United States v. Wood

207 F.3d 1222, 54 Fed. R. Serv. 108, 2000 Colo. J. C.A.R. 1645, 2000 U.S. App. LEXIS 5475, 2000 WL 328142

• Court: Court of Appeals for the Tenth Circuit
• Decided: March 29, 2000
• Docket: 99-7008
• Status: Published

Opinion

LUCERO, Circuit Judge.

C. Douglas Wood is a physician who, in 1998, was indicted for the first-degree murder of Virgil Dykes, a patient under his care at the Veterans Administration hospital in Muskogee, Oklahoma. The case was tried to a jury, and submitted on charges of first-degree murder and the lesser included offenses of second-degree murder and involuntary manslaughter. A verdiet was entered acquitting Dr. Wood of the first and second-degree murder charges, but convicting him of involuntary manslaughter. The case comes to us for review on direct appeal. We exercise jurisdiction pursuant to 28 U.S.C. § 1291 and 18 U.S.C. § 3231. Appellant advances several claims of. error: (1) denial of his motion for judgment of acquittal as to all charges; (2) wrongful prosecution; (3) delay of prosecution; (4) denial of jury instructions on his theory of the case; (5) admission of expert testimony that stated an opinion as to the his mental state; and (6) admission of unduly prejudicial testimony. Concluding that Dr. Wood was denied a fair trial because of cumulative error, we reverse and remand for retrial on the charge of involuntary manslaughter.

I

We begin with a recitation of the events leading to Dr. Wood’s indictment. Virgil Dykes was an 86-year-old man, suffering from severe abdominal pain when he arrived at the Veterans Administration hospital in Muskogee, Oklahoma, on February 5, 1994. Dr. Wood had not previously treated Dykes, but because he was the attending physician that day, he diagnosed and operated on Dykes to repair a perforation in the proximal duodenum which had led to diffuse perotonitis. Over the next eight days, patient Dykes remained under Dr. Wood’s care in the surgical intensive care unit. On the morning of February 13, Dr. Glen Lytle, the chief of surgery, Dr. Rocky Morgan, a fifth-year resident, and Dr. Randall Bass, a first-year intern, made hospital rounds. A morning blood test revealed that patient Dykes had an extracellular potassium level of 3.2 millie-quivalents (mEq) per liter. That level was below the lower limit of the range the hospital considered normal — 3.3 to 5.5. At around 9 a.m., Dr. Morgan ordered that patient Dykes be given 40 mEq of potassi um chloride (KC1) to raise his potassium level. Nurse Kinsey administered the KC1 in an elixir form through a nasogastric tube directly into Dykes’s stomach. Based on a discussion with Dr. Morgan, Dr. Bass prescribed Lasix, a diuretic drug commonly used to remove fluid from a patient’s lungs, which has the side effect of reducing potassium levels. Soon after giving patient Dykes the elixir of KC1, Nurse Marla Kinsey withdrew 170 ccs of fluid from his stomach, which indicated he was not absorbing the KC1. She then returned approximately 100 ccs of the fluid to the patient’s stomach.

Dr. Wood arrived at the hospital at 11 a.m. and, upon reviewing patient Dykes’s chest x-rays from earlier that morning, stated that he was drowning from pulmonary edema — excess fluid in his lungs. Dr. Wood then ordered Nurse Kinsey to prepare 40 mg of Lasix and an IV bag with 40 mEq of KC1 in 100 ccs of saline solution. When Nurse Kinsey informed Dr. Wood that the most rapid rate at which she could administer the KC1 solution was over the course of an hour, he ordered her to draw up a syringe of 40 mEq of KC1 in 30 to 50 ccs of saline. Nurse Kinsey prepared the KC1 solution in a 60 cc syringe, but refused to administer it, believing it to be dangerous. Nurse Martha Hardesty, who was also present, told Dr. Wood that hospital policies permitted a maximum dosage of 40 mEqs of KC1 over one hour. Dr. Wood then took the syringe from Nurse Kinsey and administered the KC1 himself. Dr. Bass, Nurse Kinsey, Nurse Hardesty, and Dr. Wood gave conflicting testimony regarding how much KC1 was administered how quickly.

During the injection the heart monitor flat-lined and patient Dykes stopped breathing. Dr. Wood stopped injecting and made one or two precordial thumps (sharp concussions to the chest) in an effort to restart Dykes’s heart. Dr. Bass and Nurse Kinsey also engaged in resuscitation efforts, including chest compressions.. After two to four minutes, Dr. Wood pronounced Dykes dead.

Following nearly four years of investigation, the government obtained an indictment charging Dr. Wood with first-degree murder in violation of 18 U.S.C. § 1111(a). After the district court denied Dr. Wood’s motion to dismiss for prosecutorial delay, the case went to trial. At the close of the government’s case, Dr. Wood moved for judgment of acquittal pursuant to Fed. R.Crim.P. 29(a). The district court denied the motion. Dr. Wood made another motion for judgment of acquittal at the close of all evidence, at which time the district court exercised its discretion to reserve its ruling until after the jury returned its verdict. See Fed. R. Crim P. 29(b). The district court instructed the jury on first-degree murder as well as the lesser included offenses of second-degree murder and involuntary manslaughter.. The jury returned a verdict of guilty to the charge of involuntary manslaughter and not guilty on the greater offenses. Following the. jury verdict, Dr. Wood renewed his Rule 29 motion, which the district court again denied.

Departing downward from the sentencing range applicable under the United States Sentencing Guidelines, the district court sentenced Dr. Wood to 5 months imprisonment, 36 months supervised release, a $100 assessment, and a $25,000 fine. Dr. Wood appeals his conviction. 1

II

Dr. Wood alleges the district court erred in denying his motions for judgment of acquittal pursuant to Fed. R.Crim.P. 29(a) because there was insufficient evidence of premeditation and malice aforethought to sustain the murder charges and the evidence of gross negligence is insufficient to support his conviction for involuntary manslaughter. In reviewing both the sufficiency of the evidence to support a conviction and a denial of a motion for judgment of acquittal, this court must review the record de novo to determine whether, viewing the evidence in the light most favorable to the government, any rational trier of fact could have found the defendant guilty of the crime beyond a reasonable doubt. See United States v. Schluneger, 184 F.3d 1154, 1158 (10th Cir.1999), cert. denied, 120 S.Ct. 800 (2000); United States v. Voss, 82 F.3d 1521, 1524-25 (10th Cir.1996). Contrary to Dr. Wood’s assertions, the evidence necessary to support a verdict “ ‘need not conclusively exclude every other reasonable hypothesis and need not negate all possibilities except guilt.’” United States v. Wilson, 182 F.3d 737, 742 (10th Cir.1999) (quoting United States v. Parrish, 925 F.2d 1293, 1297 (10th Cir.1991)). When reviewing the denial of a motion for judgment of acquittal made at the close of the government’s case-in-chief, we look only to evidence entered into the record at the time of the motion, that is, when the government rested. See Fed. R.Crim.P. 29(b). This is true even though the district court could have reserved ruling on the motion until the jury returned its verdict. See id.

A

Before addressing the merits of these arguments,, we pause to delineate the elements of the three degrees of homicide submitted to the jury: first-degree murder, second-degree murder, and involuntary manslaughter. All three offenses involve the unlawful killing of a human being. See 18 U.S.C. §§ 1111(a), 1112(a). The difference between them is the requi-' site mens rea. First-degree murder, as applicable to this case, requires both malice aforethought, see 18 U.S.C. § 1111(a), and the specific intent to commit an unlawful killing, see United States v. Sands, 968 F.2d 1058, 1064 (10th Cir.1992). A killing is committed with the requisite specific intent if it is “willful, deliberate, malicious, and premeditated.” 18 U.S.C. § 1111(a). Second-degree murder, by contrast, is a general intent crime, see United States v. Soundingsides, 820 F.2d 1232, 1242 (10th Cir.1987), that requires only malice aforethought, see 18 U.S.C. § 1111(a). “[Second degree murder’s malice aforethought element is satisfied by: (1) intent-to-kill without the added ingredients of premeditation and deliberation; (2) intent-to-do-serious-bodily-injury; (3) depraved-heart; or (4) commission of [certain felonies].” United States v. Pearson, 159 F.3d 480, 486 (10th Cir.1998) (citing Wayne R. La-Fave & Austin W. Scott, Jr., Criminal Law at 648 (2d ed.1986)). We have also held that malice aforethought “may be established by evidence of conduct which is reckless and wanton, and a gross deviation from a reasonable standard of care, of such a nature that a jury is warranted in inferring that defendant was aware of a serious risk of death or serious bodily harm.” Soundingsides, 820 F.2d at 1237 (citing United States v. Black Elk, 579 F.2d 49, 51 (8th Cir.1978)); see also United States v. Joe, 8 F.3d 1488, 1500 (10th Cir.1993). The concepts of “depraved heart” and “reckless and wanton, and a gross deviation from a reasonable standard of care” are functionally equivalent in this context. See United States v. Houser, 130 F.3d 867, 871 n. 3 (9th Cir.1997).

Involuntary manslaughter, as applicable to this case, is the unlawful killing of a human being without malice in the commission, without due caution and circumspection, of a lawful act which might produce death. See 18 U.S.C. § 1112(a). The defendant’s acts must amount to “ ‘gross negligence,’ ” defined as “ “wanton or reckless disregard for human life.’” United States v. Bryant, 892 F.2d 1466, 1470 (10th Cir.1989) (quoting United States v. Benally, 756 F.2d 773, 776 (10th Cir.1985)). “Gross negligence” describes a degree of culpability far more serious than tort negligence. See United States v. Browner, 889 F.2d 549, 553 (5th Cir.1989); United States v. Escamilla, 467 F.2d 341, 346 (4th Cir.1972).

The distinction between involuntary manslaughter and second-degree murder is that the former does not require malice aforethought. The definitions of “malice aforethought” and “without due care and circumspection” developed in our case law, however, use overlapping terminology: both refer to “reckless and wanton” behavior. See Bryant, 892 F.2d at 1470 (affirming involuntary manslaughter instructions requiring “wanton or reckless” conduct); Soundingsides, 820 F.2d at 1237 (defining malice aforethought as reckless and wanton conduct). The substantive distinction is the severity of the reckless and wanton behavior: Second-degree murder involves reckless and wanton disregard for human life that is extreme in nature, while involuntary manslaughter involves reckless and wanton disregard that is not extreme in nature. See Houser, 130 F.3d at 872; United States v. Sheffey, 57 F.3d 1419, 1430 (6th Cir.1995); United States v. One Star, 979 F.2d 1319, 1321 (8th Cir.1992); see also United States v. Fleming, 739 F.2d 945, 948 (4th Cir.1984) (noting that the difference between murder and manslaughter is “one of degree rather than kind”).

B

Dr. Wood moved for judgment of acquittal after the close of the government’s case-in-chief pursuant to Fed. R.Crim.P. 29(a) on the grounds that the evidence presented failed to prove thé requisite intent to kill — premeditation or malice aforethought. 2 The district court denied the motion, but conceded it was troubled by the implications of that result. When ruling on a motion for judgment of acquittal, a district court should consider not only whether the evidence would be sufficient to sustain a conviction of the offense charged, but also whether it would be sufficient to sustain a conviction on a lesser included offense. See Fed. R.Crim.P. 31(c) (“The defendant may be found guilty of an offense necessarily included in the offense charged.... ”); United States v. Cavanaugh, 948 F.2d 405, 409 (8th Cir.1991) (holding that if a court grants judgment of acquittal following a jury verdict of guilty, it may enter judgment of conviction on a lesser included offense); 2 Charles Alan Wright, Federal Practice and Procedure: Criminal 2d § 467 (2d ed. 1982) (“[0]n a motion for judgment of acquittal the court must consider whether the evidence would be sufficient to sustain a conviction of [ ] a lesser offense.”). If the evidence is sufficient to sustain a conviction on the lesser but not the greater offense, the judge may submit only the lesser charge to the jury. See United States v. LoRusso, 695 F.2d 45, 52 (2d Cir.1982).

We conclude that the evidence supports the district court’s instinct rather than its result. The evidence introduced during the government’s case-in-chief, taken in the light most favorable to the government, was insufficient to prove that Dr. Wood killed Dykes with premeditation or malice aforethought. The government presented no direct evidence of a specific intent to kill. Nor could a rational juror infer from the circumstantial evidence that Dr. Wood acted with premeditation or in a manner that was extremely reckless, wanton, and a gross deviation from the reasonable standard of care when he injected Dykes with potassium chloride. Because the evidence could not have sustained a conviction on the charged offense of first-degree murder or the lesser included offense of second-degree murder, the district court erred when it denied Dr. Wood’s motion for judgment of acquittal as to those offenses.

1. The Government’s Portrayal of Evidence is Misleading

The government’s characterization of the evidence to support an inference of premeditation and malice aforethought is misleading and unpersuasive. It points to the following facts: (1) the amount, speed, and method (i.e. injection into a central vein) of administration of KC1 were contrary to accepted medical practice; (2) Dykes was stable that morning and did not require emergency treatment; (3) nurses warned Dr. Wood that the proposed injection of KC1 was dangerous; (4) Dr. Wood ordered Nurse Kinsey to prepare 40 mEq of KC1 even though he said that he only intended to inject 10 mEq; and (5) Dr. Wood continued injection despite Dykes’s worsening condition and then took inadequate measures to resuscitate him. A careful review of the record reveals that the evidence does not support such a stark picture, nor does it support an inference of premeditation or malice aforethought.

The prosecution’s evidence establishes general standards for the administration of potassium and indicates that Dr. Wood’s treatment of Dykes did not comport with those standards. All of the government’s experts concurred with the following propositions: potassium is essential to life and proper heart functioning, but when administered at a high concentration over a short period of time it can be lethal; administering KC1 by intravenous injection (rather than titration from an IV bag) into a central vein was appropriate only in extreme situations; and, in general, a dosage of 40 mEq of KC1 in an hour was appropriate, though one mEq per minute was acceptable in emergency situations. An FBI agent testified that Dr. Wood responded affirmatively when the agent asked whether 10 mEq in one minute would be too fast. As to the dosage administered in the instant case, the conflicting evidence, 3 taken in the light most favorable to the government, indicates that Dr. Wood injected Dykes with approximately 20 mEq of KC1 over 30 seconds.

All of the government’s eyewitnesses conceded that Dykes was very sick on the morning of the thirteenth and required at least some KC1, though they disagreed as to whether his condition required Dr. Wood’s immediate treatment. Early that morning, Dr. Morgan reviewed a recent x-ray of Dykes’s lungs and made a tentative diagnosis of adult respiratory distress syndrome (ARDS), a serious and often preter-minal condition involving leakage of fluid into the lungs. Dr. Lytle testified, based on the same x-rays, that there was no significant change in the amount of fluid in Dykes’s lungs between 1 a.m. and 7 a.m. that morning. Dr. Lytle also testified that Dykes was extremely ill and his condition was not improving, but he was unlikely to die in the next forty-eight hours and that, while Dykes’s potassium level was low and “of concern,” he saw nothing that morning that would justify a rapid injection of KC1. Dr. Morgan and Dr. Lytle agreed that Dykes’s condition required an adjustment to his treatment and prescribed Lasix and KC1. In Nurse Kinsey’s opinion, Dykes was “unresponsive to verbal stimuli,” “very sick,” (III R. at 196), and in “critical condition,” (IV R. at 268), with serious cardiac and respiratory problems, but that his health had not “changed dramatically,” (IV R. at 228), his potassium level was low but not critical, and emergency measures were not justified. On cross-examination she stated that administration of potassium was medically indicated. Nurse Hardesty said Dykes was stable, though conceded he had multiple “extremely severe” ailments, including low potassium, (IV R. at 392), and that Dykes needed potassium.

Government witnesses not present at the hospital that day support the government’s contention that Dr. Wood’s chosen course-of-treatment was inappropriate, but they belie the suggestion that Dykes’s was anything other than very ill. Pathologist Dr. Michael Baden testified that Dykes was suffering “from various severe chronic diseases,” (IV R. at 433), but stable, and the rapid injection of KC1 described by the government eyewitnesses “was not warranted from a clinical point of view, was extremely dangerous, was beyond reasonable medical treatment.... It was reckless.” (IV R. at 446.) He also classified Dykes’s death as a homicide. In the opinion of nephrologist Dr. James Knochel, Dykes was not suffering from ARDS or pulmonary edema, and Dr. Wood should have obtained a new blood test before taking any action. He further stated that Lasix would not have reduced Dykes’s potassium level enough to warrant an injection of potassium, and that rapid injection was never justified because the predictable result was cardiac arrest. In the opinion of radiologist Dr. Max Walter, Dykes was suffering from pulmonary edema, but his condition was stable save for a worsening of pneumonia in the left lung. Cardiologist Dr. Eliot Schechter thought that on the thirteenth Dykes was in poor but stable condition and did not have ARDS, that it was unlikely that an additional dose of Lasix would drop Dykes’s potassium to a dangerous level, and that the amount of potassium administered by Dr. Wood was not “appropriate.” (VI R. at 885). Dr. Schecter conceded on cross-examination that Dykes required potassium although he did not believe a rapid injection was necessary.

The government accurately notes Dr. Wood performed the injection himself after both nurses warned the proposed dosage exceeded the amount permitted by hospital policy and recommended by a drug manual. Dr. Wood also ordered Nurse Kinsey to prepare 40 mEq of KC1 even though he later informed an FBI investigator that he only intended to inject Dykes with 10 mEq.

The government, however, mischaracter-izes the evidence as to Dr. Wood’s reaction to Dykes’s worsening condition during the course of the injection. Dr. Bass did not testify that Dr. Wood continued injection even after the heart monitor emitted a loud alarm. Rather, he testified that Dr. Wood continued to inject while Dykes’s “cardiac rhythm changed,” (V R. at 646), and after the patient gasped, but that Dr. Wood stopped injecting when the monitor flat-lined and Dykes stopped breathing.

The government also miseharacterizes the efforts to resuscitate Dykes following the injection. It was uncontroverted that Dr. Wood gave Dykes one or two precordial thumps. Nurse Hardesty testified that the purpose of precordial thumps is to revive the patient. Dr. Bass then began chest compressions, which Dr. Wood interrupted for a moment. Contrary to the government’s implication that this action was nefarious, it is uncontroverted that chest compressions must be paused in order to get an accurate reading from the heart monitor. At some point Nurse Kinsey began to oxygenate Dykes using an ampu bag. Initially Dr. Wood told Nurse Hardesty not to call a code, but eventually had her call for the assistance of respiratory therapy. 4 Dr. Lytle and Nurse Hardes ty testified that when the condition of a full-code patient like Dykes becomes critical and the attending physician is present, it is that physician’s responsibility to determine what resuscitative efforts are appropriate, including whether to call a code. Through expert testimony, the government established that a prompt injection of calcium may reverse the effects of excessive potassium and that Dr. Wood did not use the calcium available in the ICU. Dr. Schechter opined that the attempts at resuscitation were inadequate. Nurse Har-desty testified that Dr. Wood pronounced Dykes dead and ceased resuscitation efforts after two to four minutes.

2. The Evidence Fails to Establish Premeditation or Malice Aforethought

This scenario plainly is not one of first-degree murder. Well-intentioned but inappropriate medical care, standing alone, does not raise an inference that a killing was deliberate, willful, and premeditated. Cf. Kansas v. Naramore, 25 Kan. App.2d 302, 965 P.2d 211, 223-24 (1998) (reversing a physician’s convictions for attempted murder and second-degree murder). Rather, specific intent is properly inferred where the apparent purpose of the lethal act is to cause the victim’s death. See United States v. Downs, 56 F.3d 973, 975 (8th Cir.1995) (finding premeditation established by evidence of extensive planning, lying in wait, and shooting the victim multiple times as she pleaded for her life); United States v. Treas-Wilson, 3 F.3d 1406, 1409-10 (10th Cir.1993) (finding that premeditation was based on evidence that the defendant injured his victim and then dragged her outside prior to inflicting a “precise and fatal” knife wound); United States v. Sides, 944 F.2d 1554, 1558 (10th Cir.1991) (finding evidence that the victim was shot while sitting passively, crying, and crossing herself was sufficient to prove premeditation). The evidence presented during the government’s case-in-chief, taken in the light most favorable to the government, does not support an inference of lethal purpose. Rather, the only reasonable inference a jury could draw is that Dr. Wood acted in good faith with the intent to save or prolong Dykes’s life in what he believed to be an emergency situation. 5

Likewise, Dr. Wood’s treatment of Dykes does not support an inference of malice aforethought sufficient to sustain a conviction of second-degree murder. There is evidence Dr. Wood knew the course-of-treatment he chose — intravenous injection of substantial quantities of KC1 — was unusual and presented a risk, even a “serious risk of death or serious bodily harm.” Soundingsides, 820 F.2d at 1237. However, the evidence also demonstrates that the administration of KC1 in some manner was medically indicated. Dr. Wood’s treatment of Dykes involved a choice between several courses of action, some of which were more risky, but perhaps more efficacious, than others. A physician cannot be convicted of murder simply for adopting, in an emergency setting, a risky course of action intended to prolong life that, when carried out, fails to forestall or even hastens death. Cf. Naramore, 965 P.2d at 223-24. Instead, to permit a charge of murder with malice aforethought to go to the jury, that choice must be not only a gross deviation from a reasonable standard of care, but also extremely reckless and wanton. See Soundingsides, 820 F.2d at 1237. Dr. Wood’s good-faith efforts at treatment simply do not rise to the “extreme” disregard for human life necessary to satisfy the malice aforethought standard. See Houser 130 F.3d at 872. 6

This conclusion is confirmed by comparing the instant facts with those in other cases examining the sufficiency of the evidence to support second-degree murder convictions. Our research uncovered only three cases, state or federal, in which an appellate court was confronted with the sufficiency of the evidence to support a charge or conviction of second-degree murder stemming from a licensed physician’s treatment of a patient. In two cases, the reviewing court affirmed a physician’s conviction for second-degree murder. See Einaugler v. Supreme Court of New York, 109 F.3d 836, 840-41 (2d Cir.1997) (denying a petition for a writ of habeas corpus based on a claim of insufficient evidence to sustain a second-degree murder conviction because of evidence that the physician had ordered the patient to be fed through a kidney dialysis catheter and, upon discovering the mistake, waited ten hours before transferring the patient to the hospital despite being told by a specialist that prompt treatment was necessary); People v. Klvana, 11 Cal.App.4th 1679, 15 Cal. Rptr.2d 512, 527 (1992) (affirming a physician’s conviction of second-degree murder for the death of nine infants because a reasonable jury could find implied malice based on the facts that the defendant repeatedly ignored obvious signs of respiratory distress, advised parents not to take their children to the hospital despite apparent problems, and induced vaginal births in entirely inappropriate circumstances, all after having been warned on numerous occasions that his treatment was dangerously substandard). These cases involved treatment that had no conceivable clinical benefit and was entirely outside the proper standard of care. By contrast, Dr. Wood gave a medically indicated drug to a very ill man, but, construing the evidence in the light most favorable to the government, gave an inappropriate dosage. When it became apparent that his chosen course-of-treatment had failed, he then took resuscitative measures.

Similarly, in cases sustaining a second-degree murder conviction under 18 U.S.C. § 1111, the defendants’ conduct does not reflect a choice between inherently risky courses of action undertaken in a good faith effort to prolong the victims’ life. Rather, the defendants’ conduct generally displays an extreme disregard for the well-being of others. See, e.g., United States v. Eder, 836 F.2d 1145, 1149 (8th Cir.1988) (finding the evidence sufficient to support malice aforethought where the defendant inflicted a head wound on his daughter and then denied her medical attention, despite numerous pleas from third parties that he seek treatment for her); Fleming, 739 F.2d at 945-48 (finding evidence that the defendant had an extremely high blood alcohol content level, was driving at double the posted speed limit, and frequently drove on the wrong side of the road was sufficient to prove malice aforethought). When compared with the actions of physicians and others that courts have found sufficient to sustain a conviction for second-degree murder, Dr. Wood’s treatment of Dykes clearly does not amount to the extremely reckless and wanton disregard for life necessary to support an inference of malice aforethought.

Dr. Wood’s actions are even less culpable than those of the defendant in United States v. Millen, 594 F.2d 1085, 1087 (6th Cir.1979), in which the court held the district court erred in denying the defendant-physician’s motion for a directed verdict on a charge of second-degree murder. In that case, the defendant, Dr. Millen, prescribed Demerol to a known drug addict who subsequently overdosed on the drug. While the Sixth Circuit did not state why the evidence did not support a charge of second-degree murder, a comparison is nevertheless enlightening: Dr. Wood’s administration of KC1 to Dykes was medically indicated and had the goal of prolonging life, whereas Dr. Millen’s prescription of Demerol was contraindicated given the patient’s known history of serious drug abuse. Thus, not only do Dr. Wood’s actions fall far short of the extremely reckless disregard for human life evident in cases sustaining second-degree murder convictions against physicians and others, but his actions are even less reckless than the actions of Dr. Millen, whose conviction was reversed.

We conclude that, taking the evidence presented during the government’s casein-chief in the light most favorable to the government, no rational trier of fact could have found malice aforethought, let alone premeditation, beyond a reasonable doubt. Therefore, the district court erred when it denied Dr. Wood’s motion for judgment of acquittal on the charges of first and second-degree murder.

C

In contrast, the evidence presented at trial, taken in the light most favorable to the government, was sufficient to support Dr. Wood’s conviction for involuntary manslaughter. As discussed above, the distinction between depraved-heart second-degree murder and involuntary manslaughter is whether the reckless conduct is so extreme in nature as to permit an inference of malice. Although the evidence was not sufficient to demonstrate that Dr. Wood’s actions were so extremely reckless that a juror could infer malice aforethought, we conclude that a jury could find beyond a reasonable doubt that the manner in which Dr. Wood performed that injection was reckless enough to constitute a lack of “due caution and circumspection.” 18 U.S.C. § 1112. Viewed in the light most favorable to the government, the evidence establishes that the quantity and speed of the injection of KC1 exceeded the consensus as to the maximum beneficial dosage. Given Dr. Wood’s therapeutic purpose and the uncontradict-ed propriety of administering some KC1, the excessive dosage does not support an inference of malice aforethought. It suffices, however, to prove the reduced degree of recklessness necessary to sustain a conviction for involuntary manslaughter. See Pennsylvania v. Youngkin, 285 Pa.Super. 417, 427 A.2d 1356, 1361 (1981) (finding the evidence was sufficient to support a conviction for involuntary manslaughter where the physician prescribed inordinate amounts of barbituates to a patient he knew was abusing those drugs); Utah v. Warden, 813 P.2d 1146, 1152 (Utah 1991) (affirming a negligent homicide conviction where the evidence that the physician failed to treat a newborn clearly suffering from respiratory distress syndrome and discouraging the baby’s mother from seeking treatment elsewhere demonstrated a “wide gulf’ between the defendant’s actions and the appropriate standard of care).

Ill

Before determining whether the erroneous denial of Dr. Wood’s motion for judgment of acquittal on the charges of first and second-degree murder requires reversal despite the sufficiency of the evidence to sustain Dr. Wood’s conviction for involuntary manslaughter, we turn to Dr. Wood’s other assignments of error: prejudicial delay of prosecution, failure to instruct the jury on his theory of defense, and improper admission of several pieces of evidence.

Whether a defendant’s “due process rights were denied by a delay in bringing an indictment is a question of fact, which this court reviews for clear error.” United States v. Trammell, 133 F.3d 1343, 1351 (10th Cir.1998). When seeking dismissal of an indictment based on pre-indictment delay, a defendant must establish the government intentionally delayed for tactical reasons and the delay caused him actual prejudice. See id. “Vague and conclusory allegations of prejudice resulting from the passage of time and the absence of witnesses are insufficient to constitute a showing of actual prejudice.” Id.

Dr, Wood fails to demonstrate either intentional delay or prejudice. He points to no evidence from which this court could infer that the government’s delay was intentional for the purpose .of gaining tactical advantage. Furthermore, the only specific example of prejudice he alleges concerns the autopsy of Dykes. The autopsy was conducted by court order approximately 42 months after Dykes’s death, at which time the body was putrefied. Dr. Wood claims that, as a result, “objective evidence available to the accused to refute the allegations was lost.” (Appellant’s Br. at 42.) Dr. Wood fails, however, to , clarify what evidence might have been available. The evidence at trial demonstrated that potassium levels cannot be determined in an autopsy. Dr. Wood’s claim of delay, therefore, is too vague and conclusory to establish prejudice under the circumstances.

At trial, the district court refused to give Dr. Wood’s proffered jury instruction to the effect that it is not unlawful for a physician to adopt one of several proper methods of treatment. “Although a criminal defendant is éntitled to an instruction regarding his theory of the case, a trial judge is given substantial latitude and discretion in tailoring and formulating ’the instructions so long as they are correct statements of law and fairly and adequately cover the issues presented.” United States v. Pack, 773 F.2d 261, 267 (10th Cir.1985). The instructions correctly advised the jury that if it found the defendant acted in “good faith” it must return a verdict of not guilty. (VIII R. at 1408-09.) This is an adequate formulation of the defendant’s theory in light of existing case law. 7

A district court’s decision to admit evidence, including expert testimony, is reviewed under an abuse of discretion standard. See United States v. Rice, 52 F.3d 843, 847 (10th Cir.1995). Dr. Wood alleges that two parts of Dr. Baden’s testimony violate Fed.R.Evid. 704(b), which “prevents experts ' from expressly stating the final conclusion or inference as to the defendant’s' actual mental state.” United States v. Richard, 969 F.2d 849, 854 (10th Cir.1992); see also United States v. Orr, 68 F.3d 1247, 1252 (10th Cir.1995). 8

Dr. Baden first testified that ,“[m]y opinion would be the manner of [Dykes’s] death is homicide.” (IV R. at 443) (emphasis added). He had earlier defined homicide as occurring “when the death is caused at the hands of another person,” (IV R. at 441), and gave the example of “where a doctor, for one reason or another, has intentionally injected potassium to end a person’s life as a euthanasia type of situation,” (Id.) (emphasis added). He thus ruled out the other manners of death: natural, accident, suicide, or undetermined. Elaborating on this conclusion, Dr. Baden testified that a death classified as a homicide is distinct from an accident because the former .involves intentional action “that was reckless and that would predictably cause the death of the patient.” (IV R. at 483.) He then went on to make the following statement:

My opinion is that the injection of potassium, the way it was done, was not warranted from a clinical point of view, was extremely dangerous, was beyond reasonable medical treatment because it can cause death, and it did in fact cause death. It was reckless in the sense of not being a reasonable action on the part of the physician, which was fraught with the perils of causing death.

(Appellant’s Br. at 40) (citing VI R. 4 at 446) (emphasis added).

These statements “expressly draw the conclusion or inference” that Dr. Wood acted with the necessary mens rea when he caused Dykes’s death. Richard, 969 F.2d at 855. The district court instructed the jury that to convict Dr. Wood on charges of first or second-degree murder, it was required to find that he possessed the specific intent to cause harm or death to Dykes. 9 By his testimony that Dr. Wood’s actions caused Dykes’s death, Dykes’s death was a homicide, and a homicide involves the intentional taking of a person’s life, Dr. Baden expressly inferred that Dr. Wood acted with specific intent to kill Dykes.

Similarly, Dr. Baden’s testimony that Dr. Wood’s actions were reckless specifically describes the mens rea for involuntary manslaughter. As discussed, the mens rea for involuntary manslaughter is “without due care and circumspection,” 18 U.S.C. § 1112, or “gross negligence,” Bryant, 892 F.2d at 1470. In its instructions to the jury, the district court accurately defined “gross negligence” as “a wanton, reckless, indifferent or conscious disregard for human life or the safety of others.” (VIII R. at 1406) (emphasis added). This testimony by Dr. Baden, while not cast in precisely the same terminology as the statute, case law, or instruction, recites the critical components they identify. He states in no uncertain terms that Dr. Wood’s actions were “reckless” because they were “fraught with the perils of causing death.” (IV R. at 446.) This is substantively indistinguishable from the instruction that the mens rea for involuntary manslaughter is met if the defendant’s actions demonstrate a “reckless ... disregard for human life.... ” (VIII R. at 1406.) Therefore, Dr. Baden’s testimony does not merely provide “the facts or opinions from which the jury could conclude or infer the defendant had the requisite mental state.” Richard, 969 F.2d at 855. If believed, his testimony necessarily dictates the final conclusion that Dr. Wood possessed the requisite mens rea for involuntary manslaughter. See United States v. Morales, 108 F.3d 1031, 1037 (9th Cir. 1997) (“A prohibited ‘opinion or inference’ under Rule 704(b) is testimony from which it necessarily follows, if the testimony is credited, that the defendant did or did not possess the requisite mens rea.”) This intrusion into the province of the jury is precisely the sort of testimony Rule 704(b) is designed to prevent. See United States v. Dennison, 937 F.2d 559, 565 (10th Cir. 1991) (holding the trial court properly excluded testimony from an expert that a hypothetical person suffering from the same mental disorder as the defendant could not form the specific intent to commit assault); United States v. Windfelder, 790 F.2d 576, 582 (7th Cir.1986) (holding that it was error to admit into evidence an IRS agent’s testimony that the defendant intentionally understated his income). 10

D

Dr. Wood alleges that the district court also erred in admitting certain additional testimony: Dr. Knochel and Dr. Baden’s statements that KC1 is used to execute criminals; Dr. Baden’s statement that KC1 is the drug most commonly used to euthanize animals; and Dr. Bass’s statement that he had recurring nightmares of Dykes gasping before he died. He contends that this testimony should have been excluded because “its probative value is substantially outweighed by the danger of unfair prejudice, confusion of the jury, or misleading the jury....” Fed.R.Evid. 403.

The expert testimony concerning the use of KC1 in executions and the euthanasia of animals clearly prejudiced the defendant by creating the impression that KC1 is purely an instrument of death and Dr. Wood was acting as an executioner. While probative of KCl’s lethal effect, the dangers and benefits of KCL in a clinical setting were adequately established by the balance of the experts’ testimony. Accordingly, the district court should have excluded these statements.

Dr. Bass’s testimony that he had recurring nightmares of Dykes’s “last gasp” is more probative than prejudicial and therefore was properly admitted. Because the defense placed at issue Dr. Bass’s ability to accurately recall the details of Dykes’s death, the nightmare testimony was probative of why one event, occurring four years earlier, remained fresh in the witness’s memory. Furthermore, this testimony is not prejudicial. Dr. Wood argues that the prosecutor relied heavily on the “last gasp” during closing argument, even going so far as to reenact the event in an effort to shock the jury. These re-enactments were indeed egregious. Yet Dr. Wood does not object to Dr. Bass’s testimony that he heard a gasp, only to his testimony that he continues to hear it in his nightmares. The prosecution relied primarily on the former testimony, not the latter.

IV

Although the evidence taken in the light most favorable to the government is sufficient to sustain Dr. Wood’s conviction for involuntary manslaughter, and while the errors identified above might be harmless when viewed individually, under the particular circumstances of this case the erroneous denial of Dr. Wood’s first motion for judgment of acquittal on the charges of first and second-degree murder and the erroneous admission of testimony in violation of Fed.R.Evid. 403 and 704(b) constitute cumulative error requiring reversal.

A cumulative error analysis aggregates all the errors that individually might be harmless, “and it analyzes whether their cumulative effect on the outcome of the trial is such that collectively they can no longer be determined to be harmless.” United States v. Rivera, 900 F.2d 1462, 1470 (10th Cir.1990) (en banc). “The harmlessness of cumulative error is determined by conducting the same inquiry as for individual error — courts look to see whether the defendant’s substantial rights were affected.” Id. (citing United States v. Kartman, 417 F.2d 893, 894, 898 (9th Cir.1969)). Thus our cumulative error analysis must focus on “ ‘the underlying fairness of the trial.’ ” Id. at 1469 (quoting Delaware v. Van Arsdall, 475 U.S. 673, 681, 106 S.Ct. 1431, 89 L.Ed.2d 674 (1986) (citation omitted)). The standard for this analysis in cases of non-constitutional error is set forth in Kotteakos v. United States, 328 U.S. 750, 762, 66 S.Ct. 1239, 90 L.Ed. 1557 (1946): “Necessarily the character of the proceeding, what is at stake upon its outcome, and the relation of the error[s] asserted to casting the balance for decision on the case as a whole, are material factors in judgment.” The Kotteakos Court elaborated that “if one cannot say, with fair assurance, after pondering all that happened without stripping the erroneous action from the whole, that the judgment was not substantially swayed by the error[s], it is impossible to conclude that substantial rights were not affected.” Id. at 765. The applicable standard is thus different than that applied when determining if the evidence is sufficient to sustain a verdict: “The inquiry cannot be merely whether there was enough to support the result, apart from the phase effected by the error[s]. It is rather, even so, whether the error[s] [themselves] had substantial influence.” Id.

After examining the proceedings in their entirety, we conclude that Dr. Wood’s right to a fair trial was substantially impaired. By denying Dr. Wood’s motion for judgment of acquittal at the close of the government’s case-in-chief, the district court permitted the jury to consider the charges of first and second-degree murder. We recognize that a district court is permitted to reserve'ruling on a motion for judgment of acquittal until after the jury returns its verdict, see Fed. R.Crim.P. 29(b), 11 and that in this case the jury in fact acquitted Wood of those offenses regarding which his motion was improperly denied. Nonetheless, we cannot ignore that the jury, confronted with such serious charges, might have felt compelled to convict Dr. Wood of the lesser offense whether or not the government had proven each element of the crime beyond a reasonable doubt. See Millen, 594 F.2d at 1087. Such a compulsion could have been exacerbated by Dr. Baden’s inadmissible expert opinion that Dykes’s death was a homicide and that Dr. Wood possessed the requisite mens rea for the lesser offense, and the prejudicial comparison of Dr. Wood’s course-of-treatment with the execution of prisoners and euthanasia of animals. We cannot say with fair assurance that, given the profound seriousness of the charges the jury was asked to resolve, its judgment was not substantially swayed by these improper, inflammatory, and prejudicial statements. See Kotteakos, 328 U.S. at 762, 66 S.Ct. 1239. Accordingly, we reverse Dr. Wood’s conviction and remand the case for a new trial on the charge of involuntary manslaughter. 12

V

We REVERSE the judgment of the district court and REMAND for a new trial on the charge of involuntary manslaughter.

5 . Defendant argues that the district used the lower base offense level because of uncertainties raised by Dr. Lott's report. The district court, however, made no findings to that ef-feet. In any event, the sentencing guidelines are clear that the district court must impose the higher sentence.

1 . The district court granted Dr. Wood's motion for release and stay of fine and assessment pending appeal.

2 . On appeal, Dr. Wood also argues there was insufficient evidence that his actions caused Dykes’s death. The evidence, viewed in the light most favorable to the government, is sufficient to establish causation: Dykes died immediately after receiving the injection, and two government experts testified that Dykes died as a result of the KC1 injection.

3 . On direct examination, Nurse Kinsey stated that Dr. Wood injected 32 ccs of fluid (approximately 21.3 mEq of KC1) over the course of what she determined to have been 15-20 seconds. On cross-examination she conceded that on the day of the incident she noted that Dr. Wood had injected 25 ccs (approximately 16.7 mEq) and had told the FBI the dosage ■ was injected over a period of two or three minutes. Nurse Hardesty also testified on direct examination that Dr. Wood injected 25 ccs in a little more that 30-40 seconds, but on cross-examination she acknowledged that she had testified before the grand jury that only 20 ccs (approximately 13.3 mEq) had been used. Dr. Bass testified that he and the nurses later determined that Dr. Wood injected about 30 ccs (approximately 20 mEq) of KC1 over approximately 30 seconds, though he was unable to recall the duration of injection with certainty and it may have been as long as ten minutes. An FBI officer testified that Dr. Wood told him he had injected as much as 10 mEq of KC1 over seven minutes.

4 . Dykes's code status — the measures that should be taken were his condition to become critical' — was the subject of extensive testimony. Following Dykes's surgery, Dr. Atchinson made an initial determination that only limited resuscitative efforts should be taken because he felt Dykes would not survive his hospital stay. Dr. Wood modified this order to permit greater resuscitative efforts. On February 12, 1999, Dr. Morgan and Dr. Lytle ordered that no extraordinary efforts should be taken to resuscitate Dykes. Dr. Lytle withdrew that order the morning of the thirteenth because hospital policy required additional procedural steps before making such a determination for a patient like Dykes who lacked next-of-kin, not because of any change in Dykes’s condition.

5 . Nurse Kinsey testified that she believed Dr. Wood gave the potassium chloride to accommodate the Lasix and that Dykes's death was an "unexpected event” for Dr. Wood. (IV R. at 308-09.)

6 . Although we find that the evidence presented during the government’s case-in-chief was insufficient to prove premeditation or malice aforethought, Dr. Wood’s claim that the government wrongfully prosecuted the first-degree murder charge lacks merit. In Costello v. United States, 350 U.S. 359, 363, 76 S.Ct. 406, 100 L.Ed. 397 (1956), the Court held "[a]n indictment returned by a legally constituted and unbiased grand jury, ... if valid on its face, is enough to call for trial of the charges on the merits.” Dr. Wood has not alleged that the grand jury was biased or illegally constituted or that the indictment is facially invalid. Rather, he argues there is no evidence of premeditation to support the charge. The Court in Costello soundly rejected a similar argument, holding that indictments are not open to challenge "on the ground that they are not supported by adequate or competent evidence." Id. at 363-64, 76 S.Ct. 406.

7 . All of the cases cited by Dr. Wood in support of his proffered instruction involve civil malpractice claims and therefore are not controlling.

8 . Rule 704(b) states in full:

No expert witness testifying with respect to the mental state or the condition of a defendant in a criminal case may state an opinion or inference as to whether the defendant did or did not have the mental slate or condition constituting an element of the crime charged or of a defense thereto. Such ultimate issues are matters for the trier of fact alone.

9 . The instructions correctly clarified that, in the case of second-degree murder, the requisite intent can be established by evidence of conduct that is such a reckless, wanton, and gross deviation from the reasonable standard of care that a jury may infer the defendant was aware of a serious risk of death or serious bodily harm.

10 . This Circuit also prohibits experts from testifying as to ultimate issues of law in civil cases. See Specht v. Jensen, 853 F.2d 805 (10th Cir. 1988). Other circuits applying the same rule have excluded testimony in civil cases similar to that of Dr. Baden. See Woods v. Lecureux, 110 F.3d 1215, 1219-20 (6th Cir. 1997) (holding an expert could not testify that the defendant acted with deliberate indifference because that mental state was an element of the alleged statutory violation); Hygh v. Jacobs, 961 F.2d 359, 363-64 (2d Cir.1992) (holding an expert improperly testified to a legal conclusion when he described the defendant's actions as unjustified under the circumstances, unwarranted, and totally improper).

11 . Neither this nor any other circuit has decided, in a published opinion, whether a district court’s decision to reserve ruling on a motion for-judgment of acquittal is reviewable and, if it is, under what standard. Two unpublished opinions reach conflicting results. Compare United States v. Brown, 142 F.3d 446, No. 97-30082, 1998 WL 225042, at *1 (9th Cir. 1998) (holding that the decision to reserve ruling was not plain error because is was authorized by Fed.R.Crim.P. 29(b)), with United State v. Davis, 74 F.3d 1241, 1996 WL 15622, (6th Cir. Jan. 16, 1996) (holding a district court’s reservation is reviewed for abuse of discretion). We need not resolve this issue here because we find that the district court’s denial of Dr. Wood’s first motion for judgment of acquittal, combined with the eviden-tiary errors, requires reversal.

12 . We need not reach Dr. Wood's arguments that the district court committed plain error by instructing the jury on the greater offenses because the result, were we to find in his favor, would be the same.