Allen Nicklasson has had a temporary reprieve. Scheduled to be executed by lethal injection in Missouri on 23 October, the convicted killer was given a stay of execution by the state’s governor, Jay Nixon, on 11 October — but not because his guilt was in doubt. Nicklasson will live a while longer because one of the drugs that was supposed to be used in his execution — a widely used anesthetic called propofol — is at the center of an international controversy that threatens millions of US patients, and affects the way that US states execute inmates.

Shortages of anesthetic drugs usually used in lethal injection, the most common method of execution, are forcing states to find alternative sedatives. Propofol, used up to 50 million times a year in US surgical procedures, has never been used in an execution. If the execution had gone ahead, US hospitals could have lost access to the drug because 90% of the US supply is made and exported by a German company subject to European Union (EU) regulations that restrict the export of medicines and devices that could be used for capital punishment or torture. Fearing a ban on propofol sales to the United States, in 2012 the drug’s manufacturer, Fresenius Kabi in Bad Homburg, ordered its US distributors not to provide the drug to prisons.

This is not the first time that the EU’s anti-death-penalty stance has affected the US supply of anesthetics. Since 2011, a popular sedative called sodium thiopental has been unavailable in the United States. The manufacturer, US company Hospira, abandoned plans to produce the drug at its plant in Italy after regulators in the country required that the thiopental never be used in executions. The drug, which is difficult and costly to make, was already in short supply because of manufacturing problems.

“There has been a collision of the politics of capital punishment in the United States and Europe, forcing us to hopscotch around looking for suitable methods for anesthesia,” says Jerry Cohen, a former president of the American Society of Anesthesiology.

“The European Union is serious,” says David Lubarsky, head of the anesthesiology department at the University of Miami Miller School of Medicine in Florida. “They’ve already shown that with thiopental. If we go down this road with propofol, a lot of good people who need anesthesia are going to be harmed.”

The loss of thiopental from the anesthesia arsenal was a relatively minor inconvenience, says Cohen, because propofol provided an alternative. But if propofol is used for executions in Missouri or any other state, it could disappear too, leaving hospitals in a serious bind. “Propofol has a lot of uses for which there are no substitutes,” says Cohen. It is the preferred way to sedate people who have breathing tubes because it acts quickly and does not cause vomiting. Federal regulations make propofol difficult to manufacture in the United States.

The 35 US states with prisoners on death row were already scrambling to find effective drugs for lethal injection, which was used for 43 executions last year. The procedure previously relied on a course of three injections: thiopental to sedate the prisoner, muscle relaxant pancuronium bromide to induce paralysis, and potassium chloride to stop the heart. As supplies of thiopental ran low in 2009 and 2010, many states started stockpiling pentobarbital, another sedative. But in 2011, Lundbeck, a drug company in Copenhagen and sole US supplier of pentobarbital, banned it from use in executions because of Danish and EU human-rights laws. Texas’s supply of pentobarbital expired in September, but the state obtained more from unregulated compounding pharmacies, which tailor-make drugs. Pentobarbital is not “especially” useful as a surgical anesthetic, says Lubarsky, so its shortage has little impact on patient care.

On 15 October, after running out of pentobarbital, Florida executed William Happ using midazolam as the sedative. But midazolam, which is similar to diazepam (Valium), had never been used in an execution, and, according to media reports, Happ was still blinking and moving his head minutes after the injection.

Nobody knows whether midazolam is appropriate for lethal injections, says Lubarsky. “We’ve turned this into a circus of experimenting on prisoners,” he says. “The state is playing doctor without any regard for efficacy. It changes protocols willy-nilly.” The drug is not a good anesthetic, he says, and it may not shield prisoners from the pain of the final injection.

Although midazolam has now entered the realm of capital punishment, it is unlikely that surgical supplies will be affected. Hospira is one of many companies that makes midazolam and has no plans to stop, says Dan Rosenberg, a company spokesman. Rosenberg would not say where Hospira makes midazolam, but he says that European regulations “aren’t an issue”.

Meanwhile, Missouri has suspended another execution, scheduled for 20 November, while it tries to find an alternative to propofol. Lubarsky notes that although a single, large dose of propofol could work as a method of execution, its use in US prisons would be problematic because it could be complex to administer and physicians are generally not willing to participate in the process (see Nature 441, 8–9; 2006). “Putting together a foolproof protocol that could be carried out by prison guards with high-school educations is another matter entirely,” he says.

This article is reproduced with permission from the magazine Nature. The article was first published on October 22, 2013.