctor-shopping, or struggling with opioid addiction. This evidence overwhelmingly established that Parasmo subjectively knew the prescriptions were issued without a legitimate medical purpose and outside the usual course of professional practice. It is therefore “clear beyond a reasonable doubt that a rational jury would have found [Parasmo] guilty absent the error.” Ng Lap Seng, 934 F.3d at 129 (internal quotation marks omitted).²

Accordingly, we conclude that the district court's erroneous jury instruction constituted harmless error.³

B. The District Court Did Not Err by Permitting Dr. Waldman to Opine on Medical Standards in New York.

Parasmo next contends that the district court erred by permitting Dr. Waldman to opine on generally accepted medical standards in New York at the time of Parasmo's offense conduct and on whether the prescriptions at issue were written in accordance with those standards. According to Parasmo, Dr. Waldman's testimony, coupled with the district court's good-faith instruction, improperly usurped the role of the jury by embracing the ultimate issue in the case – that is, whether Parasmo acted “as authorized” under an objective standard. We review a district court's decision to admit expert testimony for abuse of discretion, United States v. Romano, 794 F.3d 317, 330 (2d Cir. 2015), which we will find only where the admission of expert testimony was “manifestly erroneous,” United States v. Jones, 965 F.3d 149, 162 (2d Cir. 2020) (internal quotation marks omitted).

Federal Rule of Evidence 704(a) provides that “[a]n opinion is not objectionable just because it embraces an ultimate issue.” The Rule includes a narrow exception, prohibiting experts from testifying about whether a criminal “defendant did or did not have a mental state or condition that constitutes an element of the crime charged or of a defense.” Fed. R. Evid. 704(b). But as we have previously explained, “Rule 704(b) does not prohibit all expert testimony that gives rise to an inference concerning a defendant's mental state.” United States v. DiDomenico, 985 F.2d 1159, 1165 (2d Cir. 1993). So long as the expert does not “expressly state the inference” and leaves it, even if obvious, “for the jury to draw,” id. (internal quotation marks omitted), his testimony is permissible.

Dr. Waldman's testimony falls squarely within the confines of Rule 704. As Ruan makes clear, the scope of a practitioner's prescribing authority is defined by federal regulation with “reference to objective criteria such as ‘legitimate medical purpose’ and ‘usual course’ of ‘professional practice,’ ” and evidence of such objective criteria – when considered in conjunction with the defendant's conduct – is relevant to the jury's evaluation of the defendant's subjective intent. 597 U.S. at 467, 142 S.Ct. 2370 (quoting 21 C.F.R. § 1306.04(a)). Accordingly, expert testimony is appropriate to explain the usual course of professional practice and whether certain practices fall within those parameters. It bears noting that Dr. Waldman offered no opinion about Parasmo's subjective mental state when he wrote the prescriptions at issue, or whether Parasmo had the intent required by the statute. See DiDomenico, 985 F.2d at 1165; see also Diaz v. United States, 602 U.S. 526, 528, 144 S.Ct. 1727, 219 L.Ed.2d 240 (2024) (“Because the expert witness did not state an opinion about whether [the defendant himself] had a particular mental state, we conclude that the testimony did not violate Rule 704(b).”). Based on this record, we see no abuse of discretion in the district court's admission of Dr. Waldman's testimony.

C. The District Court Properly Admitted Evidence of New York State's Medical Standards.

For similar reasons, the district court did not err in permitting the government to introduce evidence of New York regulations concerning the legitimate medical purposes for which a doctor may write prescriptions for opioids. Parasmo contends vaguely that “defining the ‘usual course of professional practice’ based upon state regulations invites disparate interpretations of [section 841(a)(1)].” Parasmo Br. at 61. Because Parasmo did not make this objection below, we review this challenge for plain error. See United States v. Al Kassar, 660 F.3d 108, 126 (2d Cir. 2011).

As noted above, federal regulations provide that an “authorized” prescription is one “issued for a legitimate medical purpose by an individual practitioner acting in the usual course of his professional practice.” 21 C.F.R. § 1306.04(a). The Controlled Substances Act defines “practitioner,” in turn, by reference to “the jurisdiction in which he practices.” 21 U.S.C. § 802(21) (defining “practitioner” as a physician who is “licensed, registered, or otherwise permitted, by the ․ jurisdiction in which he practices ․ to distribute [or] dispense ․ a controlled substance in the course of professional practice”). Once again, Ruan explicitly contemplates the use of “objective criteria such as ‘legitimate medical purpose’ and ‘usual course’ of ‘professional practice’ ” as probative of the doctor's knowledge and subjective intent concerning the legality of his prescribing practices. 597 U.S. at 467, 142 S.Ct. 2370 (quoting 21 C.F.R. § 1306.04(a)). It thus follows that, in this case, the medical standards relevant to defining a “legitimate medical purpose” and “usual course of professional practice” are those of New York – the state in which Parasmo was, at all relevant times, licensed to practice medicine. Id. (internal quotation marks omitted). We therefore see no error – let alone plain error – in the district court's admission of evidence regarding the New York regulations.

D. Parasmo's Ineffective-Assistance-of-Counsel Claim is Better Suited for Collateral Review.

Finally, Parasmo contends that his trial counsel provided ineffective assistance by failing to introduce evidence of New York regulations and laws that clarified when a doctor may legally prescribe opioids to persons suffering from addiction. When an ineffective-assistance claim is raised on direct appeal, we may either (1) decline to hear the claim so that it may be raised in a habeas petition brought pursuant to 28 U.S.C. § 2255, (2) remand to the district court for further factfinding, or (3) decide the claim on the record before us. See United States v. Adams, 768 F.3d 219, 226 (2d Cir. 2014). Nevertheless, we are “generally disinclined to resolve ineffective assistance claims on direct review ․ because the district court is ‘best suited to developing the facts necessary to determining the adequacy of representation.’ ” United States v. Gaskin, 364 F.3d 438, 467–68 (2d Cir. 2004) (quoting Massaro v. United States, 538 U.S. 500, 505, 123 S.Ct. 1690, 155 L.Ed.2d 714 (2003)). We see no reason to deviate from that practice here, especially since Parasmo “did not raise these contentions in the district court, [so] there is no record that would permit them to be assessed on this appeal.” United States v. Laurent, 33 F.4th 63, 97 (2d Cir. 2022). We therefore decline to address the merits of this claim, which Parasmo may raise in a motion to vacate his conviction or set aside his sentence pursuant to section 2255. See United States v. Morris, 350 F.3d 32, 39 (2d Cir. 2003).

III. CONCLUSION

For the foregoing reasons, we AFFIRM the judgment of the district court.

As the “heart and lungs” of liberty, juries play a crucial role in our criminal justice system. C. Bradley Thompson, ed., The Revolutionary Writings of John Adams 55 (2000). In criminal cases, the jury is the “oracle of the citizenry in weighing the culpability of the accused.” United States v. Gilliam, 994 F.2d 97, 101 (2d Cir. 1993).¹ The jury's constitutional responsibilities include determining the facts, applying the law to those facts, and drawing the ultimate conclusion of guilt or innocence. To aid the jury in fulfilling its duty, trial courts instruct jurors on the relevant principles of law necessary to reach a conclusion as to each element of the charged offense.

Here, the court's instruction misinformed the jury as to the only element of the charged offense that Dr. Parasmo contested—his subjective intent. See Ruan v. United States, 597 U.S. 450, 467, 142 S.Ct. 2370, 213 L.Ed.2d 706 (2022) (to prove the elements necessary to support a conviction under § 841, the government must prove that a defendant “knew or intended that his or her conduct was unauthorized”). The government's theory of the case was that Dr. Parasmo didn't write the victims’ prescriptions for a legitimate medical purpose, and he knew that. Dr. Parasmo's theory of the case was that, though he was aware of and concerned about the signs that his patients were abusing or misdirecting their medications, he subjectively believed, in good faith, that his prescriptions were appropriate to manage his patients’ pain arising from genuinely serious medical conditions, and that he was taking reasonable steps to address the possibility that patients were abusing the controlled substances. The court's instruction invited the jury to convict even if it accepted Dr. Parasmo's view of the case. Like the majority, I conclude the court's instruction was error. Majority Op. at ––––.

Unlike the majority, I conclude the error was not harmless. In conducting a harmless error review, I'm especially mindful of the perils of our weighing the evidence regarding Dr. Parasmo's subjective intent—the only real issue in this case, and one the jury itself never got to consider. After all, ordinarily, “[t]he question of whether criminal intent is inferable from the facts proved is a question for the jury.” United States v. Speare, 297 F.2d 408, 410 (2d Cir. 1962).

True, an instructional error, including the omission or misdescription of a critical element of the charge, is subject to harmless-error analysis. Neder v. United States, 527 U.S. 1, 15, 119 S.Ct. 1827, 144 L.Ed.2d 35 (1999). But where an element of an offense is contested, we should be extremely wary of treating the failure to require the jury to make a finding as to that element as a harmless error, lest we usurp the jury's role. Cf. id. at 17, 119 S.Ct. 1827 (concluding that the failure to instruct on an element of the charged crime was harmless where the “element was uncontested and supported by overwhelming evidence” (emphasis added)); see also United States v. Kahn, 58 F.4th 1308, 1319 (10th Cir. 2023) (“Where an element of an offense is contested at trial, as it was here, the Constitution requires that the issue be put before a jury—not an appellate court.”). To “safeguard[ ] the jury guarantee” in a case like this, a court sitting in review must “conduct a thorough examination of the record.” Neder, 527 U.S. at 19, 119 S.Ct. 1827. “If, at the end of that examination, the court cannot conclude beyond a reasonable doubt that the jury verdict would have been the same absent the error ․ it should not find the error harmless.” Id. (emphasis added). That's a tall order.

And the government, not Dr. Parasmo, carries the burden of demonstrating that the jury instruction error was harmless beyond a reasonable doubt. Gutierrez v. McGinnis, 389 F.3d 300, 303 (2d Cir. 2004) (“The burden of proving the error's harmlessness falls to someone other than the person prejudiced by it.”).

With these principles in mind, I cannot conclude beyond a reasonable doubt that, if properly instructed, the jury would have reached the same conclusion. There is ample evidence from which the jury could have concluded that the government failed to prove beyond a reasonable doubt that Dr. Parasmo subjectively believed he was acting as a dealer not a doctor, and the failure to require a jury to make a finding as to Parasmo's subjective good faith undermines the fairness of this trial. My conclusion is supported by the record and caselaw in analogous cases from this Circuit and beyond.

1. The Record

As emphasized above, the question before us is not whether there was sufficient evidence from which a jury could conclude that Dr. Parasmo had a guilty state of mind. There was. The question here is whether we can conclude beyond a reasonable doubt that the jury would have reached that conclusion if properly instructed. Neder, 527 U.S. at 19, 119 S.Ct. 1827. That is, whether the evidence “all flow[s] in one direction.” United States v. Tureseo, 566 F.3d 77, 86 (2d Cir. 2009).

It does not. For one thing, Dr. Parasmo took copious and detailed notes of his patients’ symptoms and the treatments he prescribed. He generally conducted physical examinations before prescribing the controlled substances. His office also ordered drug screens for patients and meticulously recorded the results of each. This is not the conduct one would expect from a “pill mill” doctor.

Plus, the jury could no doubt consider the absence of evidence to support the theory that Dr. Parasmo wrote the prescriptions for purposes other than medical treatment. As Dr. Parasmo accurately argues, “[t]here were no allegations of cash payments to Dr. Parasmo, no ‘runners,’ fake patients, or systematic diversion.” Appellant's Br. at 6.

The majority emphasizes that the government was not required to prove that Parasmo operated a “pill mill.” Majority Op. at ––––. That's true. But in assessing whether the jury necessarily would have reached the same verdict if properly instructed as to Parasmo's subjective state of mind, it's significant that Parasmo's practice did not bear the typical hallmarks of a pill mill as described by the government's own investigating detective—including prescription without examination, fake patients, cash payments, lines of people waiting to get into the mill, and cash kickbacks from a pharmacy. In the face of such evidence, a harmless error analysis might be more tenable.

And finally, in treating individual patients, Dr. Parasmo ordered numerous tests, referred his patients to a wide variety of specialists to treat their pain-inducing conditions, tried to wean them off the controlled substances, and looked for alternative treatments to control his patients’ pain, including prescribing medications other than a controlled substance—all behaviors suggestive of a good faith effort to help his patients.

Take, for example, patient Frankie Campanelli, who suffered from bilateral carpal tunnel syndrome, scoliosis, and a prior spinal surgery that left her with two metal rods in her back. In August 2013, Dr. Parasmo recorded that he significantly reduced her dosage and noted that “she was supposed to wean” the medication and if the urine screen tested positive for heroin “no more meds from me.” App'x 977. Dr. Parasmo's notes from October 2013 indicate that she was reporting severe back pain and that he was searching for a viable alternative to pain medication: “To do decompression for two weeks. If it does not work detox.” App'x 1160. At her next appointment, Dr. Parasmo's notes reflect that F.C. “[had] been crying for days ever since she went through detox” and that her “[p]ain is incredible.” App'x 978. Dr. Parasmo continued to prescribe her oxycodone, but he also referred F.C. to pain management in November 2013. The next month, he wrote a note saying, “wean meds.” App'x 979. He continued to prescribe her pain medication; however, he noted that F.C.’s “[m]other goes to pharmacy to get meds and controls them.” App'x 979. Six months after he wrote his final oxycodone prescription for F.C., he again referred her to pain management.

Or consider Summer Ferro, who suffered from severe stomach pain and anxiety disorder. Her anxiety was so severe that she had “palpitations and air hunger” and “turn[ed] blue.” App'x 696. He referred her to a psychiatrist, gynecologist, gastroenterologist, and pain management doctor. He cut her prescription in half. And when she tested positive for cocaine, Dr. Parasmo told her that he couldn't prescribe her controlled substances and gave her a list of pain management specialists.

Evidence regarding other patients, including those identified by the majority, likewise paints a more complex picture than the majority suggests. In discussing Dr. Parasmo's prescriptions for other patients, the majority doesn't acknowledge the broader contexts—including evidence of the patients’ medical histories of chronic pain. John Lettenberger suffered from traumatic brain injuries and complex fractures after a significant car accident that left him hospitalized for four months and in a coma for three weeks. Maria Scalcione suffered from a significant degenerative disk disease that was so severe a neurosurgeon concluded she needed lumbar surgery. Because of her severe chronic pain she couldn't drive, had trouble sitting or bending, and had to quit her job. Leslie Finnegan-Andrews suffered from severe migraines and pain in her right hip. And Maurice Milano, who worked in construction, suffered from chronic pain because of damage to his spine and bullet fragments in his neck due to prior motor vehicle accidents and a gunshot wound. In each case, Dr. Parasmo was confronted with a real patient suffering from severe chronic pain alongside red flags suggesting the possibility, or even likelihood, that they were misusing their medications.

And in each instance, Dr. Parasmo tried to treat the patient's underlying condition and to find alternative ways to control their pain. He encouraged Milano to move to a warmer climate to help alleviate his constant pain and referred him to pain management multiple times.

Dr. Parasmo refused to prescribe medication to Lettenberger, despite his reports of severe pain, after he tested positive for cocaine and the results of the drug test showed that he wasn't taking his prescribed medication. Instead, he prescribed an anti-inflammatory, a mood stabilizer, and an antidepressant.

Rocco Oliveri went to pain management, a neurologist, and a psychiatrist. Dr. Parasmo attempted to wean Oliveri off oxycodone, referred him to PT, and worked closely with the psychiatrist Oliveri was seeing.

And Dr. Parasmo referred Scalcione to a neurosurgeon who concluded that she needed surgery. But the surgeon didn't take her insurance. Dr. Parasmo then referred her to pain management, and after Scalcione reported that it wasn't working, he prescribed her pain medication and tried to find her a different neurosurgeon. One of the several neurosurgeons Dr. Parasmo consulted with told him that Scalcione would require three surgeries to correct her back problem, but that she should wait until the pain became unbearable.

From this evidence, a reasonable jury could readily infer that Dr. Parasmo did not intend to illicitly prescribe controlled substances, and that he subjectively believed he was prescribing controlled substances for the legitimate medical purpose of treating his patients’ pain.

The majority asserts that I present a “highly selective” view of the record. Majority Op. at ––––. But we should focus on the evidence that supports Parasmo's defense. The central question in our harmless error analysis is whether the evidence presented to the jury “all flow[s] in one direction.” Tureseo, 566 F.3d at 86. The above evidence highlights that it does not.

2. The Caselaw

My view is consistent with this Court's conclusions in analogous cases, as well as that of other circuits.

Most squarely on point is United States v. Pabisz, 936 F.2d 80 (2d Cir. 1991). In Pabisz, we considered a defendant's conviction for willfully attempting to evade federal income taxes, in violation of 26 U.S.C. § 7201. Pabisz contended that he did not willfully attempt to evade his taxes because he believed, in good faith, that he was not required to pay them. As here, the district court's instruction regarding the defendant's good faith defense misled the jury “into believing that [the defendant's] good faith beliefs could negate the element of willfulness only if those beliefs were objectively reasonable.” Id. at 83. We concluded that the instruction undermined “the fundamental fairness of the trial” and amounted to plain error. Id. Significantly, in reaching this conclusion, we emphasized that the potential for misleading the jury was heightened by the government's summation in which the prosecutor repeatedly urged the jury to consider the reasonableness of the defendant's beliefs. Id.

So it was in this case. In summation, the government repeatedly (and wrongly) emphasized “that this is an objective standard,” App'x 2065, and asserted that Dr. Parasmo's belief that he was prescribing controlled substances for a legitimate medical purpose “has to be reasonable, and is not subjective,” App'x 2372 (emphasis added). The government argued that it didn't matter whether Dr. Parasmo “thought the prescription was okay.” Id.

Pabisz is on all fours with this case. The majority's distinction—that Dr. Parasmo didn't testify in his own defense or introduce direct evidence of his subjective beliefs—is unpersuasive. Majority Op. at ––––. As noted above, there is plenty of circumstantial evidence to undermine the government's suggestion that Dr. Parasmo had the necessary guilty intent, and the good faith defense was the centerpiece of Dr. Parasmo's defense. Suggesting that the improper instruction did not undermine the fairness of Parasmo's trial because Parasmo himself did not take the stand (1) fails to recognize that it is the government's burden to prove Parasmo had a guilty state of mind, not Parasmo's burden to prove his good faith; and (2) burdens Parasmo's Fifth Amendment right not to testify.²

Likewise, in Tureseo, an aggravated identity theft case under 18 U.S.C. § 1028A, we concluded that the district court erred when it failed to instruct the jury that in order to convict it had to find that Tureseo knew that the identifying documents at issue belonged to an actual person. 566 F.3d at 85–86. And we concluded the error was not harmless. Id. at 86. Although Tureseo's use of another's birth certificate constituted substantial evidence that he knew the document belonged to an actual person, we concluded that “the evidence does not all flow in one direction.” Id. We posited, based on the identity theft victim's testimony that he had never met Tureseo, that a jury could conclude that the defendant did not know of the victim's actual existence. Id. If anything, the inference that Dr. Parasmo acted in good faith here flows even more naturally from the record evidence than the inference we relied on in Tureseo in concluding the instructional error was not harmless.

And this case is readily distinguishable from one in which we have concluded that a Ruan error was harmless. The government in United States v. Belfiore had far more evidence of the defendant's guilty intent than the mere fact that the defendant continued to prescribe controlled substances to patients despite evidence of their illicit drug use and failed tests. No. 22-20, 2024 WL 2075128, at *2 (2d Cir. May 9, 2024) (summary order). The evidence in Belfiore included a recording of the defendant warning an undercover officer that he should not share narcotics with others because someone may turn out to be an undercover officer. Id. It included a different recording of the defendant telling an undercover officer that the defendant “was set up for disaster” because the undercover officer's drug tests were negative for oxycodone and there were no pharmacy records of the officer having the prescriptions filled. Id. It included evidence that the defendant falsified medical records and urged a social worker to submit a false affidavit in a medical malpractice case stating that the defendant had arranged an emergency appointment on behalf of a patient who died from an overdose of a drug he had prescribed. Id. And there was evidence that, when the undercover officer told the defendant that he was sharing the prescribed oxycodone with his girlfriend, the defendant warned that ongoing sharing would “screw up” the pill count. Id.

Here, there was no evidence of Dr. Parasmo instructing patients as to how to avoid detection, nor was there evidence of him falsifying any records. In fact, Dr. Parasmo kept meticulous records—including recording each of his patients’ failed drug tests. And, as noted above, there was considerable evidence that Dr. Parasmo prescribed alternative non-narcotic therapies, made referrals, and tried to wean patients off the controlled substances.

The majority's position here is also squarely at odds with considered decisions of other circuits in similar cases. I find support for my view in an Eleventh Circuit decision holding that the omission of a subjective intent instruction in a similar case affected a defendant's substantial rights. See United States v. Duldulao, 87 F.4th 1239 (11th Cir. 2023). In Duldulao the Court concluded that the instructional error was not harmless because “the jury could have rested its convictions solely on an impermissible theory of liability: that [the defendant]’s actions did not comply with objective professional norms of medicine.” Id. at 1259. The Court reached this conclusion even though in Duldulao, in contrast to here, there was evidence that clinic staff falsified medical records, patients were “shooting up” in the parking lot, the clinic barely had any medical equipment or supplies, the doctors spent very little time with patients, the defendant doctors’ boss made it clear that patients expected to receive controlled substances during their visits, and untrained front desk staff wrote prescriptions for controlled substances for the doctor to sign after each patient's brief visit. Id. at 1247–49. The Court emphasized two factors that are also present here: at trial, the government stressed that the defendant's actions deviated from objective professional norms of medicine, and the jury issued a split verdict. Id. at 1259.

The Tenth Circuit has likewise concluded that instructions allowing for a conviction under § 841 without a jury determining whether the defendant knowingly or intentionally acted without authorization wasn't harmless. Kahn, 58 F.4th at 1317. Central to the Court's reasoning was that the defendant's “intent was in dispute throughout his trial and was the centerpiece of his defense.” Id. at 1319. The Court did not wade through the evidence to determine whether it might nevertheless support a conviction. Recognizing that the instructional error regarding the defendant's intent “went directly to the heart of the trial,” the Court concluded: “The jury did not make the required mens rea finding, and ‘to hypothesize a guilty verdict that was never in fact rendered—no matter how inescapable the findings to support that verdict might be—would violate the jury-trial guarantee.’ ” Id. at 1320 (citing Sullivan v. Louisiana, 508 U.S. 275, 279, 113 S.Ct. 2078, 124 L.Ed.2d 182 (1993)).³

As in the above cases, Dr. Parasmo's intent was fiercely contested and was his sole defense. Cf. Neder, 527 U.S. at 17, 119 S.Ct. 1827 (concluding that the failure to instruct on an element of the charged crime was harmless where the “element was uncontested and supported by overwhelming evidence” (emphasis added)). The government relied heavily upon the error in its closing. See Pabisz, 936 F.2d at 83. And the evidence doesn't “all flow in one direction,” Tureseo, 566 F.3d at 86, such that we can conclude nothing in the record would lead a rational juror to acquit, Neder, 527 U.S. at 19, 119 S.Ct. 1827.

* * *

This was not a “slam dunk” case. In fact, even with the erroneous instruction, the jury acquitted Dr. Parasmo on several counts. Based on the evidence at trial, there is a substantial likelihood that a properly instructed jury would have concluded that the government failed to prove that Dr. Parasmo acted with a guilty intent. I certainly cannot conclude otherwise beyond a reasonable doubt. And the erroneous jury instruction that took the central contested factual issue in this case away from the jury undermined the fundamental fairness of his trial. For these reasons, I respectfully dissent.

FOOTNOTES

1.   Parasmo had instead requested an alternative good-faith instruction, which provided in relevant part:The government must prove beyond a reasonable doubt that Dr. Parasmo did not act in good faith. A medical professional's good faith is relevant to your determination of whether Dr. Parasmo knowingly and intentionally acted outside the bounds of usual professional practice and without a legitimate medical purpose. A doctor distributes a drug in [g]ood faith when he believes he is medically treating a patient for a legitimate medical purpose and in the usual course of medical practice. Good faith in this context means good intentions and the honest exercise of professional judgment as to a patient's needs.App'x at 112–13.

2.   The dissent also points to two out-of-circuit opinions, United States v. Duldulao, 87 F.4th 1239 (11th Cir. 2023), and United States v. Kahn, 58 F.4th 1308 (10th Cir. 2023), in an attempt to demonstrate that other courts have vacated convictions in “similar cases.” Dissent at –––– – ––––. But these cases are also readily distinguishable. In Duldulao, the Eleventh Circuit held that a district court's omission of the subjective intent instruction was plain error because “the jury could have rested its convictions solely” on the “impermissible” objective theory of liability. Dissent at –––– (quoting Duldulao, 87 F.4th at 1259). Yet there, as the court noted, the government had relied mainly on circumstantial, “general” evidence of unlawful activity, which made the risk of an erroneous conviction particularly high. Duldulao, 87 F.4th at 1261. In Parasmo's trial, by contrast, the government relied on evidence of the defendant's specific interactions with patients, conversations with doctors, and medical records detailing “discrete prescriptions” – precisely the kind of evidence the Eleventh Circuit indicated would have changed its assessment in Duldulao. Id. Similarly in Kahn, the Tenth Circuit's vacatur was premised on the fact that this was “not a case in which the element of the crime that was impacted by the invalid jury instruction was ․ supported by overwhelming evidence.” 58 F.4th at 1319 (internal quotation marks omitted). The clear implication is that in a case like Parasmo's – where the evidence truly is “overwhelming” – the court would agree with our conclusion regarding the harmlessness of the instruction. See id.

3.   For the same reasons discussed above, we also reject Parasmo's request that we dismiss the indictment based on the legal insufficiency of the evidence.

1.   In quotations from caselaw and the parties’ briefing, this dissent omits all internal quotation marks, footnotes, and citations, and accepts all alterations, unless otherwise noted.

2.   United States v. Heaton, 59 F.4th 1226 (11th Cir. 2023) doesn't help the majority. It's true that the defendant in that case didn't testify in his own defense, id. at 1238, and the Eleventh Circuit found the evidence of the defendant's subjective intent “overwhelming,” id. at 1242. But in that case the defendant prescribed medication without documenting the patient's response, taking a medical history, or conducting a full physical examination; had sexual relationships with multiple patients; and misled the relevant medical board. Id. at 1232, 1242–44. The evidence here is far more equivocal.

3.   The majority attempts to distinguish Kahn and Duldulao on the basis that the government's evidence in those cases was weaker. Majority Op. at –––– n.2. But that's wrong. In contrast to this case, Duldulao involved a classic pill mill. 87 F.4th at 1247. There was no evidence that the doctors pursued the kinds of alternate treatments Parasmo recommended for some patients here, nor that they endeavored to wean patients off the drugs. In addition to evidence that the defendant doctors prescribed controlled substances notwithstanding red flags, the government established the following: the practice operated on cash or credit only, “liberally dispensed controlled substances,” had little to no medical equipment or supplies, and falsified drug test records; untrained administrative staff “wrote prescriptions for controlled substances for the doctor to sign after each patient's brief visit”; people were “nodding out” in the waiting room and “shooting up” in the parking lot, where patients left behind baggies, blunt wrappers, and syringes; and one defendant admitted to his girlfriend that he worked at a “pain mill.” Id. at 1247–48. Nevertheless, on plain error review, the Eleventh Circuit vacated the convictions, rejecting the argument that the defendants’ guilt was so clear that they were not prejudiced. Id. at 1261.The court in Kahn relied very little on the evidence presented to the jury; its focus was on the fact that, as here, the instructional error “went directly to the heart of the trial: Dr. Kahn's intent.” 58 F.4th at 1320. It concluded that the instructional error was not harmless because “the element of the crime that was impacted by the invalid jury instruction”—the defendant's intent—was not “uncontested and supported by overwhelming evidence”—a touchstone repeated at least five times in the Tenth Circuit's harmless error analysis. Id. at 1318, 1319, 1320 (emphasis added). The majority's abridged version of this formulation garbles the Tenth Circuit's reasoning by omitting the court's reference to the fact that the defendant's intent was not “uncontested.”

Richard J. Sullivan, Circuit Judge:

Judge Robinson dissents in a separate opinion.