ALLIANCE FOR HUMAN RESEARCH PROTECITON

OHRP DISCUSSION OF THE ARDSNetwork TRIALS
JUNE 10, 2003

Critique of the Ethics: Legally invalid consent, failure to minimize risks

Vera Hassner Sharav
President, AHRP

Thank you for the opportunity to express our concerns about an experiment that (we believe) crossed the boundaries of permissible human research. The Alliance for Human Research Protection (AHRP) is a national network of citizens dedicated to advancing responsible and ethical medical research practices to ensure that the rights, dignity and welfare of human subjects are protected. My first encounter with an unethical medical experiment was in 1993 when I discovered that 28 veterans were subjected to a relapse inducing experiment at the Bronx VA. Since then the list of unethical experiments has, unfortunately, grown. Our efforts led the director of the NIMH to suspend 29 clinical trials at the Institute for ethical and/or scientific violations. Our interest in the ARDS trials was aroused when we were made aware of the extraordinary lobbying effort by the University of California to pass an amendment to weaken the California Medical Experimentation Act.

The experiments conducted by the ARDS Network reveal that a fundamental moral conflict exists between current research practice and the physician's personal and professional ethical, scientific, and legal obligation toward human subjects who are also patients. The following three moral imperatives are not options:

"First, do no harm" [the Hippocratic Oath];

"The voluntary consent of the human subject is absolutely essential." [Nuremberg Code]

"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]

Yet, some influential scientists, including the editor of the New England Journal of Medicine, seem to disagree.[1] They believe that they have a right to determine whether ethical standards can be circumvented to expedite research. However, such moral decisions are not for scientists alone to make.

The public expects physicians to be guided by their oath to "do no harm" which is the cornerstone of medicine. The expectation that doctors will proceed from that precautionary principle is the basis for public trust in medicine. The foremost ethical principle of medical research is the inalienable human right to informed consent. The doctrine of informed consent to research is enshrined in the 1947 Nuremberg Code - the foremost authoritative reference source for evaluating the ethical, scientific and legal legitimacy of any medical experiment conducted on human beings. In a recent article in Critical Care Medicine[2] two anesthesiologists acknowledge the preeminence of the Nuremberg Code and its legal standing in American jurisprudence. They cite the 2001 precedent setting decision by the Maryland Court of Appeals which held:

"The Nuremberg Code is the most complete and authoritative statement of the law of informed consent to human experimentation. It is also part of international common law and may be applied, in both civil and criminal cases by state, federal and municipal courts in the United States." [3]

The Nuremberg Code laid down 10 principles by which physicians who conduct human research can be held accountable. First, and foremost, is the inalienable right of every human being not to be subjected to medical experiments without their voluntary informed consent - the Nuremberg Code allows no exceptions. In addition to being voluntary, consent must be competent, informed, and comprehending. The 1979 Belmont Report and the Code of Federal Regulations (CFR) incorporated both of these fundamental ethical principles. Thus, under the CFR, physicians who conduct human experiments are obligated to disclose fully all the "reasonably foreseeable risks or discomforts" to the subject prior to obtaining legally valid informed consent. [45 CFR 46.116] American citizens, therefore, do not suspect that upon admission to a major hospital when they are incapacitated, they may be diverted from the best available treatment, and without their informed consent, may be enrolled by others in hazardous experiments that put their lives at increased risk. Yet, that is what happened at the participating ARDS Network hospitals.

Both the Nuremberg Code and the 1964 Declaration of Helsinki, place the primary responsibility for the protection of human subjects on the physician who must ensure that "the experiment should be so designedĚ that the anticipated results will justify the performance;" [NC, 3] that the human subject is protected from "all unnecessary physical and mental suffering or injury" [NC, 4] and that "proper preparations should be made to protect the experimental subject against even the remote possibilities of injury, disability, or death." [NC, 7] The Declaration of Helsinki, which added a clinical care dimension by focusing on patient need rather than the rights of healthy volunteers, affirmed the doctor's primary responsibility "to remain the protector of the life and health of that person on whom clinical research is being carried out" [III - 1], requiring: "In any medical study, every patient including those of a control group, if any, should be assured of the best proven diagnostic and therapeutic method."(II-3) [4] Helsinki reminds physicians that: "Concern for the interests of the subject must always prevail over the interests of science or society" (I-5); "the interests of science and society should never take precedence over considerations related to the well being of the subject" [III-4].

Federal policy for the protection of human subjects is laid out in 45 CFR 46. The protection afforded under this policy cannot be diminished or waved by any other legal or regulatory body - neither by Federal, State or local laws or regulations. Federal regulations place oversight responsibility for the protection of human subjects on institutional review boards. IRBs are entrusted with reviewing the research design, to ensure that "risks to subjects are minimized [and] reasonable in relation to anticipated benefits, if any, to subjects " [45 CFR 46.111 (1)] The Belmont Report set forth three broad ethical principles to guide the conduct of medical and behavioral research - respect for person, beneficence, and justice. Belmont requires that: "relevant risks and benefits must be thoroughly arrayed in documents and procedures used in the informed consent process." [C-2 (v)] Indeed, Belmont emphasized the need for disclosure of risks beyond the standard applied in malpractice law:[5] "the research subject, being in essence a volunteer, may wish to know considerably more about risks gratuitously undertaken than patients" who seek needed care. [C-1] Beneficence is defined in the Belmont Report by two rules: (1) "do no harm" and (2) "maximize possible benefits and minimize possible harms," noting that vulnerable persons "are in need of extensive protection." [C-1] Federal regulations instruct IRBs: "In evaluating risks and benefits, the IRB should consider only those risks and benefits that may result from the research should not consider possible long-range effects of applying knowledge gained in the research " [45 CFR 46.111 (2)] Thus, the IRBs sole responsibility is to protect the human subjects - not to overestimate the potential benefits that the research may (or may not) yield for others.

Surrogate consent for medical experiments raises serious ethical concerns.

When the subject is incapacitated or incapable of giving informed consent, Helsinki, Belmont, and the CFR recognize consent to research from a "legally authorized" surrogate who is chosen "to act in that person's best interest"[Belmont C-1; 45 CFR 46.116] A discussion of the ARDSNet experiment at the 15th Annual Congress of the European Society of Intensive Care Medicine[6] revealed that in Europe surrogate consent is not accepted. In Italy, surrogate consent for research is strictly forbidden; the United Kingdom has no legal system for recognizing surrogate consent for either treatment or research. But doctors have the authority to act in the patient's best interest in those situations where patients are not able to decide for themselves. In Germany, informed consent for medical treatment of incompetent patients can only be given by a legally appointed representative. However, unlike North America, even these legal representatives do not have the authority to consent to human subjects research unless that research holds the promise of direct medical benefit to the patient [6] (This would exclude enrolling any incompetent patient in a randomized clinical trial). A critical care researcher from Israel indicated that an Israeli court recently invalidated all forms of surrogate consent.[6]

The ALI / ARDS experiment was initiated to answer a legitimate scientific question: would "the use of lower tidal volume during mechanical ventilation [ ] improve important clinical outcomes in patients" with acute lung injury (ALI) and acute respiratory distress syndrome (ARDS)?[7] However, it has been criticized for failure to abide by the precautionary "do no harm" principle, for exposing patients to unnecessary risk of harm, its flawed design [8] and wide-ranging violations of informed consent.[9] The experiment is a disturbing case study revealing current unethical practices at major universities entrusted with a government license and government funding. OHRP's investigation raised the possibility that the investigators had failed to conduct a systematic analysis of the range and frequency of tidal volumes and plateau air pressures most commonly used in routine clinical practice before proceeding with this large, multi-site experiment. Instead, they tested two rarely used fixed ventilation settings, disregarding patients' condition, medical need, and best interest.8 Their condition rendered these critically ill subjects especially vulnerable, as they were incapable of exercising their right to voluntary informed consent.

OHRP's investigation found that most of the 861 subjects were enrolled without their informed consent, and without legally effective consent by a legally authorized surrogate. [9] The ARDS Network physicians, therefore, bear great responsibility for the conduct of this experiment and the ethical, scientific and legal violations.

Federal regulations require that researchers "minimize possible harms by using procedures which do not unnecessarily expose subjects to risk" [45 CFR 46.111 (a)(1)]. However, as Eichacker[8]et al noted, the ARDSNet investigators used procedures that exposed subjects unnecessarily to increased risks of harm. Best practice for ARDS patients whose condition can rapidly progress to the highly lethal level, requires airway pressure to be as low as necessary to achieve acceptable air pressure and oxygenation. However, upon enrollment in the experiment, 861 critically ill ARDS patients were abruptly withdrawn from their individualized life-support treatment that had been prescribed by the treating physician on the basis of assessed individual patient need. They were then randomly switched (within 4 hours) to one of two fixed ventilation settings selected from opposite poles of the range used in standard care - high tidal volume (12 ml / kg PBW) and low (6 ml / kg PBW). Both of these ignored patient need, ignored the most frequently used ventilation strategies in standard practice, and ignored earlier studies that have raised substantial concern in the literature about increased risk of lung injury and pneumothorax[8] when high tidal volume ventilation was used in ARDS patients.[8] [10] Low tidal volume had been associated with an increased incidence of renal failure, and the need for dialysis.[11] Prior studies by Amato[12] compared the same two fixed ventilation methods resulting in alarmingly high death rates in patients ventilated with 12ml / kg - mortality rates of 54%, 71% and 91% compared to 33%, 38%, and 43% in the 6ml / kg arm.

Assuming the ARDS Network investigators read the literature, given these serious risks, it remains unclear why they chose to compare fixed high and low settings? The investigators refer to 12 ml /kg as "traditional," but that setting is used in only 20% of standard practice.[13] [14] That means that 80% of patients randomized to that high setting had their tidal volume raised without medical justification - therefore against their best chance of survival. The experimental low tidal volume setting selected by the ARDSNet (6 ml/kg) was used even less frequently in standard practice - only 3% of patients were receiving that ventilation setting. Indeed, a patient chart analysis by Weinert[15] found that less than 1% of ARDS patients at teaching hospitals not in the ARDS Network received 6 ml/ kg tidal volume between 1994 and 2001. Ricard11 noted that tidal volumes and airway pressure used in standard care were significantly lower by 1993--(9 ml /kg) compared to 1991 (13 ml/kg). This demonstrates that physicians in routine practice had adopted an air ventilation strategy guided by the precautionary, "do no harm" principle. It remains unclear why the ARDNet selected an abandoned high ventilation strategy in 1996-1999?

The human cost ignoring best standard practice and subjecting critically ill patients to extreme treatment methods: of the 861 patients recruited for the ARDSNet experiment, 429 were randomly assigned to high tidal volume ventilation - of these 171 died (39.8%) compared to 432 patients assigned to low tidal volume of whom 134 who died (31%). At best, the experiment corroborated prior findings that the use of high tidal volume increases mortality. Is that a sufficient justification for exposing these patients' to radical ventilation methods that increased risk of death?

Eichacker, Banks and Natanson,8 14 (and others) pointed out that the ARDSNet experiment failed to answer the scientifically valid question because the design was flawed and failed to detect small differences in mortality rates, resulting in many preventable deaths. Therefore, they argue, the claimed beneficial "finding" is invalid. Dr. John Noble's[16] statistical analysis shows that the experiment lacked scientific rigor by failing to consider the numerous possible confounding variables that would offer rival explanations for the results. He provides a protocol design that can detect small size effects while controlling for confounding co-variables.

Ricard[11] notes that evidence existed showing a consensus about best mechanical ventilation procedures, and that critical care physicians had reduced tidal volume to a moderate, precautionary level by 1993.[17] [18] Hudson found that by 1996, as the mid-way range of mechanical ventilation became widely used in standard practice, the mortality rate of ARDS patients decreased to less than 30%.[19] Clearly, the investigators and the IRBs should have carefully considered the fact that critical care physicians rarely used the two extreme ventilation methods called for by the ARDSNet protocol. A thorough IRB evaluation of the risks and alternatives is of vital importance for patients whose chance for survival is at stake. It does not appear that either the investigators or the participating ARDSNet IRBs conducted a thorough examination of the existing information about the best methods for increasing survival among ARDS patients prior to approval. Inasmuch as there was no medical rationale, and therefore no ethical justification, for exposing critically ill patients to increased risk in one of two outlying, rarely used methods of ventilation, one wonders what other factors entered into the choice made by the ARDS Network.

The ARDSNet experiment was conducted on vulnerable patients with high mortality rates.[20] It raises profound questions about how it was designed and approved under the current research review and oversight system that had been established in the aftermath of the Tuskegee revelations. IRBs were established for the purpose of protecting vulnerable human beings from exploitation in unsafe medical experiments whose risks do not justify the possible benefits. However, our examination of the IRB approved informed consent documents, the OHRP letters of determination, the investigators' published report in The New England Journal of Medicine (NEJM),7 and published criticism by various critical care specialists, leads us to conclude that this is a case of colossal systemic failure. Not a single participating ARDSNet IRB protected the lives of 861 vulnerable, in this case, mostly non-consenting human subjects from preventable harm. IRBs at all participating ARDSNet institutions and various and sundry NIH committees approved an experiment that was fatally flawed in its design; that violated the precautionary "do no harm" principle embodied in national and international codes of ethics, by abruptly withdrawing critically ill patients from "the best current therapeutic method," and exposing them to unnecessary, indeed to life-threatening risk of death. The ARDSNet IRBs approved a process of subject recruitment and enrollment that violated fundamental human rights of helpless patients who were conscripted as experimental subjects contrary to their best interest - without legally effective informed consent.

The documents we have examined reveal that informed consent was largely nonexistent at premier research institutions participating in the ARDSNet. The risks do not appear to have been considered by the IRBs that approved the experiment, nor were the risks disclosed to patients or their surrogates in the ARDSNet consent documents.

The cursory IRB review, in some cases unexamined approval of the experiment, the slapdash enrollment on the basis of illegal consent by unauthorized surrogates, the brevity of the consent documents and absence of essential information needed to make an informed decision, are prima facie evidence for indictment of the current research protection system. In their published report the ARDSNet7 investigators acknowledge that high tidal volume ventilation (which they call "traditional") " may exacerbate or perpetuate lung injury in patients with acute lung injury and the acute respiratory distress syndrome and increase risk of nonpulmonary organ or system failure." [p. 1301] Very low tidal volume, they acknowledge: "may cause respiratory acidosis [retention of carbon dioxide] and decrease arterial oxygenation and may therefore require changes in the priority of some objectives in the care of these patients." [ARDS p. 1302]

The data safety monitoring board (DSMB) was no more reliable in meeting its responsibility to protect human subjects from preventable death than the IRBs. Although the investigators' published report claimed that patient safety was assured by the interim DSMB analysis of patient outcome throughout the trial, OHRP raised concerns about whether the DSMB complied with the interim analysis protocol.[21] Failure by all individual and institutional participants whose responsibility was to protect the subjects may have contributed to the preventable loss of human lives.

Subject recruitment and Informed consent:[22]

Although Federal regulations recognize surrogate consent by a "legally authorized" representative for a subject incapacitated or incapable of giving informed consent, surrogate consent for research raises profound, unresolved ethical problems. The ARDSNet experiment provides compelling evidence for disapproving surrogate consent for research because it has been abused. An examination of the consent documents and OHRP letters, reveal that ALL participating ARDS Network medical centers violated fundamental informed consent requirements with impunity.9 Not a single consent document approved by the IRBs of participating premier academic institutions, even remotely meets federal disclosure requirements for informed consent - as outlined in 45 CFR 46.116. They practiced what Jay Katz calls "discretionary subterfuge."

  1. Consent in most cases was legally invalid, obtained from a legally unauthorized surrogate;
  2. Consent for 38 subjects was obtained from a family member or an unidentified "impartial third party" who OHRP found to be a social worker employed by the institution (Thomas Jefferson University, a subcontractor of the University of Pennsylvania (U Penn);
  3. Some consents were obtained via telephone, without written consent (University of California, San Francisco (UCSF), Duke, Michigan, Maryland);
  4. Some consents were obtained by a "nod" from the incapacitated, sedated subject, with no written informed consent (UCSF);
  5. Some consents were obtained from subjects who did not speak English, and in some cases these subjects and their families were not furnished with a consent document in their own language (UCSF);
  6. In at least one case involving 16 subjects, the IRB of the University of Washington, St. Louis, illegally waived Federal informed consent requirements entirely;
  7. In all cases the consent documents failed to disclose to patients and surrogates the experimental nature of the ventilation procedures used.
  8. In all cases, the informed consent documents "failed to adequately describe the reasonably foreseeable risks and discomforts " of either arm of the ventilation experiment. Namely, the potential for dyspnea, agitation, need for higher doses of sedatives and paralytics, volume overload, and hypernatremia associated with the low tidal-volume; and the risk of exacerbating or perpetuating lung injury, risk of pneumothorax, and increased risk of nonpulmonary organ or system failure in the high tidal volume arm;
  9. All consent documents misled subjects and surrogates with false statements such as: "Both ways of inflating your lungs are safe and are commonly used in medical practice."
  10. All consent documents failed to disclose the finding reported by the principle investigator in 1996: "in the first 100 subjects enrolled into the study, some patients randomized to the low tidal volume group ╬became very dyspneic and agitated." (IRBs at U Penn, Michigan, and Hopkins recorded the protocol amendment but failed to incorporate it into the consent document);
  11. Data were collected from 260 patients who had refused to participate in the study - without their informed consent; and IRBs at all but 3 sites waived informed consent for the collection of blood and tissues for future genetic research;
  12. None of the consent documents disclosed the risks inherent in abrupt treatment change in critically ill patients; nor that the benefits of alternative, precautionary mid range ventilation methods, may carry less risk and less discomfort with survival rates not worse - and possibly better - than either non-standard ventilation arms.

Since the ventilation change was not dictated by patient's condition, but rather by the protocol the ARDSNet experiment put the interests of science above the interest of patient safety - thereby disregarding the precautionary "do no harm" principle requiring investigators to minimize risks. [45 CFR 46.111(1)]

The University of Pennsylvania (U Penn) IRB approved the experiment under expedited review procedures in March 1997, 1998 and 1999, violating Federal regulations, inasmuch the experiment involved greater than minimal risk. [45 CFR 46.108 (b)] In 1996, the IRB approved a request to offer "a $50 gift certificate as a financial incentive to individuals referring potential subjects to the investigators."[23] This information was withheld from the consent document. Instead, the consent document drew attention to the bold underlined statement: "NO ADDITIONAL COSTS will be incurred to the patient as a result of participating in this study." Was this highlighted statement intended as an inducement to influence poor families into "volunteering" their critically ill relative? Despite the 1996 report about the adverse events suffered by the first 100 patients enrolled in the low tidal volume group, the U Penn IRB approved a consent document in 1999, that misinformed patients and surrogates by falsely stating: "No serious safety problems with either method have been identified in the first 600 patients studied."[23]

The Vanderbilt University consent document does not even hint that there are any risks associated with either high or low pressure ventilating methods. The University of Michigan consent document falsely stated: "Of the procedures listed above only the administration of lisofylline for acute lung injury is experimental. The small volumes used to inflate your lungs are standards of care." Johns Hopkins' consent document made a statement that suggested patients in the trial would be "more closely monitored" (get better care) than those not: "Some aspects of your care will be monitored more closely if you participate, especially the functioning of your ventilator." The University of Utah consent document makes a preposterous claim: "The computer protocols which will be used to standardize care in both treatment groups have been extensively tested and been associated with a four-fold increase in survival for patients with severe lung disease." An element of coercion is clear in the following statement: "alternative to participating in this trial would be similar therapy this therapy may, however, not be controlled using a computer protocol."

Only one ARDSNet IRB raised substantive concerns about the inability of eligible subjects to give informed consent, and questioned the ventilator strategy: "Do you anticipate that eligible subjects will be able to give consent ? What if a subject assigned to the lower tidal volume group cannot tolerate this level and needs to go to the higher tidal volumes? Would such a person be dropped from the study?" Though the University of Washington IRB raised the questions, it did not defer its approval of the experiment - subsequently it approved waiver of informed consent.

Common elements--Tuskegee syphilis experiment (1932-1972) and ARDSNet experiment (1996-1999):

Much as the 400 subjects in Tuskegee were conscripted to a government sponsored experiment without their informed consent, without the risks and alternative treatments disclosed to them (or their families), and without safeguards to protect their welfare, so were many of the 861 subjects in the ARDS experiment enrolled without legally valid consent, without disclosure of the known and "reasonably foreseeable risks and discomforts" or the "alternative procedures or courses of treatment that might be advantageous to the subject." [45 CFR 46.116 (a) (4)]. Much as in the Tuskegee experiment, patients' best interest, indeed their survival, was not protected. Much as in the Tuskegee syphilis experiment, risks for ARDS subjects were increased, and most commonly used treatment methods (approximating best standard of care) were withheld. In both experiments human lives were sacrificed. Both the Tuskegee researchers and the ARDSNet investigators put the interest of science above the safety of the human subjects. The Tuskegee research team was severely criticized for subjecting syphilitic patients to an experiment that withheld standard treatment for experimental purposes. Twenty years after federal regulations were in force, the ARDSNet investigators, with approval by the participating IRBs subjected critically ill lung injured patients to an experiment that withheld standard treatment and put their lives at increased risk by testing two extreme air ventilation methods for experimental purposes. The Tuskegee syphilis experiment was conducted in 1932-1972. The ARDSNet experiment was conducted in 1996-1999.

When OHRP questioned the participating ARDSNet institutions about their failure to obtain legally effective consent, they argued that they applied State laws governing surrogate consent to medical treatment. But this is a specious argument since medical interventions provided during clinical care are prescribed solely for the patient's best interest. Research often involves conflicting interests and subjects who assume risks, and are unlikely to benefit personally must be in a position to freely volunteer - which is why informed consent to research bears little relationship to consent to treatment. State laws may not diminish federal policy for the protection human subjects; such laws or regulations may be applicable only if they "provide additional protections for human subjects." [45 CFR 46.116 (f)(g)(h) ] [45 CFR 46.116(e)] A legally authorized surrogate is defined in federal regulations as "an individual or judicial or other body authorized under applicable law " [45 CFR 46.102 (c)].

The California Medical Experimentation Act provided added protections for incapacitated persons by requiring that an authorized surrogate must be appointed by a court. However, following a complaint filed with OHRP, the agency investigated and found UCSF in violation of Federal regulations and the State statute. Following OHRP's letter of determination outlining the violations,[24] UCSF used its considerable political influence to change California law.6 [25] University representatives testified before the legislature in favor of an amendment eliminating court authorization requirement (AB 2328). They failed to inform the State that they were involved with and aware of the Federal complaint against the UCSF. The amendment passed in August 2002, without a public hearing, and without testimony by representatives who opposed the bill. The University acted to protect its considerable financial interests in research involving incapacitated human beings. It had received $54 million from the National Institute of Aging to coordinate a national Alzheimer's network.[26] UCSF maneuvered to change the law when it was caught violating the law - even as the Executive Vice Chancellor of the University of California announced a moratorium on surrogate consent to research, advising the deans of all of the campuses on April 16, 2002, that "surrogate or proxy informed consent for human subjects research is inconsistent with California law."[27]

It is clear to us, that the ARDS Network investigators at major universities flagrantly violated fundamental ethical, clinical, scientific, and legal requirements of conducting permissible human research. They failed to take "proper preparations to protect the experimental subject against even the remote possibilities of injury, disability, or death." [NC 7]. They failed to provide all subjects "the best proven therapeutic method."(DOH II-3) And they failed to "minimize possible harms." Neither the investigators nor the National Heart Lung and Blood Institute, nor the IRB, nor the NEJM saw anything wrong with disregarding the rights of helpless, critically ill patients and subjecting them to increased risks and pain. Informed consent was debased to a mere signature on a form disclosing a few minor risks but withholding disclosure about serious - even life-threatening - risks from patients or their legally authorized surrogate. In this climate the ethical principles that govern medical research have been discarded as if they were an option. In a recent report, the Institute of Medicine Committee noted: "responsibility for ensuring that protective rules are followed should rest with the leadership of the organization doing the study."[28]

It is unclear how many of the acknowledged 305 patients who died in the experiment might have lived, had they received standard ventilation treatment.

Clinicians have raised concern that patients with high mortality rates may be at increased risk of death in clinical trials.20 [29] Therefore, the precautionary principle would limit research involving such vulnerable patients - who are unlikely to be capable of making informed decisions. Parshuram and Kavanagh12 raised a legitimate question about the ARDSNet trial: "does entry into the trial confer a disadvantage to patients randomized to the control group (versus not being in the trial)?" It is essential for the complete ARDSNet data to be reanalyzed by OHRP and its independent consultants to answer that question and to ensure that other investigators take seriously the ethical requirements to: "minimize possible harms by using procedures which do not unnecessarily expose subjects to risk" [45 CFR 46.111 (a)(1)]. Otherwise, lingering doubts will further erode public trust in the conduct of clinical trials generally.

[1] Drazen, J. 2003. Controlling research trials. NEJM, April 3, 348:1377-1379.

[2] Kofe, WA, Rie, MA. 2003. Research ethics and law of healthcare system quality improvement: the conflict of cost containment and quality. Critical Care Medicine. 31(3)(Suppl.): S143-S152. p. S151

[3] Grimes, Higgins v. Kennedy Krieger Institute, Maryland Court of Appeals. Aug. 16, 2001. Online at: http://www.courts.state.md.us/opinions/coa/2001/128a00.pdf p. 49

[4] Declaration of Helsinki, October 2000 version states: "The benefits, risks, burdens and effectiveness of a new method should be tested against those of the best current prophylactic, diagnostic, and therapeutic methods." (#29)

[5] The most recent revision of the DOH requires disclosure of "the aims, methods, sources of funding, any possible conflicts of interest, institutional affiliations of the researcher, the anticipated benefits and potential risks of the study and the discomfort it may entail." [# 22]

[6] Truog RD. Round Table on Ethics: Informed consent for research in intensive care. Program and abstracts of the 15th Annual Congress of the European Society of Intensive Care Medicine; September 29-October 2, 2002; Barcelona, Spain.

[7] The Acute Respiratory Distress Syndrome Network. 2000.Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injur and the acute respiratory distress syndrome. NEJM. 342:1301-1308.

[8] Eichacker, PQ, Gerstenberger, EP, Banks, SM, Cui, X, Natanson, C. 2002. A metaanalysis of ALI and ARDS trials testing low tidal volumes. American Journal of Respiratory and Critical Care Medicine. 166:1510-1514.

[9] OHRP. Letter. AHRP Aug 28, 2002. http://www.ahrp.org/ethical/CaromeARDS082802.php and OHRP. Letter. AARDS Network. October7, 2002. http://www.ahrp.org/ARDSpages/ARDSindex.php

[10] Parker JC, Henandez LA, Peevy KJ. 1993. Mechanisms of ventilator-induced lung injury. Crit Care Med. 21:131-143. Dreyfus D, Saumon G. 1998. Ventilator-induced injry: lessons from experimental studies. Am J Respir Crit Care Med. 157:294-323.

[11] Ricard, JD. 2003. Are we really reducing tidal volume - and should we? Amer J of Respiratory and Critical Care Medicine. 167: 2002-2003. See also, Stewart TE, Meade MO, cook, DJ, Granton JT, Hodder RV, Lapinsky SE, Mazer CD, McLean RF, Rogovein TS, Schouten BD, et al. 1998. Evaluation of a ventilation strategy to prevent barotraumas in patients at high risk for acute respirtory distress syndrome. Pressure-and Volume-Limited Strategy Group. N Eng J Med. 338: 355-361.

[12] See, Amato, MB, et al. 1995. Beneficial effects of the "open lung approach" with low distending pressures in acute respirtory distress syndrome. A prospective randomized study on mechanical ventilation. American Journal of Respiratory and Critical Care Medicine. 152:1835-1846 and Amato, MB, et al. 1998. Effect of a protective-ventilation strategy on mortality in the acute respiratory distress syndrome. NEJM. 338:347-354. Cited by: Parshuram C and Kavanagh B. 2003. Meta-analysis of tidal volumes in ARDS. Letter. Am J Respir Crit Care Med.167:798.

[13] Brower RG, Mathay M, Shoenfeld D. 2002. Metaanalysis of acute lung injury and acute respiratory distress syndrome trials (letter). Am J Respir Crit Care Med, 166:1515-1516.

[14] Eichacker PQ, Banks, SM, Natanson C. letter 2003. Am J of Resp Crit Care Med. 167: 798-800; Brower, et al. 2002

[15] Weinert CR, Gross CR, Marinelli WA. 2003. Impact of randomized trial results on acute lung injury ventilator therapy in teaching hospitals. Am J Respir Crit Care Med. 167:1304-1309.

[16] See Noble JH. 2003. Critique of ARDSNetwork experimental design, implementation, and reporting. Comments submitted to OHRP, June 10.

[17] Slusky AS. 1994. Consensus conference on mechanical ventilation - January 28-30, 1993 at Northbrook, Ill. Intensive Care Med, 20:64-79.

[18] Jardin F, Fellahi JL, Beaucher A et al, 1999. "Improved prognosis of acute respiratory distress syndrome 15 years on. Intensive Care Medicine. 25:936-941. Cited by Ricard, 2003.

[19] Hudson LD, Steinberg KP. Epidemiology of ARDS. Incidence and occurrence: a changing picture. In: Marini JJ, Evans TW. Editors. Acute lung injury. Berlin: Springer, 1998, p. 3-15. Cited by Eichacker 2003.

[20] FreemanBD, Danner RL, Banks SM, Natanson C. 2001. Safeguarding patients in clinical trials with high mortality rates. Am J Respir Crit Care Med. 164:190-192.

[21] OHRP. Letter. ARDS Network: Mass General hospital, Vanderbilt University; Cleveland Clinic Foundation. October, 7, 2002.

[22] Informed consent documents were obtained under the Freedom of Information Act.

[23] OHRP letter. University of Pennsylvania. Jan 30, 2002].

[24] OHRP letter. University of California, San Francisco, February 8, 2002. http://ohrp.osophs.dhhs.gov/detrm_letrs/YR02/feb02l.pdf

[25] Tansey B. 2002. UCSF violated patients' rights: doctors improperly got consent for study, feds say. San Francisco Chronicle. July 28. Online at: http://www.ahrp.org/infomail/0702/29.php and Abate T. 2002. Special report: Experiments on humans - Rules for clinical trials are confusing, inconsistent - No single agency regulates medical research on humans, putting millions of volunteers at risk. San Francisco Chronicle. August 5, 2002.

[26] National Institutes of Health. Neww Release. Alzheimer's disease clinical trials expanded, expedited. September 20. 2001. http://www.nia.nih.gov/news/pr/2001/0920.htm

[27] Hume W. Executive Vice Chancelor. 2002. Moratorium on IRB Approval of Surrogate or Proxy Informed Consent for Human Subjects Research. University of California. http://www.oprs.ucla.edu/human/NewsLetters/041602.htm

[28] Institute of Medicine. Responsible Research: a systems approach to protecting research participants. October 3, 2002, p. vii.

[29] Morgenweck CJ. 2003. Innovation to research: some transitional obstacles in critical care units. Crit Care Med, 31(3)(Suppl.): S172-S177.