- Email: [email protected]
Consent in resuscitation trials: Benefit or harm for patients and society?
Resuscitation (2006) 70, 360—368
REVIEW PAPER
Consent in resuscitation trials: Benefit or harm for patients and society?夽 Graham Nichol a,b,∗, Judy Powell a,c, Lois van Ottingham a,c, Ron Maier d, Tom Rea e, Jim Christenson f, Al Hallstrom a,c a
University of Washington Clinical Trial Center, Seattle, WA, United States Emergency Services, Seattle, WA, United States c Department of Biostatistics, University of Washington, Seattle, WA, United States d Department of Surgery, Harborview Medical Center, Seattle WA, United States e Department of Medicine, University of Washington, Seattle, WA, United States f Department of Surgery, University of British Columbia, Vancouver, Canada b
Received 10 October 2005 ; received in revised form 30 January 2006; accepted 30 January 2006 KEYWORDS Emergency treatment; Resuscitation; Clinical trials; Ethics
Summary Context: Research in an emergency setting is challenging because there may not be sufficient opportunity or time to obtain informed consent from the patient or their legally authorized representative. Such research can be conducted without prior consent if specific criteria are met. However consent is sometimes required for continued participation and may bias the results of the study. Objective: To review regulations related to waiver of consent in emergency research, and evidence of whether such regulations introduce bias. Results: Emergency research can be conducted without consent, either through community disclosure and consultation followed by patient or family notification and consent for continued participation after the intervention was applied, or under a minimal risk waiver. Review of the clinical record is necessary to determine important outcomes such as survival to discharge. If consent is required for this review but not granted, then these data are missing during analysis. If seriously ill or disadvantaged patients are less likely to assent, then investigators cannot determine reliably whether these vulnerable patients were harmed by the intervention. If missing data are different from complete data, then the analysis is susceptible to bias, and the conclusions could be misleading. Extrapolation from non-consent rates in resuscitation studies to results from the DAVID trial demonstrates that the rate of absence of data and information due to lack of assent can influence whether there is a significant difference between treatment groups (survival of control versus intervention: p = 0.04 for complete data; p = 0.08 for 10.8% lack of assent; p = 0.40 for 19.7% lack of assent).
夽
A Spanish translated version of the summary of this article appears as Appendix in the online version at doi:10.1016/j.resuscitation.2006.01.020 ∗ Corresponding author. 1107 NE 45th Street, Suite 505, Seattle, WA 98105, United States. Tel.: +1 206 685 1302. E-mail address: [email protected] (G. Nichol). 0300-9572/$ — see front matter © 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.resuscitation.2006.01.020
Consent in resuscitation trials
361
Conclusions: Exception from consent for emergency research should extend to review of the hospital record as the standard in emergency research. The only potential risk to patients associated with review of the clinical record after the intervention is loss of privacy and confidentiality. Appropriate safeguards can be taken to minimize this risk. © 2006 Elsevier Ireland Ltd. All rights reserved.
Contents Introduction ................................................................................................ Regulations governing emergency research.................................................................. Regulations governing privacy of individually-identifiable health information ............................... Need for hospital information to assess outcome............................................................ Bias .................................................................................................... Potential impact of consent process upon the estimate of the effect of treatment ..................... Discussion................................................................................................... Risks to individuals ..................................................................................... Lack of verification of harm ............................................................................ Risks to society with under-enrollment of vulnerable populations ...................................... An ethical approach to potential bias in the consent process ........................................... A regulatory approach to potential bias in the consent process ......................................... Implications for future emergency research studies .................................................... Conclusion .................................................................................................. Acknowledgments ......................................................................................... References ................................................................................................
Introduction Research involving human subjects usually requires informed consent to protect their autonomy and prevent their exposure to undue risks.1,2 Ideally this consists of patients being informed of the benefits and risks of the experimental interventions and their legal rights, before recruitment into the study or administration of the interventions. Surrogate consent may be obtained from the patient’s family or legally authorized representative. The Institutional Review Board (IRB) overseeing the conduct of a study is responsible for enforcing these standards. Research in an emergency setting is challenging because the window of opportunity to treat is short and often precludes time to obtain informed consent from the patient or their representative. Once enrolled in an emergency research trial, patients die in the field, die in the hospital or survive the event. Consent is sought from patients or their representatives for ongoing study participation, sometimes including review of the clinical record, but not always obtained. Review of the clinical record is necessary to verify important outcomes such as hospital discharge status. The need to protect potential research subjects who are vulnerable due to their relative lack of power because of the acuteness of their illness
361 361 362 363 363 364 365 365 365 366 366 366 367 367 367 367
(or other factors such as socioeconomic status) deserves special attention. While there is a strong need for ethical research, ‘‘balance is required between the competing values of access to experimental treatments that may benefit (or harm) the subject and the social need to conduct this research.3 ’’ Careful consideration of the principle of distributive justice proposed by the Belmont Report4 requires protection of individual patients and vulnerable groups. Since vulnerable populations are at particular risk of cardiac arrest and traumatic injury,5—11 any biases in the conduct of resuscitation trials would particularly affect these communities. In this review we consider the American regulations related to consent in emergency research, the impact these have on the ability to follow a patient’s course through hospitalization, the need for hospital information to assess outcome, the evidence of whether these requirements introduce bias, and their implications for two planned emergency research projects.
Regulations governing emergency research Research that involves human subjects in an emergency setting in the United States must protect
362 their autonomy (45 CFR 46 and 21 CFR 50 and 56)a ; especially when they are not competent (21 CFR 50); as well as their privacy and confidentiality (Health Insurance Portability and Accountability Act of 1996 [HIPAA], 67 CFR 157). Patients or their representatives must provide informed consent prior to being enrolled in a research study unless the study involves minimal risk or meets criteria for exception from consent. Studies conducted under minimal risk are those that have an anticipated probability and magnitude of harm that is not greater intrinsically than those ordinarily encountered in daily life or during the performance of routine care. This risk is determined by the risk of the primary disease state, not the risk of the intervention. Since cardiac arrest is a high-risk disease, diagnostic or therapeutic research in cardiac arrest is usually excluded from the minimal risk category. The concept of a risk threshold such as minimal risk applies only to nontherapeutic interventions used to collect data or to basic science research.12,13 In recognition of the difficulties inherent in trying to obtain informed consent from a patient with an acute life-threatening illness, the Food and Drug Administration (FDA) and Department of Health and Human Services developed regulations that provide an exception to the requirement to obtain informed consent prior to enrollment in a clinical investigation. For research that is subject to FDA regulation (i.e. conducted under an Investigational Device Exemption (IDE) or Investigational New Drug (IND) permit), exception from consent can be applied to emergency research if (a) an IDE or IND is in effect, (b) the research involves human subjects who cannot consent because of their emerging lifethreatening medical condition for which available treatments are unproven or unsatisfactory, (c) the intervention must be administered before informed consent from the patient’s legally authorized representative is feasible, (d) the sponsor has prior written permission from the FDA, (e) an independent data monitoring committee exists, and (f) the relevant IRB has documented that these conditions were met. For research that is not supported by federal funding or subject to FDA regulation (e.g. a trial of a novel method of manual cardiopulmonary resuscitation), IRBs generally apply similar criteria to a Emergency research is governed by similar FDA (21 CFR) and Department of Health and Human Services Regulations (45 CFR). The former refers to exception from consent for emergency research whereas the latter refers to waiver of consent. For clarity, we refer to exception from consent. Both agencies allow waiver of documentation of informed consent in instances of minimal risk.
G. Nichol et al. determine whether exception from consent is warranted for an emergency research study. In either case, research that is conducted under an exception from consent for emergency research must hold out the prospect of direct benefit to the individual patient. Community notification and public disclosure to appropriate risk populations must be provided prior to study initiation, including an opportunity for the community to discuss the proposed study, its risks and potential benefits, and provide feedback to the IRB prior to study initiation; as well as to be informed of the study results following its completion. The investigator must provide information about the emergency research study at the earliest feasible opportunity to the patient or their representative after administration of the intervention. This should include notice that the patient may withdraw or discontinue participation in the study without penalty. If the patient dies, then procedures must be in place to provide information to the family member or legally authorized representative. The IRB determines whether it is desirable to have an actual document signed by the patient or their representative for continued participation in a study. This does not constitute informed consent for what has already occurred, but verifies agreement for continued participation in the study. The regulations lack specificity about whether data can be retained if it was collected prior to the patient or their representative declines to assent for continued participation. Such determination is left up to the local IRB.
Regulations governing privacy of individually-identifiable health information The Privacy Rule of the Health Insurance Portability and Accountability Act (HIPAA) of 1996 require efforts to reduce the risk of a breach of patients’ privacy or confidentiality in an emergency or other setting (67 CFR 157). Health care organizations must make reasonable efforts to limit the disclosure of protected health information to the minimum necessary to accomplish the intended purpose. Protected health information can be disclosed without consent or authorization for public health purposes; for activities related to the quality, safety and effectiveness of an FDA-regulated product or activity; or for research purposes. Disclosure to an individual who is subject to the jurisdiction of FDA is limited to (a) reporting adverse events or problems, or biological product deviations; (b) track-
Consent in resuscitation trials ing products for recalls, repairs or replacement; and (c) conducting post-marketing surveillance as directed by the FDA. Disclosure for research purposes requires documentation of waiver of authorization from an IRB or Privacy Board, including documentation that certain criteria have been met. Disclosure of protected health information is allowable with no more than a minimal risk to the privacy of patients if there is an adequate plan to protect the identifiers from improper use and disclosure; an adequate plan to destroy the identifiers at the earliest opportunity feasible, unless there is a health or research justification for retaining the identifiers or such retention is otherwise required by law; and there are adequate written assurances that the protected health information will not be reused or disclosed to anyone else. Collectively regulations governing research in an emergency setting are intended to protect patient autonomy, assumption of undue research risks and breaches of privacy or confidentiality. Sometimes IRBs require consent before or after administration of the study intervention in order for researchers to assess outcome.
Need for hospital information to assess outcome The goal of research in emergency conditions is to identify treatments that improve meaningful survival. There are several approaches to overcome the potential inability to follow patient’s course through hospitalization due to lack of consent. One approach would be to use hospital/emergency department admission status, which is typically available to those enrolling patients in the study. However, a substantial portion of the mortality occurs after hospitalization. In two trials, patients with ventricular fibrillation who were randomized to amiodarone were more likely to survive to hospitalization but not to hospital discharge.14,15 Thus although use of admittance status as a surrogate outcome can provide some guidance, it is not sufficient to change clinical practice. An alternative would be to use publicly available records such as death certificates to determine vital status and location of death among those who survive to hospitalization. Data from this approach to determining survival status frequently is not available in a timely manner. In summary, review of a clinical record to verify discharge status is necessary in emergency research since the use of intermediate outcomes or vital status data may not provide sufficient information to assess an intervention’s effectiveness.
363
Bias If consent is required to review a clinical record after administration of a study intervention but consent is not obtained, then the researcher is obliged not to include these data in analysis of the study results. If patients with missing data are different from those with complete data, then the analysis is susceptible to bias in the estimate of the effectiveness in the overall population or in vulnerable subgroups, and the conclusions are likely to be misleading. Furthermore, if seriously ill or disadvantaged patients are less likely to assent, then lack of review of their charts means that investigators cannot determine reliably whether these vulnerable patients were harmed by the intervention. This implies that the consent process may increase the bias of a study and the risk to participating for future patients. A study with such bias would neither benefit society nor the patients who received the study intervention. Several recent studies demonstrate whether the consent process introduces bias into research conducted in emergency settings. Before publication of the regulations governing emergency research with exception from consent, enrollment in a randomized trial to evaluate the use of hypothermia in patients with acute brain injury was voluntarily suspended.16 Pilot studies that sought to obtain consent from patient’s family members or authorized representatives skewed the study population by excluding minority patients whose proxies did not have telephones and could not be notified that their family member had been injured.17 The investigators recognized the sensitive nature of enrolling minorities, and the selection bias that would result from their lack of enrollment. They obtained permission from the Secretary of Health and Human Services to conduct the study with a waiver of informed consent. Minorities were underrepresented in the trial before but not after the waiver of consent was granted. The Public Access Defibrillation (PAD) trial enrolled patients who were treated by lay responders trained in either standard cardiopulmonary resuscitation (CPR) or those trained in CPR plus the use of an automated external defibrillator (AED).18 Consent was sought for follow-up prior to or shortly after discharge of survivors from hospital. There was a trend towards a higher rate of consent in the treatment arm (75% of CPR-only survivors consented, 86% of CPR + AED survivors consented) (Table 1). Since the primary outcome of the study was survival to discharge, lack of consent for follow-up did not bias the study.
364 Table 1
G. Nichol et al. Consent outcomes in PAD trial CPR-only
CPR + AED
No consent 4 (25.0%)
Consent 12 (75.0%)
No consent 5 (14.3%)
Consent 30 (85.7%)
3 (75.0%) 0 (0.0%) 3 (75.0%) 63.5 (11.0)
10 (83.3%) 0 (0.0%) 7 (58.3%) 61.8 (13.4)
4 (80.0%) 1 (20.0%) 3 (60.0%) 65.0 (13.3)
25 (83.3%) 2 (15.4%) 17 (56.7%) 56.1 (18.3)
CPC score at discharge, estimated from patient/family interactions 1 3 (75.0%) 8 (66.7%) ≥2 0 (0.0%) 4 (33.3%) Unknown 1 (25.0%) 0 (0.0%)
2 (40.0%) 3 (60.0%) 0 (0.0%)
23 (76.7%) 7 (23.3%) 0 (0.0%)
Sex, number (%) male Race, number (%) minority Race, % unknown Age, mean (S.D.)
The ASPIRE trial evaluated a band compression CPR device in patients with out-of-hospital cardiac arrest (personal communication, A. Hallstrom, 7 July 2005). In this study consent was sought from survivors for follow-up either prior to or shortly after discharge from hospital. Since the primary results are unpublished, we report here percentages but not raw numbers. To date, 77.4% of control survivors and 76.9% of intervention survivors consented. Although the overall consent rate was similar between control and intervention, there was a trend towards fewer women and fewer minorities consenting in the control arm (Table 2). In one of the above studies, a participating site initially assumed that obtaining oral telephone consent from survivors was sufficient to permit review of the clinical record to determine neurological outcome. The site’s IRB subsequently determined that this was insufficient, the survivors could not be contacted again to seek written consent, and the data from these nine survivors had to be excluded from all analyses. One of the study interventions in the trial was associated with significant harm (p = 0.005) before but not after (p = 0.1) these data were excluded. The Inflammation and Host Response to Injury, Large-Scale Collaborative Research Program (Glue Grant, http://www.gluegrant.org/) is an ongoing prospective cohort study of patients with major traumatic injury to identify gene expression profiles and leukocyte phenotype patterns from whole
Table 2
blood that may identify which traumatized or burned patients will develop multisystem organ failure. Consent is sought for ongoing participation prior to discharge of survivors from hospital. The overall rate of non-consent to date is 10.7% in the Seattle site (personal communication, R. Maier, July 19, 2005).
Potential impact of consent process upon the estimate of the effect of treatment The consent rates observed in these resuscitation studies were extrapolated onto results from the DAVID trial to estimate the impact of lack of consent to review a clinical record upon the power of a study to detect a treatment effect.19 The DAVID trial evaluated the efficacy of dual-chamber pacing (DDDR-70) compared with backup ventricular pacing (VVI-40) in patients with standard indications for implantable defibrillator implantation but without indications for pacing for bradycardia. About 506 patients were included. Health-related quality of life was assessed by using the Minnesota Living with Heart Failure Questionnaire (MLHFQ) at baseline and periodically after enrollment.20 Patients who received VVI-40 had greater survival compared to those who received DDDR-70 (log rank p-value 0.04). Assuming the very seriously ill patients are likely to decline because they feel too sick and the very well are likely to decline because they
Consent outcomes in the ASPIRE Trial Manual CPR
Male sex, number (%) Minority race, number (%) Age in years, mean (S.D.)
Manual CPR and autopulse
No consent (22.6%)
Consent (77.4%)
No consent (23.1%)
Consent (76.9%)
42% 33% 65 (20)
71% 20% 59 (12)
56% 11% 59 (17)
50% 20% 62 (13)
Consent in resuscitation trials
365
Discussion
Figure 1 Effect of low dropout rate on estimate of survival in DAVID trial.
Current American requirements for emergency research can introduce bias and expose participating and future patients to harm. If an IRB determines that an intervention has enough scientific merit and life-saving potential to allow its evaluation under an exception from informed consent and the consultation process determines support for such research in the community then the IRB should require that in-hospital data be reviewed without consent in order to reduce bias and increase protection of human subjects.
Risks to individuals
Figure 2 Effect of moderate dropout rate on estimate of survival in the DAVID trial.
have recovered from their illness and moved on to other activities, so patients with the bottom 7.6% and top 3.2% baseline MLHFQ scores were missed due to lack of assent,b then there would be no significant difference between treatment groups (Figure 1, p = 0.08). If those with the bottom 13.6% and top 6.1% MLHFQ scores were missed due to lack of assent,c then there would be no significant difference between treatment groups (Figure 2, p = 0.40). The sample size of a trial can be increased to account for such missing data if there is no treatment-related bias, but it is inefficient to do so.d Furthermore, if there is treatment-related bias, increasing the sample size will only result in greater confidence in an ‘‘incorrect’’ result. In summary, although the population studied in the DAVID trial did not require acute resuscitation, extrapolation from the results of this study demonstrate that lack of consent to review clinical data in a resuscitation trial can threaten its validity.
b This non-consent rate is similar to that observed in the GLUE grant. c This non-consent rate is less than that in the PAD trial. d If 10% of data are missing, then the sample must be increased by 1/0.92 to maintain power.21
After administration of the study intervention, the primary risk to individual patients if their clinical record is reviewed without consent is a breach in privacy and confidentiality. A signed consent does not entitle researchers to breach confidentiality. Avoiding a breach of patient confidentiality is thus not a consent issue but a technical issue that should be addressed by the appropriate institutional practices and enforced by the IRB. For example, all subject identifiers in any study database should be removed and replaced by a coded link known only to the local providers of patient care. In addition, all study personnel involved in data collection and analysis should be required to sign a confidentiality agreement.
Lack of verification of harm Studies that are conducted under FDA jurisdiction, and by extension studies conducted under any IRB jurisdiction, require complete and timely reporting of all potential adverse events. In the emergency setting, this requires that both prehospital and hospital records be reviewed. Since it is difficult to predict which outcomes should be considered as adverse events in advance, data describing all potentially relevant outcomes should be collected throughout the subjects’ hospital stay. If consent is required to review the clinical record and not granted, then researchers will be unable to ascertain whether an individual patient was harmed by the experimental intervention. Although an individual clinician may be able to ascertain known adverse events, only by looking at the data in aggregate will significant22 or idiosyncratic adverse events23 be identified. Such incomplete assessment of adverse effects in a trial is undesirable as it exposes both study patients and society to undue risk. Specifically, it is our duty to the participants
366 and the community to collect this data since we undertook the study without consent.
Risks to society with under-enrollment of vulnerable populations Vulnerable populations, including those who are minorities or economically disadvantaged, are at greater risk of need for resuscitation and often have poorer outcomes after attempts to do so,5—11 as well as being more likely to not consent.16,17 The resulting increased uncertainties in treatment benefit or harm in these vulnerable populations reduces the likelihood that health disparities could be reduced by applying effective interventions or not applying ineffective interventions.
An ethical approach to potential bias in the consent process How to best protect individuals and vulnerable populations in a research study can be evaluated by using an ethical framework. If informed consent is seen, as it is in the traditional protective conception of justice in the research setting, as an absolute demand, then neither intervention administration nor clinical record review without consent is just. That rules for research under exemption exist implies that an alternative view prevails justice consists of balancing independent values, none of which are absolute. These values include the societal need for research in an emergency setting to test treatments for patients presenting with acute crises; the potential benefit to some patient (those in the treatment group who receive new therapies if those new therapies are successful); the need to protect these individuals from being exploited by researchers and harmed by new therapies that turn out to be harmful; the right of all individuals not to be used as research subjects without consent; and the potential distress inflicted on patients or their representatives after the actual intervention has occurred. In some cases these principles may be able to be satisfied jointly. Community consultation addresses the first two of these values; patient notification and opportunity to withdraw address the last. However, ascertainment of potential harm, a key component of the research, is not feasible if consent is required to review the clinical record. Potential harm to individuals and society due to the consent process is rarely considered but should be if we are to achieve the optimal balance between risk and benefit. To achieve this balance, we propose that review of the complete hospital record should be the stan-
G. Nichol et al. dard in emergency research in order to protect individual subjects as well as the well-being of vulnerable groups. Review of the clinical record after administration of the study intervention constitutes minimal risk, is required in any trial conducted under the jurisdiction of the FDA, and can be conducted under a waiver of documented informed consent if ‘‘the research presents no more than minimal risk of harm to subjects and involves no procedures for which written consent is normally required outside of the research context.’’ (CFR 46 118(C)(2) appropriate safeguards must be used to ensure that identifiable data are protected.
A regulatory approach to potential bias in the consent process An alternative approach to the potential lack of verification of harm in emergency research would be to remove the current regulations that permit this waiver of consent in emergency research. However we believe that our approach is consistent with the existing regulations as described in the preamble to the final rule for 21 CFR 50: ‘‘The agency thinks that it may not always be possible to develop a meaningful informed consent document for continued participation in the research, because the relevant information may vary significantly depending upon when it becomes feasible to provide the information to the subject or legally authorized representative. The agency is therefore, not requiring that such a form be developed. The agency notes however that Sec. 50.24 (a)(6) places the responsibility on the IRB to review and approve ‘informed consent procedures and an informed consent document’ for use with subjects or their legal representatives, and procedures and information to be used in consultations with family members, in situations where use of such procedures is feasible.’’ [Page 51520] During the comment period for these regulations, the agency received feedback that the subject should be able to choose to continue to participate fully in a study, to continue the intervention but not have their data included in the research database or results, or to discontinue the intervention and use of the subject’s data. This was rejected on the following grounds: ‘FDA regulations. . . require investigators to prepare and maintain adequate case histories recording all observations and other data pertinent to the investigation on each individual treated with the drug or exposed to the device. The agency needs all such data in order to be able to determine the safety
Consent in resuscitation trials
367
and effectiveness of the device. The fact of having been in an investigation cannot be taken back. Also if a subject were able to control the use (inclusion and exclusion) of his or her data, and particularly if the clinical investigation were not blinded, the bias potential would be immense.’ [Page 51520]
adverse effects, reduces the potential for bias and constitutes (only) minimal risk due to breach of confidentiality. Hence we propose that exception from consent should extend through in-hospital clinical record review and be the standard in emergency research so long as identifiable data are protected.
Implications for future emergency research studies
Acknowledgments
The Resuscitation Outcomes Consortium was funded by the National Institutes of Health and other agencies to conduct a series of trials of interventions in life-threatening emergencies in the out-of-hospital setting (http://www.uwctc.org/ accessed on July 20, 2005). As such, all of the protocols will be conducted under an exemption from informed consent. Two studies are proposed, the first will use a single bolus of 250 cc of hypertonic saline, with and without dextran, compared to normal saline in patients with severe hypotension or traumatic brain injury (TBI). The second study will evaluate interventions to improve blood flow during CPR for patients in cardiac arrest. Both of these studies will be conducted in the field without further intervention during the hospital stay. For the hypertonic saline study, the only additional participation for those with hypotension includes contacting and consenting patients who were discharged from the hospital prior to 28 days to check on vital status and rehospitalization. For brain injured patients, consent will be sought to verify neurological status at 6 months post event by telephone interview. In the cardiac arrest study, the primary outcome is survival to hospital discharge with consent sought for follow-up after discharge. For each of these studies, an ability to ascertain hospital outcomes on every enrolled patient would reduce the potential for bias and would improve the protection of vulnerable individuals who have already received study interventions. Therefore waiver of informed consent to review a clinical record in both of these studies seems justifiable and indicated. We shall report our experience in seeking review of the clinical record without consent upon completion of these studies.
Conclusion Research involving human subjects in an emergency setting is challenging. Consent for ongoing participation in an emergency research study can introduce bias. Extension of an exception from consent for emergency research to include hospital clinical record review assures an assessment of survival and
Supported in part by Contracts #N01-HC-95177 and 1 U01 HL077863 from the National Heart, Lung, and Blood Institute, Bethesda, MD, and 1 U54 GM6211901A1 from the National Institute of General Medical Sciences, Bethesda, MD; and grants from St. Jude Corporation, Sylmar, CA and ZOLL Medical Corporation, Chelmsford, MA.
References 1. US Department of Health and Human Services. Protection of Human Subjects. 45 CFR 46116;1997. p. 115. 2. Food and Drug Administration of the United States. Protection of Human Subjects. 21 CFR 5023 (1979);1998. p. 246—7. 3. Kahn JP, Mastroianni AC, Sugarman J. Changing claims about justice in research: an introduction and overview. In: Kahn JP, Mastroianni AC, Sugarman J, editors. Beyond consent seeking justice in research. New York: Oxford University Press; 1998. p. 4. 4. The National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. The Belmont Report: Ethical Principles and Guidelines for the Protection of Human Subjects. Washington, DC;1979. 5. Becker LB, Han BH, Meyer PM, et al. Racial differences in the incidence of cardiac arrest and subsequent survival. The CPR Chicago Project. N Engl J Med 1993;329(9):600—6. 6. Iwashyna TJ, Christakis NA, Becker LB. Neighborhoods matter: a population-based study of provision of cardiopulmonary resuscitation. Ann Emerg Med 1999;34(4 Pt 1):459—68. 7. Sampson RJ, Raudenbush SW, Earls F. Neighborhoods and violent crime: a multilevel study of collective efficacy. Science 1997;277(5328):918—24. 8. Sampson RJ, Morenoff JD, Raudenbush S. Social anatomy of racial and ethnic disparities in violence. Am J Public Health 2005;95(2):224—32. 9. Dunn L, Henry J, Beard D. Social deprivation and head injury: a national study. J Neurol Neurosurg Psychiatry 2003;74(8): 1060—4. 10. Cubbin C, LeClere FB, Smith GS. Socioeconomic status and the occurrence of fatal and nonfatal injury in the United States. Am J Public Health 2000;90(1):70—7. 11. Cubbin C, LeClere FB, Smith GS. Socioeconomic status and injury mortality: individual and neighbourhood determinants. J Epidemiol Community Health 2000;54(7):517—24. 12. McRae AD, Weijer C. Lessons from everyday lives: a moral justification for acute care research. Crit Care Med 2002;30(5):1146—51. 13. Anonymous. Chapter 4: assessing risks and potential benefits and evaluating vulnerability. Ethical and policy issues in research involving human participants. Betheseda, MD: National Bioethics Advisory Commission; 2001. p. 69—96.
368 14. Kudenchuk PJ, Cobb LA, Copass MK, et al. Amiodarone for resuscitation after out-of-hospital cardiac arrest due to ventricular fibrillation. N Engl J Med 1999;341(12):871—8. 15. Dorian P, Cass D, Schwartz B, Cooper R, Gelaznikas R, Barr A. Amiodarone as compared with lidocaine for shockresistant ventricular fibrillation. N Engl J Med 2002;346(12): 884—90. 16. Clifton GL, Miller ER, Choi SC, et al. Lack of effect of induction of hypothermia after acute brain injury. N Engl J Med 2001;344(8):556—63. 17. Clifton GL, Knudson P, McDonald M. Waiver of consent in studies of acute brain injury. J Neurotrauma 2002;19(10): 1121—6. 18. Anonymous. Public-access defibrillation and survival after out-of-hospital cardiac arrest. N Engl J Med 2004;351(7): 637—46. 19. Wilkoff BL, Cook JR, Epstein AE, et al. Dual-chamber pacing or ventricular backup pacing in patients with an implantable
G. Nichol et al.
20.
21.
22.
23.
defibrillator: the Dual Chamber and VVI Implantable Defibrillator (DAVID) trial. JAMA 2002;288(24):3115—23. Rector TS, Cohn JN. Assessment of patient outcome with the Minnesota living with heart failure questionnaire: reliability and validity during a randomized double-blind, placebo controlled trial of pimobendan. Am Heart J 1992;124:1017—25. Lachin JM. Introduction to sample size determination and power analysis for clinical trials. Controlled Clin Trials 1981;2:93—114. Anonymous. Preliminary report: effect of encainide and flecainide on mortality in a randomized trial of arrhythmia suppression after myocardial infarction. The Cardiac Arrhythmia Suppression Trial (CAST) Investigators. N Engl J Med 1989;321(6):406—12. Pfeffer MA, Keech A, Sacks FM, et al. Safety and tolerability of pravastatin in long-term clinical trials: prospective Pravastatin Pooling (PPP) Project. Circulation 2002;105(20): 2341—6.
REVIEW PAPER
Consent in resuscitation trials: Benefit or harm for patients and society?夽 Graham Nichol a,b,∗, Judy Powell a,c, Lois van Ottingham a,c, Ron Maier d, Tom Rea e, Jim Christenson f, Al Hallstrom a,c a
University of Washington Clinical Trial Center, Seattle, WA, United States Emergency Services, Seattle, WA, United States c Department of Biostatistics, University of Washington, Seattle, WA, United States d Department of Surgery, Harborview Medical Center, Seattle WA, United States e Department of Medicine, University of Washington, Seattle, WA, United States f Department of Surgery, University of British Columbia, Vancouver, Canada b
Received 10 October 2005 ; received in revised form 30 January 2006; accepted 30 January 2006 KEYWORDS Emergency treatment; Resuscitation; Clinical trials; Ethics
Summary Context: Research in an emergency setting is challenging because there may not be sufficient opportunity or time to obtain informed consent from the patient or their legally authorized representative. Such research can be conducted without prior consent if specific criteria are met. However consent is sometimes required for continued participation and may bias the results of the study. Objective: To review regulations related to waiver of consent in emergency research, and evidence of whether such regulations introduce bias. Results: Emergency research can be conducted without consent, either through community disclosure and consultation followed by patient or family notification and consent for continued participation after the intervention was applied, or under a minimal risk waiver. Review of the clinical record is necessary to determine important outcomes such as survival to discharge. If consent is required for this review but not granted, then these data are missing during analysis. If seriously ill or disadvantaged patients are less likely to assent, then investigators cannot determine reliably whether these vulnerable patients were harmed by the intervention. If missing data are different from complete data, then the analysis is susceptible to bias, and the conclusions could be misleading. Extrapolation from non-consent rates in resuscitation studies to results from the DAVID trial demonstrates that the rate of absence of data and information due to lack of assent can influence whether there is a significant difference between treatment groups (survival of control versus intervention: p = 0.04 for complete data; p = 0.08 for 10.8% lack of assent; p = 0.40 for 19.7% lack of assent).
夽
A Spanish translated version of the summary of this article appears as Appendix in the online version at doi:10.1016/j.resuscitation.2006.01.020 ∗ Corresponding author. 1107 NE 45th Street, Suite 505, Seattle, WA 98105, United States. Tel.: +1 206 685 1302. E-mail address: [email protected] (G. Nichol). 0300-9572/$ — see front matter © 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.resuscitation.2006.01.020
Consent in resuscitation trials
361
Conclusions: Exception from consent for emergency research should extend to review of the hospital record as the standard in emergency research. The only potential risk to patients associated with review of the clinical record after the intervention is loss of privacy and confidentiality. Appropriate safeguards can be taken to minimize this risk. © 2006 Elsevier Ireland Ltd. All rights reserved.
Contents Introduction ................................................................................................ Regulations governing emergency research.................................................................. Regulations governing privacy of individually-identifiable health information ............................... Need for hospital information to assess outcome............................................................ Bias .................................................................................................... Potential impact of consent process upon the estimate of the effect of treatment ..................... Discussion................................................................................................... Risks to individuals ..................................................................................... Lack of verification of harm ............................................................................ Risks to society with under-enrollment of vulnerable populations ...................................... An ethical approach to potential bias in the consent process ........................................... A regulatory approach to potential bias in the consent process ......................................... Implications for future emergency research studies .................................................... Conclusion .................................................................................................. Acknowledgments ......................................................................................... References ................................................................................................
Introduction Research involving human subjects usually requires informed consent to protect their autonomy and prevent their exposure to undue risks.1,2 Ideally this consists of patients being informed of the benefits and risks of the experimental interventions and their legal rights, before recruitment into the study or administration of the interventions. Surrogate consent may be obtained from the patient’s family or legally authorized representative. The Institutional Review Board (IRB) overseeing the conduct of a study is responsible for enforcing these standards. Research in an emergency setting is challenging because the window of opportunity to treat is short and often precludes time to obtain informed consent from the patient or their representative. Once enrolled in an emergency research trial, patients die in the field, die in the hospital or survive the event. Consent is sought from patients or their representatives for ongoing study participation, sometimes including review of the clinical record, but not always obtained. Review of the clinical record is necessary to verify important outcomes such as hospital discharge status. The need to protect potential research subjects who are vulnerable due to their relative lack of power because of the acuteness of their illness
361 361 362 363 363 364 365 365 365 366 366 366 367 367 367 367
(or other factors such as socioeconomic status) deserves special attention. While there is a strong need for ethical research, ‘‘balance is required between the competing values of access to experimental treatments that may benefit (or harm) the subject and the social need to conduct this research.3 ’’ Careful consideration of the principle of distributive justice proposed by the Belmont Report4 requires protection of individual patients and vulnerable groups. Since vulnerable populations are at particular risk of cardiac arrest and traumatic injury,5—11 any biases in the conduct of resuscitation trials would particularly affect these communities. In this review we consider the American regulations related to consent in emergency research, the impact these have on the ability to follow a patient’s course through hospitalization, the need for hospital information to assess outcome, the evidence of whether these requirements introduce bias, and their implications for two planned emergency research projects.
Regulations governing emergency research Research that involves human subjects in an emergency setting in the United States must protect
362 their autonomy (45 CFR 46 and 21 CFR 50 and 56)a ; especially when they are not competent (21 CFR 50); as well as their privacy and confidentiality (Health Insurance Portability and Accountability Act of 1996 [HIPAA], 67 CFR 157). Patients or their representatives must provide informed consent prior to being enrolled in a research study unless the study involves minimal risk or meets criteria for exception from consent. Studies conducted under minimal risk are those that have an anticipated probability and magnitude of harm that is not greater intrinsically than those ordinarily encountered in daily life or during the performance of routine care. This risk is determined by the risk of the primary disease state, not the risk of the intervention. Since cardiac arrest is a high-risk disease, diagnostic or therapeutic research in cardiac arrest is usually excluded from the minimal risk category. The concept of a risk threshold such as minimal risk applies only to nontherapeutic interventions used to collect data or to basic science research.12,13 In recognition of the difficulties inherent in trying to obtain informed consent from a patient with an acute life-threatening illness, the Food and Drug Administration (FDA) and Department of Health and Human Services developed regulations that provide an exception to the requirement to obtain informed consent prior to enrollment in a clinical investigation. For research that is subject to FDA regulation (i.e. conducted under an Investigational Device Exemption (IDE) or Investigational New Drug (IND) permit), exception from consent can be applied to emergency research if (a) an IDE or IND is in effect, (b) the research involves human subjects who cannot consent because of their emerging lifethreatening medical condition for which available treatments are unproven or unsatisfactory, (c) the intervention must be administered before informed consent from the patient’s legally authorized representative is feasible, (d) the sponsor has prior written permission from the FDA, (e) an independent data monitoring committee exists, and (f) the relevant IRB has documented that these conditions were met. For research that is not supported by federal funding or subject to FDA regulation (e.g. a trial of a novel method of manual cardiopulmonary resuscitation), IRBs generally apply similar criteria to a Emergency research is governed by similar FDA (21 CFR) and Department of Health and Human Services Regulations (45 CFR). The former refers to exception from consent for emergency research whereas the latter refers to waiver of consent. For clarity, we refer to exception from consent. Both agencies allow waiver of documentation of informed consent in instances of minimal risk.
G. Nichol et al. determine whether exception from consent is warranted for an emergency research study. In either case, research that is conducted under an exception from consent for emergency research must hold out the prospect of direct benefit to the individual patient. Community notification and public disclosure to appropriate risk populations must be provided prior to study initiation, including an opportunity for the community to discuss the proposed study, its risks and potential benefits, and provide feedback to the IRB prior to study initiation; as well as to be informed of the study results following its completion. The investigator must provide information about the emergency research study at the earliest feasible opportunity to the patient or their representative after administration of the intervention. This should include notice that the patient may withdraw or discontinue participation in the study without penalty. If the patient dies, then procedures must be in place to provide information to the family member or legally authorized representative. The IRB determines whether it is desirable to have an actual document signed by the patient or their representative for continued participation in a study. This does not constitute informed consent for what has already occurred, but verifies agreement for continued participation in the study. The regulations lack specificity about whether data can be retained if it was collected prior to the patient or their representative declines to assent for continued participation. Such determination is left up to the local IRB.
Regulations governing privacy of individually-identifiable health information The Privacy Rule of the Health Insurance Portability and Accountability Act (HIPAA) of 1996 require efforts to reduce the risk of a breach of patients’ privacy or confidentiality in an emergency or other setting (67 CFR 157). Health care organizations must make reasonable efforts to limit the disclosure of protected health information to the minimum necessary to accomplish the intended purpose. Protected health information can be disclosed without consent or authorization for public health purposes; for activities related to the quality, safety and effectiveness of an FDA-regulated product or activity; or for research purposes. Disclosure to an individual who is subject to the jurisdiction of FDA is limited to (a) reporting adverse events or problems, or biological product deviations; (b) track-
Consent in resuscitation trials ing products for recalls, repairs or replacement; and (c) conducting post-marketing surveillance as directed by the FDA. Disclosure for research purposes requires documentation of waiver of authorization from an IRB or Privacy Board, including documentation that certain criteria have been met. Disclosure of protected health information is allowable with no more than a minimal risk to the privacy of patients if there is an adequate plan to protect the identifiers from improper use and disclosure; an adequate plan to destroy the identifiers at the earliest opportunity feasible, unless there is a health or research justification for retaining the identifiers or such retention is otherwise required by law; and there are adequate written assurances that the protected health information will not be reused or disclosed to anyone else. Collectively regulations governing research in an emergency setting are intended to protect patient autonomy, assumption of undue research risks and breaches of privacy or confidentiality. Sometimes IRBs require consent before or after administration of the study intervention in order for researchers to assess outcome.
Need for hospital information to assess outcome The goal of research in emergency conditions is to identify treatments that improve meaningful survival. There are several approaches to overcome the potential inability to follow patient’s course through hospitalization due to lack of consent. One approach would be to use hospital/emergency department admission status, which is typically available to those enrolling patients in the study. However, a substantial portion of the mortality occurs after hospitalization. In two trials, patients with ventricular fibrillation who were randomized to amiodarone were more likely to survive to hospitalization but not to hospital discharge.14,15 Thus although use of admittance status as a surrogate outcome can provide some guidance, it is not sufficient to change clinical practice. An alternative would be to use publicly available records such as death certificates to determine vital status and location of death among those who survive to hospitalization. Data from this approach to determining survival status frequently is not available in a timely manner. In summary, review of a clinical record to verify discharge status is necessary in emergency research since the use of intermediate outcomes or vital status data may not provide sufficient information to assess an intervention’s effectiveness.
363
Bias If consent is required to review a clinical record after administration of a study intervention but consent is not obtained, then the researcher is obliged not to include these data in analysis of the study results. If patients with missing data are different from those with complete data, then the analysis is susceptible to bias in the estimate of the effectiveness in the overall population or in vulnerable subgroups, and the conclusions are likely to be misleading. Furthermore, if seriously ill or disadvantaged patients are less likely to assent, then lack of review of their charts means that investigators cannot determine reliably whether these vulnerable patients were harmed by the intervention. This implies that the consent process may increase the bias of a study and the risk to participating for future patients. A study with such bias would neither benefit society nor the patients who received the study intervention. Several recent studies demonstrate whether the consent process introduces bias into research conducted in emergency settings. Before publication of the regulations governing emergency research with exception from consent, enrollment in a randomized trial to evaluate the use of hypothermia in patients with acute brain injury was voluntarily suspended.16 Pilot studies that sought to obtain consent from patient’s family members or authorized representatives skewed the study population by excluding minority patients whose proxies did not have telephones and could not be notified that their family member had been injured.17 The investigators recognized the sensitive nature of enrolling minorities, and the selection bias that would result from their lack of enrollment. They obtained permission from the Secretary of Health and Human Services to conduct the study with a waiver of informed consent. Minorities were underrepresented in the trial before but not after the waiver of consent was granted. The Public Access Defibrillation (PAD) trial enrolled patients who were treated by lay responders trained in either standard cardiopulmonary resuscitation (CPR) or those trained in CPR plus the use of an automated external defibrillator (AED).18 Consent was sought for follow-up prior to or shortly after discharge of survivors from hospital. There was a trend towards a higher rate of consent in the treatment arm (75% of CPR-only survivors consented, 86% of CPR + AED survivors consented) (Table 1). Since the primary outcome of the study was survival to discharge, lack of consent for follow-up did not bias the study.
364 Table 1
G. Nichol et al. Consent outcomes in PAD trial CPR-only
CPR + AED
No consent 4 (25.0%)
Consent 12 (75.0%)
No consent 5 (14.3%)
Consent 30 (85.7%)
3 (75.0%) 0 (0.0%) 3 (75.0%) 63.5 (11.0)
10 (83.3%) 0 (0.0%) 7 (58.3%) 61.8 (13.4)
4 (80.0%) 1 (20.0%) 3 (60.0%) 65.0 (13.3)
25 (83.3%) 2 (15.4%) 17 (56.7%) 56.1 (18.3)
CPC score at discharge, estimated from patient/family interactions 1 3 (75.0%) 8 (66.7%) ≥2 0 (0.0%) 4 (33.3%) Unknown 1 (25.0%) 0 (0.0%)
2 (40.0%) 3 (60.0%) 0 (0.0%)
23 (76.7%) 7 (23.3%) 0 (0.0%)
Sex, number (%) male Race, number (%) minority Race, % unknown Age, mean (S.D.)
The ASPIRE trial evaluated a band compression CPR device in patients with out-of-hospital cardiac arrest (personal communication, A. Hallstrom, 7 July 2005). In this study consent was sought from survivors for follow-up either prior to or shortly after discharge from hospital. Since the primary results are unpublished, we report here percentages but not raw numbers. To date, 77.4% of control survivors and 76.9% of intervention survivors consented. Although the overall consent rate was similar between control and intervention, there was a trend towards fewer women and fewer minorities consenting in the control arm (Table 2). In one of the above studies, a participating site initially assumed that obtaining oral telephone consent from survivors was sufficient to permit review of the clinical record to determine neurological outcome. The site’s IRB subsequently determined that this was insufficient, the survivors could not be contacted again to seek written consent, and the data from these nine survivors had to be excluded from all analyses. One of the study interventions in the trial was associated with significant harm (p = 0.005) before but not after (p = 0.1) these data were excluded. The Inflammation and Host Response to Injury, Large-Scale Collaborative Research Program (Glue Grant, http://www.gluegrant.org/) is an ongoing prospective cohort study of patients with major traumatic injury to identify gene expression profiles and leukocyte phenotype patterns from whole
Table 2
blood that may identify which traumatized or burned patients will develop multisystem organ failure. Consent is sought for ongoing participation prior to discharge of survivors from hospital. The overall rate of non-consent to date is 10.7% in the Seattle site (personal communication, R. Maier, July 19, 2005).
Potential impact of consent process upon the estimate of the effect of treatment The consent rates observed in these resuscitation studies were extrapolated onto results from the DAVID trial to estimate the impact of lack of consent to review a clinical record upon the power of a study to detect a treatment effect.19 The DAVID trial evaluated the efficacy of dual-chamber pacing (DDDR-70) compared with backup ventricular pacing (VVI-40) in patients with standard indications for implantable defibrillator implantation but without indications for pacing for bradycardia. About 506 patients were included. Health-related quality of life was assessed by using the Minnesota Living with Heart Failure Questionnaire (MLHFQ) at baseline and periodically after enrollment.20 Patients who received VVI-40 had greater survival compared to those who received DDDR-70 (log rank p-value 0.04). Assuming the very seriously ill patients are likely to decline because they feel too sick and the very well are likely to decline because they
Consent outcomes in the ASPIRE Trial Manual CPR
Male sex, number (%) Minority race, number (%) Age in years, mean (S.D.)
Manual CPR and autopulse
No consent (22.6%)
Consent (77.4%)
No consent (23.1%)
Consent (76.9%)
42% 33% 65 (20)
71% 20% 59 (12)
56% 11% 59 (17)
50% 20% 62 (13)
Consent in resuscitation trials
365
Discussion
Figure 1 Effect of low dropout rate on estimate of survival in DAVID trial.
Current American requirements for emergency research can introduce bias and expose participating and future patients to harm. If an IRB determines that an intervention has enough scientific merit and life-saving potential to allow its evaluation under an exception from informed consent and the consultation process determines support for such research in the community then the IRB should require that in-hospital data be reviewed without consent in order to reduce bias and increase protection of human subjects.
Risks to individuals
Figure 2 Effect of moderate dropout rate on estimate of survival in the DAVID trial.
have recovered from their illness and moved on to other activities, so patients with the bottom 7.6% and top 3.2% baseline MLHFQ scores were missed due to lack of assent,b then there would be no significant difference between treatment groups (Figure 1, p = 0.08). If those with the bottom 13.6% and top 6.1% MLHFQ scores were missed due to lack of assent,c then there would be no significant difference between treatment groups (Figure 2, p = 0.40). The sample size of a trial can be increased to account for such missing data if there is no treatment-related bias, but it is inefficient to do so.d Furthermore, if there is treatment-related bias, increasing the sample size will only result in greater confidence in an ‘‘incorrect’’ result. In summary, although the population studied in the DAVID trial did not require acute resuscitation, extrapolation from the results of this study demonstrate that lack of consent to review clinical data in a resuscitation trial can threaten its validity.
b This non-consent rate is similar to that observed in the GLUE grant. c This non-consent rate is less than that in the PAD trial. d If 10% of data are missing, then the sample must be increased by 1/0.92 to maintain power.21
After administration of the study intervention, the primary risk to individual patients if their clinical record is reviewed without consent is a breach in privacy and confidentiality. A signed consent does not entitle researchers to breach confidentiality. Avoiding a breach of patient confidentiality is thus not a consent issue but a technical issue that should be addressed by the appropriate institutional practices and enforced by the IRB. For example, all subject identifiers in any study database should be removed and replaced by a coded link known only to the local providers of patient care. In addition, all study personnel involved in data collection and analysis should be required to sign a confidentiality agreement.
Lack of verification of harm Studies that are conducted under FDA jurisdiction, and by extension studies conducted under any IRB jurisdiction, require complete and timely reporting of all potential adverse events. In the emergency setting, this requires that both prehospital and hospital records be reviewed. Since it is difficult to predict which outcomes should be considered as adverse events in advance, data describing all potentially relevant outcomes should be collected throughout the subjects’ hospital stay. If consent is required to review the clinical record and not granted, then researchers will be unable to ascertain whether an individual patient was harmed by the experimental intervention. Although an individual clinician may be able to ascertain known adverse events, only by looking at the data in aggregate will significant22 or idiosyncratic adverse events23 be identified. Such incomplete assessment of adverse effects in a trial is undesirable as it exposes both study patients and society to undue risk. Specifically, it is our duty to the participants
366 and the community to collect this data since we undertook the study without consent.
Risks to society with under-enrollment of vulnerable populations Vulnerable populations, including those who are minorities or economically disadvantaged, are at greater risk of need for resuscitation and often have poorer outcomes after attempts to do so,5—11 as well as being more likely to not consent.16,17 The resulting increased uncertainties in treatment benefit or harm in these vulnerable populations reduces the likelihood that health disparities could be reduced by applying effective interventions or not applying ineffective interventions.
An ethical approach to potential bias in the consent process How to best protect individuals and vulnerable populations in a research study can be evaluated by using an ethical framework. If informed consent is seen, as it is in the traditional protective conception of justice in the research setting, as an absolute demand, then neither intervention administration nor clinical record review without consent is just. That rules for research under exemption exist implies that an alternative view prevails justice consists of balancing independent values, none of which are absolute. These values include the societal need for research in an emergency setting to test treatments for patients presenting with acute crises; the potential benefit to some patient (those in the treatment group who receive new therapies if those new therapies are successful); the need to protect these individuals from being exploited by researchers and harmed by new therapies that turn out to be harmful; the right of all individuals not to be used as research subjects without consent; and the potential distress inflicted on patients or their representatives after the actual intervention has occurred. In some cases these principles may be able to be satisfied jointly. Community consultation addresses the first two of these values; patient notification and opportunity to withdraw address the last. However, ascertainment of potential harm, a key component of the research, is not feasible if consent is required to review the clinical record. Potential harm to individuals and society due to the consent process is rarely considered but should be if we are to achieve the optimal balance between risk and benefit. To achieve this balance, we propose that review of the complete hospital record should be the stan-
G. Nichol et al. dard in emergency research in order to protect individual subjects as well as the well-being of vulnerable groups. Review of the clinical record after administration of the study intervention constitutes minimal risk, is required in any trial conducted under the jurisdiction of the FDA, and can be conducted under a waiver of documented informed consent if ‘‘the research presents no more than minimal risk of harm to subjects and involves no procedures for which written consent is normally required outside of the research context.’’ (CFR 46 118(C)(2) appropriate safeguards must be used to ensure that identifiable data are protected.
A regulatory approach to potential bias in the consent process An alternative approach to the potential lack of verification of harm in emergency research would be to remove the current regulations that permit this waiver of consent in emergency research. However we believe that our approach is consistent with the existing regulations as described in the preamble to the final rule for 21 CFR 50: ‘‘The agency thinks that it may not always be possible to develop a meaningful informed consent document for continued participation in the research, because the relevant information may vary significantly depending upon when it becomes feasible to provide the information to the subject or legally authorized representative. The agency is therefore, not requiring that such a form be developed. The agency notes however that Sec. 50.24 (a)(6) places the responsibility on the IRB to review and approve ‘informed consent procedures and an informed consent document’ for use with subjects or their legal representatives, and procedures and information to be used in consultations with family members, in situations where use of such procedures is feasible.’’ [Page 51520] During the comment period for these regulations, the agency received feedback that the subject should be able to choose to continue to participate fully in a study, to continue the intervention but not have their data included in the research database or results, or to discontinue the intervention and use of the subject’s data. This was rejected on the following grounds: ‘FDA regulations. . . require investigators to prepare and maintain adequate case histories recording all observations and other data pertinent to the investigation on each individual treated with the drug or exposed to the device. The agency needs all such data in order to be able to determine the safety
Consent in resuscitation trials
367
and effectiveness of the device. The fact of having been in an investigation cannot be taken back. Also if a subject were able to control the use (inclusion and exclusion) of his or her data, and particularly if the clinical investigation were not blinded, the bias potential would be immense.’ [Page 51520]
adverse effects, reduces the potential for bias and constitutes (only) minimal risk due to breach of confidentiality. Hence we propose that exception from consent should extend through in-hospital clinical record review and be the standard in emergency research so long as identifiable data are protected.
Implications for future emergency research studies
Acknowledgments
The Resuscitation Outcomes Consortium was funded by the National Institutes of Health and other agencies to conduct a series of trials of interventions in life-threatening emergencies in the out-of-hospital setting (http://www.uwctc.org/ accessed on July 20, 2005). As such, all of the protocols will be conducted under an exemption from informed consent. Two studies are proposed, the first will use a single bolus of 250 cc of hypertonic saline, with and without dextran, compared to normal saline in patients with severe hypotension or traumatic brain injury (TBI). The second study will evaluate interventions to improve blood flow during CPR for patients in cardiac arrest. Both of these studies will be conducted in the field without further intervention during the hospital stay. For the hypertonic saline study, the only additional participation for those with hypotension includes contacting and consenting patients who were discharged from the hospital prior to 28 days to check on vital status and rehospitalization. For brain injured patients, consent will be sought to verify neurological status at 6 months post event by telephone interview. In the cardiac arrest study, the primary outcome is survival to hospital discharge with consent sought for follow-up after discharge. For each of these studies, an ability to ascertain hospital outcomes on every enrolled patient would reduce the potential for bias and would improve the protection of vulnerable individuals who have already received study interventions. Therefore waiver of informed consent to review a clinical record in both of these studies seems justifiable and indicated. We shall report our experience in seeking review of the clinical record without consent upon completion of these studies.
Conclusion Research involving human subjects in an emergency setting is challenging. Consent for ongoing participation in an emergency research study can introduce bias. Extension of an exception from consent for emergency research to include hospital clinical record review assures an assessment of survival and
Supported in part by Contracts #N01-HC-95177 and 1 U01 HL077863 from the National Heart, Lung, and Blood Institute, Bethesda, MD, and 1 U54 GM6211901A1 from the National Institute of General Medical Sciences, Bethesda, MD; and grants from St. Jude Corporation, Sylmar, CA and ZOLL Medical Corporation, Chelmsford, MA.
References 1. US Department of Health and Human Services. Protection of Human Subjects. 45 CFR 46116;1997. p. 115. 2. Food and Drug Administration of the United States. Protection of Human Subjects. 21 CFR 5023 (1979);1998. p. 246—7. 3. Kahn JP, Mastroianni AC, Sugarman J. Changing claims about justice in research: an introduction and overview. In: Kahn JP, Mastroianni AC, Sugarman J, editors. Beyond consent seeking justice in research. New York: Oxford University Press; 1998. p. 4. 4. The National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. The Belmont Report: Ethical Principles and Guidelines for the Protection of Human Subjects. Washington, DC;1979. 5. Becker LB, Han BH, Meyer PM, et al. Racial differences in the incidence of cardiac arrest and subsequent survival. The CPR Chicago Project. N Engl J Med 1993;329(9):600—6. 6. Iwashyna TJ, Christakis NA, Becker LB. Neighborhoods matter: a population-based study of provision of cardiopulmonary resuscitation. Ann Emerg Med 1999;34(4 Pt 1):459—68. 7. Sampson RJ, Raudenbush SW, Earls F. Neighborhoods and violent crime: a multilevel study of collective efficacy. Science 1997;277(5328):918—24. 8. Sampson RJ, Morenoff JD, Raudenbush S. Social anatomy of racial and ethnic disparities in violence. Am J Public Health 2005;95(2):224—32. 9. Dunn L, Henry J, Beard D. Social deprivation and head injury: a national study. J Neurol Neurosurg Psychiatry 2003;74(8): 1060—4. 10. Cubbin C, LeClere FB, Smith GS. Socioeconomic status and the occurrence of fatal and nonfatal injury in the United States. Am J Public Health 2000;90(1):70—7. 11. Cubbin C, LeClere FB, Smith GS. Socioeconomic status and injury mortality: individual and neighbourhood determinants. J Epidemiol Community Health 2000;54(7):517—24. 12. McRae AD, Weijer C. Lessons from everyday lives: a moral justification for acute care research. Crit Care Med 2002;30(5):1146—51. 13. Anonymous. Chapter 4: assessing risks and potential benefits and evaluating vulnerability. Ethical and policy issues in research involving human participants. Betheseda, MD: National Bioethics Advisory Commission; 2001. p. 69—96.
368 14. Kudenchuk PJ, Cobb LA, Copass MK, et al. Amiodarone for resuscitation after out-of-hospital cardiac arrest due to ventricular fibrillation. N Engl J Med 1999;341(12):871—8. 15. Dorian P, Cass D, Schwartz B, Cooper R, Gelaznikas R, Barr A. Amiodarone as compared with lidocaine for shockresistant ventricular fibrillation. N Engl J Med 2002;346(12): 884—90. 16. Clifton GL, Miller ER, Choi SC, et al. Lack of effect of induction of hypothermia after acute brain injury. N Engl J Med 2001;344(8):556—63. 17. Clifton GL, Knudson P, McDonald M. Waiver of consent in studies of acute brain injury. J Neurotrauma 2002;19(10): 1121—6. 18. Anonymous. Public-access defibrillation and survival after out-of-hospital cardiac arrest. N Engl J Med 2004;351(7): 637—46. 19. Wilkoff BL, Cook JR, Epstein AE, et al. Dual-chamber pacing or ventricular backup pacing in patients with an implantable
G. Nichol et al.
20.
21.
22.
23.
defibrillator: the Dual Chamber and VVI Implantable Defibrillator (DAVID) trial. JAMA 2002;288(24):3115—23. Rector TS, Cohn JN. Assessment of patient outcome with the Minnesota living with heart failure questionnaire: reliability and validity during a randomized double-blind, placebo controlled trial of pimobendan. Am Heart J 1992;124:1017—25. Lachin JM. Introduction to sample size determination and power analysis for clinical trials. Controlled Clin Trials 1981;2:93—114. Anonymous. Preliminary report: effect of encainide and flecainide on mortality in a randomized trial of arrhythmia suppression after myocardial infarction. The Cardiac Arrhythmia Suppression Trial (CAST) Investigators. N Engl J Med 1989;321(6):406—12. Pfeffer MA, Keech A, Sacks FM, et al. Safety and tolerability of pravastatin in long-term clinical trials: prospective Pravastatin Pooling (PPP) Project. Circulation 2002;105(20): 2341—6.