A shortage of sodium thiopental, a fast-acting barbiturate and general anesthetic used in lethal injections of death-row convicts, has delayed several such executions throughout the U.S. and reignited a long-standing debate over the combination of chemicals used to carry out capital punishment. Most recently, Arizona inmate Jeffrey Landrigan was executed Tuesday night only after a delay caused by a legal battle over the source and quality of the sodium thiopental used as part of the lethal injection.

Lethal injection is used for capital punishment by the federal government and 36 States, at least 30 of which use the same combination of three drugs: sodium thiopental (a barbiturate to induce anesthesia), pancuronium bromide (a muscle relaxant that paralyzes all the muscles of the body) and potassium chloride (a salt that speeds the heart until it stops). This protocol was developed in 1977 for the state of Oklahoma by then–Chief Medical Examiner Jay Chapman, but it has never been codified or sanctioned by the U.S. Food and Drug Administration (FDA).

The only U.S. maker of sodium thiopental, Hospira, Inc., in Lake Forest, Ill., has reported a shortage of the raw materials needed to make the drug. When Arizona law enforcement last week declined to identify its source of the sodium thiopental intended for Landrigan, his lawyers seized on the opportunity. On October 21, they requested a stay of execution, contending that their client could suffocate painfully if the anesthetic comes from an unknown source and did not work properly before the pancuronium bromide and potassium chloride kicked in. U.S. District Judge Roslyn Silver initially blocked the execution, but she was overruled Tuesday by the U.S. Supreme Court, which voted 5-4 that a condemned prisoner does not need to know the source of the drugs used in his execution.

The Superior Court of Maricopa County, Ariz., originally sentenced Landrigan to death in 1990 after convicting him of first-degree felony murder in the strangulation and stabbing death of Chester Dean Dyer of Phoenix. At the time of the murder Landrigan was an escapee from an Oklahoma prison, where he was serving time following a conviction in that state of assault and battery with a deadly weapon, second-degree murder, and possession of marijuana.

Regardless of whether Landrigan's legal team was simply using the drug shortage as stalling tactic, their legal maneuvering brings to the fore a contentious dispute over the science (or some would say lack thereof) behind lethal injection executions in the U.S. For more than two decades, it has been argued that the FDA should be required to certify the safety and effectiveness of drugs used to carry out executions (as it does for drugs used to euthanize animals). The FDA, wanting to stay out of the capital punishment debate, disagrees.

In 2008 the U.S. Supreme Court (pdf) upheld a lower court ruling that the state of Kentucky's three-drug method of lethal injection did not constitute "cruel and unusual punishment," as defined by the Eighth Amendment. Some scientists disagree. Scientific American spoke with University of Miami Miller School of Medicine molecular biologist Teresa Zimmers about this controversial topic.

[An edited transcript of the interview follows.]

You and a group of colleagues in 2007 published a report in PLoS Medicine that examined public records of executions. What was the purpose, and what sort of reaction did you receive?
We were actually trying to look at whether there was any evidence that the three-drug protocol—sodium thiopental, pancuronium bromide and potassium chloride—acted in the way it was supposed to act. We analyzed the time to death or the time to different events, such as cardiac arrest, in order to understand what might be the mechanism of death. We found no evidence to support the use of this protocol, the dosage of the drugs or the order in which the drugs were administered in executions.

A lot of responses to the study were negative—people assumed that we had a specific political agenda. This began to change as people looked at our data more closely. We were invited to the March 2008 Fordham Urban Law Journal Symposium, "The Lethal Injection Debate: Law and Science," by Fordham University School of Law professor Deborah Denno to represent the medical and scientific aspects of it. Supreme Court Justice Stephen Breyer also used our studies as part of his research for Baze v. Rees in 2008, which upheld an earlier ruling in Kentucky that the state's approach to administering lethal injections does not violate the "cruel and unusual punishments" ban promised in the Eighth Amendment. Breyer's research—our paper—did not uphold the constitutionality, but rather argued for the likelihood of unrecognized pain and suffering.

What concerned you and your colleagues about the way lethal injections are administered?
There's no record of a medical or scientific inquiry into whether this would be the best method. And there isn't any medical evidence to support this approach. Part of the paradox is that it looks like a medical procedure, but it hasn't been rigorously tested. There are no controlled trials, data collection, analysis or peer review of the processes to determine whether it works the way it's been said to work.

Why is sodium thiopental used as part of a lethal injection execution?
Sodium thiopental was chosen to render the person deeply unconscious and unable to feel the paralysis brought on by the pancuronium bromide, which causes the person to lose the ability to breathe. And the potassium chloride is extremely painful. Some people have said that three to five grams of sodium thiopental alone should be enough to induce death. [In December 2009 Ohio became the first state to use a single dose of sodium thiopental to execute death-row inmates.] We looked at whether inmates died reliably after the sodium thiopental, and it's not clear this is the case. We also determined that the doses of sodium thiopental used are not always as "massive" as claimed. It's not even clear how much a massive dose is in this context. We found that, at most, the highest doses were two times the lethal dose for animals, regardless of the inmate's weight.

It has been reported that in addition to a shortage of sodium thiopental, the doses that some states stockpile are set to expire before scheduled executions can be carried out. What sort of shelf life does sodium thiopental have?
Sodium thiopental has quite a long shelf life—up to 48 months in its unconstituted form. Once you add liquid, it's been reported to be stable for 24 hours or, if it's kept cold, it can last for seven days. They typically prepare it on the day of execution. Shelf life may be a problem because states perform executions infrequently and now don't have a supply of new doses.