Ethics of ARDS lung experiment debated

AHRP old LogoAlliance for Human Research Protection 
Fri, 18 Jul 2003

Dear Dr. Campbell and others concerned about ethical research:

This is in response to your July 11 comments objecting to AHRP’s criticism of the ethics of two controversial ARDS experiments sponsored by the National Heart Lung and Blood Institute conducted at participating ARDSNetwork centers.

Your comments did not address either the ethical issues that raised our concerns (and others) about patient safety, or the findings of OHRP’s investigation. Like you, AHRP does not advocate to “arbitrarily exclude access to projects which provide promise to the participants and to our society.”

However, the researchers and institutional review boards (IRBs) of the participating ARDS Network centers violated the moral and legal imperative of ethical research involving human beings: “The voluntary consent of the human subject is absolutely essential.” “Proper preparations should be made…to protect the experimental subject against even remote possibilities of injury, disability, or death.” [Nuremberg Code]

Similarly, federal regulations require that: “risks to subjects are minimized by using procedures which are consistent with sound research design and which do no unnecessarily expose the subjects to risk” [45 CFR 46.111(a)]

OHRP’s investigation found that MOST of the 861 subjects in the completed ventilation experiment (ARMA) were enrolled without their informed consent, many without legally effective consent by a legally authorized surrogate, and some were enrolled without anyone’s consent.

These facts contradict your assertion that this case is about “several isolated instances of IRB failure.” As I said in my testimony (June 10), “The experiment is a disturbing case study revealing current unethical practices at major universities entrusted with a government license and government funding.”  ( Read, Sharav tesimony)

Indeed, OHRP’s investigation found: “that at nearly all study sites the informed consent documents failed to include death as one of the risks of the research. In particular, there was no statement that the subject could have a higher risk of death depending on which of the experimental groups he or she was assigned to, in comparison to the other experimental groups and in comparison to not entering the trial and thereby receiving individualized care based upon the best clinical judgment of the subject’s physicians.” (Read, OHRP determination)

The consent document approved by the IRB of Vanderbilt University (the leading ARDS Network center whose doctors in residence serve at St. Thomas Hospital, the hospital you are employed at), gives no indication at all that there are any risks associated with either high or low pressure lung ventilating methods. The 2 1/2 page consent document does not inform prospective subjects. It is an example of what Jay Katz calls, “discretionary subterfuge”– emphasis is given to blood and urine tests, but nothing is disclosed about the serious risks involved.

OHRP’s July 3 letter notes that: “subjects in each experimental group of the ARMA and FACTT trials potentially may have been disadvantaged compared to patients treated according to concurrent routine clinical practice.” The “disadvantage” refers to an increased risk of death. Putting patients in harm’s way is a clear violation of international codes of ethics and federal regulations which require that risks to subjects be minimized.

Among the numerous ethical violations cited by OHRP, the ARDS Network participating centers failed to comply with the ethical requirement for research involving patients, as set forth in the Declaration of Helsinki: “The benefits, risks, burdens and effectiveness of a new method should be tested against those of the best current prophylactic, diagnostic, and therapeutic methods.” [Helsinki, Article 29]

Inasmuch as the ARMA experiment failed to compare the experimental lower tidal volume against routine standard of care, the investigators’ published recommendation- “this lower-tidal-volume protocol should be used in patients with acute lung injury and the acute respiratory distress syndrome”-has not been substantiated. (ARDSNetwork, New England Journal of Medicine, 2000, 342:1307; Eichacker et al, AJRCCM, vol 167, 778-800)

It is shocking to read a recent letter by Dr. Roy G. Brower (Johns Hopkins University); Dr. Gordon Rubenfeld (University of Washington), and Dr. B. Taylor Thompson (Massachusetts General Hospital) on behalf of the ARDS Network / NIH / and NHLBI, in which they cite 5 clinical trials involving controlled lung ventilation procedures that were conducted since 1998 in violation of the Helsinki ethical standard: “None of the trials included a standard or routine care study group because they were not designed to compare outcomes of patients who received lower tidal volumes with those who received routine care.” (American Journal of Respiratory & Critical Care Medicine, 2003, vol 168: 255-6]

Under whose authority did the investigators who conducted those trials disregard the ethical requirement to include a comparison control group that receives standard of care treatment in experiments involving human subjects?

Furthermore, what justification was given for exposing critically ill patients to a high ventilation setting (12 ml/kg)in the ARDSNetwork trial (1996-1999) when it has been pointed out by several critical care specialists that tidal volumes in standard practice were already significantly lower in 1993 (9 ml/kg)? [Jardin et al, 1999, Intensive Care Medicine, 25:936-941]

This $37 million dollar ARMA experiment did NOT prove that 6 ml/kg tidal volume is better than routine care in which tidal volumes are selected according to patient need. The experiment proved only that assigning patients to high (12 ml/kg) tidal volume beyond their need is harmful-even fatal.

You state, “the lower tidal volumes are our new treatment paradigm.” If that is the case, then, as Dr. Meryl Nass pointed out in her June 10 testimony, the ARDSNet recommendations ” could potentially worsen the care of future patients, by wrongly altering the standard of care.” (Read Dr. Nass testimony)

According to several reports (which we cite in our June 10 testimonies) few clinicians have accepted the ARDS “new treatment paradigm.” For example, Dr. Jean-Damien Ricard’s editorial in the Amer J of Resp and Crit Care Med cited several preliminary reports that found the ARDS Network recommendation to lower tidal volumes to 6 ml/kg has NOT been adopted by clinicians–neither at teaching hospitals (not affiliated with the ARDS Network) nor even at some ARDSNet participating hospitals: “a reduction in tidal volume was applied in less than 5% of patients with ARDSor acute lung injury despite their institution being one of the centers participating in the ARDS Network trial.” [AJRCCM 2003, vol 167: 1297-98]

However, as Dr. Nass points out, “clinical practice could have changed in response to this trial, and had it done so, following the ARDSnet guidance to use very low tidal volumes, would our patients be better or worse off? As one example, would we be seeing increased use of sedatives and muscle relaxants in ARDS patients, with resultant increases in post-ventilator myopathies?” (See: Neil McIntyre presentation: Mechanical ventilation strategies for lung protection, May 1999)

Neither the ARDSNet investigators nor the IRBs can evade the reality of the high mortality rate in the ARMA and FACTT experiments. For those unlucky patients in the ARMA trial who were randomized to the high tidal volume group (39.8%) compared with those who were eligible but refused to participate and received stanadard care(31%). Critical care physicians whom we consulted indicated that no one in respiratory therapy raises tidal volumes independent of patient need. Yet, 80% of the patients who were randomized to the 12ml/kg arm of the experiment had their tidal volumes raised.

How can anyone justify the death of a patient who was enrolled as a subject– without legally valid consent? As a physician, how can you justify the death of a patient who was withdrawn from routine care which offered a survival rate of 70%, but after being randomized into an experimental high tidal volume setting (12 ml/kg) his /her chance of survival dropped to 60%?

The aborted FACCT trial relied on the unsubstantiated claims about the low tidal volume approach (which you call a “new treatment paradigm”). In the FACTT trial ventilation settings were restricted to 6 ml/ kg without medical justification. Before it was suspended, 400 critically ill patients were subjected to three experiments conducted simultaneously in the FACTT trial. Two catheter methods, two fluid control methods and two experimental drugs were tested simultaneously-each of these experimental treatments involved increased risks and discomfort for the human subjects.

OHRP’s letters of finding (October 7, 2002 and July 3, 2003) reveal that the ARDS investigators and their institutions’ IRBs failed utterly to meet their regulatory responsibility. The IRBs have but one mission-it is to protect the safety and dignity of the subjects. But the ARDS Network IRBs acted irresponsibly by approving the ARMA and FACTT experiments without a credible review of the research design; without assessing the risks, benefits (if any) of each of the experimental interventions relative to alternatives in routine clinical practice. The silent complicity of the IRBs facilitated the recruitment of 1261 patients for two ARDS Network trials without safeguards. The research approval process at some of the nation’s leading research institutions was revealed to be an exercise in rubber-stamping–1261 patients were recruited for experiments that exposed them to experimental interventions that posed greater risk of harm than routine care.

The magnitude of the human casualties and the violations uncovered by OHRP in these ARDS trials are not yet fully comprehended–the public has yet to find out. In human terms: 171 patients died in the 12/ ml/kg treatment group (N = 429) and 134 patients died in the 6 ml/ kg treatment group (N = 432). It is unclear how many of the patients who died in the experiment might have lived had they received standard treatment.

OHRP’s June 3 letter to the ARDS Network is unique among the many that I have personally examined over the years. The finding–“failed”–appears twenty times in the letter–which is most unusual. In a July 19, 2001, letter informing Johns Hopkins University that “all Federal supported research projects” at JHU were suspended following an investigation of the death of Ellen Roche, OHRP used the term “failed” four times.

Unless the suspended FACTT experiment is completely transformed and redesigned– with full consideration given to the safety of the human subjects–unless it undergoes a full review by a competent IRB, unless each aspect of the experimental procedures is justified-not to mention reformulating the informed documents so that they fully disclose the risks and alternatives, this experiment cannot possibly be allowed to go forth.

OHRP noted that: “the interests of future human subjects would be served best by further discussion within the scientific and bioethics communities about issues relating to appropriate research design…”

Unfortunately, the bioethics and research communities have been swept up by a culture of opportunism that encourages the forked tongue approach to regulatory compliance. A reflection of that culture can be gleaned from the ARDS investigators’ response to OHRP’s questions about the “misleading and inaccurate” ARDS Network consent forms, should lead Congress to: “While the specific interventions in the management strategies are considered standard of care, the actual strategies themselves are experimental.”

In July 1991, bioethicist Arthur Caplan, University of Pennsylvania, told MSNBC: “the suspension of clinical research at Hopkins is a symptom of a much deeper disease – the collapse of adequate protections for those involved in research at every American medical center, clinic, testing facility and hospital.” Furthermore, he said, “The system for protecting human subjects research is not simply sick – it is dead.” [Read MSNBC News]

Both Johns Hopkins and U of Pennsylvania are ARDS Network centers.

AHRP offers an ethical and statistically valid research design that would answer an important question for clinicians: What is the safest method for ventilating ARDS patients to improve their survival rate? (Read testimony of John H. Noble, Jr., PhD)

OHRP noted in its July 3 letter that “it is incumbent upon the ARDS Network investigators to provide in their written protocols a more expansive, substantive discussion of the multiple complex ethical and regulatory issues related to the protection of human subjects that must be addressed by the IRBs reviewing such research.”

I believe it is incumbent upon scientists and institutions that receive millions of taxpayer dollars, and given the privilege of conducting research on human beings, should be held accountable to at least comply with federal regulations whose standards provide minimal safeguards for human subjects.

Vera Sharav


—–Original Message—–
From: Dr. Bart Campbell
Sent: Friday, July 11, 2003 4:26 PM
To: veracare
Subject: RE: AHRP Response to criticism re: fatal ARDS lung experiment

Thank you for your reply.

As I stated “two extreme, rarely used methods” of ventillation were not tested. The 12 ml/kg. tidal volume is the method most commonly used in this country (see Marini, Evoloving concepts in the ventillatory management of acute respiratory distress syndrome Clin. Chest Med 1996;17:555). The plateau pressures which you mention are, of course, a reflection of the chest compliance/volume relationship. The uncommonly used 6ml/Kg. volume (being compared in this study to the usual approach) clearly showed notably lessened mortality. This group did have , as expected, lower plateau (and peak) pressures due to the lower ventillatory volumes.

Loss of compliance is, as I’m sure you know, part of this syndrome. This observation of lessened mortality in this group is an important advance in our ability to care for these people and clearly – given the data at hand- the lower tidal volumes are our new treatment paradigm. (As you point out, prior to this landmark study, physicians were rarely using such low volumes.) Your observation that people with ARDS who have very high peak ventillatory pressures have higher mortality rates is well known. This is the very group who have more loss of pulmonary compliance–ie. a sicker group– and of course will have a poorer outcome. (A Bayes theorum problem, as you can see.) This observation in no way alters the very appropiate observation that we improve mortality with the lower tidal volumes. It only supports the conclusion. (The Washington Manual speculation that some ARDS patients might do better with TVs less than 12ml/Kg. was simply that– speculation–. they had no data. Now we have data.)

Your observation that people who were not included in the trial had “a lower mortality rate than both arms of the ARDSNet Trial” is not supported by your reference (American Jour. Resp and Critical Care Med ’00;161: A210). Indeed, the mortality was 33.4% for the entire study group and 31.3% for the non-participents ( No Significant Difference). (Note that this provides no data re: the etiology in the non-participent group and as no randomization occurred, this can not be considered a meaningful comparison.) The mortality in the low tidal volume group was strikingly below the best mortality in the four small trials cited in the meta-analysis by Eichacker and collegues (AM. J. Respir Crit Care Med 166; 1510:’02) and even the high volume (control) group had a mortality below 6 of the 8 groups in these trials (and comparable to the other two at 38/40%). Therefore, there is nothing to suggest that this trial– given the pre-hoc data– increased risks for any participent over and above the risks of treatment approaches being widely employed at the time; ie it did not, as you alledge, “put critically ill patients at increased risk of death”. The good news is that we will be able to reduce mortality as a direct result of these observations.

You cite several isolated instances of IRB failure. As IRB Chair at Nashville’s St.Thomas Hospital (where I am employed), I am keenly aware of our intermittent failures in spite of all the hard working, well meaning, concerned people who truly try to do their best for those who place their trust in our human subject’s protection programs. We must keep stiving for improvement and your organization provides us all with an important perspective. I do not confuse surrogate consent for routine treatment with that for research (as you suggest). IRBs have a fiduciary obligation to evaluate the “risk/reward” balance. It would be a shame if poeple who had high liklihood of reaping benefit from inclusion in a research protocol were excluded due to their in ability to speak for themselves. I have read the OHRP letter to ARDSNet. We must together keep trying to improve but not arbitrarily exclude access to projects which provide promise to the participents and to our society. In my strong opinion, as a dispassionate observer, the ARDSNet ventillation study is just such a project.

W. Barton Campbell

Please Note: The views expressed above are entirely my own and do not reflect the position of any institution with which I am affiliated.