Jump to comment:

• ARDS Network experiment defied elementary ethical standards
  Vera H. Sharav
  Published on: 21 September 2005
• Missing the point
  John H Noble Jr
  Published on: 6 September 2005

• Published on: 21 September 2005

  ARDS Network experiment defied elementary ethical standards

  • Vera H. Sharav, public advocate

  Dear Editor

  Dr Howard Mann[1] discusses the "intense controversy about the scientific and ethical validity" that followed the publication of a government sponsored lung ventilation experiment conducted on 861 critically ill patients suffering from acute respiratory distress syndrome (ARDS) or acute lung injury. However, he does not address the violation of the first and foremost tenet of acceptable research involvin...

  Show More

  Dear Editor

  Dr Howard Mann[1] discusses the "intense controversy about the scientific and ethical validity" that followed the publication of a government sponsored lung ventilation experiment conducted on 861 critically ill patients suffering from acute respiratory distress syndrome (ARDS) or acute lung injury. However, he does not address the violation of the first and foremost tenet of acceptable research involving human subjects: "the voluntary, informed consent of the human subject is absolutely essential." (Nuremberg Code).

  In their published report in The New England Journal of Medicine (2000) the ARDS Network[2] investigators stated: "informed consent was obtained from the patients or surrogates at all but one hospital, where this requirement was waived. " [p. 1302]

  However, an investigation by the federal Office of Human Research Protection (OHRP) [3] determined that "at nearly all" the prestigious ARDS Network sites informed consent standards had been violated. And OHRP found that most of these patients were enrolled without legally effective consent by a legally authorized surrogate. Some patients were enrolled without anyone's consent. For example, at the University of Washington[4] (Seattle), 71 subjects 'were unable to provide legally effective informed consent, and consent for 55 of these subjects was obtained and documented from another individual." And, OHRP acknowledged that for 16 of the 71 subjects, the IRB waived consent altogether.

  OHRP found that ARDS Network sites had not only failed to obtain informed consent but when consent was obtained it violated essential requirements:

  "at nearly all study sites the informed consent documents failed to adequately describe the purpose of the research... failed to adequately describe the nature of the experimental design...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 not entering the trial and thereby receiving individualized care based upon the best clinical judgment of the subject?s physicians. " And " subjects in each experimental group" of the trial "may have been disadvantaged compared to patients treated according to concurrent routine clinical practice." [3]

  The ARDS Network IRBs approved informed consent documents that "failed to describe the reasonably foreseeable risks and discomforts." Patients and families were misled with false reassurances. For example, the University of Utah consent document made unsubstantiated claims: ?risks will be minimized by close monitoring and by a computer protocol." 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."[5] The most disturbing aspect of this is that the exact opposite was shown to be true.[6] Who should bear responsibility for such violations?

  An article co-authored by a principle ARDS investigator from the University of Utah, Dr. Alan Morrison,[7] and two pediatric critical care physicians affiliated with Harvard (the site of the ARDSNet statistics analysis) argued in 1999, in the NEJM that explicit informed consent for research should not be required in experiments such as the ARDS trial:

  "We believe that as with clinical care, in the case of many randomized, controlled trials, the patient's participation can and should be considered to be authorized by his or her general consent for treatment and that specific consent should not be required...In addition, there is a price that is paid when one insists on specific informed consent for all randomized, controlled trials. Many worthwhile studies will not be conducted if investigators are required to obtain specific informed consent."

  Although their utilitarian argument was roundly criticized by federal officials, physicians and ethicists,[8] it is most disturbing that OHRP?s investigation revealed that the ARDS Network had, in fact, largely dispensed with federal informed consent requirements.

  The controversial design of this multi-site experiment, in which patients were withdrawn from the individualized ventilation settings prescribed by a critical care specialist (i.e. following standard practice), then randomly assigned to one of two fixed ventilator settings- -a low 6 ml/kg vs. a high 12 ml/kg--is being heatedly debated. The American Journal of Respiratory Critical Care Medicine and Critical Care Medicine, have seen fit to give equal opportunity to both sides of the debate--but not The New England Journal of Medicine.

  In this contested trial design, the investigators substituted individualized standard treatment with a randomly assigned, fixed ventilation setting, while failing to include a control group against which the two tested methods could be compared. But what is the value (or justification) for conducting an experiment that puts human beings at increased risk if the findings will not answer the important clinical question: Which of the two tested methods of ventilation is better than current practice?

  The appropriate control group for ARDS patients would have been a group of patients receiving standard care which allows physician to select a tidal volume setting from a range of possible settings, depending upon the severity of injury to the lung and condition of the patient. At the time of the ARDS experiment that range in standard care was between 6.5 ml/kg and 14.5 ml/kg. [See: Mann, Fig 1]

  Prior to being randomized to either fixed tidal volume, most patients in the ARDS Network were receiving 9.5 ml/kg to 11.5 ml/kg tidal volume. Without a control group, it is not possible to know whether one or the other tested method of ventilation--i.e., 6 ml/kg or 12 ml/kg--is as good as, better than, or worse than the setting prescribed in current practice.. As Dr. Mann correctly points out, "An inappropriate choice may nullify the clinical value of a trial that otherwise meets the requirement of internal validity." [p. 550]

  Dr. Mann makes three recommendations: first, a systematic review of the relevant literature. It is hard to believe that the standards of approval for clinical trial have not yet reached the point at which we can reasonably expect that IRBs would have ensured 'prior to approval' that investigators have conducted a systematic review of the relevant literature to ascertain risk.

  Two additional steps recommended prior to IRB approval include a formal survey of relevant clinical experts to document current practices; and publication of the trial's protocol to solicit critical appraisal of its rationale and design. To ensure against bias, flawed trial design, as well protecting human subjects from unjustified risks in invalid experiments will require accountability.

  The ARDS Network experiment is a disturbing case study revealing current unethical practices at major US universities, and thereby demonstrating the ineffectiveness of the regulatory protection system that is supposed to prevent experiments that unnecessarily expose human subjects to increased risk of death. Only one IRB in the ARDS Network raised substantive concerns about the inability of eligible patients to give informed consent. The University of Washington (Seattle) IRB questioned the ventilator trial design: "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?" Yet, after raising substantive concerns, the IRB approved the research and approved a waiver of informed consent. This government sponsored trial cost taxpayers $48 million.

  References

  [1] Mann, H. Controversial choice of a control intervention in a trial of ventilator therapy in ARDS: standard of care arguments in a randomised controlled trial. Journal of Medical Ethics, 2005; 31:548-553.

  [2] The Acute Respiratory Distress Syndrome Network. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. New Engl J Med 2000;342:1301?8.

  [3] See, OHRP Letter of determination to ARDS Network. July 3, 2003. Online at: http://www.hhs.gov/ohrp/detrm_letrs/YR03/jul03a.pdf

  [4] See OHRP letter to the University of Washington (Seattle) Feb 4, 2002. Online at: http://www.hhs.gov/ohrp/detrm_letrs/YR02/feb02g.pdf

  [5] Informed consent documents were obtained by the Alliance for Human Research Protection (AHRP) under the Freedom of Information Act.

  [6] Mechanical ventilation in ARDS: One size does not fit all*, Editorial by Katherine J. Deans, Peter C. Minneci, Steven M. Banks, Charles Natanson, Peter Q. Eichacker, Critical Care Medicine, 2005; 33(5):1141.

  [7] Truog, RD, Robinson, W., Morris, AH. Is informed consent always necessary for randomized, controlled trials? The New England Journal of Medicine, 1999; 340 (10):804-807.

  [8] New England Journal of Medicine, 1999; 341:448-450, Aug 5, 1999. Correspondence

  Show Less

  Conflict of Interest:
  None declared.

  • Back to top
• Published on: 6 September 2005

  Missing the point

  • John H Noble Jr, Policy Analyst

  Dear Editor,

  Dr. Mann's 20/20 hindsight discloses the availability of some of the alternative research designs that might have answered the question addressed by the ARDSNET researchers. However, the choice among alternatives in face of uncertainty is not itself controversial. What is and should be controversial in all such cases is the failure to disclose to the unknowing patients or their surrogates the nature o...

  Show More

  Dear Editor,

  Dr. Mann's 20/20 hindsight discloses the availability of some of the alternative research designs that might have answered the question addressed by the ARDSNET researchers. However, the choice among alternatives in face of uncertainty is not itself controversial. What is and should be controversial in all such cases is the failure to disclose to the unknowing patients or their surrogates the nature of the risk into which the researchers would put them. Interestingly, Dr. Mann's focus and survey of the expert clinicians to establish the "probative" value of an experimental design lacking representation of the applicable, albeit putative, standard of care in the situation entirely misses the point that the ARDSNET researchers failed to satisfy the ethical and legal requirement pursuant to 45 CFR 46 of obtaining the informed consent of the patients or their surrogate decision-makers.

  The Alliance for Human Research Protection (AHRP) that Dr. Mann cites as one of the critics of the ARDSNET trial complained to the Office of Human Research Protection (OHRP) about the lack of informed consent for a substantial number of the human subjects, which the OHRP itself had drawn to the attention of the ARDSNET researchers.

  The ordinary "standard of care" requires the physician to explain the benefits and risks of the intended treatment before undertaking it. This essential step makes the difference at law between authorized treatment and a battery committed on the person of the patient.

  The question that remains to be addressed in all of this is "If the researchers were so muddled and uncertain about what they were doing, how could they have possibly conveyed the possible benefits and risks to their patients or surrogate decision-makers?" That this should have been done is uncontroversial, and fixing this deficiency was one of the stipulations in the OHRP permission for the ARDSNET researchers to resume the suspended trial in question.

  Show Less

  Conflict of Interest:
  None declared.

  • Back to top