Four Arkansas Executions Are Tied to the Expiration Date of a Drug That Does Not Work in Lethal Injections

An October 2012 photograph of the lethal injection room at San Quentin State Prison. Arkansas is executing several death row inmates timed with the pending expiration date of midazolam, a component of the lethal injection. Crime Law Society/Reuters

According to Arkansas Governor Asa Hutchinson, the lives of several death row inmates hinge on chemistry. Earlier this year, Hutchinson scheduled eight executions during 10 days in April because, he said, the state's supply of midazolam expires at the end of the month. This chemical is one of three that Arkansas uses for lethal injections, the current standard for enacting the death penalty in U.S. states that have not outlawed the practice. Courts have blocked four of the executions, but the others are proceeding. All this because of an expiration date on a drug that many experts believe should never have been included in the execution cocktail.

In several states, midazolam is the first of the two or three drugs used for lethal injections. It belongs to a family of chemicals called benzodiazepines, psychoactive drugs used primarily for treating anxiety. Midazolam is basically like valium. "It mellows you out," says Lee Cantrell, who teaches clinical pharmacy at the University of California, San Francisco, and directs the San Diego division of California's poison control system. After getting the midazolam, prisoners in Arkansas are injected with pancuronium bromide, which paralyzes them, and then potassium chloride, which stops the heart.

But the use of midazolam is extremely controversial, mainly because of its inadequate effects. The original first component in chemical executions was a more powerful anesthetic, such as sodium thiopental. This drug is a barbiturate, the kind used for general anesthesia during surgery. Many states still use sodium thiopental or pentobarbital, another drug in the same family, to put inmates under prior to killing them. With a barbiturate, the individual is unconscious when the heart-stopping chemical is injected into his or her bloodstream. That does not happen with benzodiazepines.

Both types of chemicals rely on a protein called GABA, short for gamma-aminobuteric acid. In its natural state, GABA is a so-called inhibitor transmitter; it sends signals that stop activity rather than excite activity. "Its whole job is to slow you down and put you to sleep," says Chris McCurdy, professor of medicinal chemistry at the University of Florida.

Benzos and barbiturates have a stark difference, however. The latter family causes the GABA receptor to work continually without stopping, like opening the tap and never reaching the maximum flow capacity. The more barbiturate you take, the more your GABA receptors signal.

Drugs like midazolam don't do that. They have what pharmacists call a "ceiling effect." A certain amount of benzodiazepine exhausts the GABA-enhancing effects. That amount may be different for different people, but it's never unlimited. Once a person hits that ceiling, the sedative effect cannot be increased. With benzos, you can turn the tap all you want, but you're never going to get more water to come out of the faucet. That ceiling effect makes it impossible to guarantee sedation prior to death. Midazolam, says Cantrell, "doesn't do anything for pain."

But states that have run out of barbiturates for executions have turned to midazolam as a substitute. And indeed, the replacement has proved extremely problematic. In 2014, Oklahoma death row inmate Clayton Lockett awoke 10 minutes after he'd been declared unconscious following a midazolam injection, before the heart-stopping chemicals could kill him. Witnesses described him writhing on the gurney. He died of a heart attack 40 minutes later, a complete departure from the quick and painless process that lethal injections are supposed to follow.

In 2015, the U.S. Supreme Court considered the use of midazolam in lethal injections in Glossip v. Gross. Richard Glossip was sentenced to death in Oklahoma for commissioning a murder (Glossip replaced the first petitioner, Charles Frederick Warner, after he was killed by the state in January 2015). Glossip's execution was stayed three times because the lethal injection practices in that state raised several issues, including the inclusion of midazolam.

An amicus brief by 16 pharmacology professors for this case argued against the use of midazolam in lethal injections. "Everyone was in agreement," says McCurdy, one of the authors of the brief. Because midazolam works differently from thiopental, the pharmacologists did not consider it an acceptable substitute. "You can't achieve the general anesthesia that you can with a barbiturate." The scientists did not convince the court, which ruled 5-4 against Glossip, who remains on death row in Oklahoma. Florida, Alabama and Virginia also use midazolam in executions.

Death row drugs are extremely hard to acquire, and manufacturers either refuse to provide their pharmaceuticals to states for this purpose or refuse to admit they've done so. The refusal of companies to provide barbiturates or benzodiazepines to states for lethal injections is why Arkansas is rushing the current executions through this month. The impending expiration date of Arkansas's midazolam would not be an issue if it could easily buy more. But it can't.

Arkansas will not say how it acquired its stash of midazolam, impeding discussion of whether the executions could be delayed. The Associated Press reported that the state purchased the drugs from West-Ward Pharmaceuticals, a subsidiary of the British company Hikma. But Brooke Clarke, a spokesperson for West-Ward, says that is not so, and that company controls prevent such sales. "If the State of Arkansas was able to procure any of our U.S.-manufactured drugs for use in lethal injections despite these controls—which it will not confirm or deny to us—it was not directly from us, nor with our knowledge," says Clarke.

Questionable Timings

The assertion that the executions must beat the expiration date is also scientifically dubious. Scientific data show no evidence of sudden changes in a pharmaceutical immediately following the date stamped on the side of a bottle. While it's true that studies of the validity of expiration dates have focused on whether the prescription drugs become more dangerous over time, they have also shown a gradual, not abrupt, decline in the potency of an active chemical. Cantrell and colleagues recently analyzed several old medications to see if the expiration date was relevant. Some of the drugs were 40 years old. "Many retained the potency stated on the label," says Cantrell.

Chemicals like benzodiazepine will degrade over time. And as they do so, they weaken. But the timing is uncertain. The U.S. Food and Drug Administration instructs drug-makers on how much active ingredient must be present and expiration dates are shaped around that measurement. But estimating when a given drug will dip below that level is just that, an estimate. "It's sort of an arbitrary date," says McCurdy. The amount of active ingredient can be easily determined in a laboratory analysis, were someone to bother taking that step. "It's not run of the mill, but it can be done," says McCurdy. At the very least, there is no reason to believe that Arkansas's midazolam will curdle on May 1, rendering it any more ineffective than it already is.

For now, though, Arkansas is continuing its race against the expiration-date clock. Last week, Ledell Lee was executed. On Monday night, Jack Jones and then Marcel Williams were executed an hour apart. But not before nurses tried for 45 unsuccessful minutes to insert a central line into Jones's neck for the injections. Eventually, they gave up and placed the tube elsewhere on his body. A court appeal filed just a few minutes before Williams's execution noted that Jones gulped for air as he died, "evidence," the filing stated, "of continued consciousness."