Acta Anaesthesiol Scand 2013; 57: 408–416
Printed in Singapore. All rights reserved
An ethical analysis of proxy and waiver of consent in
critical care research
R. M. G. Berg1, K. Møller1,2 and P. J. H. Rossel3
1
Centre of Inflammation and Metabolism, Department of Infectious Diseases, Rigshospitalet, Copenhagen Ø, Denmark, 2Neurointensive Care
Unit 2093, Rigshospitalet, Copenhagen Ø, Denmark and 3Unit of Medical Philosophy and Clinical Theory, Institute of Public Health,
University of Copenhagen, Copenhagen K, Denmark
Abstract
It is a central principle in medical ethics that vulnerable patients
are entitled to a degree of protection that reflects their vulner-
ability. In critical care research, this protection is often estab-
lished by means of so-called proxy consent. Proxy consent for
research participation constitutes a substituted judgement by a
close relative or friend, based on knowledge of patient’s values,
preferences, and view of life. For the consent to be genuine, the
proxy must be informed of and understand three fundamental
aspects of research practice: (1) that participation is voluntary
and the consent can be withdrawn at any time; (2) that the
research is designed to benefit future patients and society as a
A major challenge in critical care research is that
critically ill patients are a particularly vulner-
able population of potential research subjects that
are temporarily unable to provide informed consent.
This reflects a fundamental conflict in medical
ethics: how can we involve humans in research for
the benefit of future patients and society as a whole,
without violating the basic rights of the individual?
The basic rights of the individual are encom-
passed in the so-called categorical imperative as stated
by Immanuel Kant in his formula of humanity:1
‘Act in such a way that you treat humanity,
whether in your own person or in the person of
any other, never merely as a means to an end, but
always at the same time as an end.’
This implies that any human being possesses an
inherent value and dignity and must be treated with
respect. Several rights can be derived from this fun-
damental guiding principle, and in the context of
medical research, these include the right to be pro-
tected from harm, exploitation, coercion, and deception.
Research should therefore attain beneficence, justice,
408
bs_bs_banner
© 2013 The Acta Anaesthesiologica Scandinavica Foundation
Published by Blackwell Publishing Ltd.
ACTA ANAESTHESIOLOGICA SCANDINAVICA
doi: 10.1111/aas.12083
Review Article
whole, and not the individual study participant; and (3) that
participation involves an incremental non-therapeutic risk. If this
is not fulfilled because the research is to be conducted under
circumstances where the proxy is unavailable, adequate protec-
tion of the patient must be ensured by other means. Thus, the
research must be designed specifically to benefit critically ill
patients, and the incremental non-therapeutic risk must only
comprise a minimal risk.
Accepted for publication 13 January 2013
© 2013 The Acta Anaesthesiologica Scandinavica Foundation
Published by Blackwell Publishing Ltd.
and respect for persons, and this is ensured through
independent research ethical committees that
review research protocols.2 Beneficence requires that
potential benefits for research subjects are maxim-
ised while potential harm is minimised.2 Justice
requires that research ‘does not prey on the vulner-
able’;2 hence, vulnerable populations must not be
exploited as easily accessible research subjects. They
may only be used as research subjects if the research
is designed specifically to benefit this particular
group, and the research cannot be conducted on less
vulnerable populations with a similar scientific
gain. Respect for persons requires that persons do not
serve as research subjects against their will or
knowledge, that is, they must not be coerced or
deceived to participate in research; research subjects
must therefore be treated as autonomous individu-
als and provide consent for research participation,2
and vulnerable populations with diminished
autonomy must be protected in a manner that
reflects their vulnerability.2 In critical care research,
this protection will often be established by proxy
consent.3 The necessity of immediate medical inter-
 Consent in critical care research

Table 1
Key concepts.
Capacity to consent: Ability to understand three fundamental aspects of research practice: (1) that participation is voluntary
and that the consent may be withdrawn at any time, (2) that the consent relates to a research protocol
designed to benefit future patients and society as a whole, and (3) that research participation involves
an incremental non-therapeutic risk.
Categorical imperative: Any human being possesses an inherent value and dignity and must be treated with respect.
Clinical equipoise: A state of honest, professional disagreement in the community of expert practitioners as to the preferred
treatment or procedure for a given clinical condition.
Clinical practice: The conduct of activities that have the prospect of benefiting the individual patient, whether it is in the
form of standard therapy or non-validated innovative therapy (the latter is sometimes designated
‘experimental therapy’).
Deferred consent: Consent obtained after the patient has been included in an ongoing research protocol. The consent
relates specifically to the patient’s continued participation in the study, and is not relevant if the
intervention is not ongoing at this time. If the intervention has been completed, a waiver of consent
has been effectuated, regardless of whether the approval of the patient or a proxy is subsequently
attained.
Genuine proxy consent: A consent based on a substituted judgement from a proxy who has the capacity to consent.
Informed consent: A procedure that serves to protect from coercion or deception by permitting an individual to accept or
refuse a proposed action. Informed consent presupposes (1) that the person who is to make the
decision has received adequate information regarding the proposal, (2) that the person has
understood the information delivered, (3) that the person is competent to make a decision, and (4) that
the decision is voluntary, in that the person is neither being forced or manipulated to make a decision
against his or her will (that is, the decision must be free from coercion and deception).
Non-therapeutic risk: Risk that is specifically encountered in research practice because of interventions that are solely
instituted to answer the specific research question. Such interventions do not have the prospect of
benefiting the individual patient, and the non-therapeutic risk is thus justified by the expected benefit of
the research for future patients and society as a whole. Non-therapeutic risks may for example be the
risks associated with additional blood sampling for pharmacokinetic studies and the recording of
biological data that are not normally considered in clinical practice.
Minimal risk: Risk of a magnitude that is common to us all in our everyday lives, and roughly equivalent to the risk of
a general medical examination. In research, this could be certain non-therapeutic risks, such as
additional analyses made on blood from blood samples that have already been obtained, or the
recording of data on a monitor.
Proxy consent: An informed consent provided on behalf of a patient by a proxy (sometimes designated a surrogate
decision maker).
Research practice: The conduct of activities that have the prospect of benefiting future patients and society as a whole,
but not the individual patient. May or may not involve therapeutic risks and always involves
non-therapeutic risks.
Substituted judgement: Judgement of what another person would have decided in a given situation had he or she been given
the choice, based on knowledge of the person’s values, preferences, and view of life.
Therapeutic risk: Risk imposed by procedures that have the prospect of benefiting the individual patient in both clinical
and research practice. These may for example be the administration of a drug, a particular ventilation
strategy, or insertion of a catheter that may improve monitoring for other therapeutic interventions.
Waiver of consent: A waiver of consent is present when a given research intervention is initiated without a prospective
genuine consent from either the patient or a proxy.

vention in emergencies may, however, render even
proxy consent unachievable, and if such research is
to ensue, a waiver of consent is inevitable.
In the present review, we highlight and discuss
the key concepts pertaining to the ethics of proxy
and waiver of consent in critical care research
(Table 1), in order to infer how the basic rights of
critically ill patients can be preserved, while serving
as research subjects for the benefit of future patients
and society as a whole. A thorough discussion of the
legislative issues pertaining to this area does,
however, reach beyond the scope of the present
paper, which focuses specifically on the normative
and descriptive ethical aspects.
Informed and proxy consent
Informed consent is a procedure that permits an
individual to accept or refuse a proposed action. In
this regard, it has four principal prerequisites:4 (1)
that the person who is to make the decision has
received adequate information regarding the pro-
posal, (2) that the persons has understood the infor-
mation delivered, (3) that the person is competent to
make a decision, and (4) that the person is neither
being coerced or deceived. Thus, informed consent
cannot be obtained from someone who is vulnerable
or dependent. This problem is encountered on an
everyday basis in the intensive care unit (ICU) and

409
R. M. G. Berg et al.
emergency department, where patients commonly
exhibit an altered mental state, ranging from confu-
sion and agitation to unconsciousness and coma.
When it comes to medical decisions, we therefore
rely on proxy consent. In principle, proxy consent
can take two different forms;3 it is either based on a
substituted judgement of what the patients might have
chosen in the given situation or on what is deemed
to be in the patient’s best interest.
Substituted judgement requires intimate knowl-
edge of the patient’s values, preferences, and view of
life, and can in reality be exercised only by close
relatives or friends.3 Although the best interest
principle is subordinate to substituted judgement,
they both apply to medical treatment decisions.
However, if the proxy consent concerns a research
protocol, it is to be based exclusively on a substi-
tuted judgement. This rests on the distinction
between clinical practice and research practice.5 Clini-
cal practice may both involve standard therapy
and/or innovative, non-validated therapy (some-
times designated ‘experimental therapy’), as long as
the given intervention is executed to benefit the par-
ticular patient. In contrast, research practice involves
interventions that are solely implemented to benefit
future patients and society as a whole. Although
research must not violate the best interest of the
individual patient, that is, it must not subject the
patient to deliberate harm or exploitation, it is not a
priori in the best interest of the individual patient.
This assessment rests not with close relatives or
friends, but with the independent research ethical
committee that reviewed and approved the research
protocol. Thus, a proxy consent for research partici-
pation specifically constitutes a substituted judge-
ment from close relatives or friends, as to whether
the patient would have chosen to participate in the
particular research study for the benefit of future
patients and society as a whole, had he or she been
given the choice.
Capacity to consent
A recurrent matter in critical care research relates to
the proxy’s mental capacity and ability to make a
conscious and reflected choice relating to research
participation under the given circumstances. Close
relatives to critically ill patients are in a state of acute
crisis, and they have to cope with the fear of loss as
well as the intimidation of the unfamiliar and bewil-
dering environment in the emergency department
or ICU. Under such circumstances, one may be less
able to refuse other’s demands and cannot be said
410
with certainty to be free from duress. This motivates
two perhaps somewhat provocative questions: do
relatives have the mental capacity to consent, and is it
at all possible to obtain genuine proxy consent in the
context of critical care research? To answer these
questions adequately, it is necessary to address what
constitutes genuine consent.
Informed consent is not an ethical principle in its
own right, but is rather an important procedure that
has traditionally been considered a means of ensur-
ing that patient autonomy is respected.4 However,
autonomy solely refers to individuals who can
understand and choose what they do.6 Because
both medical treatment and, consequently, medical
research also extend to those with impaired decision-
making capacity, respect for autonomy cannot be
considered the only or most important ethical prin-
ciple.6,7 Indeed, as argued by the British philosopher
Onora O’Neill, it is the rule rather than the exception
that the mental capacity to consent, in the sense of
being able to fully understand all potential risks and
benefits of an intervention, is limited when it comes
to medical decisions.6,7 Furthermore, such consent is
always more or less opaque, simply because it is
impossible for anyone to see through all potential
consequences of the consent, regardless of how
extensive and elaborate the information is.6,7 It is thus
neither reasonable to seek consent for every detail of
a proposed research protocol or to assume that
consent to one proposal implies consent to all its
logical consequences. The purpose of the informed
consent procedure is rather to protect the individual
from coercion and deception.6,7 Hence, proxy consent is
genuine only when the proxy provides a substituted
judgement on a voluntary basis without being
deceived in the process. This does not require that the
proxy knows of and comprehends every detail of the
protocol. Indeed, it may be argued that only three
fundamental aspects of research practice are required
for the proxy to provide genuine consent; these
aspects deserve specific comment.
Firstly, the proxy must be informed and compre-
hend that participation is voluntary, that the consent
is rescindable, that is, it may be withdrawn at any
time, and that the cost of refusal or withdrawal is
only exclusion from a study, with no other conse-
quences for treatment and care.7 Secondly, the proxy
must understand that the consent relates to a
research protocol designed to benefit future patients
and society as a whole. Thus, the proxy must have
the mental capacity to comprehend that the research
does not comprise a ‘special treatment’ designed
specifically to help this individual patient, that is, the

proxy must not be subject of the so-called therapeutic
misconception. Lastly, the proxy must understand
that research participation will subject the patient to
some incremental risk when compared with clinical
practice, and he or she has must have access to
extensive information on this matter at his or her
own discretion. By basing the consent on these three
fundamental aspects of research practice, the proxy
is neither deceived or coerced;6 if the proxy knows
he or she has access and that the consent may be
withdrawn, the proxy does in effect have a veto over
what is done.7 If, on the other hand, the proxy does
not understand these fundamental aspects, either
because he or she is in a state of mental crisis, or
because there is not sufficient time before interven-
tion is required, the proxy does not have the capacity
to consent (Fig. 1A).
Componential risk assessment
One may rhetorically ask whether the relatives of a
critically ill patient can be expected to truly under-
stand and appreciate the incremental risk imposed
by research participation. This stresses the impor-
tance of the research ethical committee to ensure that
a research study is only permitted when agreeing to
enter the study would in fact be a reasonable choice.8
Furthermore, it necessitates a componential analysis
of risk, as advocated for by the American physician
and ethicist Charles Weijer.8 In this approach, Weijer
argues that therapeutic and non-therapeutic risks of
research participation should be considered sepa-
rately. The focus is consequently shifted away from
the life-threatening complications of the patient’s
illness, which are present regardless of whether the
patient participates in research or not, to the incre-
mental risk posed by study participation.9
Therapeutic risks are those imposed by proce-
dures that are expected to benefit the particular
study patient.8,9 Such therapeutic procedures, when
tested in clinical research, must meet the standard
of clinical equipoise, which requires that there exist a
state of honest, professional disagreement in the
community of expert practitioners as to the pre-
ferred treatment or procedure.10 This may both
involve standard and innovative non-validated
therapies; the latter are commonly attempted
outside research protocols in emergencies, where
any intervention that has a theoretical prospect of
benefiting the patient may be attempted when the
patient is on the brink between life and death. Any
critically ill patient may thus be subjected to these
therapeutic risks in clinical as well as research
Consent in critical care research
practice, and in both instances, it is justified by the
prospect of benefit for the individual patient. It is
the responsibility of the research ethical committee
to judge whether the requirement of clinical equi-
poise is fulfilled.8 This is not necessarily straight-
forward because two therapeutic procedures may
have competing benefits that relate to different out-
comes (for example survival vs. a full neurocogni-
tive recovery), and the individual patient’s
preferences in this regard are not taken into
account. If clinical equipoise is deemed to be
present by the research ethical committee, there
are only two obvious differences between clinical
and research practice with regard to therapeutic
risk. Firstly, in research, the particular therapy or
procedure may be applied after randomisation
rather than at the discretion of the individual phy-
sician’s personal preferences and experience. Sec-
ondly, any personal preference that the patient may
have with regard to therapeutic procedures is not
taken into account. However, randomisation does
not pose any incremental risk per se. If it did, it
would be because the relative efficacy of interven-
tions was already known, rendering the research
unnecessary and unethical.11,12 When it comes to
therapeutic procedures in the critical care setting, it
is furthermore rarely possible or relevant to take
the patient’s personal preferences into account,
simply because immediate intervention is required
within seconds to minutes. Thus, neither of these
differences between clinical and research practices
violate the basic rights of the patient.
In contrast to the therapeutic risks, which are
always present in clinical practice and may be
present in research practice, the non-therapeutic risks
exclusively constitute those risks that the patient
would not be subjected to, did he or she not partici-
pate in the research study.8 These procedures are
performed without therapeutic warrant, solely to
answer the study question, and as opposed to thera-
peutic risks, they are present in any study involving
human subjects. These risks are therefore justified
by the importance of the knowledge to be gained
from the research study.8 It is the responsibility of
the research ethical committee to ensure that the
research protocol obeys the principle of beneficence,
in the sense that the risks of non-therapeutic proce-
dures are minimised and that they are reasonable in
relation to the knowledge to be gained.8 When
informing a proxy of a critically ill patient of the
fundamental aspects of research practice, the focus
of the risk assessment should be on the incremental
non-therapeutic risk (Fig. 1A).
411
R. M. G. Berg et al.
Waiver of consent and deferred consent
Even when the consent procedure is shifted to focus
on the three fundamental aspects of research prac-
tice, it will be practically impossible to obtain
genuine proxy consent in a number of situations in
412
Fig. 1. Protection of vulnerable research
subjects. (A) Proxy consent in critical care
research. (B) Waiver of consent in critical
care research.
the critical care setting. For example, the researcher
will not have the time to reach and inform the rela-
tives of the fundamental aspects of research practice
during emergencies or resuscitation efforts. Is it,
then, at all permissible to use patients as research
subjects under these circumstances?

In essence, this question relates to whether
research can be conducted without any consent,
while preserving the basic rights of the patient. This
is sometimes acceded to by requiring that third
parties, such as health professionals or lawyers that
are not involved in the study, provide proxy consent
on behalf of the patient. These do, however, not have
sufficient knowledge of the patient’s values, prefer-
ences, and view of life to provide a substituted
judgement, and therefore, such consent is not ethi-
cally valid. At best, they may assess whether the
protocol violates the patient’s best interest from a
clinical point of view, an assessment that has already
been made by the research ethical committee. Thus,
so-called proxy consent in the form of a
signature from a health professional or a lawyer
represents a legal formalism that fails to protect
the patient’s basic rights, while impeding critical
care research where immediate intervention is
required.13
In this context, it is important to stress again that
informed consent is not an ethical principle per se
and not a goal in its own right. It must be obtained if
the individual has the capacity to consent; if not, the
person is entitled to protection to a degree that
reflects his or her vulnerability.2 If this protection is
not achievable by proxy consent, other means
should be instituted. To obey the categorical impera-
tive, the focus must therefore be shifted from the
consent procedure to other ethically relevant aspects
of the research practice, namely the incremental risk
imposed by research participation and the selection
of research subjects.7,8,13 Thus, the non-therapeutic
risk imposed by research participation must com-
prise no more than a minimal risk.14 A minimal risk
refers to the risks that are common to us all and can
pragmatically be considered to be equal of that of a
general medical examination, including blood test-
ing.8,14 Furthermore, the study should be done only
in this particular group of patients if the circum-
stances that prevent consent constitute a necessary
characteristic that defines the research population,
that is, if it makes no sense to conduct the study in a
population that can provide informed consent.
Moreover, the researcher must have no reason to
believe that the patient would object to being
exposed to a minimal risk for the benefit of future
patients and society as a whole. Under these specific
circumstances, the patient’s basic rights may be pro-
tected despite a waiver of consent (Fig. 1B). Accord-
ing to national legislation in some countries, a
waiver is only permitted if the therapeutic interven-
tion in the study has the prospect of benefiting the
Consent in critical care research
individual patient. However, based on the compo-
nential risk assessment, it is clear that such a distinc-
tion is ethically irrelevant, as long as the non-
therapeutic risk constitutes no more than a minimal
risk.8,9,13
The researcher is obliged to inform the patient
and/or the relatives of the inclusion of the patient in
a research study when a waiver of consent is
exerted. If the study is ongoing at this time, the
researcher must attempt deferred consent, based on
the same fundamental aspects of research practice as
during prospective proxy consent. Under these cir-
cumstances, the patient and/or the relatives must be
presented with the option of withdrawing from the
study. A common practice is furthermore to permit
the patient and/or the relatives to retract all data
obtained from the patient from the study. This
conduct is, however, highly criticisable. The option
of retracting patient data does nothing to protect the
patient from the putative non-therapeutic risks
imposed by study participation. Rather, it prevents
the researchers from monitoring the patient with
regard to adverse effects because of the study inter-
ventions, which may render both the individual par-
ticipant and other patients at an increased risk,
compared with those who opted to remain in the
study. It furthermore introduces significant bias in
prospective studies and violates the foundations of
the intention-to-treat analysis. The basic rights of the
patient are not protected by attempting to eliminate
any experience and knowledge already gained and
recorded from the individual patient, nor do the
basic rights provide the individual patient with the
privilege of impeding dissemination of knowledge
that may benefit others.
Empirical studies of proxy and waiver
of consent
In this section, we move from the normative ethical
aspects of proxy and waiver of consent to descrip-
tive ethics. It is of key importance to discuss not only
how these procedures should be conducted, but also
how they are perceived by patients and their rela-
tives, as well as to which extent relatives are in fact
capable of providing genuine proxy consent. The
empirical studies reviewed were identified through
a systematic search strategy conducted in PubMed
in January 2012 and repeated in July 2012. The fol-
lowing terms were used in various combinations:
‘deferred consent,’ ‘informed consent’, ‘proxy
consent’, ‘waiver of consent’, ‘surrogate decision
maker’, ‘critical care research’, and ‘resuscitation
413
R. M. G. Berg et al.
research’. All abstracts were reviewed, and only
original articles written in English that specifically
reported empirical data on proxy consent for
research participation in the critical care setting
were included. The search strategy yielded a total of
10 articles.15–24
In a population survey from France, it was found
that the majority of respondents would want a
proxy to make medical decisions on their behalf if
they were temporarily incapacitated and admitted
to an ICU;15 almost 90% would want their spouse or
cohabite as a proxy under these circumstances. Like-
wise, survivors from critical illness generally prefer
proxy consent as the means to protect their basic
rights in critical care research,16,17 and close relatives
are the preferred proxies.16,18 Close relatives gener-
ally concur with this.16,19 Moreover, they are gener-
ally positive towards research conducted in this
patient population.19 The majority of patients in a
tertiary care emergency department were in favour
of a waiver of consent if they were critically ill and
temporarily incapacitated, as long as the research
study only involved a minimal risk.20
A recent questionnaire survey of relatives who
had been approached for the possible participation
of a critically ill patient in research suggested that
many chose to enrol the patient based on a substi-
tuted judgement. Thus, many respondents reported
that the decision was based on the perception that
the patient would have agreed to participate in the
study had he or she been given the choice.21 But to
which extent do the decisions of the proxy actually
reflect the wishes of the patients? When presented
with two hypothetical studies, survivors of critical
illness that had regained capacity were more likely
to accept inclusion than the relatives that had acted
as proxies during the admission.22 An interview
survey of high-risk elective cardiac surgery patient
and their potential proxies should they be admitted
to the ICU post-surgery, likewise compared their
decisions on the patient’s participation in two hypo-
thetical studies. The proxy predicted the patient’s
decision with 80% accuracy when the patient
accepted and with 50% accuracy when the patient
declined participation.22 The spouse or cohabite was
superior to other relatives at predicting the
patient’s decision, an ability that largely appeared
to depend on whether they had previously dis-
cussed their attitudes towards medical research in
humans.18,22 Thus, the ability of relatives to predict
the patient’s preferences, and thus to provide a
proxy consent based on a substituted judgement, is
highly variable.
414
Although many relatives may be able to predict
the patient’s preferences regarding hypothetical
studies, they may not necessarily have the mental
capacity to do so when it comes to actual studies in
the critical care setting. Accordingly, and perhaps
obviously, the majority of relatives present on scene
during out-of-hospital cardiac arrest have been
found to be entirely incapable of comprehending the
informed consent procedure for research participa-
tion.23 Indeed, a recent questionnaire-based study
suggests that the most common reason for declining
consent for research participation among relatives to
patients admitted to the ICU is that they are too
worried to consider a research study.21 In another
questionnaire survey, it was furthermore found that
the relatives of unconscious, critically ill patients
thought that drug trials were designed to benefit the
individual study participant,19 suggesting the pres-
ence of the therapeutic misconception. Paradoxi-
cally, many respondents at the same time recognised
that drug trials are necessary to achieve new scien-
tific knowledge and improve care of future
patients.19 A similar paradox appears to be equally
evident among relatives that have provided proxy
consent for research participation; thus, the two
major reasons for providing consent were ‘to help
the patient recover faster’ and to ‘improve patient
care in the future’.21 In this study, a common reason
for declining consent was that the proxy did not
want the patient to receive experimental therapy.21
The therapeutic misconception was also evident in a
study of relatives that were approached for deferred
consent in the Brain Resuscitation Clinical Trial II, in
which patients suffering from cardiac arrest were
randomised to receive a calcium entry blocker
or placebo within 30 min after the restoration of
spontaneous circulation.24 Thus, relatives consented
while expressing a desire ‘to take any chance to save
a loved one’, ‘to do whatever is best to help the
patient’, and many felt relieved that their relative
might get ‘special’ treatment.24 Together, these
studies suggest that many relatives fail to appreciate
the fundamental aspects of research practice, and
thus to provide genuine proxy consent.
Conclusion
Proxy consent for research participation is genuine
only when it is provided by someone who has the
time and mental capacity to understand and appre-
ciate the fundamental aspects of research practice,
and subsequently to make a decision based on
knowledge of the patient’s values, preferences, and

view of life. The three fundamental aspects of
research practice are (1) that participation is volun-
tary and that the consent may be withdrawn at any
time; (2) that the research is designed to benefit
future patients and society as a whole, and not the
individual study participant; and (3) that participa-
tion imposes an incremental non-therapeutic risk on
the patient.
Although empirical studies indicate that both
patients and relatives are in favour of proxy consent
in critical care research, the proxies’ ability to
predict patients’ preferences are highly variable, and
many proxies fail to appreciate these fundamental
aspects of research practice. This may, however, be
alleviated by avoiding information overload and
focusing the informed consent procedure more spe-
cifically on these three fundamental aspects.
It must nevertheless be stressed that it is the
moral responsibility of the researcher to ensure that
the proxy has the capacity to provide genuine
consent in the given situation, and that there must
be no reason to believe that the patient would object
to participation. In case of any uncertainties relating
to these matters, the patient should be excluded
from research participation. If the circumstances
dictate that intervention is required before the proxy
can be contacted, the research ethical committee
must furthermore ensure that the non-therapeutic
risk associated with research participation com-
prises no more than a minimal risk. In this regard, it
is ethically irrelevant whether the research study
involves therapeutic interventions or not.
If the abovementioned precautions are taken, and
the research study is specifically designed to benefit
similar, critically ill patients, the patient is ensured a
degree of protection that reflects his or her vulner-
ability and cannot be said to be treated ‘merely as a
means to an end’. This enables critical care research
for the benefit of future patients and society as a
whole while preserving the basic rights of the indi-
vidual patient.
Author contributions: R. M. G. Berg drafted the
manuscript. K. Møller and P. J. H. Rossel made criti-
cal revisions. All authors read and approved the
final manuscript.
Conflicts of interest: None.
Funding: The study was supported by The P. Carl
Petersen Foundation, University Hospital Rigs-
hospitalet and the Faculty of Health Sciences,
University of Copenhagen. The study was further
supported by the Danish Council for Independent
Research – Medical Sciences, the Commission of
the European Communities (Grant Agreement no.
Consent in critical care research
223576 – MYOAGE). Centre of Inflammation and
Metabolism (CIM) is part of the UNIK Project: Food,
Fitness & Pharma for Health and Disease, sup-
ported by the Danish Ministry of Science, Technol-
ogy, and Innovation. CIM is a member of DD2 – the
Danish Center for Strategic Research in Type 2 Dia-
betes (the Danish Council for Strategic Research,
grant no. 09–067009 and 09–075724).
References
1. Kant I. Grundlegung zur Metaphysik der Sitten. JF Hartknoch,
1785.
2. The National Commission for the Protection of Human Sub-
jects of Biomedical and Behavioral Research. Belmont report:
ethical principles and guidelines for the protection of human sub-
jects of research. US Government Printing Office, 1978.
3. Buchannan AE, Brock DW. Deciding for others. The ethics of
surrogate decision making. New York: Cambridge Univer-
sity Press, 1990.
4. Faden RR, Beauchamp TL, King NMP. A history and theory
of informed consent. New York: Oxford University Press,
1986.
5. Levine RJ. Clarifying the concepts of research ethics. Hast-
ings Cent Rep 1979; 9: 21–6.
6. O’Neill O. Paternalism and partial autonomy. J Med Ethics
1984; 10: 173–8.
7. O’Neill O. Some limits of informed consent. J Med Ethics
2003; 29: 4–7.
8. Weijer C. The ethical analysis of risk. J Law Med Ethics 2000;
28: 344–61.
9. Weijer C. The ethical analysis of risk in intensive care unit
research. Crit Care 2004; 8: 85–6.
10. Freedman B. Equipoise and the ethics of clinical research. N
Engl J Med 1987; 317: 141–5.
11. Abramson NS, Meisel A, Safar P. Informed consent in resus-
citation research. JAMA 1981; 246: 2828–30.
12. Abramson NS, Meisel A, Safar P. Deferred consent. A new
approach for resuscitation research on comatose patients.
JAMA 1986; 255: 2466–71.
13. Berg RMG, Møller K, Rossel PJH. European legislation
impedes critical care research and fails to protect patients’
rights. Crit Care 2011; 15: 148.
14. Freedman B, Fuks A, Weijer C. In loco parentis. Minimal risk
as an ethical threshold for research upon children. Hastings
Cent Rep 1993; 23: 13–9.
15. Azoulay E, Pochard F, Chevret S, Adrie C, Bollaert PE, Brun
F, Dreyfuss D, Garrouste-Orgeas M, Goldgran-Toledano D,
Jourdain M, Wolff M, Le Gall JR, Schlemmer B. Opinions
about surrogate designation: a population survey in France.
Crit Care Med 2003; 31: 1711–4.
16. Chenaud C, Merlani P, Verdon M, Ricou B. Who should
consent for research in adult intensive care? Preferences of
patients and their relatives: a pilot study. J Med Ethics 2009;
35: 709–12.
17. Scales DC, Smith OM, Pinto R, Barrett KA, Friedrich JO,
Lazar NM, Cook DJ, Ferguson ND. Patients’ preferences for
enrolment into critical-care trials. Intensive Care Med 2009;
35: 1703–12.
18. Coppolino M, Ackerson L. Do surrogate decision makers
provide accurate consent for intensive care research? Chest
2001; 119: 603–12.
19. Perner A, Ibsen M, Bonde J. Attitudes to drug trials among
relatives of unconscious intensive care patients. BMC
Anesthesiol 2010; 10: 6.
415
R. M. G. Berg et al.

20. Smithline HA, Gerstle ML. Waiver of informed consent: a 24. Abramson NS, Safar P. Deferred consent: use in clinical
survey of emergency medicine patients. Am J Emerg Med resuscitation research. Brain Resuscitation Clinical Trial II
1998; 16: 90–1. Study Group. Ann Emerg Med 1990; 19: 781–4.
21. Mehta S, Quittnat Pelletier F, Brown M, Ethier C, Wells D,
Burry L, MacDonald R. Why substitute decision makers
provide or decline consent for ICU research studies: a ques-
tionnaire study. Intensive Care Med 2012; 38: 47–54. Address:
22. Ciroldi M, Cariou A, Adrie C, Annane D, Castelain V, Cohen Ronan M. G. Berg
Y, Delahaye A, Joly LM, Galliot R, Garrouste-Orgeas M, Centre of Inflammation and Metabolism, Department of
Papazian L, Michel F, Barnes NK, Schlemmer B, Pochard F, Infectious Diseases, Section M7641
Azoulay E, Famirea study group. Ability of family members Rigshospitalet
to predict patient’s consent to critical care research. Intensive 2100 Copenhagen Ø
Care Med 2007; 33: 807–13. Denmark
e-mail: ronan@dadlnet.dk
23. Hsieh M, Dailey MW, Callaway CW. Surrogate consent by
family members for out-of-hospital cardiac arrest research.
Acad Emerg Med 2001; 8: 851–3.

416