Cook, Deborah MD; Burns, Karen MD; Finfer, Simon MD; Kissoon, Niranjan MD, FCCM; Bhagwanjee, Satish MD; Annane, Djillali MD; Sprung, Charles L. MD, JD, FCCM; Fowler, Rob MD; Latronico, Nicola MD; Marshall, John MD
From International Forum for Acute Care Trialists (InFACT) Ethics Committee, Departments of Medicine (DC), Clinical Epidemiology, and Biostatistics, McMaster University, Hamilton, Canada; Interdepartmental Division of Critical Care (KB, RF, JM), University of Toronto, Toronto, Canada; Department of Intensive Care Medicine (SF), University of Sydney, Sydney, Australia; Department of Pediatrics (NK), University of British Columbia, Vancouver, Canada; Department of Intensive Care Medicine (SB), University of Witwatersrand, Johannesburg, South Africa; Department of Intensive Care Medicine (DA), University of Versailles, Garches, France; Department of Anesthesiology and Critical Care Medicine (CS), Hadassah Hebrew University Medial Center, Jerusalem, Israel; Department of Anesthesia (NL), Intensive Care and Perioperative Medicine, University of Brescia, Brescia, Italy.
All the authors care for critically ill patients in practice, and collaborate on, or lead clinical or health services research for these patients.
Dr. Burns has received grants from the Canadian Institutes for Health Records, Physician Services Inc., and the Public Health Agency of Canada. The remaining authors have not disclosed any potential conflicts of interest.
For information regarding this article, E-mail: firstname.lastname@example.org
The H1N1 2009 pandemic poses unprecedented research challenges in critically ill patients. The severe acute respiratory syndrome outbreak of 2003 demonstrated how difficult it is to develop and implement studies of critical illness during epidemics when clinical services are overwhelmed (1). Early reports of the recent influenza season in the southern hemisphere indicate that young, otherwise healthy adults are predisposed to H1N1-related critical illness, characterized by refractory, life-threatening respiratory failure. Given the 20% risk of death, the possibility for improved outcomes will not be realized without collaborative national and international investigations. Accordingly, as we approach influenza season in the northern hemisphere, many fundamental clinical and health services research questions remain unanswered, leaving clinicians caring for the most seriously ill patients without current, valid evidence. Investigations will need to be developed, implemented, and completed in a timely manner to maximize the chance of informing clinicians and public health policy. Essential investigations during the pandemic will be impossible without advanced planning and rapid implementation strategies.
Research is crucial to understanding existing and potential critical care services, and the consequences of excessive demands placed on them. The anticipation of increased need for intensive care resources during the H1N1 pandemic raises concerns that the capacity to deliver basic and advanced life support in the intensive care unit (ICU) will be exceeded. Although plans to expand critical care resources have been made in some countries (e.g., United Kingdom) (2), other jurisdictions are not as prepared. If local capacity is overwhelmed, then some critically ill patients may be subject to triage and denied ICU admission. Care for patients without H1N1 may also be curtailed as scheduled elective surgery is postponed. If and when triaging protocols are invoked, real-time analysis of patient outcomes will be ethically imperative. Accurate timely clinical research regarding epidemiology, diagnosis, and treatment will be crucial for critically ill patients presenting with potential H1N1 infection. Predictors of early respiratory deterioration will help to inform rational use of monitoring devices and critical care. The sensitivity and specificity of initial and sequential screening tests in the ICU setting are unknown. False-negative test results could prompt the premature lifting of isolation precautions and dangerously delay in treatment. False-positive test results could waste scarce drug supplies and increase demands on resources for isolation. Pharmacologic and technological therapies that favorably influence morbidity and mortality and decrease nosocomial transmission to other critically ill patients and health workers will require rapid identification and implementation.
Barriers to research implementation in the ICU setting during the H1N1 pandemic in the northern hemisphere will be legion. These include the need to test for and treat H1N1 under emergency conditions, study an acute illness with its attendant high morbidity and mortality, and recruit patients within narrow time windows. The rapid onset and dissipation of the pandemic may preclude coordinated efforts, and research staff may be deployed to provide clinical care to a surge of critically ill patients. Some clinicians may be reluctant to enroll gravely ill patients into randomized trials, while resorting to unproven or potentially harmful treatments. Other clinicians may view the H1N1 pandemic as an incomparable opportunity, or a mandate, to answer urgent research questions that might otherwise never be answered. For instance, when several thousand patients with severe acute respiratory distress syndrome need mechanical ventilation within in a short time frame, readiness with a large, simple, international trial could evaluate the impact of systemic corticosteroids, antivirals, or high-frequency oscillatory ventilation.
Just as specialists in public health, family, emergency, pulmonary, and critical care medicine are planning the appropriate clinical response to the pandemic, investigators and institutions need to plan their research response. Herein, we propose recommendations to clinical researchers and research ethics boards (REBs) preparing for H1N1-related critical illness.
Institutional Review Boards
The need for research oversight by REBs is no less essential during a pandemic than in less emergent circumstances. REB oversight ensures that the rights, safety, and well-being of vulnerable participants are protected. Today, REBs face increasing demands because of an increasing number of studies and volume of regulations (3). Redundancy in the approval process for multicenter studies and variation in REB responses to the same study (4, 5) prompt the credo of “do it once and do it well” (6).
H1N1 research will escalate REB workload, particularly for multicenter protocols, which are typically reviewed by each participating center. Experience in the winter months of Australia, New Zealand, and South America portends a sudden influx of patients with peak ICU occupancy occurring within approximately 6 wks, decreasing quickly thereafter. Thus, adherence to the usual timelines for investigators to prepare and submit proposals to REBs, for multicenter REB review, and for investigator responses to REB conditions would mean that the pandemic will have passed before any research can be initiated.
We believe that during a pandemic, expedited and preferably centralized full review will help to ensure patient safety while avoiding delays that could block investigations necessary to advance knowledge and improve outcomes. A centralized process starts with one authorized national, provincial, state, or regional in-depth REB review; after conditions are met, the (potentially revised) protocol, central REB documents, and approval letter are submitted to all participating local REBs. Local REBs then consider and typically endorse the “central approval” and more rapidly review the protocol to primarily provide guidance to investigators on local adaptations and implications. When performed well, central REB review can expedite the process of research ethics oversight and obviate replication at each participating site. However, if central review merely duplicates full reviews, this could waste time or create confusion over differing opinions between central and local REBs. In Table 1, we present recommendations for REB approval processes for clinical research protocols during the H1N1 pandemic.
Critically ill patients invariably lack decision-making capacity, rendering first-person consent for research participation a rare event. Therefore, a priori consent is usually sought from surrogates who are often under considerable emotional stress. However, a priori consent from surrogates is not possible for adult patients in some countries (7). Furthermore, relying exclusively on a priori surrogate informed consent could preclude clinical research during the pandemic, thereby slowing enrollment into studies, denying potential participants the opportunity to benefit from research participation, limiting the generalizability of study results, delaying the identification of treatments as effective, ineffective, or harmful, and potentially foregoing the acquisition of new knowledge on the pandemic (8).
There is a growing recognition that alternative consent models require consideration in the setting of critical illness (Table 2) (9). The Declaration of Helsinki allows research involving individuals from whom it is not possible to obtain consent if the condition that prevents obtaining informed consent is a key characteristic of the research population (10). Many national research councils have attempted to strike a balance between the need for clinical research and the challenge of involving patients in this process at times of serious threat. For example, the Canadian Tri-Council Policy (11) permits research to be conducted in emergencies “without the free and informed consent of the subject in the presence of a serious threat requiring immediate intervention, where no efficacious standard of care exists or research offers a real possibility of direct patient-benefit and the risk of harm is not greater than that of standard care or is clearly justified by the direct benefits to the subject.” The U.S. Food and Drug Administration similarly allows for exceptions to informed consent (Table 3) (12). Despite this directive, waived consent is unusual for studies of critically ill patients, and individuals other than family members seldom provide surrogate consent in this setting (13).
Conditions have been proposed under which informed consent could be waived; we believe these should be invoked during the pandemic. These include when all treatments offered in the trial do not involve more than minimal additional risk in comparison with alternatives, and when treatments offered in the trial could be offered outside the trial without informed consent. Furthermore, genuine clinical equipoise needs to exist among the treatments offered, and patient preference for one treatment over any other needs to be unlikely. Finally, patients should be informed of the guidelines for waiver of informed consent so that they have the opportunity to seek additional information or care elsewhere (14).
Abuse of vulnerable persons under the auspices of research has been documented throughout history. Today, cultural shifts are occurring in the way that the public considers research participation. Biomedical knowledge obtained through clinical research is a public good available to benefit an individual, even if that individual does not contribute to it. The “public good argument” suggests that individuals at least consider research participation when approached unless they have a good reason not to, rather than the opposite, i.e., participate only if they have a good reason to do so (15).
Informed Consent for Observational Studies
Available evidence indicates that mandating traditional a priori first-person or surrogate informed consent sometimes may be contrary to the public good. We believe that this applies to research involving critically ill patients during the H1N1 pandemic regarding a priori informed consent for registries, audits, retrospective chart reviews, and prospective observational studies that do not influence patient care. For chart reviews (16), authorization bias can result in statistically significant differences in prognostic variables between participants and nonparticipants, threatening the validity and generalizability of study results (17).
Privacy legislation may lead to requests to obtain informed consent for enrollment in registries, which could also threaten the credibility of registry results. For example, mandatory a priori first-person consent for enrollment in the Canadian Stroke Registry initially resulted in inclusion of only 39% of eligible patients, which increased to 51% when a dedicated research nurse worked at each site. Mortality was significantly lower in enrolled vs. eligible patients (7% vs. 22%, p < .001), thereby creating a selection bias. Unsuccessful attempts to obtain a priori informed consent for the Canadian Stroke Registry led to temporary cessation of enrollment (18).
Evidence: Informed Consent for Randomized Trials
Providing there has been peer scientific and REB review, many examples of alternate consent models exist for randomized trials of emergent or urgent conditions or treatments that permit clinical research on incapacitated patients. For example, a randomized trial of urgent albumin resuscitation was recently performed with deferred consent (19). For a randomized trial of emergency corticosteroids for head injury, different consent models were used (deferred, waived or a priori surrogate consent) in different jurisdictions (20). Within a randomized trial of intensive insulin therapy in which early implementation was considered important to rigorously evaluate the treatment effect, deferred consent was used in some centers in Australia and New Zealand, whereas a priori surrogate consent was used in Canada (21). When the consent model in a trial of corticosteroids for severe sepsis was changed from a priori surrogate consent to waived consent if no relative was located when inclusion criteria were met, enrollment increased from four to ten patients each month (22). In Table 4, we present recommendations for informed consent for clinical studies on critically ill patients during the H1N1 pandemic.
The H1N1 influenza pandemic is projected to precipitate the need for increased critical care resources for many patients worldwide; between 10% and 25% of these patients may die. We believe that the global research community has an ethical obligation not only to document the pandemic and understand its epidemiology but also to conduct high-quality diagnostic, therapeutic, and health services research that maximizes the chance of improving outcomes. The ethical framework for pandemic research, similar to pandemic treatment, should hinge not only on the individual but also on the entire population of those actually or potentially afflicted.
We believe that clinical critical care research during the H1N1 pandemic must be approached differently from research conducted under nonemergent circumstances. Investigators need to prepare methodologically sound protocols carefully and quickly, thoughtfully suggesting appropriate approaches to informed consent. REBs will need to provide research oversight but avoid delaying the approval of sound protocols, unnecessarily duplicating the oversight process, or requiring a priori consent for all study designs. The H1N1 pandemic offers an unparalleled opportunity to use reduce redundancy, develop common REB review processes, and facilitate research on critically ill patients.
Acknowledging the different cultures and contexts into which healthcare systems are embedded, there are lessons to be learned from studying the effect of differing approaches to consent in various jurisdictions. Deferred or waived consent models will result in research questions being answered more quickly, thereby producing results more easily generalized than would otherwise be the case. Such approaches may hasten the acquisition of new knowledge, allow effective treatments for use during the pandemic, and liberate scarce resources for other patients. Increasing societal appreciation that research participation is recognized as a societal good is particularly germane under these circumstances. Failure to improve outcomes through rigorous efficient investigations during the pandemic is as ethically irresponsible as failing to provide care itself. We call on those whose responsibilities and actions influence the conduct of critical care research to take the steps necessary to make such research a reality.
We thank members of several clinical research consortia participating in the International Forum for Acute Care Trialists (InFACT) Group for discussions related to these issues. We appreciate suggestions from Dr. John Granton on earlier drafts of this manuscript. Dr. Cook is a Research Chair of the Canadian Institutes for Health Research. Dr. Burns holds a Clinician Scientist Award from the Canadian Institutes for Health Research. Dr. Fowler holds a Clinician Scientist Award from the Ontario Ministry of Health and Long-term Care and is a Phase II Clinician Scientist Recipient from the Heart and Stroke Foundation of Canada.
3.Ashcroft RE, Newson AJ, Benn PM: Reforming Research Ethics Committees. BMJ 2005; 331:587–588
4.Druml C, Wolz M, Pleiner J, et al: Research ethics committees in Europe: Trials and tribulations. Intensive Care Med 2009; 35:1636–1640
5.Silverman HJ, LeMaire F: Ethics and research in critical care. Intensive Care Medicine 2006; 32:1697–1702
6.Hebert PC, Saginur R: Research ethics review: Do it once and do it well. Can Med Assoc J 2009; 180:597
7.Zamperetti N, Latronico N: Clinical research in critically ill patients: the situation in Italy. Intensive Care Med 2008; 34:1330–1332
8.Cook DJ, Moore-Cox A, Xavier D, et al: Randomized trials in vulnerable populations. Controlled Clin Trials 2008; 5:61–69
9.Burns KEA, Zubrinich C, Marshall J, et al: The “consent to research” paradigm in critical care: Challenges and potential solutions. Intensive Care Med 2009; 35:1655–1658
12.Biros MH: Research without consent: Current status, 2003. Ann Emerg Med 2003; 42:550–564
13.Mitka M: Aiding emergency research aim on report on exceptions to informed consent. JAMA 2007; 298:2608–2609
14.Truog RD, Robinson W, Randolph A, et al: Is informed consent always necessary for randomized controlled trials? N Engl J Med 1999; 10:804–807
15.Schaefer GO, Emanuel EJ, Wertheimer A: The obligation to participate in biomedical research. JAMA 2009; 302:67–72
16.Al Shahi R, Warlow C: Using patient-identifiable data for observational research and audit. BMJ 2000; 321:1031–1032
17.Tu JV, Willison JD, Silver FL; for the Investigators in the Registry of the Canadian Stroke Network: Impracticability of informed consent in the registry of the Canadian Stroke Network. N Engl J Med 2004; 350:1414–1421
18.Kho ME, Duffett M, Willison DJ, et al: Written informed consent and selection bias in observational studies using medical records: systematic review. BMJ 2009; 338:b866
19.The SAFE Study Investigators: A comparison of albumin and saline for fluid resuscitation in the intensive care unit. N Engl J Med 2004; 350:2247–2256
20.CRASH Trial Collaborators: Effect of intravenous corticosteroid on the death within 14 days in 10,008 adults with clinically significant head injury (MRC CRASH): Randomized placebo controlled trial. Lancet 2004; 364:1321–1328
21.The NICE-SUGAR Study Investigators: Intensive versus conventional glucose control in critically ill patients. N Engl J Med 2009; 360:1283–1297
22.Annane D, Outin H, Fisch C, et al: The effect of waiving consent on enrolment in a sepsis trial. Intensive Care Med 2004; 30:321–324
© 2010 by the Society of Critical Care Medicine and Lippincott Williams & Wilkins