AHRP Response to criticism re: fatal ARDS lung experiment
- To: Dr. Bart Campbell
Cc: jyg2@...; LAUGHREN@...; murphyd@...;
robertsr@...; RUMBLEW@...; temple@...;
robert.roswell@...; afauci@...; cl63c@...;
da43w@...; ez26y@...; ti4g@...; BSCHWETZ@...;
dlepay@...; dtroy@...; LCRAWFORD@...;
ALLIANCE FOR HUMAN RESEARCH PROTECTION (AHRP)
Contact: Vera Hassner Sharav
Dear Dr. Campbell,
In response to your objection to AHRP's June 26 Infomail
about our testimonies criticizing the ARDSNetwork experiment
that tested two mechanical tidal volume ventilation settings in
critically ill patients with ARDS, we note the following:
First, you wrote: The trial did not, as alleged, "test two extreme,
rarely used methods of mechanical lung ventilation"...
AHRP response: The experimental method used low tidal volumes (6ml/kg)
which were shown in at least one survey to only be used by no more than
3% of treating physicians--that's extremely rare.
The other arm of the trial, referred to as the control, used 12 ml/kg TV
and often plateau pressures higher than 35 although the ARDS literature
is fairly clear that such pressures, though they may not necessarily be
associated with grossly visible barotrauma, have resulted in higher
rates. The 1995 Washington Manual warned that ARDS patients might need TVs
lower than other ventilator patients, who would often be started on 10-12
(The ARDSnet trial began in 1996.) Simply put, the control value chosen (12
ml/kg TV) was not the gold standard treatment (or the standard of care) it
should have been.
As we note in our testimonies, the ARDSNet experimental design failed to
include tidal volumes "in the range wherein 77 percent of ARDSNetwork
study center practice was occurring prior to random assignment
into one or other of the two extremes."
Prior to enrollment in the ARDSNet experiment, patients had been under
the care of pulmonologists at leading institutions such as yours,
Dr. Campbell. Indeed, Vanderbilt University, with which you are
affiliated, is the primary ARDSNetwork center.
It is significant that only 20% or less of ARDS patients received 12 ml/kg
tidal volumes in clinical practice. But in the ARDSNet trial, 80% of
randomized to the 12ml/kg control arm, had their tidal volume raised to meet
the protocol requirements--in some patients TV were doubled. In accepted
clinical practice TV would not be raised without medical justification.
Second, you wrote: "the subjects...were fortunate to be in a trial that
clearly lessened mortality in the treatment group...."
AHRP response: On the contrary, patients who declined to participate were
the fortunate ones as they had a lower mortality rate than both arms of the
ARDSNet trial. Furthermore, those patients who were unlucky enough to have
been randomized to receive 12ml/kg TV had a 40% mortality rate compared to
30% of those who declined.
Even the ARDSNet investigators acknowledged:
"...critically ill eligible patients who participated in this trial did not
fare better overall than eligible patients who declined to participate."
[See: Krishnan JA, Hayden D, Schoenfeld D, Bernard G, Bower R (for the
NHLBI ARDSNetwork Investigators). Outcome of participants vs eligible
nonparticipatns in a clinical trial of critically ill patients. [Abstract]
American Journal of Respiratory Critical Care Medicine, 2000, 161:A210.]
Third, you justified surrogate consent to research involving critically ill
patients, stating "surrogate consent was commonly accepted (just as it was
in providing standard care for potentially life-saving therapies in those
whose loved ones gave surrogate consent)."
AHRP response: Dr. Cambell, you are blurring the issue between surrogate
consent for treatment which is to provide needed treatment for the benefit
of critically ill patients with the entirely different issue of surrogate
consent for research. The ARDSNET experiment put critically ill patients at
increased risk of death--as the survival rates of those who refused
Furthermore, in the case of the ARDSNet experiment, some "surrogates" turned
to be hospital employees such as social workers, not "loved ones." At least
institution's IRB waived informed consent entirely for 16 subjects.
Fourth, you wrote, "Surrogates of these desperately ill people were informed
the risks of ARDS..."
AHRP response: The informed consent documents which AHRP examined and the
letters demonstrate the contrary. In our testimonies we cite specific "sins
omission and sins of commission."
For example, "In all cases the consent documents failed to disclose to
and surrogates the experimental nature of the ventilation procedures used.
In all cases, the informed consent documents "failed to adequately describe
reasonably foreseeable risks and discomforts " of either arm of the
experiment. Namely, the potential for dyspnea, agitation, need for higher
of sedatives and paralytics, volume overload, and hypernatremia associated
the low tidal-volume; and the risk of exacerbating or perpetuating lung
risk of pneumothorax, and increased risk of nonpulmonary organ or system
in the high tidal volume arm."
Indeed, OHRP's letter to AHRP (Aug. 28, 2002) states:
"In reviewing the above letters, you will note that in all cases OHRP found
the informed consent documents approved by the institutional review boards
for the research failed to describe adequately the reasonably foreseeable
and discomforts of the research."
As we note in our testimonies, in some cases grossly inaccurate information
provided in the informed consent document, misleading subjects and
with false statements such as:
"...both treatment groups have been extensively tested and been associated
with a four-fold increase in survival for patients with severe lung
See AHRP testimonies:
As noted by OHRP the ARDSNet consent process was coercive in nature.
Fifth, you state, "Society is greatly benefited by this study and the
AHRP response: How could society benefit from a trial that increased
mortality rates? How could society benefit from a human experiment that
provided no useful information to guide clinicians in selecting the
safest method of ventilating the lungs of ARDS patients?
The researchers, who were using human subjects, had an obligation to
design the trial in such a way that the controls were assigned the best
alternative treatment, and had a legal and ethical obligation to obtain
proper informed consent. It was not their role to decide that the laws
governing human subject research should be changed for the good of society,
and thus act accordingly.
Finally, AHRP agrees with your statement, "We must take great care to
serve well those who put their trust in us."
AHRP believes that performing research on human subjects is a privilege
and requires the highest ethical and medical standards. In this particular
case, both appeared to be lacking.
Meryl Nass, MD, board member
Vera Sharav, President
John H. Noble, Jr., Ph.D, board member
> From: Dr. Bart Campbell [mailto:BCAMPBEL@...]respiratory
> Sent: Friday, June 27, 2003 6:15 PM
> To: veracare@...; jyg2@...; LAUGHREN@...;
> murphyd@...; robertsr@...; RUMBLEW@...;
> temple@...; woodcockj@...; COMMISSIONER@...;
> nelda.p.wray@...; anthony.principi@...;
> robert.roswell@...; afauci@...; cl63c@...;
> da43w@...; ez26y@...; ti4g@...; BSCHWETZ@...;
> dlepay@...; dtroy@...; LCRAWFORD@...;
> hhsmail@...; CCURIE@...
> Subject: AHRP Testimonies re: Fatal ARDS lung experiment
> AHRP provides a service to humankind by offering alternative viewpoints in
> research and bioethics. However, as a practicing cardiologist with
> extensive experience in caring for parients afflicted with acute
> distress syndrome (since it was described and named by my mentors, TomPetty
> and David Ashbaugh), I must object to AHRP's vilification of the ARDSNetused
> report in the May 4,'00 NEJM. Several AHRP allegations are totally devoid
> of veracity. The trial did not, as alleged, " test two extreme, rarely
> methods of mechanical lung ventillation" but compared instead (and forsound
> reasons) the traditional (and most widely used) method of ventillation inthat
> these extremely ill people to a lower ventillatory volume method. The
> question; "how many o
> f the subjects who died might have lived had they continued to receive
> standard, individualized care" suggests the study resulted in adverse
> outcomes when in reality, mortality in the group that did not receive
> "standard" care had a mortality strikingly lower that any of the five
> trials in the Eichacker et al post hoc meta-analysis. The conclusion
> "the published ARDSNet findings in The New England Journal of Medicinecould
> potentially jeprodize the lives of future patients" could not be furthermany
> from the truth. This was a necessary and landmark trial that is likely to
> significantly lessen mortality in a catestrophic condition that has for
> years claimed an extremely high mortality. These "subjects were (indeed)treatment
> fighting for their lives" and were
> fortunate to be in a trial that clearly lessened mortality in the
> Regarding the allegation that the trial was ethically flawed, AHRP is
> technically correct that some patients did not have "legally valid
> consent". As IRB chair in a State that does not provide a law validatingwhose
> surrogate consent, I have had great concern about patients who are too ill
> to give consent for trials that might provide significant benefits. Prior
> to OHRP's decision that there was no way to include these patients,
> surrogate consent was commonly accepted in such States (just as it was in
> providing standard care for potentially life-saving therapies in those
> loved ones gave surrogate consent). Unfortunately, until a law is inplace
> providing access, these people cannot participate even when theircaregivers
> (and loved ones) realizeavailable
> they are, in some cases, being denied access to potentially life saving
> therapies. Is this ethically appropiate? The allegation by AHRP that
> "informed consent was largely nonexistent" and "Neither the known nor
> foreseeable risks were disclosed to any of the patients or their
> surrogates is simply not true and is inflammatory. Surrogates of these
> desparately ill people were informed of the risks of ARDS and of the high
> liklihood of not surviving. They are fortunate to have the study
> to them.is
> The ARDS investigators are to be congratulated, not castigated. Society
> greatly benefitted by this study and the ARDSNet. We must support themand
> urge them to get on with the business of reducing mortality in this dread
> disease. AHRP is right to ask us to think carefully about the ethical
> nature and validity of research. We must take great care to serve well
> those who put their trust in us. In this instance however, AHRP has sadly
> missed the mark.
> W. Barton Campbell