NEW YORK — Dennis McGuire clearly knew something was wrong. At 10:34 a.m. on Jan. 16, as a crowd at the Southern Ohio Correctional Facility looked on, the convicted murderer began gasping for air. Then McGuire began to make snorting and choking sounds. For the next 10 minutes, as a combination of midazolam (a relaxant similar to Valium) and hydromorphone (an analgesic related to morphine) coursed through his veins, McGuire’s chest and stomach heaved as the oxygen in his blood dwindled. Death was approaching, but slowly.

Watching a man gradually suffocate may have come as a surprise to some people in the gallery, but it didn’t surprise David Waisel, an associate professor of anesthesia at Harvard Medical School, who had predicted this would happen. Ten days earlier Waisel had presented U.S. district court judge Gregory Frost with a nine-page declaration explaining that the state of Ohio planned to use an improper dose of midazolam — a short-acting benzodiazepine that’s often used to induce sedation and amnesia before a medical procedure — to kill McGuire. “In light of the insufficient dose of midazolam,” Waisel wrote, “it is substantially likely that McGuire will be aware of this agony and horror.” Based on his expertise, he felt there was a “substantial, palpable, objectively intolerable risk of experiencing the agony and horrifying sensation of unrelenting air hunger” during the execution, suggesting that “McGuire will remain awake and actively conscious for up to five minutes, during which he will increasingly experience air hunger as the drugs suppress his ability to breathe.” It turns out Waisel may have undershot things; Dennis McGuire took nearly 30 minutes to die.

When I spoke with Waisel about his testimony, he explained that he had used a standard, simple set of criteria called STOP-Bang to determine that McGuire’s airway would likely become obstructed shortly after the medications were administered, causing him to slowly suffocate. He would die, certainly, but not in the manner intended. Medications do different things at different doses, and the amount of midazolam that McGuire received caused his throat to partially close, as though his body was slowly strangling itself from the inside, rather than causing him to drift off to sleep.

In a matter of minutes, it turns out, a physician with even minimal information (gender, neck size, blood pressure) can determine whether an inmate sentenced to death is likely to suffer. The problem, of course, is that the state is not compelled to listen to the physician. The other issue is that the American Medical Association’s code of ethics bars members from participating in executions. This creates a troubling paradox: The people most knowledgeable about the process of lethal injection — doctors, particularly anesthesiologists — are often reluctant or unable to impart their insights and skills.

Indeed, most of the anesthesiologists I spoke with declined to comment on the record about lethal injection. One professor of anesthesiology wrote to me, “Obviously it’s a sensitive and complicated subject and I suspect you will find few anesthesiologists who will want to be interviewed about it, even those who support the death penalty. Naturally, our profession does not want to see an erosion of confidence in general anesthesia if there is a public association with capital punishment. Many of us could provide you with optimal technical details on how to more efficiently kill someone, but I don’t think we should be doing it at all, particularly given the flaws in the justice system.”

‘Incompetent’ executioners

Without an expert in the room, states often rely on executioners who don’t really know what they’re doing. As one anesthesiologist told me, “the executioners are fundamentally incompetent. They have neither the technical skill nor the cognitive ability to do this properly.” Another added, “In medicine, the burden of proof is on the doctor to show that something is safe. We would never give a new drug to a patient until it’s been tested, approved by the FDA, etc. With the death penalty, the burden of proof has been inverted. These compounds, which are clearly causing patients to suffer, are deemed safe until proven otherwise. Yet the department of corrections prevents the release of information pertaining to how the lethal injection is carried out, making it impossible for a lawyer to make a strong case that this method is cruel and unusual.” Georgia is in fact working on a Lethal Injection Secrecy Act.

As our understanding of cruelty continues to evolve — let’s not forget that drawing and quartering was once an acceptable method of execution — future generations may wonder why lethal injection was performed so poorly and carelessly, and with so little oversight. Part of the problem is the terminology: Words like injection and cocktail and gurney give the illusion that this form of capital punishment is civil. This allows, regrettably, for a softening of the perception of what is actually happening: Medications that were designed to heal have been repurposed to kill.

And it’s not just the wrong doses — it’s the wrong drugs. A professor of anesthesiology at a large academic medical center said, “We have the drugs to do it in a way that doesn’t cause suffering. I read the doses they were using and thought, ‘That’s not enough! Who is coming up with this? Whoever did certainly doesn’t do this for a living.’ You need two components for lethal injection: amnesia and analgesia. This ensures the person is not aware and not in pain. Drugs like potassium chloride and pancuronium (a paralytic) — the drugs approved by the Supreme Court — are unnecessary. When they euthanize a dog, they don’t use potassium or a paralytic. You don’t even need an anesthesiologist! Any physician could look up the proper dosing in a textbook.”

While I was researching this piece and discussing with friends the nuances of optimizing lethal injection, a number of them stopped me midsentence and asked, “Who cares?” Should it be our concern that a monster may have experienced profound discomfort in his or her final minutes? Recounting precisely what happened to Dennis McGuire — who was convicted of the 1989 rape and murder of 22-year-old Joy Stewart, who was about 30 weeks pregnant at the time — led some to express the hope that he did suffer. But regardless of your stance on the death penalty, the story of McGuire’s slow asphyxiation should lead you to wonder whether it violated our Constitution’s ban on cruel and unusual punishment.

The Supreme Court has spoken. In the 2008 case Baze v. Rees, the court ruled that the cocktail used in Kentucky — sodium thiopental (an amnestic), pancuronium bromide (a paralytic), and potassium chloride (designed to stop the heart) — was not in violation of the Eighth Amendment. So despite what you read about inmates suffering — Florida convict Angel Diaz took 34 agonizing minutes to die after executioners mistakenly inserted needles into his flesh instead of his veins — the United States considers lethal injection in its current form neither cruel nor unusual. But Deborah Denno, a law professor at Fordham University and an expert on lethal injections, recently told me that “the court’s ruling was based in part on the uniformity of drug combinations across the states.” But as the drugs have become less available, that’s no longer the case. “This is a very different world in 2014,” she said, “than it was in 2008.”

European ban

Many of the most effective drugs — including propofol, which contributed to Michael Jackson’s death — are made at compounding pharmacies in Europe, and the manufacturers are threatening to cease exportation of their products to the United States if they are used for lethal injection, citing the European Union’s ban on the death penalty. This roadblock has led some states to halt executions to consider their options or to search for other ways to end their inmates’ lives. But not all are. Florida executed someone just last week.

Other states are being more creative, trying to obtain lethal drugs through clandestine backchannels, a practice that has called into question the purity of the medications being used. On March 26 a judge in Oklahoma ruled that the state could no longer keep secret the identities of the suppliers of lethal injection drugs, acknowledging that without knowing the source of the drugs, prisoners were denied the right to know, and potentially question, how they would be put to death.

As prisoners continue to die with varying degrees of suffering and companies restrict the use of their compounds, the lethal injection cocktail will continue to evolve, leading some to wonder if this slapdash mixing and matching of chemicals constitutes impermissible research on inmates. Seema Shah, a lawyer and bioethicist at the National Institutes of Health, pointed out in her 2008 study that gathering systematic data about lethal injection would require setting criteria for relevant evidence, measuring variables that may make the condemned more or less likely to experience a quick and painless death, and using statistical methods to determine the relevant number of test subjects. Put simply, to adequately reform lethal injection, it has to be studied — but given our country’s troubled history of experimenting on inmates, doing so is ethically and legally problematic.

Should we let physicians help carry out capital punishment? As a doctor, I believe the answer is no.

From 1913 to 1951, Leo Stanley, chief surgeon at San Quentin, performed a wide variety of experiments involving testicular implants. He would take the testicles out of executed prisoners (or goats) and surgically implant them into living prisoners. During the 1950s more than 100 inmates in the Ohio state prison system were injected with live cancer cells so researchers could study how the human body reacted to them. A decade later dozens of pharmaceutical companies tested scores of experimental drugs at Holmesburg Prison in Philadelphia. It may seem hard to believe, but as Shah writes, prior to the early 1970s, “approximately 90 percent of all pharmaceutical research was conducted on prisoners.”

As news of this reprehensible practice spread, several states and the Federal Bureau of Prisons took the drastic step of banning research on prisoners altogether, citing concerns of “exploitation, secrecy, danger and the impossibility of obtaining informed consent.” Some state laws have very broad definitions of research and may prohibit any “health-related experimental procedure.” It’s unclear if lethal injection could be successfully challenged using this precedent, but legal groups are certainly considering it.

Return to the firing squad?

Another option, which may be gaining traction, was proposed in a column for USA Today by Joel Zivot, an assistant professor of anesthesiology and surgery at Emory. He called for a moratorium on the use of all anesthetic agents for lethal injection. While that stance wasn’t particularly controversial, the line that followed was: “If the state is inclined to execute, it might be the time again to take up hanging, the electric chair, or the bullet.” At first blush I thought he was being facetious, but after speaking with him, I think he may be on to something.

Firing squads have traditionally consisted of five volunteer marksmen, four of whom receive rifles with live rounds while a fifth rifle contains a blank so no single member can know that he shot a fatal bullet. The marksmen are hidden from both the inmate, who is blindfolded and strapped to a chair, and from the witnesses. Using a stethoscope, a state official — in Utah it was a physician — finds the precise location of the prisoner’s heart and places a bull’s-eye over it. After a signal from the warden, the marksmen shoot.

Whether this is the least painful option is hard to say. In one of the few documented experiments associated with shooting, a man in Utah allowed doctors to conduct an EKG during his execution. The EKG showed that the heart was beating nearly three times its normal rate before shots were fired, and when the bullets entered his heart, it went into a four-second spasm, followed by a 15-second interval of uniform electrical activity that included a second spasm. Complete electrical silence followed.

A compelling case can be made that based on efficacy, diffusion of responsibility, and inexpensiveness, death by firing squad is a better option. (Or perhaps the guillotine.) Some organs would remain intact for donation, and although it might appear grisly, it’s quick, and it is the only method of execution for which we already train people. Interestingly, in states that have offered both shooting and hanging — which also fulfills many of the above criteria — inmates usually opt for the firing squad. One could argue that if properly done, lethal injection would be more humane than either of these methods, but we can no longer expect that it will be properly done.

McCarthy is a writer and physician based in New York. He writes the Medspin column for Deadspin.