Attorneys and their allied experts ventured deep into the weeds of the available medical literature on the drug midazolam today in federal court, where they are arguing for and against a lawsuit brought by Arkansas death row prisoners seeking to make the case that the state’s abbreviated execution schedule — which would see seven men executed this month — would violate the Eighth Amendment and their right to effective counsel. Midazolam is a sedative that is the first of a three-drug cocktail the state plans to use to execute seven inmates this month. Today is the final day of four days of hearings in the matter.

Read our coverage of day two and day three here.


UPDATE: Thursday evening, AP reported:

“Two pharmaceutical companies asked a federal judge Thursday to prevent Arkansas from using their drugs to execute seven inmates by the end of the month, saying they object to their products being used for capital punishment.”


The companies think the state obtained two drugs, including midazolam, outside approved supply chains.

In February, Governor Hutchinson scheduled double executions for each of four nights, April 17, 20, 24 and 27. A federal judge later stayed the execution of one of the men scheduled to die, Jason McGehee, after the state Parole Board recommended his clemency.


On the stand this morning was Dr. Joel Zivot, associate professor of anesthesiology and surgery at Emory University School of Medicine. Attorneys walked Zivot through a long list of television and print interviews in which the doctor appeared to talk about the death penalty and his public opposition to lethal injection. Zivot also said that he also isn’t in favor of physician assisted suicide, because “killing is not healing.”

Zivot said that, in his opinion, midazolam is an inappropriate drug for execution because it is unable to render an inmate deeply anesthetized to the point of being unable to feel pain. The drug wasn’t formulated to carry out executions, Zivot said. Under questioning by state’s attorney Jennifer Merritt, Zivot discussed the “black box warning” on the manufacturer’s label for midazolam. Zivot said that midazolam is sometimes used “as an adjunct” as the first of many medications used to produce general anesthesia and is used as a sedative that reduces anxiety and impairs memory of painful procedures like colonoscopies and intubations. Zivot and Merritt went back and forth with a series of question in which Merritt asked him repeatedly if midazolam could be used by itself to produce general anesthesia, with Zivot saying several times that he would never use the drug by itself to do that. Eventually, Zivot said, “You are asking me about bad medical practice.”

At the same time, Merritt tried to repeatedly corner Zivot on the idea that midazolam could be deadly as a drug by itself, referring to the manufacturer’s warning label that claimed the drug could cause potentially lethal respiratory depression. In a previous report submitted for the defense, Zivot had said midazolam is not a toxic, lethal drug. Asked if a person might die if he gave them a 2 to 3 milligram dose of the drug and then “walked away,” Zivot said he would never do that, and never allows a patient to be alone and unmonitored after he has administered medicine to them.

Zivot said that unresponsiveness in an anesthetized patient is impossible to verify with certainty, and that drugs affect people in different ways. “I have given a 110 pound man 10 milligrams of midazolam and nothing happened,” Zivot said at one point in questioning. He said that on a regular basis, even when using multiple drugs to achieve anesthesia, including intravenous lidocaine and a consciousness check consisting of touching the eyelashes, some surgical patients still react when they are intubated, a process that can be painful.


In his report, Zivot has said there is a risk of bacterial contamination as the drugs are drawn from the vial into the syringe. Told by the state’s attorney that because the inmate was to be executed, that wouldn’t make a difference if there was bacterial contamination, Zivot said he didn’t think Arkansas was contemplating injecting bacteria into inmates as part of the execution protocol.

On redirect, Zivot said that at the request of the defense, he had looked at the autopsy files of five Florida inmates executed using the identical drug protocol to be used by Arkansas. He said that while there is “the claim that lethal injection creates something akin to sleeping and then the person dies,” if that was the case, their organs should be pristine upon autopsy, showing no damage. In looking at the autopsies of the five executed Florida men, however, Zivot said he found fluid-filled “heavy lungs” in each. Because none of the men had complained of fluid in their lungs — which would have been uncomfortable or painful — prior to being executed, Zivot said the condition must have been somehow caused by the execution, but is “at a loss” to explain why. He said that while the paralytic vercuronium bromide creates the illusion of peace and calm as the inmate is put to death, he is “very worried … that these are really death by suffocation, or something akin to it.” If the midazolam fails to properly anesthetize the inmates, he said, they might have “an interior experience” of suffocation as they die.

Next up was Dr. Joseph Antognini, a California anesthesiologist who said he had anesthetized over 10,000 people during his nearly 30-year career. Antognini said he had previously testified in a case in which an elderly woman developed respiratory distress and later died after being giving midazolam in a California hospital. Since then, he said, he has provided reports on midazolam for several states, including Mississippi, Missouri, and Ohio, regarding the use of the drug in their execution protocol.

Antognini said it is his opinion that the 500 milligram dose of midazolam called for in the Arkansas protocol would render the inmates anesthetized and unable to sense pain. He said he believes the 500 milligram dose would render a person unable to feel pain for more than six hours, much more than is needed for the execution process. This would render inmates unable to perceive the effects of the vercuronium bromide and potassium chloride, he said.  
While he said the data is “pretty sparse” on whether midazolam can be used to produce general anesthesia in humans, he spoke of an animal study that shows the drug can achieve general anesthesia in mice, rendering them unable to move or lift their tails.

Asked about patients moving while anesthetized, Antognini said patients regularly move in the operating room when completely unconscious, which is why they are sometimes strapped down during surgery. He spoke of even brain dead patients sitting up, turning their heads or exhibiting what’s called the “Lazarus Effect” in which they cross their arms over their chests. During the harvesting of organs from the brain dead, he said, patients with no brain activity are often given anesthesia because their blood pressure can spike during the procedure — a response to the pain. He would later say that an inmate coughing or choking during the execution procedure wouldn’t indicate the inmate was conscious or feeling pain.

Antognini said the “ceiling effect” of midazolam, discussed in previous days’ testimony as a strike against the drug for use in an execution, is not relevant and is “a distraction” — a “confounder” introduced by death penalty opponents that “grew legs.” Antognini said he has personally used midazolam to induce general anesthesia while inserting breathing tubes in people’s throats.

On cross, Antognini said that he is being paid $4,000 for the day he testified in the case, and paid $1,600 per day for those days he only observed the proceedings. After a break for lunch, inmate attorney Julie Vandiver took Antognini laboriously through several scientific papers he’d used to draw his conclusions about midazolam and his belief that it can be used as a general anesthetic, including two Japanese studies, one dealing with the use of midazolam during c-section deliveries, and another about injecting the drug into the spine. Vandiver took pains to note that the circumstances in which midazolam is discussed in those papers — always as a therapeutic drug, never as an execution agent — differ from Arkansas’s midazolam-included execution protocol.    

Asked why he believes the state uses the paralytic vecuronium bromide as one of the three drugs in their execution protocol, Antognini said, “Why do they use four to five bullets in a firing squad?” a reference to the ability of vecuronium bromide to paralyze the diaphragm and potentially kill by itself through suffocation. Vandiver asked if the paralytic was unnecessary, given that inmates are strapped down, and Antognini said he had no opinion on whether it was or not, but later added that he doesn’t agree that because midazolam is a dangerous drug, there is no need for the vecuronium bromide. Of midazolam, Antognini said that if one of the inmates was rushed to a hospital for an emergency appendectomy, midazolam could be a drug used to induce general anesthesia for the operation. Because of that, he said, he doesn’t understand why people are concerned about whether the drug will sufficiently produce anesthesia for that same inmate’s execution.

After Antognini, Dr. Craig Stevens returned to the stand and defended his previous testimony, including the idea that midazolam has a “ceiling effect” beyond which a larger dose will produce no more sedation. “I seem to be the only one actually talking about the way this drug works,” Stevens said.

Stevens said that benzodiazopines, a family of drugs of which midazolam is a member, are much safer than barbiturates. He said that when Dr. Daniel Buffington had previously testified that benzodiazopines have a similar respiratory depressive effect as opioids, “I had an internal gasp at that. … He couldn’t be more wrong.” Stevens said that benzodiazepine deaths are “quite rare.”

Stevens reiterated his belief that 500 milligrams of midazolam will not render the inmates insensate to pain, saying the drug is not a good substitute for the fast-acting barbiturates used in previous execution protocol. “Using more of a drug does not change the nature of a drug,” he said, and said that it’s his belief the drug can’t produce general anesthesia.

Inmate attorneys then called Dr. Zivot back to the stand, where he talked about the rare phenomenon of “awareness” during anesthesia — being able to sense and remember things happening to your body during surgery, including hearing sounds and feeling pain. He said that the experience, which sometimes happens to people given paralytics and muscle relaxants prior to operations, can be “extremely traumatic” and “a terrifying experience” to patients, sometimes spurring long-term PTSD.  “If one adds choking to the experience of being paralyzed and awake, that would be a truly terrifying experience,” Zivot said.

ADC Director Wendy Kelley came to the stand as the hour drew down toward 5 p.m. On the stand, Kelley — who also testified last night — said she had been involved in practices for the upcoming executions. She said that she was involved in scheduling the executions, with preparations beginning after ADC officials were “asked by the governor’s office if we could get these done before the drugs expired.” The governor made the final decision on scheduling, she said. Attorneys for the state asked Kelley about a current poll which found that 61 percent of Arkansas respondents support the death penalty, 41 percent support lethal injection, and 38 percent support public hanging. Asked if ADC has a facility set up to perform hangings, or whether that would be legal under current law, Kelley said no.

Kelley, who joined the ADC in 2006, was questioned about being sued in previous lawsuits by death row inmates, including a 2015 lawsuit over using compounded drugs for lethal injection. Over the course of almost an hour, she was questioned line by line about the ADC protocol from 2008 and how it differs from the revised, midazolam-included protocol developed more recently. Asked at one point why ADC officials had deleted a requirement for there to be a heart monitor attached to the inmate in the death chamber, Kelley said they had heard about a case where someone had bumped a gurney on which a dead inmate was laying, which caused that state’s heart monitor to beep. That could make it mistakenly appear the inmate might have a heartbeat, Kelley said. Their former monitor was “an antiquated machine” so after they heard about that case, they removed the requirement for a heart monitor from the protocols.  

Asked later about news accounts that the ADC was having trouble finding enough volunteers to view the lethal injections this month — by law, at least six citizen witnesses must view every execution — Kelley said they will have enough witnesses present for each of the seven executions scheduled.

Judge Kristine Baker had previously said the hearing will conclude by 8:15 p.m. tonight.