Argued October 17, 2019; decided November 26, 2019

People v Li, 155 AD3d 571, affirmed.

Here, the Appellate Division rejected defendant's challenges to his conviction of two counts of manslaughter in the second degree. We affirm the Appellate Division order, insofar as appealed from.

Defendant, then a physician board-certified in anesthesiology and pain management, was accused of running a "pill mill" {**34 NY3d at 361}at his Queens pain management clinic. During a lengthy jury trial, the People presented evidence that defendant prescribed medically unnecessary high doses of opioids, alprazolam, and other controlled substances as a first resort. Defendant generally did not verify the source of the pain complained of by the patient for which the patient sought the controlled substances, order diagnostic tests for objective confirmation of the existence of the pain, or consider other pain management treatment options. He conducted little to no physical examination. Defendant often prescribed heavy doses of whatever medication his patients requested to alleviate their complaints of pain. He required payment in cash and charged extra for, among other things, higher doses of opioids. Several of defendant's [*2]former patients testified at trial that they were opioid addicts. They testified that they used the drugs defendant prescribed them to get high, rather than for legitimate pain management. Indeed, defendant was advised by other medical practitioners and patients' family members that several of his patients were addicted to opioids and at risk of dying from opioid abuse.

Two of defendant's patients, Joseph Haeg and Nicholas Rappold, died of overdoses caused by a combination of oxycodone and alprazolam on December 29, 2009 and September 14, 2010, respectively, shortly after filling prescriptions for such drugs issued by defendant. Pills from those prescriptions were found in their possession when their bodies were discovered. Defendant was charged with two counts of manslaughter in the second degree (see Penal Law § 125.15 [1]) for the deaths of Haeg and Rappold, along with multiple other crimes related to Haeg, Rappold, and a number of other patients.

Defendant was ultimately convicted of two counts of manslaughter in the second degree, three counts of reckless endangerment in the first degree, three counts of reckless endangerment in the second degree, 170 counts of criminal sale of a prescription, one count of scheme to defraud in the first degree, two counts of grand larceny in the third degree, nine counts of falsifying business records in the first degree, and eight counts of offering a false instrument for filing in the first degree. The Appellate Division unanimously affirmed (155 AD3d 571 [1st Dept 2017]), and a Judge of this Court granted defendant leave to appeal (31 NY3d 1119 [2018]).

On this appeal, defendant challenges only his conviction of two counts of manslaughter in the second degree. He raises{**34 NY3d at 362} two contentions. First, defendant argues that, as a matter of law, he cannot be convicted of any homicide offense for providing controlled substances that result in an overdose death. Second, defendant asserts that his conviction on the manslaughter counts is not supported by legally sufficient evidence.

Defendant is incorrect that, as a matter of law, his conduct may not be prosecuted as a homicide offense. He relies heavily on People v Pinckney (38 AD2d 217 [2d Dept 1972]), where the Appellate Division upheld the dismissal of counts of an indictment charging manslaughter in the second degree and criminally negligent homicide after the defendant sold heroin to the victim, who later died after injecting it (see id. at 218). The Appellate Division reasoned that the legislature had already provided penalties in the Penal Law for the sale of dangerous drugs but had not amended the homicide provisions of the Penal Law "to include homicide by the selling of dangerous drugs" (id. at 220-221).

This Court affirmed the Appellate Division order in Pinckney without opinion (32 NY2d 749 [1973]). The precedential value of such a ruling is minimal. An affirmance without opinion constitutes approval of only the result reached and "does not imply approval of everything contained in the opinion of the court below" (People ex rel. Palmer v Travis, 223 NY 150, 156 [1918]; see also Matter of Clark, 275 NY 1, 4 [1937]; Rogers v Decker, 131 NY 490, 493 [1892]). We disagree with our dissenting colleague that our affirmance in Pinckney, which involved an indictment alleging a one-time sale of heroin and the instruments for injecting it, forecloses the prosecution of defendant for a homicide offense under the very different factual circumstances presented here (see dissenting op at —).

Subsequent decisions from this Court refute defendant's assertion that a person who provides dangerous drugs that result in death can never, under any circumstances, be prosecuted for homicide (see People v Galle, 77 NY2d 953, 955-956 [1991]; People v Cruciani, 36 NY2d 304, 305-306 [1975]). Although in those cases, the defendants injected the victims with drugs, we did not state that this was a necessary element, as a matter of law, for homicide charges to be sustained. Rather, the defendants' injection of the drugs in those cases was one piece of evidence that supported the homicide charges and that distinguished those cases from Pinckney (see Cruciani, 36 NY2d at 305-306).

{**34 NY3d at 363}Insofar as the Appellate Division reasoned in Pinckney that the defendant could not be charged with a homicide offense because the legislature had criminalized the sale of illegal drugs but had not amended Penal Law article 125 to include a specific reference to death caused by the sale of drugs (see Pinckney, 38 AD2d at 220-221), that rationale was flawed. "As a general rule, a statutory prohibition against a particular type of conduct will not be [*3]deemed to constitute the exclusive vehicle for prosecuting that conduct unless the Legislature clearly intended such a result" (People v Duffy, 79 NY2d 611, 614 [1992]).

There is no basis to conclude that the legislature intended to exclude from the ambit of the homicide statutes the prosecution of a defendant who, with the requisite mens rea, engages in conduct through the sale or provision of dangerous drugs that directly causes the death of a person. The fact that the legislature has separately criminalized the illegal sale of controlled substances does not require a different conclusion (see id. at 614-615).^[FN*] We agree with the Appellate Division that "all that was needed for the manslaughter charge to be sustained was for the People to satisfy its elements" (155 AD3d at 574).

We further conclude that defendant's conviction of two counts of second-degree manslaughter is supported by legally sufficient evidence. "A verdict is legally sufficient when, viewing the facts in a light most favorable to the People, 'there is a valid line of reasoning and permissible inferences from which a rational jury could have found the elements of the crime proved beyond a reasonable doubt' " (People v Danielson, 9 NY3d 342, 349 [2007], quoting People v Acosta, 80 NY2d 665, 672 [1993]). "A sufficiency inquiry requires a court to marshal competent facts most favorable to the People and determine whether, as a matter of law, a jury could logically conclude that the People sustained its burden of proof" (Danielson, 9 NY3d at 349). "This deferential standard is employed because the courts' role on legal sufficiency review is simply to determine whether enough evidence has been presented so that the resulting verdict was lawful" (Acosta, 80 NY2d at 672).{**34 NY3d at 364}

Importantly, "[i]n determining the legal sufficiency of the evidence for a criminal conviction we indulge all reasonable inferences in the People's favor, mindful that a 'jury faced with conflicting evidence may accept some and reject other items of evidence' " (People v Carrel, 99 NY2d 546, 547 [2002], quoting People v Ford, 66 NY2d 428, 437 [1985]). It is the "province of the jury" to assess witness credibility (People v Calabria, 3 NY3d 80, 82 [2004]), and we therefore assume on a legal sufficiency review that the jury credited the People's witnesses (see People v Gordon, 23 NY3d 643, 649 [2014]).

To convict defendant of second-degree manslaughter, the People were required to prove beyond a reasonable doubt that defendant "recklessly cause[d] the death" of Haeg and Rappold (Penal Law § 125.15 [1]).

"A person acts recklessly with respect to a result or to a circumstance described by a statute defining an offense when [that person] is aware of and consciously disregards a substantial and unjustifiable risk that such result will occur or that such circumstance exists. The risk must be of such nature and degree that disregard thereof constitutes a gross deviation from the standard of conduct that a reasonable person would observe in the situation" (Penal Law § 15.05 [3]).

A conviction for reckless manslaughter "require[s] that there be a 'substantial and unjustifiable risk' that death . . . will occur; that the defendant engage in some blameworthy conduct contributing to that risk; and that the defendant's conduct amount to a 'gross deviation' from how a reasonable person would act" (People v Asaro, 21 NY3d 677, 684 [2013], quoting Penal Law § 15.05 [3]). The conduct must be "the kind of seriously blameworthy carelessness whose seriousness would be apparent to anyone who shares the community's general sense of right and wrong" (Asaro, 21 NY3d at 685 [internal quotation marks and citation omitted]).

The People's considerable evidence with respect to defendant's prescribing practices on a broader scale, which included testimony from several patients and their family members, was relevant as context to assess his mens [*4]rea on the manslaughter charges with respect to Haeg and Rappold. The People presented evidence that patients were not required to{**34 NY3d at 365} make appointments at defendant's clinic and were required to pay in cash. The clinic was open only on weekends, yet defendant wrote over 21,000 prescriptions for controlled substances between 2008 and 2011, most of those for a substance containing oxycodone or alprazolam (Xanax). From January 2008 to January 2011, defendant increased his prescribing of controlled substances by 683%. On a single day in January 2011, for example, defendant saw 94 patients at his clinic.

Defendant generally charged $100 per office visit, and he increased the charge to $150 if a patient returned early for more pills, had a friend or family member pick up the prescription, was obtaining prescriptions from other doctors, or wanted a higher daily dose of opioids. Patients testified that he generally prescribed whatever they requested, so long as they had the requisite cash payment. Physical examinations were either cursory or nonexistent. Defendant usually did not order diagnostic tests, and if he did, he did not change his prescribing practices if a patient failed to comply with tests that he did order.

The People's expert, Dr. Gharibo, testified that defendant made no attempt to consider non-opioid pain management treatment for his patients, despite his training in other pain management options and the addictive nature of the drugs he was prescribing. Defendant disregarded warning signs that his patients were abusing their medication and were addicted to opioids, such as early visits, obtaining prescriptions from other doctors, deterioration in physical appearance, and, in some cases, direct warnings from family members and hospitals that defendant's patients had overdosed. Defendant did not change his prescribing practices until law enforcement began investigating him in 2011. Defendant altered medical records in response to an investigation request from the New York State Health Department's Office of Professional Medical Conduct.

According to the People's witnesses, there was no basis for defendant's prescription of Xanax to a number of patients, including Haeg and Rappold, since that drug had no legitimate pain-relieving function. The People's witnesses also testified about the "synergistic respiratory depression" effect of opioids and Xanax when taken together. Those witnesses explained that oxycodone has a respiratory depression effect that, if taken in large enough doses, will cause a patient to stop breathing entirely. Xanax, also a depressant, exacerbated that effect when taken together with oxycodone, such that smaller doses of oxycodone{**34 NY3d at 366} could cause respiratory failure. Dr. Gharibo testified that Xanax is known to be highly addictive in combination with opioids, that addicts sometimes requested the medications together in order to enhance their narcotic highs, and that the combination of opioids and Xanax greatly increased the risk of a fatal overdose. Defendant was highly trained in pain management and the interaction and potential addictiveness of these drugs. Yet he frequently prescribed oxycodone and Xanax together without documentation as to why that combination of prescriptions was medically necessary for a particular patient. According to Dr. Gharibo, defendant's prescriptions of high doses of opioids and Xanax were not attempts to treat legitimate pain based on reasoned medical judgment but rather were designed to create and feed a cycle of craving and addiction. With respect to one of defendant's prescriptions written for a different patient, Dr. Gharibo testified that it was "an overdose waiting to happen."

Unlike the evidence with respect to some of defendant's surviving patients, the People did not present evidence that defendant was directly informed that the deceased patients, Haeg and Rappold, were addicts or had previously overdosed on medications he prescribed. Nevertheless, viewing the evidence in the light most favorable to the People, as we must, and giving the People the benefit of all reasonable inferences, we conclude that a rational jury could have found that defendant was aware of and consciously disregarded a substantial and unjustifiable risk that his prescription practices would result in the deaths of Haeg and Rappold.

Dr. Gharibo testified that defendant prescribed opioids to Haeg on the first visit. There was no diagnostic workup, no attempt to determine whether non-opioid treatments could be effective, and no verification of the information Haeg gave him about receiving high doses of opioids from other doctors. Although Haeg gave defendant an MRI from 2005 showing a central L 5-1 herniation, Dr. Gharibo testified that this was a "general finding" that was "not necessarily diagnostic." Defendant ordered another MRI for Haeg but did not change his prescribing practices after Haeg failed to obtain one. Instead, defendant continued to prescribe high doses of short-acting opioids, which Dr. Gharibo testified would create a cycle of craving and withdrawal. When Haeg complained of increased pain, defendant added a prescription for Percocet (oxycodone and acetaminophen), without determining the reason for the{**34 NY3d at 367} [*5]increased pain. Defendant also prescribed Xanax, an anti-anxiety medication, to Haeg without any indication that Haeg suffered from anxiety and without any other documented medical basis. Furthermore, Haeg returned early for his medications three times before he overdosed in December 2009, which Dr. Gharibo testified should have alerted defendant that Haeg had an addiction and was unlikely to take his medications as prescribed. According to Dr. Gharibo, defendant created a "prescription regimen to enhance an addict's high." On December 26, 2009, three days before he was discovered dead from an overdose, defendant prescribed Haeg oxycodone, Percocet, and Xanax, among other medications. Dr. Gharibo testified that this specific prescription defendant issued to Haeg on December 26, 2009 created a "very high" risk that "covered the whole range of morbidity and mortality," including "overdosing due to misusing [the] medication and dying from respiratory death." In addition, after the medical examiner notified defendant of Haeg's death and requested a copy of defendant's patient file, defendant made several alterations to Haeg's chart to make it appear as if he had taken a more complete patient medical history.

Rappold first saw defendant in July 2009, complaining of pain due to a fall. Defendant did not consider any non-opioid treatments for Rappold, who was then 20 years old, before prescribing a high dose of opioids to him on that first visit. Dr. Gharibo testified that defendant conducted only a cursory physical examination, ordered no diagnostic testing, and did not diagnose the source of Rappold's pain. When Rappold returned to defendant over a year later, complaining of pain from another fall, defendant again failed to order diagnostic tests to objectively assess the complaint of pain and conducted little to no physical examination. Instead, defendant prescribed oxycodone and Xanax, without any indication that Rappold suffered from anxiety or needed Xanax for any other reason. Dr. Gharibo testified that this high-dose prescription was designed to create a cycle of craving and not to treat legitimate pain. Rappold returned six days later, complaining that he had lost his prescription. Without checking to see whether Rappold had, in fact, filled that prescription, defendant prescribed Percocet and Xanax at a decreased dose, with no explanation as to the change in prescription. On September 11, 2010, three days before he died, Rappold told defendant that the Percocet and Xanax were not working, so defendant returned to the earlier{**34 NY3d at 368} prescription, which constituted a significant increase in Rappold's daily dose of oxycodone and Xanax. Dr. Gharibo opined that this prescription created a "high probability of overdose and death" even if Rappold took it exactly as prescribed.

Viewing that evidence in the light most favorable to the People, we conclude that the evidence was sufficient to support the jury's finding that defendant acted recklessly. A rational juror could have concluded, based on a valid line of reasoning and permissible inferences, that defendant was aware of and consciously disregarded a substantial and unjustifiable risk that Haeg and Rappold would take more drugs than prescribed and would die by overdose, and, given defendant's position as their medical doctor, that defendant's conduct constituted a "gross deviation from the standard of conduct that a reasonable person would observe in the situation" (Penal Law § 15.05 [3]).

We disagree with our dissenting colleague that we have created a rule whereby "a reckless doctor is criminally liable for all deaths of patients under his or her care . . . irrespective of whether the doctor knew or should have known that the deceased patient would abuse the prescription medicine and would die as a result of the abuse" (dissenting op at —). Rather, we agree with the dissent that in order to uphold defendant's conviction of two counts of manslaughter, we must conclude that the People proved, by legally sufficient evidence, that defendant was aware of and consciously disregarded a substantial and unjustifiable risk that Haeg and Rappold specifically would abuse their medications and die as a result. We simply disagree with the dissent that, viewing the evidence in the light most favorable to the People, that standard was not met here. As explained, while the record here may not contain evidence that defendant was directly told that Haeg and Rappold were abusing their prescriptions or previously had come close to death by overdose, the record does contain evidence from which the jury could infer that defendant was aware of and consciously disregarded a substantial and unjustifiable risk that Haeg and Rappold were abusing the prescription drugs that defendant provided and would die as a result. We further disagree with the dissent that defendant's prescribing practices as to other patients were irrelevant to his mens rea as it pertained to Haeg and Rappold specifically.{**34 NY3d at 369}

The People also were required to prove that defendant's conduct was a "sufficiently direct cause" of death, and that there was not "an obscure or merely probable connection" between defendant's conduct and the deaths (People v Stewart, 40 NY2d 692, 697 [1976] [emphasis and internal quotation marks omitted]). Defendant's conduct [*6]must "set[ ] in motion the events which ultimately result in the victim's death" (People v Matos, 83 NY2d 509, 511 [1994]). Nevertheless, defendant's actions "need not be the sole cause of death," and defendant "need not commit the final, fatal act to be culpable for causing death" (id.; see Matter of Anthony M., 63 NY2d 270, 280 [1984]). As we recently summarized, a defendant's conduct constitutes a sufficiently direct cause of death when the People prove "(1) that defendant's actions were an actual contributory cause of [the] death, in the sense that they forged a link in the chain of causes which actually brought about the death; and (2) that the fatal result was reasonably foreseeable" (People v Davis, 28 NY3d 294, 300 [2016] [internal quotation marks and citations omitted]).

When Haeg's body was discovered on December 29, 2009, the police also recovered prescription bottles of oxycodone and Percocet, prescribed by defendant on December 26, 2009, and filled on the same date, with dozens of pills missing from each bottle. Haeg had Xanax (among other drugs) in his system, which lowered the amount of oxycodone necessary to kill him. The People did not prove that the Xanax Haeg ingested came from defendant. Nevertheless, the toxicologist testified that Haeg's oxycodone levels were so high that it was "clearly a fatal dose." Based on this evidence, a rational juror could conclude that defendant's reckless conduct was an actual contributory cause of Haeg's death.

The issue is closer with respect to Rappold, but we conclude that the evidence of causation was legally sufficient. The evidence showed that on the night before he died, Rappold took Xanax from a bottle prescribed by defendant two days before, and when that pill bottle was recovered from his car, more than half of the pills prescribed were gone. The People demonstrated that Rappold's death on September 14, 2010 was caused by "[a]cute intoxication due to [the] combined effects of alprazolam and oxycodone," meaning that although the substances were not found in his body at "overwhelmingly high level[s]," the doses were high enough that, acting synergistically,{**34 NY3d at 370} they depressed his respiration and caused his death. Although the People failed to prove that the oxycodone that contributed to Rappold's death came from defendant, the evidence supported a conclusion that the Xanax Rappold ingested did. Thus, there is a valid line of reasoning and permissible inferences from which the jury could conclude that defendant's conduct was an actual contributory cause of Rappold's death, in the sense that it "forged a link in the chain of causes which actually brought about the death" (Davis, 28 NY3d at 300). As noted, defendant's conduct need not be the sole cause of death (see Matos, 83 NY2d at 511).

Defendant's contention that Haeg's and Rappold's ingestion of the prescribed drugs in an amount greater than he prescribed was either an intervening cause or unforeseeable is without merit. "Even an intervening, independent agency will not exonerate defendant unless the death is solely attributable to the secondary agency, and not at all induced by the primary one" (Anthony M., 63 NY2d at 280 [internal quotation marks omitted]; see Stewart, 40 NY2d at 697; People v Kibbe, 35 NY2d 407, 411-413 [1974]). With respect to foreseeability, "the People must prove 'that the ultimate harm is something which should have been foreseen as being reasonably related to the acts of the accused' " (Davis, 28 NY3d at 301, quoting Kibbe, 35 NY2d at 412).

The fact that Haeg and Rappold took the substances defendant prescribed for them in a greater dosage than prescribed is neither an intervening, independent agency nor unforeseeable. It is a direct and foreseeable result of defendant's reckless conduct. As explained, viewing the evidence in the light most favorable to the People, a rational juror could conclude that defendant was aware of and consciously disregarded a substantial and unjustifiable risk that Haeg and Rappold would take the medications he prescribed at a higher dose than prescribed in order to attain a narcotic high rather than for legitimate pain management, and that they would die as a result.

Finally, defendant argues that Dr. Gharibo's testimony was not credible or reliable, and that his own expert testified that defendant's prescriptions were well within the therapeutic range of normal dosing, supported by sound medical judgment, and could not have caused death if taken as prescribed. Defendant points to evidence, including his own testimony, that there was no reason for him to know that Haeg and Rappold were addicted to opioids, that they would misuse his prescriptions, or that they would die as a result.{**34 NY3d at 371}

These arguments, however, pertain to the weight of the evidence presented to the jury on the manslaughter counts, an issue that we have no power to review (see Danielson, 9 NY3d at 349). If the jury's verdict is supported by legally sufficient evidence, we have no power to overturn the conviction on weight grounds, "regardless of our subjective assessment of the strength of the People's case" (People v Delamota, 18 NY3d 107, 116 [2011]). The [*7]Appellate Division rejected defendant's arguments pertaining to the weight of the evidence (155 AD3d at 578), and defendant does not contend that the Appellate Division failed to conduct a weight analysis or applied an incorrect standard (see Danielson, 9 NY3d at 349; Acosta, 80 NY2d at 672). "[T]he limitations of our Court's jurisdiction prevent us from second-guessing" the Appellate Division's determination that defendant's conviction on the manslaughter counts was not contrary to the weight of the evidence (Delamota, 18 NY3d at 117).

Accordingly, the order of the Appellate Division insofar as appealed from should be affirmed.

Wilson, J. (dissenting). The central problem with the majority's decision is that it contains no limiting principle. Dr. Li was grotesquely reckless. I have no quarrel, not even a quibble, with the majority's conclusion that Dr. Li's prescription practices were reckless, contrary to sound medical practice, and unlawful. Dr. Li was convicted of 170 counts of criminal sale of a prescription (Penal Law § 220.65), as well as multiple counts of reckless endangerment in the first degree (Penal Law § 120.25) and second degree (Penal Law § 120.20), grand larceny in the third degree (Penal Law § 155.35 [1]), falsifying business records in the first degree (Penal Law § 175.10), offering a false instrument for filing in the first degree (Penal Law § 175.35), and a single count of scheme to defraud in the first degree (Penal Law § 190.65 [1] [b]). Those add up to 196 separate convictions. No doubt he is a criminal. He has not challenged any of those convictions in this appeal.

The two counts at issue are for manslaughter in the second degree (Penal Law § 125.15 [1]), arising from prescriptions Dr. Li issued to Joseph Haeg and Nicholas Rappold, who died when they consumed substantially larger doses of controlled substances than those prescribed by Dr. Li. The fundamental questions are these: under what circumstances has the legislature authorized manslaughter convictions of physicians when (a) a{**34 NY3d at 372} patient has disregarded the prescribed dosage, resulting in death; or, more generally, (b) a physician has made a reckless decision and a patient has died. Historically, the criminal prosecution of medical doctors for homicide has been exceedingly rare (see R.E. Ferner & Sarah E. McDowell, Doctors Charged with Manslaughter in the Course of Medical Practice, 1795-2005: A Literature Review, 99 J Royal Soc'y Med 309 [2006] [reporting that, in the United Kingdom from 1795-2005, only 85 doctors were charged with manslaughter for deaths resulting from their medical practice]). The United States Drug Enforcement Agency tracks nationwide criminal convictions of doctors related to the distribution of controlled substances. Of the 294 doctors criminally convicted from 2003-2018, only two were [*8]convicted of homicides and two pleaded guilty (to involuntary manslaughter and negligent homicide, respectively) (see U.S. Department of Justice, Drug Enforcement Administration, Diversion Control Division, Cases Against Doctors available at https://apps2.deadiversion.usdoj.gov/CasesAgainstDoctors/spring/main?execution=e1s1 [last accessed Nov. 19, 2019]).^[FN1] The legislature has expressly provided a circumstance in which physicians may be charged with second-degree manslaughter: assisted suicide (Penal Law § 125.15 [3]). Other instances in which doctors have been charged with manslaughter have been ones in which the doctor knew, or should have known, that the doctor's actions would likely result in that specific patient's death.

The rule implicitly adopted by the majority is quite different: a reckless doctor is criminally liable for all deaths of patients under his or her care where drugs prescribed by (or errors made by) the doctor contributed to a patient's death, irrespective of whether the doctor knew or should have known that the deceased patient would abuse the prescription medicine and would die as a result of the abuse (or error).

I assume, as the majority does, that given Dr. Li's prescribing practices and thousands of patients, it was foreseeable that{**34 NY3d at 373} some patient(s) of his were likely drug abusers, and a subset of those would die from ingesting drugs he had prescribed. But unlike the majority, I do not believe that proposition is sufficient to demonstrate the causation necessary to sustain a conviction for manslaughter. Instead, I believe our decisional law governing causation requires that Dr. Li had some basis to foresee that his prescriptions to Messrs. Haeg and Rappold were likely to cause their deaths. As to certain other of Dr. Li's patients, the record evidence would be sufficient to support a finding of causation—but those patients did not die. As to Messrs. Haeg and Rappold, the record evidence is insufficient to prove causation because it fails to establish that Dr. Li had reason to foresee the risk of their deaths in particular.

A person is guilty of manslaughter in the second degree when that person "recklessly causes the death of another person" (Penal Law § 125.15 [1]). Here, "recklessly" is indisputably satisfied; the difficult issue resides in "causes." To cause the death of another under New York law, a defendant's actions must be " 'a sufficiently direct cause' of the death so that the fatal result was reasonably foreseeable" (People v Hernandez, 82 NY2d 309, 314 [1993], quoting People v Kibbe, 35 NY2d 407, 412 [1974]). A defendant's recklessness may be established, at least in part, by proof of the defendant's reckless conduct in circumstances other than the specific conduct resulting in the injury. Proof of causation, however, is distinct from recklessness because foreseeability must be established as to the specific fatal result occasioning the charge of manslaughter. In People v Roth, a petroleum transport company and two of its managers were indicted for second-degree manslaughter, criminally negligent homicide, reckless endangerment and several other crimes, resulting from an explosion that killed an employee (80 NY2d 239 [1992]). The "Appellate Division affirmed the dismissal of the homicide charges, the reckless endangerment charge and the charge of endangering public health, safety or the environment" (id. at 243). We affirmed the dismissal of the reckless manslaughter and criminally negligent homicide charges but reinstated the reckless endangerment charge. Although all of those charges required the same proof of recklessness, we explained that the manslaughter and criminally negligent homicide charges were legally unsupportable for lack of causation:{**34 NY3d at 374}

"the People were also required to submit proof from which the Grand Jury could conclude that the actual cause of the explosion was foreseeable. For purposes of criminal liability, it was not enough to show that, given the variety of dangerous conditions existing at the site, an explosion was foreseeable; instead the People were required to show that it was foreseeable that the explosion would occur in the manner that it did.

"[*9]It was error, therefore, to instruct the Grand Jury that the defendants could be indicted 'if you find that they recklessly created unsafe conditions that led to [the victim's] death by a foreseeable event, namely, the explosion.' Moreover, even if the Grand Jury had been properly instructed in this regard, the indictment on these charges could not stand" (id. at 244 [citations omitted]).

We reinstated the reckless endangerment charge because, unlike the manslaughter and criminally negligent homicide charges, "there is no requirement that the defendants' reckless conduct cause injury or death" (id. at 245). Although recklessness—including the foreseeability of some explosion—was sufficiently proved for all three counts, causation was not for the manslaughter and criminally negligent homicide charges. Likewise, to uphold Dr. Li's manslaughter convictions, more than his rampant recklessness must be proved. There must also be sufficient evidence from which the jury could have found that Dr. Li' s treatment of Messrs. Haeg and Rappold was the direct and reasonably foreseeable cause of their deaths.

The standard to prove causation for manslaughter is different from, and more demanding than, the standard to prove causation in tort. As we explained in People v Warner-Lambert Co. (51 NY2d 295 [1980]): "[w]e subscribe to the requirement that the defendants' actions must be a sufficiently direct cause of the ensuing death before there can be any imposition of criminal liability, and recognize, of course, that this standard is greater than that required to serve as a basis for tort liability" (id. at 306 [emphasis omitted]; see also Roth, 80 NY2d at 244). In the criminal context, a general foreseeable risk and an action that ignites a chain of causation, resulting in death, are not sufficient to prove that a defendant caused a specific reckless homicide (see Warner-Lambert, 51 NY2d at 305-306). Instead, the "actual immediate, triggering cause" of the specific {**34 NY3d at 375}victim's death must be foreseeable for a defendant to be found guilty of manslaughter in the second degree (id. at 307).

Our prior cases in which a defendant was found guilty of second-degree manslaughter when the defendant had provided drugs causing an overdose, though not involving doctors, satisfy that heightened standard of causation. In People v Cruciani, we upheld the manslaughter conviction of a defendant who directly injected his girlfriend with heroin, causing her death, when the evidence established that the defendant knew his girlfriend was "completely bombed out on downs," knew she had lost her capacity to "walk or talk straight," and acknowledged his injection created a substantial possibility that she would die (36 NY2d 304, 305 [1975]). The defendant could foresee that his actions would be the immediate and triggering cause of his girlfriend's death (see also People v Galle, 77 NY2d 953 [1991]). Analogously, if Messrs. Haeg or Rappold had walked into Dr. Li's office, and Dr. Li had fed them an overdose of oxycodone and/or Xanax, causation would be firmly established. But under the rationale advanced by the majority here, every heroin dealer may be convicted of manslaughter for the deaths of all users overdosing from drugs supplied by that dealer. That, of course, is not the law, even though, just as Dr. Li should have known that some patients of his would overdose given his reckless prescription practices, heroin dealers know that some customers, too, will overdose.

The majority's treatment of a case involving a heroin dealer, People v Pinckney (38 AD2d 217 [2d Dept 1972], affd, 32 NY2d 749 [1973]), is flawed. The Appellate Division held that a heroin dealer could not be prosecuted for manslaughter when one of his customers died from an overdose of the drugs he provided, using paraphernalia he also provided. Although we affirmed Pinckney, the majority notes that our summary affirmance has "minimal" precedential value and " 'does not imply approval of everything contained in the opinion of the court below' " (majority op at — [citation omitted]). Nevertheless, our summary affirmance in Pinckney poses a severe problem for the majority and cannot be brushed aside.

CPL 470.05 (2) states:

"[f]or purposes of appeal, a question of law with respect to a ruling or instruction of a criminal court during a trial or proceeding is presented when a protest thereto was registered, by the party claiming{**34 NY3d at 376} error, at the time of such ruling or instruction or at any subsequent time when the court had an opportunity of effectively changing the same."

In Pinckney, the defendant made only two arguments in the court of instance. First, he argued that the facts in the indictment did not meet the legal definition of recklessness, and second, he argued that "the legislature intended that manslaughter in the second degree and criminal negligent homicide should not apply in this type of case" (aff of Edward S. Panzer in support of motion, record on appeal at 15). Because we are [*10]powerless to decide a criminal appeal on an issue not raised in the court of instance, our summary affirmance in Pinckney must have adopted one or the other of those grounds.

If Pinckney's holding rests on the lack of specific legislative action, it applies here as well, because the legislature has enacted no homicide statutes specifically related to the reckless prescription of medicines. If, instead, it rests on the indictment's failure to establish recklessness, that too would bar Dr. Li's manslaughter convictions. The indictment in Pinckney charged that the defendant:

"wilfully[sic], wrongfully and unlawfully did recklessly cause the death of one Francis John Muthig . . . by then and there selling to and providing said Francis John Muthig with a quantity of the narcotic drug Heroin which said Francis John Muthig immediately, with the instruments furnished to him by said William L. Pinckney, prepared for injection and did inject into his body, as a direct result of which said Francis John Muthig thereafter, on the 5th day of April, 1970, died, the said William L. Pinckney knowing the said Heroin to be a dangerous drug" (indictment, record on appeal at 7).

If that indictment is insufficient to charge Mr. Pinckney with manslaughter, Dr. Li cannot be convicted of it either.

The majority's treatment of Pinckney is also instructive as to its view of the legal standard for causation. In Pinckney, both the Appellate Division majority and concurrence noted that prior cases upheld manslaughter convictions for the sale of a poison because the "obvious result of its use [is] death, [which] was known to the seller at the time of sale" (Pinckney, 38 AD2d at 219-220, citing People v Licenziata, 199 App Div 106 [2d Dept 1921], and People v Voelker, 220 App Div 528 [4th Dept 1927]). The Appellate Division then distinguished those cases{**34 NY3d at 377} from the case of a defendant whose sale of heroin resulted in a user's death. Pinckney pointed out that, as to heroin, "[a]lthough it is a matter of common knowledge that the use of heroin can result in death, it is also a known fact that an injection of heroin into the body does not generally cause death" (Pinckney, 38 AD2d at 220; see also id. at 223-224 [Shapiro, J., concurring]). The same is true of opioids other than heroin, particularly when prescribed by a doctor: they can, but do not usually, result in death. Presumably, the majority's reason for brushing Pinckney aside is to adopt a different rule of causation, one in which the reckless distribution of a dangerous drug is itself sufficient to establish causation—including foreseeability—from deaths resulting from a defendant's distribution of that drug. Otherwise, Dr. Li could not be convicted of manslaughter.

That said, I do agree with the majority that oversight or physical administration of the fatal drugs is not "a necessary element, as a matter of law, for homicide charges to be sustained" (majority op at —). For example, evidence regarding some of Dr. Li's other patients was sufficient to allow a trier of fact to find that Dr. Li's continued prescription of controlled substances to those patients would foreseeably result in a substantial risk of their deaths. One patient explicitly told Dr. Li that she was addicted to oxycodone and Xanax; that same patient's mother also called Dr. Li and implored him not to prescribe Soma to her daughter because it "was something that really could end her life" (supplemental appendix at 1342). As to certain other patients, Dr. Li knew they had previously overdosed on the same drugs he was prescribing; that knowledge would have allowed a trier of fact to conclude that Dr. Li could foresee that those patients would overdose again, with the attendant risk of death.^[FN2] Had Mr. Rappold, or his mother, alerted Dr. Li to the fact that a few months before his death, Mr. Rappold had successfully completed an opioid detoxification program that administered Suboxone to him, that information would have been evidence suggesting that a foreseeable result of Dr. Li's subsequent provision of opioids to Mr. Rappold was his death. However, the record contains no evidence that Dr. Li was aware of that information. Such examples are{**34 NY3d at 378} not exhaustive but serve to demonstrate that the heightened standard of causation for criminal liability can be met in circumstances where the defendant, doctor or otherwise, does not directly administer the lethal dosage.

By contrast, Dr. Li's prescriptions to Messrs. Haeg and Rappold were not foreseeably the direct cause of their deaths. Mr. Haeg originally came to Dr. Li with an MRI showing a central L 5-1 herniation and complaining of chronic back pain that had persisted for 17 years, treated by opioids from his prior doctors. Dr. Li prescribed pain [*11]medication to Mr. Haeg at the same level Mr. Haeg said he had been receiving from his prior doctor—a far from fatal dosage if taken as instructed. In sharp contrast to certain other of Dr. Li's patients described above, Dr. Li had received no information to suggest that Mr. Haeg would not take his medication as instructed. Mr. Haeg's early visits for medication, alone, did not indicate that his prescriptions would be the immediate and triggering cause of his death, and on the two occasions (Sept. 26, 2009 and Dec. 5, 2009) that Mr. Haeg returned early, Dr. Li wrote prescriptions for 84 pills instead of the usual 120.

As to Mr. Rappold, Dr. Li conducted an introductory physical examination of his new patient, who complained of pain from a fall, and prescribed him pain medication within accepted therapeutic bounds. When Mr. Rappold returned early, explaining that he had lost his prescription, Dr. Li issued him a prescription with a sharply reduced dose. Dr. Li did not return Mr. Rappold's prescription to its original dosage until Mr. Rappold told Dr. Li that the pain medications were no longer working. Again, there was no indication that Mr. Rappold would not take his prescriptions as instructed and no evidence to suggest that those prescriptions would foreseeably result in Mr. Rappold's death. Indeed, the evidence at trial suggested that Mr. Rappold had ingested drugs from multiple sources on the night he died, further attenuating the claim that Dr. Li caused his death. No record evidence indicates that Dr. Li had reason to believe Mr. Rappold was obtaining drugs from others, likely illegally.

Our several cases relied on by the majority to describe the causation requirement for homicide do not support the conclusion that, absent some reason for a physician to believe that a patient will radically disregard prescribed dosages, liability for{**34 NY3d at 379} homicide attaches if the patient dies as a result of an overdose. In People v Stewart (40 NY2d 692 [1976]), the defendant stabbed his victim in the stomach; the question was whether the victim, who died during surgery, died as a result of the stab wound or as a result of a heart attack occurring when the surgeons decided to repair a hernia unrelated to the stab wound. There, we expressly stated that "something more is required" than proving that the defendant's conduct "forged a link in the chain of causes which actually brought about the death," namely, "the defendant's actions must be a sufficiently direct cause of the ensuing death before there can be any imposition of criminal liability" (id. at 697). We vacated the defendant's conviction for manslaughter, reducing it to assault.

People v Matos (83 NY2d 509 [1994]) holds that the felony murder rule applies to a defendant whose rooftop flight from the police resulted in the falling death of a pursuing officer because "it should also be foreseeable that someone might fall while in hot pursuit across urban roofs in the middle of the night" (id. at 512). Matter of Anthony M. (63 NY2d 270 [1984]) held that defendants could be held to have caused the cardiac-arrest deaths of their assault victims, who had no prior history of cardiac trouble, where medical evidence supported the conclusion that the stress of the assaults caused the fatal heart attacks days later. People v Davis (28 NY3d 294 [2016]) is the same: death by cardiac arrest following an assault. Those cases are fundamentally different from Dr. Li's. Unlike the defendants in those cases, Dr. Li had no reason to foresee that Messrs. Haeg or Rappold would so substantially deviate from the prescribed dosages as to cause their deaths.

Although the majority asserts that "a rational juror could conclude that defendant was aware of and consciously disregarded a substantial and unjustifiable risk that Haeg and Rappold would take the medications he prescribed at a higher dose than prescribed in order to attain a narcotic high rather than for legitimate pain management, and that they would die as a result" (majority op at —), the majority does not specify on what evidence a rational juror could so conclude. There are only two possibilities: either it is the overall evidence of Dr. Li's reckless pain management practice in general—which means the majority has adopted a rule that if a doctor's recklessness makes it foreseeable that some patients may die from an overdose, causation is established as to any death resulting from any patient's overdose—or else something about {**34 NY3d at 380}Dr. Li's practice with regard to Messrs. Haeg and Rappold in particular demonstrates specific foreseeability as to their deaths.

If it is the latter, the evidence is not sufficient to establish causation. Taking Mr. Haeg, as to whom the majority says the evidence of recklessness is stronger, most of what the majority describes (majority op at —) does not show any reasonable foreseeability of prescription abuse, much less death. For example, Dr. Li's failure to verify [*12]independently the doses prescribed by Mr. Haeg's former doctors, or his failure to reduce Mr. Haeg's dosages when Mr. Haeg did not provide the updated MRI that Dr. Li requested, do not bear on the question of foreseeability of death. Presumably, a doctor who prescribes Xanax to a patient without an indication that the patient has anxiety is not liable for manslaughter if the patient takes an overdose of Xanax, although perhaps the majority's new rule is otherwise.

The sole item cited by the majority that appears to bear on foreseeability is Dr. Gharibo's testimony that the prescription given to Mr. Haeg "three days before he was discovered dead from an overdose . . . created a 'very high' risk that 'covered the whole range of morbidity and mortality,' including 'overdosing due to misusing [the] medication and dying from respiratory death' " (majority op at —). But that is not what Dr. Gharibo testified. Instead, the majority has taken his statement that the prescribed dosages created a "very high risk, highly addictive, inappropriate misprescribing . . . driven to prescribe to create addiction and potentially even create diversion and death to the individual as well as to the people around them" (supplemental appendix at 0906)—which makes a clear distinction between the "very high risk" (that Dr. Li would create an addiction) and the less foreseeable result (that he could "potentially even create diversion and death")—and appended it to Dr. Gharibo's assessment of the general risk faced by opioid addicts. That generalized risk includes a smörg sbord of catastrophes ranging from "falling and injuring themselves and getting hurt in a whole variety of ways, whether it's injuring their foot or hand or getting hit by a car, to overdosing due to misusing their medication and dying from respiratory death." Dr. Gharibo's testimony as to the litany of doom that might befall any opioid addict—from stubbing one's toe to death—demonstrates the lack of foreseeability of any particular result within that range. His testimony does not meet the{**34 NY3d at 381} standard for causation as set forth in Warner-Lambert and Roth.

The majority has extended homicide liability to physicians by blurring recklessness into causation: if a doctor is generally reckless in prescribing drugs, it is foreseeable that some patient may die; if it is foreseeable that some patient may die, causation is established when any patient dies from an overdose of the prescribed drug(s). In applying a standard under which doctors can be found guilty of manslaughter if they maintain reckless prescription practices and a patient takes an overdose, the majority creates a novel and unwelcome extension of criminal liability for physicians.

The result here (holding a reckless doctor liable for homicide as a result of a patient's death) is inconsistent with the way errors—even grossly negligent, fatal errors—by doctors have been historically addressed. The state licenses doctors to make sure they are at least minimally competent, withdraws licensure if a doctor proves incompetent, and subjects all doctors, good and bad, to tort liability under well-established law governing medical malpractice.^[FN3] Doctors can also be found guilty of crimes other than manslaughter, including crimes that the legislature specifically targeted at the criminal sale of prescriptions for which Dr. Li has been convicted.

Exactly how far today's decision expands the homicide liability for doctors is unclear. If a doctor recklessly prescribes drugs that interact and cause a death, will that be considered manslaughter? If a doctor relies on a patient's self-report of medical history or allergies and the patient dies, will that be deemed reckless and also sufficient to establish causation because the doctor acted without independently verifying the information, justifying a manslaughter charge? We have previously recognized the impropriety in conflating the causation element required in typical homicides with the causation element required in commercial situations resulting in death (see{**34 NY3d at 382}People v DaCosta, 6 NY3d 181, 186 [2006]). I would tread much more gingerly in advancing homicide liability against doctors than does the majority, leaving the legislature to determine, prospectively, what criminal penalties should [*13]attach to what kinds of reckless conduct by doctors, instead of attempting to apply to the medical profession causation rules developed in cases involving garden-variety assaults and burglaries. To some extent, the legislature has already done so, as evidenced by the myriad other crimes of which Dr. Li was tried and convicted, convictions he has not appealed to this court.

Chief Judge DiFiore and Judges Rivera, Stein, Garcia and Feinman concur; Judge Wilson dissents in an opinion.

Order insofar as appealed from affirmed.