A medical watchdog group is calling on the National Institutes of Health to immediately stop the enrollment of patients in a clinical trial of sepsis treatment and launch an investigation of how the study received approval, arguing that it puts patients at risk of serious harm, including death.
The group, Public Citizen, says in a letter dated Tuesday that the study’s protocol is also so “fundamentally flawed” that it will not produce reliable findings.
The problems with the “CLOVERS” study are so serious that “it is difficult to imagine any reasonable person agreeing to enroll … if he or she were fully informed of [its] true nature and risks,” wrote Dr. Sidney Wolfe and Dr. Michael Carome of Public Citizen’s Health Research Group. They contended that patients in the trial are “unwitting guinea pigs in a physiology experiment that will not advance medical care for sepsis,” and that other lapses are “stunning in their breadth and scope.”
The basis for the central criticism is twofold. First, Carome told STAT, patients are being put at risk because one of the two sepsis treatments in the study could, depending on how seriously ill the patients are, allow their blood pressure to remain dangerously low for hours. In addition, he said, everyone will receive what he called an experimental treatment, rather than some patients receiving the standard of care, as clinical trials generally do to enable researchers to compare results for patients treated with experimental and standard treatments.
CLOVERS scientists rejected that assertion, arguing that the treatment arms are similar to two widely used approaches. “Nothing in the trial is experimental care,” said Dr. Wesley Self, vice chair of research in the department of emergency medicine at Vanderbilt University Medical Center and an investigator in the study.
Trial organizers began recruiting patients in March and aim to enroll up to 2,320 at 44 U.S. hospitals and other medical sites. All of the patients will receive an initial quantity of intravenous fluids recommended by guidelines from the “Surviving Campaign,” from the Society for Critical Care Medicine. After that, about half will get more fluids (the “liberal fluids” group) and the others (the “restrictive fluids” group) will receive vasopressors, drugs that increase blood pressure, if needed.
An institutional review board at Vanderbilt, which is charged with ensuring that human studies are ethical and scientifically rigorous, approved CLOVERS, which is funded by NIH.
The criticism comes two months after NIH terminated another clinical trial, on alcohol, for severe ethical lapses.
Each year some 1.5 million people in the U.S. develop sepsis, in which an infection cascades into a medical emergency of plunging blood pressure, racing heart, and fever. One-quarter million people die of it annually. Despite the prevalence of sepsis, there is no consensus on how best to treat it, Self said.
The main divide is between giving patients large volumes of intravenous fluids immediately and then vasopressors, or smaller quantities of IV fluids followed more quickly by vasopressors. “Both are used fairly routinely,” Self said, and earlier research, including in lab animals, suggests either can be effective.
Observational evidence has linked the “liberal fluids” approach — which some critical care physicians have adopted over the last 15 years or so — to lower mortality from sepsis than the “restrictive fluids” approach. But the decline in death rates from sepsis might also, or instead, be due to something else that hospitals are doing better, not to giving patients large quantities of fluids.
“We’ve never unbundled why mortality has gone down,” Self said. “That’s why we believe a clinical trial is necessary.” Another reason is to determine whether giving patients large volumes of IV fluids might cause edema (swelling), restricting oxygen flow to organs and thereby contributing to organ damage, as some observational evidence suggests.
The two arms of the CLOVERS (“Crystalloid Liberal or Vasopressors Early Resuscitation in Sepsis”) trial differ somewhat from the Critical Care Medicine group’s guidelines. Those call for sepsis patients to quickly (within three hours) receive IV fluids, as CLOVERS patients will. But the guidelines also call for physicians to closely monitor patients’ blood pressure, and give additional IV fluids if blood pressure continues to be dangerously low. “The use of IV fluids in the resuscitation of patients is a cornerstone of modern therapy,” the guidelines state, but must be followed by vasopressors if blood pressure remains dangerously low.
CLOVERS’ “liberal fluids” arm comes closest to today’s standard sepsis care as well as to the usual-care control arm in earlier clinical trials of sepsis treatments. But it differs from all of those in key ways. Patients in the trial will get larger-than-usual volumes of IV fluids and in less time, and will receive “rescue vasopressors” only after getting about 5 quarts of fluids. Public Citizen calls that “not usual care but unusual care.” According to Carome, “Patients could experience up to several hours of blood pressure much lower than would be tolerated in usual care before they receive vasopressors, and that is not acceptable.”
By not having a third group of patients receive usual care, the study could produce dangerously misleading results, he said. If it finds that one treatment has 20 percent mortality, and the other has 25 percent, the first would seem superior. But if a control group received usual care, like that in the Critical Care Medicine guidelines, it might have 15 percent mortality.
“You wouldn’t be able to know that from this trial,” Carome said. “That could adversely affect future patients.”
Dr. Nathan Shapiro of Beth Israel Deaconess Medical Center in Boston and co-leader of the CLOVERS trial, said in a statement that, to the contrary, “Our study will provide clinicians with authoritative data on which approach leads to better patient outcomes. The study protocol was designed by expert clinicians in emergency and critical care medicine,” and “follows a well-accepted design” that was reviewed not only by Vanderbilt’s IRB but also by a data safety and monitoring board appointed by NIH’s National Heart, Lung, and Blood Institute, the study’s funder.
Bioethicists argue that no human study should be conducted unless it is designed to yield useful results. “If a trial is bad science, it should not be done,” said clinical trials scholar and bioethicist Karen Maschke of the Hastings Center, who has not looked into the CLOVERS study. “The IRB has an ethical obligation to make sure the science is good.”
The consent form for the CLOVERS trial does not specify risks like those Public Citizen alleged. Instead, it says that any risks come from receiving extra IV fluids and vasopressors. There is no mention of possibly having severely low blood pressure for hours, a condition that can cause septic shock and organ failure.