Hospitals in medical study tested epilepsy drugs without telling patients

Grady Memorial was among 58 sites nationwide that took part in a medical study of drugs for stopping epileptic seizures

Combined ShapeCaption
Grady Memorial was among 58 sites nationwide that took part in a medical study of drugs for stopping epileptic seizures

Two Atlanta hospitals were part of nationwide study drawing criticism from watchdog group

Patients in prolonged, life-threatening epileptic seizures who came to emergency rooms at Grady Memorial Hospital or Children’s Healthcare of Atlanta at Egleston from 2015 to 2018 were placed in a medical study without their knowledge.

The hospitals were among 58 sites nationwide that randomly assigned patients, ages 2 to 94, one of three approved anticonvulsant drugs after the patients failed to respond to initial treatment. The goal was to learn which medication was best.

Doing it without patients’ consent was legal. But a watchdog advocacy group based in Washington, D.C., Public Citizen, is accusing the researchers and the regulators who approved the study of potentially endangering some patients’ health in order to let the research proceed. And they are calling for review of the system that made it possible.

In a letter to federal regulators, the group alleges multiple flaws in the study’s design. As a result of those flaws, the letter says the hospitals may have under-dosed patients who weighed more than 165 pounds; may have denied some patients the right drug at the right time for them; and did not disclose those factors in consent forms given to patients or their parents after the treatment was over.

Furthermore, the letter says, even though researchers had an obligation to ensure an equitable population in the study group, the patients enrolled in the study were more likely to be African-American than the population of patients historically hospitalized for the condition, called status epilepticus.

Grady and CHOA provided a high number of patients for the study compared to most of the other hospitals in the study: at least 62 of the patients, more than 10%. Emory University, a medical research powerhouse, staffs Grady and CHOA and does research at the hospitals.

Lives at stake

The patients, in the midst of a seizure and not responding to the first-line treatment of benzodiazepine drugs, weren’t beyond hope, with all remedies exhausted. Rather, after they first failed to respond to the benzodiazepines, their treatment was changed from the normal next steps.

Normally, a doctor would have decided which second-line treatment to give, and commonly might have settled on one of three approved drugs to try to break the seizure. The drugs are levetiracetam, fosphenytoin, and valproate.

Under this study the patient’s doctor didn’t make that decision. Instead, the study randomly assigned one of those drugs, and the doctor administered whatever that was. Both the patient and the doctor were blinded to the name of the drug for 60 minutes.

If a patient had previously had a record of success with one drug in a prolonged seizure, or a record of not responding to another drug, the doctor would not consider that. If patients had a daily prescription of a form of one of the drugs that worked for them to prevent seizures, but they had failed to take their medications, the doctor would not consider that, either.

The study protocol allowed unblinding in case of emergency but discouraged it, calling it “a deviation from protocol.”

The patients, or in children’s cases their parents, were told about their enrollment in the study only after the crisis was over. A sample letter drawn up by the researchers does not contain the points about limited dosing raised by Public Citizen.

“This is among the most concerning trials I’ve seen in my experience,” said Dr. Michael Carome, director of the health research group at Public Citizen. Previously, Carome worked at the U.S. Health and Human Services Department’s Office for Human Research Protections.

The stakes of treatment were high. Status epilepticus has a significant mortality rate, and the faster the seizure is stopped, the better. Studies across age groups and types of patients show widely varying mortality rates. In this study, called the ESETT, among 462 patients enrolled, 12 died, according to the watchdog’s analysis.

The National Institutes of Health, which funded the study, said in a statement that the study was important for the very reason that doctors need to choose the right drug. Before the study, there was no evidence-based way to know if doctors might be skipping a hidden gem among the three drugs, or unwittingly choosing a loser. The results showed they were indeed best left to their own judgment.

In addition, NIH pushed back at Public Citizen’s charges about underdosing, saying the dose amounts followed national single-dose guidelines and did not result in adverse effects. As to the racial makeup of the patients, NIH said that was simply a factor of the hospitals chosen for the study.

The Georgia health systems involved in the study did not respond directly to Carome’s allegations. Instead, in a joint statement to the AJC, Emory Healthcare, Grady and CHOA noted that the study had been accepted in two elite medical journals: the New England Journal of Medicine and The Lancet in Great Britain.

“Additionally, peer review of the studies has proven positive, and follow-up responses from participants and family members of study participants showed 90 percent were glad they/their family member participated in the ESETT study,” the statement said, citing the name of the study, the Established Status Epilepticus Treatment Trial.

In fact, some of those 90% said they were neutral, not positive. Not included in the hospitals’ statement: Not all participants and family members could be reached for the survey, and many of those 90% said that even so, it was not OK to do the study without first getting permission. Thirty of 317 respondents said they wished they or their loved one had not been in the study at all.

“These numbers are quite similar to what’s been observed in almost every other study we’ve done,” Dr. Neal Dickert said in presenting the findings. The lesson learned, he said, was to work on improving outreach.

The hospitals referred further questions to the study leaders. The doctors who led the study nationwide, Dr. Jaideep Kapur of the University of Virginia, Dr. Robert Silbergleit at the University of Michigan and Dr. James Chamberlain at Children’s National Health System, did not individually respond to emails or telephone messages, nor did the doctors in charge at the Georgia sites: Dr. Jonathan Ratcliff and Dr. Harold Simon, Emory University emergency doctors who practice at Grady and CHOA.

A federal rule

A 1996 federal rule designed to provide for better study of emergency medicine, a field where patients are often unconscious, allows researchers in limited circumstances to enroll patients without obtaining consent. The federal rule is called Exception from Informed Consent, or EFIC.

Researchers say it has been a boon for patients.

For example, the researchers who did the epilepsy drug study said, doctors in different hospitals often had feelings or prejudices about one of the three drugs that didn’t really have a basis in fact. The study put those to rest.

Critics of EFIC say abuses against patients have piled up under it, because the leaders involved in designing, approving and publishing studies are also usually researchers and people sympathetic to researchers. Their mindset inevitably tips over from care from the individual patients in the study to the public gains that might be made, said George Annas, director of the Center for Health Law, Ethics & Human Rights at Boston University.

He favors reform of Exception from Informed Consent. Harriet Washington, a medical ethicist and lecturer at Columbia University who has written on race and medical research consent, wants it repealed.

Washington said that the disproportionately large share of African-Americans in the study “violates social justice” because the risks of the study fell disproportionately on them.

With the epileptic seizure trial, the crux of the issue may have come most strongly into view in late fall of 2017, when the study reached a pre-agreed trigger to stop. Overall data showed the attempt to find a best or worst option among the three drugs was futile: They were all going to turn out pretty much equally effective.

But internally, the study authors thought they still had a chance with the data for children. They argued to safety oversight officials that they could stop studying adults, but keep going in children, with the potential for a different result. The officials agreed.

Dr. Jaideep Kapur, the principal investigator on the study, acknowledged in a presentation to peers that that decision came with risk for the children.

“The price we paid, and I think one child paid the price for it, I think was that child who received the overdose on Dec. 30, 2018,” Kapur said.

“In the end I felt pretty strongly that we had made a commitment—a very strong commitment—to have the study adequate, to leave no doubts about what is the best treatment for children and adults,” Kapur said. “...I think it was the right thing to do looking back.”

Hospitals' Response on Seizure Study

This response from Emory University, Grady Health System and Children's Healthcare of Atlanta was issued on June 7, 2021:

Emory University's Department of Emergency Medicine, with research partners Grady Memorial Hospital and Children's Healthcare of Atlanta Egleston Hospital, took part in the 58-site, NIH-funded Established Status Epilepticus Treatment Trial (ESETT) from 2015 to 2018. Study participants who arrived at emergency rooms at Grady Memorial Hospital and Children's at Egleston with life-threatening status epilepticus that did not improve after initial treatment were enrolled in the trial. Status epilepticus is a seizure of greater than five minutes, or more than one seizure within a five-minute period without the person returning to normal between them.

Patients who arrived at one of the 58 study sites were given first-line, standard-of-care treatment (benzodiazepines) for seizure activity. Because of the nature of this serious medical condition, and if the first-line treatment was not effective, participants were enrolled in the ESETT clinical trial using "exception from informed consent" to determine which of three commonly used study medications was most effective in stopping on-going seizures. Emory's Institutional Review Board (IRB) performed a rigorous review of this study, along with regulators, ethics groups and community members, before enrollment began, and throughout the conduct of the study at both Grady and Children's at Egleston.

Results of the findings were published in November 2019 in the New England Journal of Medicine and in April 2020 in The Lancet, determining "any of the three drugs can be considered as a potential first-choice, second-line drug for benzodiazepine-refractory status epilepticus."

Additionally, peer review of the studies has proven positive, and follow-up responses from participants and family members of study participants showed 90 percent were glad they/their family member participated in the ESETT study.

Response from the National Institutes of Health

ESETT was an important study precisely because there are no guidelines or evidence-based ways for doctors to pick an emergency treatment for prolonged seizures in the emergency department. Emergency treatments are usually initiated immediately, typically before doctors are able to determine what medications have been given in previous episodes of prolonged seizures. Prior responses also do not reliably predict subsequent responses. In a previous study of other prehospital treatments for prolonged seizures, researchers learned that a good or bad response to an antiseizure medicine in a prior episode does not predict well what will work best in a later episode. This study, like all research aiming to get at the best treatment for persons with life-threatening conditions that impair consciousness can only be done under Exception from Informed Consent (or EFIC). The grant application was reviewed by a standing NIH peer review study section, and EFIC details were provided in the application and reviewed. This was in addition to FDA requirements for review and approval of the EFIC plan regarding public disclosure and community consultation.

The ESETT study used exactly the same drug doses recommended by the American Epilepsy Society treatment and treatment algorithm for status epilepticus, available on their website. The doses of all three medications studied were consistent with best current clinical practices in the emergency treatment of prolonged seizures (status epilepticus). For all three medications, the study used doses at the high end of the ranges used in clinical practice to avoid any under-dosing. After the trial ended, researchers did an analysis to determine whether those weighing more than 75 kilogram (or even 90 kilogram) did worse than those weighing less than 75 kilogram. This analysis confirmed that heavier participants were not under-dosed. The three medications did not work any better in participants who had more medicine per kilogram body weight than in larger people who had less medicine per kilogram body weight. (Epilepsia. 2020 Jun;61(6):e66-e70, PMID: 32420641).

There was no negative control group (placebo group); a placebo would never be used in a person who is critically ill with prolonged seizures. There is no positive control, because nothing is already known to work better than the three medications being studied. The ESETT design is therefore not a placebo-controlled study, but rather a design called a "comparative effectiveness study" which demonstrates the relative effectiveness of different active treatments. Furthermore, doctors were only asked to enroll patients in ESETT when there they didn't know which medication to give. If a patient's medical record or emergency medical alert tag indicated any allergy or contraindication to any of the study drugs, the patient was excluded from the study. Documented preferences for a particular medication based on patterns of known prior favorable or unfavorable responses, or for other reasons, were considered a contraindication.

Sites enrolled all eligible patients presenting to their hospitals, so the rate of enrollment by race is proportional to that of those presenting to the emergency departments for prolonged seizures. Sites were those participating in the federally funded NETT and PECARN emergency research networks. These sites include many major university academic medical centers, which are often located in urban areas, and that serve patient populations with greater racial diversity than is present in aggregate nationally. In addition to Grady Hospital in Atlanta (the highest volume center) there were 57 other hospitals around the country that enrolled subjects in this study.