dissenting,
with whom MURPHY and KELLY, Circuit Judges, join.I would grant a stay of Earl Ringo’s execution. I therefore respectfully dissent from the order denying a stay.
*881The Eighth Amendment prohibits a state from proceeding with an execution at a time when a prisoner is incompetent. See Ford v. Wainwright, 477 U.S. 399, 410, 106 S.Ct. 2595, 91 L.Ed.2d 335 (1986); see also Panetti v. Quarterman, 551 U.S. 930, 957, 127 S.Ct. 2842, 168 L.Ed.2d 662 (2007) (explaining that Ford indicates the Eighth Amendment forbids the execution of a prisoner whose incompetence “prevents him from comprehending the reasons for the penalty or its implications [or who is] unaware of the punishment they are about to suffer and why they are to suffer it.”) (internal quotations and citations omitted).
Recent revelations, disclosed for the first time in the last few days,1 indicate Missouri has been intravenously injecting large doses of the drug midazolam into its death row inmates before the time at which each inmate’s death warrant becomes valid.2
In several of these executions (namely, Ferguson’s, Middleton’s, and Worthing-ton’s), the intravenous injection of these large doses of midazolam had been preceded by use of the sedative Valium in 5 mg doses a few hours before the execution. In addition, the recent revelations show that Missouri’s intravenous injections of the drug midazolam have escalated over time, beginning with 2 mg doses in the first five executions noted above, increasing to 3 mg doses in the executions of Rousan and Winfield, and increasing to a total of 6 mg in each of the most recent July and August executions of Middleton and Worthington.
The unusually large doses of midazolam Missouri has intravenously injected into inmates in its last four executions — just minutes prior to the time when the death warrants become effective — is alarming with respect to the constitutional prohibition against executing a prisoner in a state *882of incompetency. Uncontradicted medical testimony from Dr. Mark Heath, a board certified anesthesiologist, indicates that intravenous doses of 3 mg of midazolam are equivalent to oral doses of four times that amount. “A dose of three milligrams of intravenous midazolam is likely to place a typical male adult in a substantially intoxicated state characterized by marked compromise of the intellect, marked reduction in cognitive capacity, profound impairment of memory functionality, and deeply impaired judgment.” Heath Declaration dated September 7, 2014, at ¶ 13. “After receiving 3 mg of IV midazolam, it would likely be difficult for most people to speak coherently or articulate anything beyond a short sentence or simple thought. Any spoken or written statement made by a person to whom 3 mg IV midazolam was recently (within the past hour) administered would necessarily be considered unreliable and not reflective of their otherwise normal state of mind.” Id. at ¶ 15.
In addition, in those cases where Missouri administers a dose of the sedative Valium prior to the intravenous injection of high doses of midazolam (such as in the cases of Ferguson, Middleton, and Wor-thington), an inmate’s cognitive impairment increases because both Valium and midazolam “share the same mechanism of action, and accordingly the impairment produced by their combination would be additive.” Id. at ¶ 16.
The use of 6 mg doses of midazolam before an execution commences is even more alarming with respect to Ford’s prohibition against executing prisoners while in a state of incompetency. The uncontra-dicted medical evidence indicates the following:
The administration of 6 mg of intravenous midazolam, over a period of less than 30 minutes, would place the vast majority of normal adult patients or persons into a state of severe cognitive impairment. Such a person would be extremely unlikely to be capable of meaningful or accurate comprehension of the nature of their circumstance. In the specific case of an execution, a person who had received 6 mg of intravenous midazolam would almost certainly be unable to understand that they were about to be killed, that they were being killed as a lawful infliction of punishment for an action that they had committed, or that the act that they committed was a violation of the law. It is very likely that a person who had received 6 mg of intravenous midazolam would be so stuporous that virtually all ideation and mentation would have ceased. If such a person were to be conscious and mentating, it is highly likely that their thought processes and executive function would be substantially deranged and distorted.
Id. at ¶ 18.
As stated above, Missouri has presented no medical evidence to contradict the medical evidence presented by Ringo. Instead, Missouri relies upon the selfserving statement of its prison warden, Terry Russell, who claims “each offender since November 2013[] appeared awake and alert after [midazolam] was given.” Russell Affidavit dated September 5, 2014, at ¶ 13. Warden Russell is not a physician, however, and his casual observations of an inmate’s outward appearance have little bearing on whether a specific medication has impaired an inmate’s cognitive function to the point of being considered incompetent under Ford. Indeed, Dr. Heath specifically testified as follows:
A person who is significantly cognitively impaired by an intoxicant such as alcohol or a sedative, on superficial observation, may appear to be behaving normally. The degree of impairment is often *883only revealed and apparent by pointed/directed questioning and performance of tasks that are known to elicit evidence of impairment. As an example, this is why police officers are trained to deploy a specific suite of roadside tests to detect drivers who are intoxicated by alcohol or other substances. Midazolam is one of the many substances and drugs that cause an intoxication that may not be easily discerned and is only evident when specifically probed. Accordingly, a person, patient, or prisoner who is intoxicated and cognitively impaired by midazolam may appear to a casual observer, or to an observer who is not in close proximity, to have no noticeable signs of cognitive compromise.
Heath Declaration at ¶ 22.
Dr. Heath summarizes by stating it is his opinion “to a high degree of medical certainty, that [Missouri] has administered intravenous midazolam to condemned prisoners in doses that are almost certain to produce substantial impairment of cognitive faculties.” Id. at ¶ 23.
The above summary reflects the state of the medical evidence before our court. Given the state of this record, I conclude the standard for granting a stay of execution has been satisfied in this case. Mr. Ringo has made a substantial threshold showing he will be incompetent to be executed at the moment it matters, and that Missouri is the active agent in causing the incompetence through its own conduct. Missouri has not presented evidence to indicate it intends to change its pattern and practice of intravenously injecting Mr. Ringo with midazolam in the minutes before his execution warrant becomes effective, as Missouri has done in every execution carried out in the past nine successive executions.
Ringo further argues Missouri’s practice of using such large doses of midazolam in the minutes prior to an execution amounts to a de facto use of the drug as part of its actual execution protocol. I agree. I believe these new revelations need to be fully aired in court before Missouri continues executing inmates under its current practice.
Use of the drug midazolam as the actual lethal agent in states such as Ohio, Oklahoma, and Arizona has resulted in closer scrutiny of the execution protocols in those states following executions that have gone awry. All the while, Missouri has steadfastly maintained its execution protocol makes no use of the controversial drug. Yet the 6 mg doses of midazolam Missouri has used in the minutes prior to the executions of John Middleton and Michael Wor-thington approach the 10 mg dose of mi-dazolam Ohio used as the actual lethal agent to execute Dennis McGuire. Missouri’s use of such large doses of midazo-lam, just minutes prior to an execution, indicate Missouri’s claim that the drug is not part of its actual execution protocol should be viewed with a healthy dose of judicial skepticism.
Missouri has taken great pains to distance itself from the use of midazolam during executions, but without plausibly explaining how administering large doses of the drug just minutes before a death warrant takes effect is meaningfully different from using the controversial drug as part of its actual execution protocol. The explanation that the drug is merely a pre-execution sedative loses its meaning when the drug is administered at the discretion of the state instead of at the inmate’s request, and when the effects of these last minute intravenous injections last into the execution itself. In light of Missouri’s checkered past regarding its implementation of the death penalty, which I have noted before, I believe the legitimate concerns now raised about Missouri’s current *884practices need a full examination before— not after — Mr. Ringo is executed.
I respectfully dissent.
. Missouri had previously disclosed the use of midazolam as a pre-execution sedative, but also represented such use typically occurred several hours before an execution and generally at the inmate’s own request. What is new about the recent revelations is fourfold: (1) use of such large doses of midazolam— which serves as the actual lethal agent in several other states (including Arizona, Oklahoma, and Ohio); (2) injection of large doses just minutes prior to the execution; (3) use of the drug intravenously rather than orally; and (4) injections of the drug apparently exclusively at the state’s discretion rather than at the inmate's request.
. Specifically, in Joseph Franklin’s execution on November 20, 2013, Missouri intravenously injected 2 mg of midazolam into Franklin sixteen minutes before the start of his execution. In Allen Nicklasson’s execution on December 11, 2013, Missouri intravenously injected 2 mg of midazolam into Nicklasson at an undisclosed time before the start of his execution. In Herbert Smulls’s execution on January 29, 2014, Missouri intravenously injected 2 mg of midazolam into Smulls twenty minutes before the start of his execution. In Michael Taylor’s execution on February 26, 2014, Missouri intravenously injected 2 mg of midazolam into Taylor twenty minutes before the start of his execution. In Jeffrey Ferguson’s execution on March 26, 2014, Missouri intravenously injected 2 mg of midazolam into Ferguson thirty-three minutes before the start of his execution. In William Rousan’s execution on April 23, 2014, Missouri intravenously injected 3 mg of midazolam into Rou-san thirty-two minutes before the start of his execution. In John Winfield’s execution on June 18, 2014, Missouri intravenously injected 3 mg of midazolam into Winfield thirty-four minutes prior to the start of his execution. In John Middleton's execution on July 16, 2014, Missouri intravenously injected 3 mg of midazolam into Middleton twenty-six minutes prior to the start of his execution, and then another 3 mg of midazolam three minutes prior to the start of his execution. In Michael Worthington's execution on August 6, 2014, Missouri intravenously injected 3 mg of midazolam into Worthington thirty minutes prior to the start of his execution, and then another 3 mg six minutes prior to the start of his execution. See Exhibit A attached to Russell Affidavit dated September 5, 2014.