INTRODUCTION
In this document we have sought to make recommendations on how ethical issues in
relation to research involving human participants should be dealt with. On the great
majority of issues, our recommendations are in keeping with those of other parallel bodies
and, where relevant, we make reference to their reports. In all instances, we have paid
careful attention to legal requirements but there are several difficult issues in which the
legal situation remains ambiguous or uncertain and a very small number in which, in our
view, the law appears to operate on principles that run counter to either ethical standards
or good practice, or both. When that is the case, we draw attention to the discrepancies
currently involved but, nevertheless, outline what we regard as the desired ethical practice.
Similarly, at the time of our report, discussions are taking place on a possible revision of
the Declaration of Helsinki and European bodies are considering the revision of
requirements on data protection. Again, we draw attention to possible discrepancies but
focus on the ethical principles.
Guideline recommendations are displayed in purple text throughout
this document.
1. ETHICAL RESPONSIBILITES
This document is aimed at both researchers and research ethics committees for human
participants (hereafter referred to simply as ethics committees). The responsibility for the
ethical conduct of research rests firmly on the principal investigator (and, for legal reasons,
therefore on his or her employer), and not on either the research ethics committee or the
funding body, although both have a concern that the principal investigator exercises
those ethical responsibilities in an appropriate fashion. We specifically note that, although
there is a requirement to obtain the consent of participants1 in the research (we make no
distinction in that connection between patients and volunteers), the responsibility remains
with the principal investigator and is not shared with the participant.
The responsibility for the ethical conduct of research rests firmly on
the principal investigator.
1. Or, in the case of people who are incompetent to give consent, those who can give consent on
their behalf – see paragraph 5.
1
ROYAL COLLEGE
2. RISKS AND BENEFITS OF RESEARCH
OF PSYCHIATRISTS
RESEARCH
GUIDELINES:
ETHICS
2.1 Definition of research
Research can be defined in a number of ways. For purposes of ethical review, the Royal
College of Physicians (1996) defined the difference between research and practice as
one of intent. In medical practice the primary intention is to benefit the individual patient.
In research the primary intention is to advance scientific knowledge so that patients in
general may benefit.
In considering whether a procedure is research or not, local research
ethics committees and researchers are referred to the Royal College
of Physicians (1996) guidelines, paragraph 6.4:
“The distinction between medical research and innovative medical
practice derives from the intent2. In medical practice the predominant
intent is to benefit the individual patient consulting the clinician,
not to gain knowledge of general benefit, though such knowledge
may emerge from the clinical experience gained. In medical research
the primary intention is to advance knowledge so that patients in
general may benefit: the individual patient may or may not benefit
directly.”
2.2 Value of research
Many advances in clinical practice have been crucially dependent on the findings of
research. In some cases the gains have come from bringing together, over time, the results
of basic research, which have led to a much better understanding of causal processes and
of how they may be prevented or treated effectively. In other cases the gains have come
from systematic evaluations that have shown which interventions are effective and which
ones ineffective; or which patterns of service delivery are most beneficial for the most
people. Sometimes, too, the benefits have derived from systematic descriptions of the
experiences, perceptions or opinions of consumers or of staff. The benefits of research,
however, are crucially dependent on it having been of high quality. Poor research, or
biased research, by contrast, may impede improvements in health services, and is unethical.
A key issue with respect to the advances in clinical practice that should derive from
research is that it would be unethical to have a system that meant that individuals from
certain segments of the population could not participate in research that focused on the
disorders from which they suffered or for which they were specifically vulnerable. The
criterion that is relevant in this connection, however, must be the needs of that segment
of the population and not the needs of the researcher. In paragraph 5 we consider the
protections that need to be employed in that connection.
2
2. In the Royal College of Physicians’ guidelines the phrase was ‘sole intent’. Although that is
correct in most instances of medical practice, and although the benefit to the individual patient
predominates in all cases, there are important circumstances throughout medicine in which it
cannot be the sole intent. Thus, there are legal requirements to notify certain diseases because of
the risk to others; similarly, compulsion may be necessary in psychiatric practice because of risks to
others as well as to the individual. The key difference between clinical practice and research is that
the chief focus in practice revolves around needs arising from the individual, whereas in research the
focus is on extending knowledge that will provide benefits to groups of people. It is also the case that,
although the aim of research is to advance knowledge, the conduct of any trial must be undertaken
in a manner that is responsive to patient needs. Thus, if during the trial it becomes evident that the
patient is at risk, the patient will need to be withdrawn from the study because the interests of the
individual have to take precedence over the research interests.
It is in the interests of everyone that high quality research should be
fostered and supported. Ethics committees need to check that the
research they approve is of adequate quality. Because of the immense value of research, it is unacceptable that individuals from any
segments of the population be disallowed, by virtue of their being
members of that group, from participation in research that is necessary to improve the understanding of disorders from which they are
particularly likely to suffer. The difficulties that must be faced in this
connection with groups whose capacity to consent is limited are
considered further below in paragraph 5. Nevertheless, it is a basic
ethical principle that psychiatric patients, like all other patients, must
be able to benefit from the fruits of research and, hence, they must
have the opportunity to participate freely in sound research.
RISKS
AND
BENEFITS OF
RESEARCH
2.3 Risks of research
Regrettably, history shows that some research, even that which is otherwise scientifically
sound, has put participants in the research at levels of risk that are unacceptable, sometimes
without the participants being aware that it has been so (Beecher, 1966; Pappworth,
1967; Fienberg & Resnick, 1997).
Ethics committees should be required to assess the levels of risk in all
research that they review and they should not approve projects in
which the risk is regarded as excessive when considered in the light
of all the circumstances, and of the potential benefits. With research
regarded by the ecthics committee as within an acceptable risk level,
there is the additional requirement that all participants should be
adequately informed about the nature of the research, its possible
risks and potential benefits, so that they can make up their own
minds, on an informed basis, with respect to whether or not they are
willing to take part in the study.
2.4 Misuse of research
Equally regrettably, history tells us that some funding bodies have sought to misrepresent
or suppress the results of research (Barnes et al, 1995; Bero et al, 1995; Glantz et al, 1995,
1996; Hanauer et al, 1995; Slade et al, 1995; Hilts, 1996; Ong & Glantz, 2000). It is clear
that, in some instances, such bodies have sometimes supported, without constraint or
censorship, other reputable research, apparently in an attempt to conceal or downplay
the examples in which they have misused research findings.
Ethics committees should refuse to approve research where the funding body is of a kind that raises serious doubts about its track record
on abuses of research findings.
2.5 Censorship of research
It is not in the interests of the community to participate in research that is censored,
subject to suppression/distortion of some findings, or for which publication is refused or
unduly delayed by the funding body. Ethics committees should ensure that such constraints
do not exist3. Particular difficulties arise in three main circumstances. Government
departments sometimes require that they approve publications because of the policy
3. Throughout this report, we have assumed that our guidelines apply to researchers/clinicians
employed by universities, charities, governmental research funds, the National Health Service
(NHS) or other bodies in which the conditions of employment do not contain clauses requiring
secrecy or other imposed controls. Thus, for example, our remit does not cover employees of
pharmaceutical companies or governmental defence establishments. Their ethical issues are comparable
but they must fall outside the remit of university/NHS ethics committees because the conditions of
employment clash with our guidelines. Nevertheless, we recommend that no research should be
conducted without there being an ethics committee to which research proposals must be submitted.
3
ROYAL COLLEGE
OF PSYCHIATRISTS
RESEARCH
GUIDELINES:
ETHICS
considerations that may arise. It is reasonable, and acceptable, that they can require the
opportunity to comment and make suggestions in advance of publication. Nevertheless,
it cannot be ethically acceptable that they suppress or alter findings. If they do not accept
particular findings, or the interpretations put on them, it is acceptable that they require
some explicit statement to that effect to be included in the report, but this must not
involve prevention of publication (or unreasonable delay in publication). The second
circumstance arises when the findings are commercially sensitive because of the potential
financial benefits for the funding agency (such as could be the case with pharmaceutical
companies). A reasonable short delay in publication may be required to allow commercial
exploitation but it is not acceptable that findings be kept out of the public arena for other
than a strictly defined, previously agreed, time period. The third circumstance is when
Government departments consider that the publication of findings could involve national
security risks. Although it is accepted that this could be a reality in some circumstances
(rarely so in the field of biomedical research), the same principle as that relevant to
commercial interests applies. A reasonable, previously agreed, delay in publication is
acceptable, but prevention of publication is not acceptable.
Ethics committees should require applicants to state whether the
funding body (or any other interested party) places any constraints
on publication. If any such constraints exist, the ethics committee
must satisfy itself that they are reasonable in type and degree, that
they are explicit and time-limited and that they do not involve risk of
censorship or distortion of findings or prevention of publication.
4
3. SCOPE OF ETHICS COMMITTEES AND VARIETIES
OF RESEARCH
We take it as given that the role of ethics committees is to facilitate good research and
that the prime purpose of ethical guidelines is to protect those who participate in research
as ‘subjects’ and, in addition, to protect the researchers who undertake the research. All
employing institutions have an obligation to ensure that the research undertaken by its
staff meets expected ethical requirements; this obligation has implications for the remit
of ethics committees.
3.1 Scope of ethics committees’ remit
In order to provide the protection needed for participants, researchers and institutions,
the scope of the remit of ethics committees necessarily has to be broad. The question of
how the remit is dealt with in relation to different types of circumstances is considered
further below.
All research that involves human subjects directly or indirectly, and
that is undertaken by the staff of any discipline (paid, honorary or
emeritus) of an institution4, comes within the remit of its relevant
ethics committee, irrespective of whether or not the participants
are its patients, and irrespective of where the research is undertaken. Similarly, irrespective of who undertakes the research, all
research that involves the institution’s patients, clients, students
or staff as participants in the research falls within the ethics committee remit.
3.2 Clinical audit
Clinical audit is now a required routine in National Health Service (NHS) practice as a
set of procedures designed to ensure that the requisite standards of clinical care, notemaking, reporting, timeliness of communication and the like are being met. In so far as
they have a focus on advancing knowledge (about patient care, note-taking, etc.) rather
than the care of an individual, they fall within our definitions of research. As such, it is
the responsibility of ethics committees to ensure that they are conducted in an ethical
manner. Ordinarily, all regular clinical audits in a particular institution follow an accepted
routine, and once this has been agreed and established, it should not be necessary for
further ethical consideration unless and until it is proposed to alter the routine.
In addition, however, there is a range of other forms of audit that go beyond these
routines because they involve asking patients or staff to complete questionnaires or be
interviewed. For example, this would apply to studies to determine patient satisfaction
with the care they have received, or staff views on the handling of risk situations. Similarly,
studies comparing different ways of giving patients information or different forms of
preparation for clinical procedures constitute specific research projects. All such special
forms of non-routine audit constitute research that needs to be dealt with in the same
way as any other form of research.
Routine clinical audit solely based on perusal of records must be
conducted in an ethical manner and it is the responsibility of ethics
4. We recognise that non-medical employing bodies that are concerned with research that is not
funded by either biomedical charities or research councils may not have ethics committees and
procedures of a kind that we consider in this document. We regard this as unsatisfactory but its
consideration falls outside the scope of our working party. Nevertheless, as noted in footnote 3, we
recommend that all research should be subject to ethics committee approval.
5
ROYAL COLLEGE
OF PSYCHIATRISTS
RESEARCH
GUIDELINES:
ETHICS
committees to ensure that this is the case5. This will require them to
obtain written details of the procedures to be followed. It is up to
the committee to decide whether to delegate this responsibility to
someone acting on their behalf (such as the clinical director), or to
consider the procedures in committee. Either way, however, the ultimate responsibility lies with the ethics committee and it is their
duty to ensure that the necessary ethical needs have been met. Patient information sheets and consent forms should not be required
for such case note audit and, once a set of procedures has been
agreed, it should not be necessary to reconsider the procedures
until it is proposed to change them.
Research into audit that involves any form of new information from patients or staff
(such as by questionnaire or interview), any form of intrusion into routine procedures
(such as by video- or audiotape recording or observation by researchers) or any form of
participation in a comparison of procedures (whether or not by random allocation) should
be subject to the same form of individual application as required for other research
applications to the ethics committee. Patient information and consent forms should
ordinarily be required.
3.3 Other forms of research using records and archived samples
Archived records have proved invaluable in a wide range of research that has led to
important advances in the understanding of risk and protective processes for disease and
disorder (see Medical Research Council, 2000 for a discussion of the range of ethical
issues to be considered in relation to their use). In recent years, concerns over data
protection have raised the possibility that legislation might be introduced (although it has
not happened so far) to require individual consent for such use of records or to restrict
their use (Walton et al, 1999); this would effectively stop such research because of the
impracticality of contacting people on the basis of records that may have been obtained
as long as half a century ago (Melton, 1997). In any case, contact in order to ask for
consent could only be obtained through the use of the very records the use of which is the
subject of the consent being sought! There are many reasons to preserve the availability
of archived records and samples for research, there is no possible harm to the individuals
provided the necessary safeguards are maintained, and there is no reason to treat this use
of archived data differently from clinical audit. However, there are four rather different
uses that require some variations in approach (see paragraphs 3.3(a) to 3.3(d)).
As with any other type of research involving human subjects, approval of research using records and archived samples must be sought
from the appropriate ethics committee, which will need to consider
the usual ethical issues with respect to purposes, reputability of
researchers, lack of inappropriate constraints on publication, source
of funding and the other matters outlined in this document. In addition, as also recommended by the Royal College of Physicians (1999)6,
the ethics committee should have specifically agreed to exempt the
research from the general requirement for individual consent from
each research subject. In that respect, our recommendation on research using records and archived samples follows the same principles as those that apply to clinical audit.
3.3(A) GROUP ANALYSES USING ANONYMISED DATA
In cases in which the sole intention is to undertake group analyses of anonymised data,
the ethics committee should be satisfied that the required anonymisation has been
6
5. Currently, in the UK, the use of medical records and samples for audit does not require ethical
approval, despite the fact that it cannot be differentiated from research in any clear and unambiguous
fashion (Choo, 1998). We agree that individual consent should not be required but, despite the fact
that it may not be required by law, we consider that ethical consideration of the audit procedures
should be undertaken.
undertaken before the data are made available to the researchers seeking to use the
archived data. When the archived information involves pooled data-sets (as would be
the case with many DNA or brain tissue banks), the organisation having custodianship of
the databank should have its own ethics committee to review the purposes to which the
analyses are to be put, and the reputability of the researchers and the funding agency.
This is not currently standard practice and we recommend that it should become so.
SCOPE OF
ETHICS
COMMITTEES AND
VARIETIES OF
RESEARCH
Individual consent should not be necessary for group analyses of
anonymised data but the ethics committee should ensure that the
required anonymisation has been achieved before the data are made
available. Custodians of pooled data-sets should have their own ethics committees review applications for use of those sets.
3.3(B) ANALYSIS OF PERSONALISED RECORDS
In some instances it may be necessary for the analysis of archived samples or records to include
access to personalised data either because of the need to derive information from patients’
case notes of a kind that is not or could not really be computerised, or because of a need to link
data-sets. In keeping with the Royal College of Physicians’ (1994) Working Group, we agree
that this is ethical without individual patient consent provided that:
(a) access to the records be controlled by their custodian (who must not
be the investigator – in order to avoid the conflict of interest entailed
in someone granting themselves research access to the records of
patients under the care of another consultant) who must be a senior
clinician who takes the necessary steps to be satisfied on the bona
fides, professional standing of the investigator and the ability of the
investigator to handle the identifiable data securely in such a way
as to maintain confidentiality;
(b) the custodian ensures that the investigator is an appropriate senior
professional person who is bound by a code of ethical standards set
by a professional licensing body, and who could be subject to
disciplinary action in the event of a break of confidentiality;
(c) the publication of research information in no circumstances directly
or indirectly identifies individuals.
Individual consent should not be necessary for analyses of personalised records provided that no contact with participants is envisaged, that access to records is controlled by a custodian (who must
not be the investigator) who has the responsibility for checking the
details of what is proposed and that no data will be published that
could directly or indirectly identify individuals.
6. The Royal College of Physicians (1999) went on, however, to recommend that, in the future,
patients and healthy volunteers be made aware that the information and samples they provide may
be used for research and they should be given the opportunity to dissent from such use and that
such dissent would prohibit the use of their information and samples from future research, unless an
ethics committee later ruled that it was in the public interest that their objection be overruled. We
agree with the recommendations that all patients be made aware that their records may be used for
research, but we also agree with Professor Wald’s minority view that this opportunity to dissent is
illogical, would serve no useful purpose, lacks credibility and carries the risk of distorting research
findings by making samples biased (Royal College of Physicians, 1999). It seems unsatisfactory to
appear to give people a choice when that choice can be rescinded later without their knowledge or
involvement in the decision-making. If there are no risks and if the relevant ethics committee has
agreed that it is in the public interest for the research to be undertaken, there seems no point in the
introduction of a meaningless constraint that would not apply to existing data-sets. Moreover, as
such constraints on the use of anonymised data do not apply to clinical audit, it seems illogical to
apply them to research use as these involve the same considerations. It also seems unreasonable to
expect people to agree or disagree with research usage in the future when they have no idea what
purpose that will serve. It is likely to be the case that some people will dissent on the basis of their
views at the time the information or samples are obtained but, if asked some decades later, would
give agreement freely for a purpose that they could see was valuable (or vice versa).
7
ROYAL COLLEGE
OF PSYCHIATRISTS
RESEARCH
GUIDELINES:
ETHICS
3.3(C) USE OF ARCHIVED DATA FOR TRACING
In other instances, archived data may have to be used as the starting point for studies because
they require contact with individual patients or volunteers. This would be the case, for
example, with follow-up studies. In these circumstances, the archived data may be used,
without individual consent, in order to trace and make contact with the potential participant
in the research, provided the requirements above (paragraph 3.3(b)) are met, and provided
the safeguards noted for follow-up studies (see paragraph 4.13) are used. However, at the time
of contact with the potential participant, all the usual requirements for an information sheet
and informal consent should apply. The dispensing with individual consent applies only to the
use of records for tracing (or for group analyses of archived data) and participants should have
their usual freedom to agree or disagree to taking part in any research that requires further
information, samples or any other form of personal involvement.
Archived data may be used without individual consent to trace individual patients or volunteers in order to ask if they are willing to
participate in a study. However, the planned study must have received ethical approval and all the usual expectations of individual
informed consent apply to participation in any aspect of the research required for individual contact or new information/samples
from individuals.
3.3(D) USE OF OTHER PERSONAL RECORDS FOR TRACING
A related issue to that considered above is the use of other personal records (such as NHS
identification numbers) in order to trace individuals in order to ask whether they are
willing to participate in a particular study. Currently, practice on this issue seems
inconsistent and dependent on the varying views of individuals, but it is evident that
practice is becoming more restrictive. Because of the proven great importance of followup studies, and because scientific findings are likely to be seriously misleading if followup is based on incomplete biased samples, we recommend that the principles outlined
above (paragraph 3.3(c)) be extended to the use of other personal records for tracing.
The use of personal records for tracing, when the records are not
themselves the basis for identification of potential participants in
research, should be dealt with in the same way as the use of archived
records for tracing. For this to be generally possible, the legal acceptability of this procedure under data protection rules will have
to be made explicit. Ethics committees should ensure that the method
of tracing to be used is discrete (see paragraph 4.13).
3.4 Therapeutic v. non-therapeutic research
8
The distinction between therapeutic and non-therapeutic research was included in the
original Declaration of Helsinki (World Medical Association, 1989). It has particular
relevance to the inclusion of some vulnerable groups in ‘non-therapeutic’ research, for
example children, those with learning disabilities, some patients with dementia and others
without the capacity for consent to research (also see paragraph 4). However, the
Declaration of Helsinki has recently been revised (in 2000) and the therapeutic v. nontherapeutic distinction is no longer maintained.
The distinction continues to be adopted in most research ethics guidelines (British
Paediatric Association, 1992; Medical Research Council, 1992; Royal College of
Psychiatrists, 1990; Royal College of Physicians, 1996). However, it has increasingly been
criticised for excluding ‘vulnerable’ patients from benefiting from research and sometimes
for thwarting potentially valuable research altogether.
There is also a growing recognition that it tends to disguise the fact that therapeutic
research (for example, a trial of a new drug) may often be considerably more hazardous
than non-therapeutic research. Hence, although carrying ethical and legal weight, we
recommend that the distinction between therapeutic and non-therapeutic research should
be dropped in favour of a general provision that all research, whether or not potentially
benefiting individual research subjects directly, should be considered with respect to its
costs and benefits.
Local research ethics committees and researchers should be aware
of the continued ethical and legal relevance of the therapeutic v.
non-therapeutic distinction as embodied in a number of current
guidelines. However, rather than relying on the distinction as made,
they should assess each research project on its merits according to
the general principles outlined above, having regard in particular to
the risks and benefits to potential participants. We recommend that
steps should be taken to remove the distinction between therapeutic and non-therapeutic research.
SCOPE OF
ETHICS
COMMITTEES AND
VARIETIES OF
RESEARCH
3.5 Non-intrusive or non-invasive research
Although terms such as non-intrusive or non-invasive research and minor or ethically
trivial research still appear in many guidelines and codes, they are now generally considered
to be most unsatisfactory as guides to the ethical aspects of research. We agree. There are
three main problems with such distinctions. First, what is deemed non-intrusive by
professionals (because the procedures are routine and very low risk) may not be viewed
in the same way by the lay public. Such differences of perception might apply, for example,
to venesection or magnetic resonance imaging. Similarly, questionnaires that are poorly
designed may give rise to distress that was not anticipated by researchers. Second, the
distinctions are dimensional rather than categorical; that is, procedures vary in the degrees
to which they are intrusive, invasive, major or ethically problematic. It has not proved
possible, or helpful, to try to define these terms in a meaningful dichotomous fashion that
applies in a comparable way to all forms of research and all kinds of participants. The
degrees necessarily depend on a wide range of patient-related (or participant-related)
and research-related considerations. Third, there is a danger that definitions that classify
certain types of research as ‘minor’ or ‘ethically trivial’ will prove to be over inclusive, so
allowing significant ethical issues to be overlooked. Ethics committees need to consider,
for example, whether the use of non-standard interviews or questionnaires involves
unacceptable intrusion or psychological risk.
Each research proposal should be assessed on an individual basis
with respect to its individual merits and risks. The details of all procedures should be considered with regard to their possible intrusiveness, invasiveness or distress-provoking properties in relation to the
target group of participants. Researchers should be expected to
have taken appropriate steps to keep all of these potentially negative features to a minimum.
3.6 Pilot studies
All good research is likely to require the use of some form of pilot study to develop new
tests or measures, to assess the likely relevance of particular approaches, to determine the
applicability of established measures for use in new groups (differing, for example, in age,
gender or sociocultural background from those previously studied) or to determine the
timing or acceptability of combinations of measures. In practice it is often difficult to
draw a clear line between clinical innovation and pilot studies for research. The criterion
we apply is the one with which the document started, namely whether the main intent
is to help the individual or to gain knowledge applicable to groups. When the new
development constitutes a pilot study for research, it necessarily requires ethical
consideration, regardless of whether one participant or 100 are involved. This means that
the risks and benefits in relation to the kinds of participants and to the overall circumstances
need to be considered in relation to the criteria applicable to any form of research. It
should become a routine for researchers to obtain a second opinion on the level of risk
from someone who is not associated with the pilot research but who has the relevant
expertise to assess what is involved. It should also be mandatory for participants to be
informed that the procedure is research and not ordinary clinical care, and that informed
consent be obtained in an explicit fashion.
Although we consider that pilot studies of all kinds, and at all stages, fall within the
remit of ethics committees, there is a dilemma with respect to when a formal application
for approval needs to be made. Clearly it would be impractical to expect application for
9
ROYAL COLLEGE
OF PSYCHIATRISTS
RESEARCH
GUIDELINES:
ETHICS
the first initial exploratory steps. On the other hand, it is important that the ethical issues
be considered seriously. It is for that reason that we recommend that a second opinion
should be sought at the outset. If, in the view of that external reviewer, the pilot study
involves either ethical concerns or risks that might be viewed by participants (not just
researchers) as more than minor, formal application for approval by the ethics committee
should be required before proceeding further. If, however, the second opinion is that the
risks are minor, the pilot study is being handled ethically and proper informed consent is
being obtained from participants, it is acceptable that the second opinion gives approval
on behalf of the ethics committee. It is up to the ethics committee to set criteria for who
is able to act in this way. However, once there is a formulated plan for a pilot study,
application to the ethics committee should be made in the usual way. The second opinion
should have the right to specify that this must occur after a particular time period, or a
particular number of cases or after some early objective have been reached.
All pilot studies, at all stages and of all kinds, fall within the remit of
the ethics committees. It is acceptable, however, for the early stages
of pilot work to proceed without formal application provided that a
second opinion (by someone approved by the committee to act on
its behalf) is obtained and that the view is taken that the risks are
trivial (as likely to be viewed by participants), the study is being
ethically undertaken and satisfactory informed consent is being obtained from all participants.
3.7 Complaints
Anyone who has valid and serious concerns over possibly unethical research has a duty to
bring these concerns to the attention of the proper authority, the individuals who do so
should be protected and it should be the responsibility of ethics committees to investigate
all such complaints in a dispassionate manner in whatever detail seems appropriate in the
circumstances.
Ethics committees should be prepared dispassionately to investigate
any complaints over possibly unethical practice in relation to any
research falling within their area of responsibility.
3.8 Scientific fraud
Fraudulent research is unethical and, hence, concerns over possible fraud fall among the
responsibilities of ethics committees. Whether or not they constitute the most appropriate
body to investigate allegations of possible fraud is a separate matter, but they should
ensure that the organisations to which they are answerable have set up appropriate
mechanisms for dealing with concerns over fraud and that the individuals who raise such
concerns are protected. A key issue with respect to possible fraud is that researchers must
be expected to make proper records of all steps in the research and that such records be
retained for later scrutiny if required because queries about the integrity of the science
have been raised. We understand that the National Academies Policy Advisory Group
(NAPAG) is currently conducting a study of scientific fraud and a draft set of guidelines
is peing prepared.
It is the responsibility of ethics committees to ensure that appropriate mechanisms for dealing with concerns over possible fraud are
available and are operated in a fair and efficient manner.
10
4. RESEARCH GOVERNANCE AND CLINICAL CARE
SCOPE OF
ETHICS
COMMITTEES AND
VARIETIES OF
RESEARCH
4.1 Research quality
Necessarily, consideration of the quality of research has to be a concern of ethics
committees; bad or biased research is, in itself, unethical. That is because it wastes
participants’ time, it creates a false impression for participants that the findings will be
useful, it may lead to inappropriate policies or practice and in many cases it will also waste
public monies. We consider below (see paragraph 6.5) the practical implications for ethics
committees that arise from this ethical principle, but here we consider the expectations
that should be placed on researchers. In all cases, there should be a recognised, named
research leader who has the skills and experience to undertake well the particular study
proposed, and who has the time and resources necessary to provide the supervision required.
The measures used should be well designed for their purpose and should involve no more
risk (including distress) than is essential for the proper, good standard conduct of the
research. Appropriate informed consideration of statistical power should have been
undertaken so that the sample is not so small as to be useless in meeting the stated aims
of the research, or so large as to be unacceptably wasteful. When the research is of a kind
that might give rise to occasional clinical risks (such as side-effects of drugs, discovery of
child abuse or significant homicidal/suicidal risk) the organisation of the research team
operation must provide for the availability at all times of a senior clinician who can advise
on any action needed. It can never be acceptable for junior researchers to have to take
such decisions on their own without the opportunity for consultation (except when
essential because of the extreme urgency of the situation). It must be expected that
untoward incidents arising during the course of a study (irrespective of the researcher’s
judgement on whether the research procedures caused the incidents) are routinely
reported to the ethics committee for it to decide on their significance and on whether
there are any implications for the further continuation of the study. There must also be
a commitment to publishing the results of the research (see paragraph 2.5).
Quite appropriately, much research is of an exploratory or hypothesis-generating
nature and often the main purpose of a study is to train a student in research methods,
rather than to obtain a definitive answer to a research question. This is both scientifically
necessary and ethically acceptable. Nevertheless, all such studies should be undertaken
with appropriate skilled supervision; they should meet at least minimal scientific
requirements in relation to the experience of the researchers, the goals of the study and
the risk involved; and all the usual ethical requirements should be fulfilled. It is reasonable,
however, for the degree of scientific scrutiny to be proportional to the risks for both the
participants and the researchers.
Ethics committees must consider the scientific quality of the study
they are asked to approve and the adequate provision of expert
advice within the research team on the handling of risk situations.
The degree of scientific scrutiny should be proportional to the risks
involved. Untoward incidents arising during the course of a study
should be reported to the ethics committees, as a matter of routine.
4.2 Clinical care
It is a standard ethical requirement that a person’s clinical care must not be affected by
his or her unwillingness to take part in a study, or by his or her withdrawal from the study
during its course. Moreover, they must have the right to decline to take part, or to
withdraw, without giving a reason. It is also a standard expectation that participation in
a study should not result in a person receiving a worse standard of care than would
ordinarily be expected on the basis of the knowledge available at the start of the study.
We consider that these are fundamental ethical principles that must be maintained.
11
ROYAL COLLEGE
OF PSYCHIATRISTS
RESEARCH
GUIDELINES:
ETHICS
A person’s clinical care should not be affected by his or her unwillingness to take part in a study, or his or her withdrawal from the
study during its course. No reason for non-participation or withdrawal need be given. The ethics committee needs to determine
that, in these circumstances, it is practical for normal care to be
provided. Participation in a study, similarly, should not result in a
person receiving a worse standard of care than would ordinarily be
expected; the ethics committee should determine that the research
design will not have that adverse effect.
4.2(A) COMPARISON GROUPS AND PLACEBO CONTROLS
The general principles outlined in paragraph 4.2 can give rise to possible difficulties in
seven main circumstances concerned with comparison groups and placebo controls.
First, all evaluations of treatment require some form of comparison. It is therefore
necessary to ensure that the control or comparison condition should not involve a quality
of care below that which would be expected if the research were not taking place. This
means, for example, that if the control is provided by a baseline period without active
treatment, this period should not be substantially longer than the usual waiting list time
in that clinical service. Similarly, if the control is provided by random allocation to a
pattern of care other than that involved in the ‘experimental’ treatment to be evaluated,
that other pattern of care must be at least as good as that ordinarily available locally. This
means, for example, that ethically it is not acceptable for there to be a placebo control
condition if a treatment of proven efficacy is already available and would ordinarily be
provided. Occasionally, concerns are expressed that this brings about an undesirable
interference with scientific design. We do not agree. If there is already an effective
treatment, the relevant scientific question is whether the new treatment is more
advantageous in any sense (e.g. in reducing symptoms or impairment, in being less likely
to produce side-effects or in providing the same benefits in a more pleasant way) and not
whether the new treatment is better than no treatment. In any case, we consider it
ethically unacceptable that one arm of a treatment trial provides inferior care. If there is
reasonable doubt over the efficacy (and/or risks) of standard treatments, the use of a
placebo condition gives rise to no ethical problems. In that connection, it is up to the
investigator to make the case that there is doubt and it is up to the ethics committee to
decide whether that argument is valid. We accept that the case on doubt needs to be
based on the evidence as a whole (and not just one isolated study) and it needs to
consider the risks and benefits in the long-term as well as short-term.
Second, the scientific question may focus on the immediate pharmacological effects
of a new drug or psychological effects of a new psychological intervention, rather than on
the clinical benefits. In this circumstance the science may require a placebo control for
the period during which immediate effects are being examined. This is ethically acceptable
provided that:
(a) the science requires a placebo control;
(b) the time period is short;
(c) there is a demonstrated need to use patients rather than volunteers;
(d) the participant is fully informed of the details (including the
withholding of an active treatment) and agrees to take part.
The ethics committee should require a reasoned argument on the science (including
reasons why a placebo condition is required) and should satisfy itself that the study is
ethical in the light of the overall circumstances, including the current state of knowledge.
Third, there are practical difficulties in deciding during the course of a study when
sufficient evidence has accumulated to demonstrate a clear advantage of one treatment
over another (DeMets et al, 1999). History makes clear that some trials have been
terminated prematurely, whereas others have continued for an unreasonably long time
after it became obvious that one treatment was ineffective. An even more difficult
situation arises when the findings show a clear short-term benefit, but doubts remain on
long-term risks (this often applies with respect to preventive measures) – or vice versa.
There is not (and cannot be) a simple rule of thumb to apply in these circumstances.
12
Researchers have an obligation to consider carefully the balance between the scientific
and ethical considerations, and to present to the ethics committee in a fair fashion the
pros and cons of different courses of action. Data monitoring committees have a key role
in this connection.
Fourth, a dilemma arises when the clinical research team providing the experimental
and comparison treatments is not resourced to provide treatment outside the confines of
the research. In these circumstances there is an obligation on the researchers to make
explicit how ordinary clinical care will be provided if the individual refuses to take part in
the study or withdraws from it. There can be no expectation that the research approach
will enable special services outside of the research to be made available, but it must be
practical to operate the clause that participants can withdraw from the study without
detriment to their clinical care.
Fifth, there is also an ethical dilemma if (the study having been completed), when it
is demonstrated that one treatment has been effective for an individual participant and
needs to continue, there are no resources for the effective treatment to continue to be
provided after the research comes to an end. We recognise that there are cost issues
involved and that the way these are dealt with is outside the control of the researchers.
Nevertheless, we regard it as ethically dubious to stop an effective treatment that needs
to continue and we recommend that the relevant health authority considers in advance
how this eventuality would be dealt with. If there can be no expectation that an effective
treatment can continue after the end of the research, this should be made explicit in the
information sheet. There is a particular problem with research undertaken by UK
researchers in other countries where the expectation for all participants is that free services
are not available and it is unlikely that participants will be able to pay for the treatment
themselves (Nuffield Council on Bioethics, 1999). We recommend that the issue be
raised explicitly in advance with whatever pharmaceutical company is involved, asking
for continuing free treatment provision for research participants if the treatment proves
effective for the individuals and will need to be continued. Parallel issues arise with
respect to treatments other than those involving the use of drugs.
Sixth, in the course of follow-up studies it is common for research assessments to
reveal unmet clinical needs and equally common for participants to request clinical help.
By the nature of the research agreement between researchers and participants, there can
be no expectation that clinical services can be provided. Nevertheless, as a matter of good
clinical research governance, it is reasonable to expect the research team to be as helpful
as they can in informing participants how they may obtain services. The senior clinician
referred to in paragraph 4.1 should be expected to ensure that such advice is available.
Seventh, there are some circumstances in which scientific requirements would seem
to require a design in which a comparison group is subject to more than minimal risks and
that these are risks that they would not experience in the absence of the research. The
issues have been debated in relation to the clinical acceptability or otherwise of certain
forms of sham surgery (Freeman et al, 1999; Macklin, 1999). We recognise that opinions
are divided on this matter but, on the principle that a risky procedure would be against
the best interests of the individual, we recommend that the use of a risky procedure as a
control would not be ethically acceptable if it falls outside the scope of ordinarily expected
treatment. However, often it may be possible to devise sham procedures that serve the
scientific needs and yet are not risky; these may be acceptable if the participants are
offered the active treatment at a later date should it prove successful.
RESEARCH
GOVERNANCE
AND CLINICAL
CARE
Comparison groups are an essential part of research designs to evaluate the efficacy of treatment. When a treatment known to be effective is ordinarily available, this treatment should usually be chosen as
the comparison for the new treatment being investigated; this is
both scientifically and ethically appropriate. Placebo controls may,
however, be justified and ethically acceptable if the science requires
it and their use is not against the best interests of the individual.
4.2(B) TERMINATION OF THERAPEUTIC TRIALS
As discussed in the previous section, there are real dilemmas in decisions on when a
therapeutic trial should be terminated. These arise because of uncertanties over when
13
ROYAL COLLEGE
OF PSYCHIATRISTS
RESEARCH
GUIDELINES:
ETHICS
there is acceptable sufficient evidence that a treatment either is or is not effective, and
over the balance between risks and benefits (in the short-term and long-term) of the new
treatment being evaluated. Nevertheless, however diffucult the individual decision may
be, the ethical princlple is clear. Therapeutic trials must be terminated when it is clear
that one of the treatments being evaluated is ineffective or when a treatment is associated
with an unacceptable risk of harm.
Therapeutic trials must be undertaken with attention to the necessary steps required to provide adequate assessment of risks and benefits. Trials must be terminated when findings show either that a
treatment is ineffective or that it is associated with an unacceptable
risk of harm. For the latter to be apparent, there must be a systematic monitoring of untoward effects; these must be reported to the
ethics committee.
4.3 Payment to research participants
Although there has been a good tradition of altruism in clinical research, it is appropriate that
research participants should be reimbursed for their time, travel/subsistence expenses and
inconvenience. It is also appropriate that there be parity between payments made to patients
and payments made to non-patient participants. Nevertheless, it would be unethical for such
payments to be at such a high level that they either are, or could be interpreted to be,
inducements to take part in research to which participants would not otherwise agree. In
order to safeguard propriety, ethics committees should consider the size of reimbursements
made to participants. There is a lack of consensus in the scientific community on the principles
to be considered in deciding when and in what manner it is ethical to pay research participants.
The pros and cons of different models are helpfully discussed by Dickert and Grady (1999)
and we follow the approach that they recommend as providing the best balance. That is,
payment should cover expenses and should recompense people for their time and inconvenience
but, in order to avoid differing levels of payment according to earning power, this should be at
a standard unskilled labour level.
It is reasonable that participants in research should be reimbursed for their time, expenses and inconvenience. Ethics committees, however, need to ensure that the payments are not at
such a high level that they constitute an inducement to participate.
4.4 Payment to researchers
There are no ethical problems in research workers being employed (full-time or part-time) on
research grants or contracts. There is also no ethical objection to research workers whose
salaries are otherwise covered being reimbursed for their time, expenses and inconvenience;
the situation is directly comparable with payment to research participants. However, some
pharmaceutical companies have offered very substantial payments per patient recruited (or
per patient who completes the trial), the level being so high that it can only be interpreted as
being intended to provide an inducement. Similarly, unduly lavish entertaining may serve a
similar purpose. Application forms to ethics committees should include a specific question on
the level and nature of all payments to researchers, and ethics committees should satisfy
themselves that these are in keeping with the guidelines of the relevant professional bodies. It
should be a standard requirement that there are explicit statements on how much researchers
are being paid per patient, whether the payment is to the individual researcher or the institution
and what use is being made of the money.
The guideline on payment to researchers is the same as that on payment to participants.
14
4.5 Deception of research participants
RESEARCH
GOVERNANCE
It is unethical for patients or other participants to be included without their knowledge
and agreement in a study involving personal contact of any kind. Also, it is a standard
expectation that participants should be informed about the overall purpose of the research.
This need not always specify the hypothesis to be tested or the details of the sampling
because such information could create bias in reporting. If the study is a randomised
controlled trial, it is, however, essential that the information sheet provided to participants
be explicit that there will be random allocation to conditions, including a clear explanation
of what this involves and why it is scientifically necessary.
Ordinarily it is ethically unacceptable to deceive participants deliberately about the
nature or purpose of a study. Nevertheless, there are some types of research in which the
essence of the experimental design requires participants to be misled for the brief periods
required for the experiment. Care needs to be taken in giving approval to such studies but
there are circumstances in which this is both scientifically necessary and ethically
acceptable. However, in order to be acceptable, it is required that the participants be
appropriately debriefed after the experiment has been completed. It is never justifiable to
deceive participants without discussing with them afterwards what was done and why it
was necessary.
AND CLINICAL
CARE
Participants should not be included without their knowledge or
agreement in a study involving personal contact (see paragraph 3.3
regarding the exceptions with respect to group analyses of archived
data). Ordinarily, participants must also be informed about the purposes of the research in which they are being asked to participate.
There are occasional cases in which the essence of the scientific
design requires a degree of deception. Such research needs to be
carefully considered with regard to its ethical acceptability, but it
may be acceptable if scientifically essential and if there is appropriate debriefing at the conclusion of the experiment.
4.6 Safety of participants
The broad issue of confidentiality is discussed in paragraph 4.8 and paragraph 4.12, below, but
specific consideration needs to be given to whether or not the patient’s general practitioner (or
hospital or clinic professional if ongoing care is provided there) should be informed about
either their participation in a study or findings from the research that have medical implications.
We consider that all personal medical information belongs to the patient and it is ethically
unacceptable to pass on such information without the patient’s consent. Permission may be
withheld unreasonably but there are valid reasons why patients may consider that they will be
disadvantaged if information is passed on. Either way, the responsibility for the decision on
disclosure of information lies with the patient.
The issue is somewhat different when the research involves an intervention (such as
drug treatment) that could be dangerous if other medical treatments were given in
ignorance of the research intervention. When this is a significant risk, it would be unethical
to include the person in the study if permission to disclose the information on the research
participation was withheld. It is the responsibility of the ethics committee to decide
whether the risk is sufficient to require that the patient should not enter the trial.
All personal medical information belongs to the patient who has the
right to refuse this information (either with respect to participation
in a study or to findings from the research) being passed on to his or
her medical carer unless the safety of others is in jeopardy (see paragraph 4.8). If such refusal constitutes a sufficient danger to the participant (because of other treatments given in ignorance of the research intervention), it would be unethical to put the patient at risk
by inclusion in the study.
15
ROYAL COLLEGE
OF PSYCHIATRISTS
RESEARCH
GUIDELINES:
ETHICS
4.7 Safety of researchers
It is unethical to undertake research that puts research staff under avoidable, or
unreasonable, risk. The ethics committee needs to make clear to researchers that they
have a responsibility for the safety of their research staff. All good research teams should
have written guidelines and set procedures for dealing with whatever risks are likely to
arise in the context of the particular type of research being undertaken.
Acceptable research governance requires that research teams have agreed
procedures for minimising the risks to staff involved in the research.
4.8 Safety of others
The confidentiality of research findings should be regarded as directly comparable with the
confidentiality of clinical findings. This means that, in all ordinary circumstances, findings
should not be passed on to anyone else without the participant’s explicit permission. There are,
however, legal limits, so that, for example, if evidence is obtained that there is current child
abuse, researchers are required by law to inform social services. When this occurrence is at all
likely because of the nature of the research, the information sheet should be explicit that
confidentiality can only be maintained in so far as the law allows. In practice, if such evidence
is dealt with sensitively by experienced clinical researchers it is extremely rare for it to be
necessary to break confidentiality without the participant’s consent.
A more common situation, posing a difficult ethical dilemma, is when genetic
information about the participant has serious risk implications for other members of the
family (Nuffield Council on Bioethics, 1993). In these circumstances, the seriousness of
the situation must be discussed between the participant and a senior clinician (either
within or outside the research team) knowledgeable about the specifics of the risk. If the
participant still refuses permission for relevant information to be given to other family
members, consideration needs to be given to whether the risks are such that the
confidentiality should be breached. In these circumstances it would be good clinical
practice to seek the advice of experienced colleagues and, if appropriate, of the ethics
committee. The criteria to be applied should be regarded as exactly comparable to those
expected if the same information arose in the course of clinical care. Exactly the same
applies with respect to very serious risks of other kinds (such as homicidal threats). In
practice, it is exceedingly rare not to be able to handle the risks effectively without
breach of confidentiality (through discussion with the participant), but both the law and
good clinical and ethical practice recognise that there can be situations in which
confidentiality must be breached.
When the nature of research means that it is likely that findings relevant to other
family members could arise (as would be the case with some types of genetic studies), the
contingency should be discussed at the point of obtaining informed consent with the aim
of obtaining agreement for disclosure. If this is refused, the researcher needs to consider
whether it is ethically justifiable to include that participant in the research.
Research findings, like clinical findings, should in all ordinary circumstances be regarded as confidential and should not be passed
on to others without the participant’s explicit permission. There
are, however, rare circumstances in which the law (and/or good
clinical and ethical practice) requires that confidentiality be
breached. The criteria for when this rare occurrence can be considered to arise are the same in research as in clinical practice. When
the nature of the research means that it is likely that findings could
have implications for other family members, this contingency should
be discussed at the point of obtaining informed consent, with the
aim of obtaining agreement for disclosure.
4.9 Safety of children in relation to researchers
16
Currently, it is an expectation that university and NHS employed researchers working
directly with children have a police check before starting such work. This is not, so far,
required of those employed by the Medical Research Council (MRC) or any of the
charities. The police check is focused exclusively on crimes related to children and the
only information passed to the employing authority is whether or not there is a relevant
risk. Clearly, this provides a highly imperfect screening, but, in our view, it is better than
no screening and we recommend that it should be mandatory for all researchers working
with children. Ordinarily, the personnel department should have the responsibility for
monitoring that the police check has been made.
RESEARCH
GOVERNANCE
AND CLINICAL
CARE
All researchers working directly with children should have a police
check for crimes relating to children before starting such work.
4.10 Feedback of general research findings to participants
Participants in research have the right to expect to be informed about the general results
of research in which they have participated. The practice of providing feedback in the
form of succinct newsletters (written in readily understandable language) summarising
the key findings from the research and their implications is growing and should be
strongly encouraged. It would be reasonable for this to become a standard requirement
for all major studies in the future.
Researchers should be strongly encouraged to provide feedback to
participants on the general findings and implications of research in
which they have participated.
4.11 Feedback of specific research findings applicable to the individual
When research involves clinically skilled individual assessments, or when tests giving rise
to individually valid findings are undertaken, it should ordinarily be expected that
participants will be informed of findings that have substantial medical significance for
them as individuals. This would apply, for example, to the discovery of hitherto undetected
diseases such as a brain tumour revealed on a scan obtained for some other research
purpose. It is essential that when feedback on specific research findings is provided that
this be done in person (and not by letter), that the professional giving the feedback
understands the clinical implications and that the information be given in a clinically
sensitive manner with time available for discussion.
There are two main circumstances, however, in which it might well be unethical
to provide feedback. First, feedback would be inappropriate when the clinical
significance of the ‘abnormalities’ found is unknown or uncertain. This would apply,
for example, to some forms of chromosomal anomaly for which there is no known
functional consequence, or some laboratory test findings that are outside the supposed
normal range but appear to have no clinical significance. In many instances further
testing will resolve the uncertainties but, when this is not the case, it is not clinically
appropriate or helpful to tell the participant that they have some ‘abnormal’ finding
but it is not known what it means. Second, in many large scale epidemiological
studies the testing is undertaken by people who are not clinically trained and/or the
tests used are satisfactory for detecting group differences but not satisfactory for
individual diagnosis. It has become standard practice in many such studies to make
it explicit on the information sheet provided at the time of obtaining informed consent
that individual findings will not be fed back to participants. That is sound ethical
practice.
When research involves clinically valid assessments relevant for individual diagnosis, it should be expected that participants will be informed on all findings that have substantial medical significance for
them as individuals. This must be done by someone who understands
the clinical implications of the findings and the information must be
given in a clinically sensitive manner. However, when the findings are
of unknown significance or when they have meaning only at a group
level, feedback should not be expected. For studies from which the
findings are likely to be of this kind, the information sheet should be
explicit that individual feedback will not be provided.
17
ROYAL COLLEGE
OF PSYCHIATRISTS
RESEARCH
GUIDELINES:
ETHICS
4.12 Data protection and sharing
Data collected during the course of research should be protected and held securely and
confidentially under the control and supervision of the research group originally acquiring the
data. These requirements apply equally to written records, video- or audiotaped recordings,
biological samples and computerised data; the expectations with respect to security and
confidentiality are the same as those that apply to clinical information (Caldicott Committee,
1998). There should be a record of the name of a given patient/participant that is linked to
their data; this is necessary for the purpose of corroborating data and also in case of an urgent
need to contact the person if information is obtained on a condition requiring intervention. So
far as possible, however, all records (in whatever form) should be stored using identification
numbers, rather than names, with the link between number and name kept separately. Ethics
committees have a responsibility to ensure that research leaders for all studies are aware of
their responsibilities with respect to data protection, and operate systems that provide adequate
safeguards with respect to security and confidentiality.
At the time of recruitment into a study, participants are (through the information
sheet and consent form – see paragraph 5.3) entering into an agreement on the use of the
data they provide. This involves two key elements:
(a) the purpose of the research (and the uses to which information will
be put);
(b) the research group to whom responsibility for the data is being given.
So far as the latter is concerned, the agreement is, and needs to be, with a research
group; the composition of the group may change over time, but there must be a leader and
contact person (with work address and telephone numbers) specified on the information
sheet. Much research now is, and needs to be, collaborative and, hence, research data will
need to be shared with a broader range of researchers at other centres, some of which may
be abroad. In most cases, the sharing can be of anonymised data, and where this is possible,
it should be the rule. In some cases, however, the sharing of personalised data may be
necessary (as, for example, in order to check the reliability of ratings). This is ethically
acceptable, under the concept of extended confidentiality (Wald et al, 1994) provided that:
(a) the research leader exercises scrupulous research governance
regarding data security and confidentiality;
(b) the fact that this may happen is made explicit in the information sheet;
(c) the data are shared with researchers working in a country with a
data protection regime at least as rigorous as that in the UK.
The expectation that the consent agreement is on the basis of a known defined
purpose poses more problems when archived data of any kind may, for very good reasons,
be used for purposes not envisaged at the time consent was obtained. This is particularly
likely in the case of databanks that pool data-sets and make them available to a wider
group of researchers. We consider that this is ethically acceptable provided that:
(a) it is made explicit in the information sheet;
(b) the databanks operate ethics committees to check both the bona
fides of researchers applying to use databanks and the scientific
legitimacy and ethical propriety of the proposed research use.
Research data should be subject to the same safeguards on security
and confidentiality as clinical data. The custodianship of the data
should be the responsibility of the named lead researcher for the
study that collected the data initially but, given appropriate safeguards, the data may be shared with collaborating researchers under the concept of extended confidentiality.
4.13 Letters and telephone calls to participants
18
Researchers need to assume that initial letters and telephone calls to potential participants
(particularly in follow-up studies based on addresses and telephone numbers that are
long out of date) may be intercepted by third parties and, therefore, that such
communications should not implicitly or explicitly provide personal information. Thus,
for example, postal franking and letter heading should not be such as to identify the
communication as coming from a psychiatric facility (or any other sensitive source) and
the content of the letter should not provide any information about individuals that could
be regarded as confidential.
RESEARCH
GOVERNANCE
AND CLINICAL
CARE
Researchers need to be aware that letters and telephone calls to potential
participants, especially when using addresses and telephone numbers
that may no longer be applicable, may be intercepted by third parties. This
possibility requires attention to the need to avoid initial communications
that, implicitly or explicitly, include personal information that could inadvertently provide a breach of confidentiality.
4.14 Biological samples
Permission to use biological samples of any kind should be obtained in the same way as for any
other form of research information, and the materials (and associated identifying data) should
be subject to the same safeguards. It is an ethical requirement that consent for research use be
obtained at the time of sample collection if research use is expected or intended, and that this
is required even if most of the sample needed is to be obtained anyway for clinical purposes.
Reference should be made to the MRC guidelines for collection of human tissue samples
(Medical Research Council, 2001)for details on handling.
Situations arise, however, when unused portions of samples obtained with the sole
intention of clinical use are kept and an unanticipated research possibility arises later.
The issues that pertain to this situation were discussed in paragraph 3.3.
Informed consent must be obtained if it is envisaged that a portion
of a biological sample obtained for a clinical purpose may also be
used for research.
4.15 Avoidance of overload on participants
Particularly in very active research centres, there is a potential danger that some participants
may be overloaded with research requests. In some cases, this may involve significant risks –
such as the repeated exposure to radiation (as a result of people being in several studies with
this feature, without the researchers being aware of the other investigations). In other instances,
it may involve no more than an unreasonable imposition with respect to time and inconvenience.
It is part of good research governance to ensure that such overload does not occur. It would not
be feasible to expect ethics committees to monitor load on participants, but they can expect
their organisations (trusts, universities, etc.) to seek to ensure that researchers coordinate their
studies in relation to load on participants, and they can (and should) alert lead researchers to
their responsibilities in this connection. Also, when significant potential risks are involved
(such as through repeated irradiation) they should require appropriate questions on past
exposure to be included on the information sheet. In some instances, it may be appropriate to
ask participants for permission to inform their general practitioner about their research
involvement so that the general practitioner can serve as a check on possible over-involvement.
Ethics committees should emphasise to their organisations and to
the applicants asking for ethical approval of studies the need to
avoid undue overload on research participants.
4.16 Collaboration and consultation with participant groups
It is good research (and ethics) practice to consult members of the community to be
studied prior to finalising research plans (Patel, 1999). Such consultation might, as
appropriate to the individual study, involve consumer groups, or professional groups (who
will be asked to provide information or participate in other ways), or community groups
(including religious or ethnic minority groups). The several purposes of consultation
19
ROYAL COLLEGE
OF PSYCHIATRISTS
RESEARCH
GUIDELINES:
ETHICS
include the likelihood of better participation rates if consumer concerns have been
addressed, alerting to research issues relevant to measurement and to the meaning of
measures, and attention to sensitivities regarding possible misinterpretations or misuses of
the research project or its findings as they apply to groups as well as to individual
participants.
It is good research (and ethics) practice to consult members of the
community to be studied when planning studies. As appropriate to
the individual study, such consutation might involve consumers, professionals, community members or ethnic/religious groups.
20
5. INFORMED CONSENT, CAPACITY AND COMPETENCE
RESEARCH
GOVERNANCE
AND CLINICAL
CARE
The ethical acceptability of research requires that it be based on the freely given informed
consent of participants (exceptions in relation to archived data were discussed in paragraph
3). This is no different in the case of participants who may suffer from psychiatric disorder,
but the presence of mental disorder or impairment raises certain special issues of ethical
concern. The ethical acceptability of research also resides on the premise that it has a
sound methodological basis, uses methods that are adequate to provide valid results, is
undertaken by researchers with the necessary skills and experience and involves a degree
of risk that is commensurate with the benefits and that has been kept as low as possible in
relation to the scientific needs.
5.1 Opting in and opting out of participation
Ordinarily, it is necessary that participants should actively opt in to research, rather than
have researchers assume their agreement provided that they do not actively opt out. This
provides a problem, however, in the case of large scale epidemiological studies (which may
require samples of tens, or even hundreds, of thousands) when parents are consenting on
behalf of children. Experience shows that opting in on the basis of a posted letter (or letter
sent via the child) often results in very low response rates, which is likely to bias or even
nullify the research. Furthermore, when samples of non-responders have been contacted
personally it has been evident that the great majority were quite happy for their children
to be involved; it was just that they had not got round to replying or had mislaid the letter
of invitation. It has become accepted now in several major studies in the UK that an ‘opt
out’ strategy is ethically acceptable and practically feasible provided that:
(a) all parents are sent an information sheet about the study and are
given full details of whom to contact if they do not wish their child
to take part, or if they wish for further information before deciding;
(b) the research procedures are of a kind that are a normal part of
ordinary routine practice (such as the use of educational tests) or
involve no significant risk (such as the use of a well tested
questionnaire or interview);
(c) the children are given appropriate information about the study, are
able themselves to give or withhold their consent and that their
refusal is respected;
(d) the details of the study have been scrutinised and approved by the
appropriate ethics committee. An opt out strategy is not ethically or
legally acceptable for any medically intrusive procedures (such as
venesection) even if the risks are minimal.
Ordinarily, it is necessary that potential participants actively opt in
to a study. Opt out procedures with respect to parents consenting
on behalf of children may be ethically acceptable, however, for large
scale studies involving routine type procedures carrying no significant risk so long as certain safeguards are provided and the details of
what is proposed have been approved by the appropriate ethics
committee. It will always be necessary for the children to be given
appropriate information, for them to give or withhold their consent, and their refusal should be accepted.
5.2 Vetoing of access to patients
It is a fundamental principle that patients should have the freely exercised right to
participate in sound research. Accordingly, family doctors, or other professionals, do not
21
ROYAL COLLEGE
OF PSYCHIATRISTS
RESEARCH
GUIDELINES:
ETHICS
‘own’ patients on their lists and should not be able to refuse access if the potential
participants are known to the researcher through some other route (such as through
membership of a user organisation, attendance at some other facility or a case register).
Similarly, the issue applies to the provision of addresses to enable potential participants to
be contacted to ask if they are willing to take part in a study (see paragraph 3.3). An
exception to this guidance arises when the patient is currently receiving treatment and
when the carer considers that the research could jeopardise the care being provided.
Participants should ordinarily have the right to decide for themselves whether or not they wish to participate in a study, and professionals should not have the right to refuse access when the potential
participant is known to the researcher through some other route.
An exception arises when the patient is currently under active treatment and the carer considers that the research could jeopardise the
care being provided.
5.3 Information sheets
22
It is crucial that respondents’ agreement (or otherwise) to participate in research be based
on an adequate understanding of the purpose of the study and what will be entailed if he
or she agrees to take part. The research purpose and procedures must be expressed in
terms that have meaning to a lay person, must be in a language that is likely to be
understandable to all respondents and specific aims must be made reasonably explicit.
General lofty aspirations (such as ‘to find the causes of schizophrenia’ or ‘to learn about
the best treatment for depression’) are not acceptable. Similarly, the information sheet
should be explicit about what is required of the participants with respect to procedures
(e.g. complete questionnaires, be interviewed, have a blood test, etc.) and give a reasonable
estimate of the time likely to be involved. It is not necessary to give details of all questions
to be asked in a questionnaire or interview, but a clear notion of the types of topics to be
covered should be given. Significant risks must be noted, together with an indication of
how they will be dealt with. The information is given for the potential participant to come
to an informed decision on whether to agree to take part; it is not intended as a legal
protection for the researcher in relation to unlikely hazards. Thus, for example, if a
treatment is thought to carry an important risk, this must be stated, but it is not necessary
(or desirable) to list all possible remote or minor side-effects that could occur. Researchers
should, however, be ready and willing to provide more detailed information if the
participant requires it. The guiding principle is that the consent be based on a full and
frank disclosure of the relevant facts. The information sheet should provide a succinct
summary but it should not be a substitute for personal discussion between the researcher
and the prospective participant (see paragraph 5.4).
It is helpful to all concerned for the ethics committee to develop a set of recommended
standard descriptions of standard procedures and their risks. Thus, these might be made
available for the different types of brain scan, for randomisation and for the various forms
of genetic study. Some illustrative examples are given in Appendix 1. The form of words
should have been checked with user representatives. Jargon terms should always be
avoided. Care should be taken to ensure that the language used is neither discriminatory
nor derogatory. Applicants should be instructed to have checked the reading level
required for the information sheet (using appropriate computer software).
The information sheet should have a readily understandable title for the study and
should give the name, work address and contact telephone numbers of the lead researcher
responsible for the study and the person to contact with respect to queries or practicalities
(if different).
Application forms should include a mandatory question on possible conflicts of interests
(these go beyond the source of funding of the particular study that is the subject of the
application) and this should be parallel to the same issue with respect to committee
members (see paragraph 6.7). They should also include questions on payments to
researchers and to participants.
If some of the sample require special consideration, there should be separate versions
of the information sheet tailored to their needs (but what is said must always be compatible
across versions). Thus, if the sample includes children, there should be consideration of
whether there should be separate versions of the information sheet for parents and children.
Similar considerations may apply when some of the sample are patients and some are
recruited as volunteers.
In multi-centre studies, the information sheet and consent form should be standardised
in their important details in order to ensure comparability in samples and procedures
across centres.
The information sheet must state explicitly that respondents have a choice on whether
or not to participate and that declining to take part will not affect the availability of
normal treatment. Similarly, it should be stated that participants may withdraw from the
study at any stage and need not give a reason for doing so.
Despite the need for the information sheet to provide key details about the study, it
is essential that it be in a form, and of a length, likely to be read and understood by the
respondent. Thus, it must be a succinct statement and not a discussion essay. Ordinarily
it should be confined within a single page of A4 paper.
INFORMED
CONSENT,
CAPACITY AND
COMPETENCE
Researchers have a responsibility to ensure that all respondents have
a clearly written, readily understandable information sheet that is
explicit on the purpose of the study, the procedures involved, what
is required of the participants, details of the lead researcher and
contact person, any significant risks and the right to decline to take
part (or to withdraw during the study) without giving a reason and
without detriment to normal treatment.
5.4 Consent forms
It is a requirement that participants have the opportunity to ask questions or raise concerns
and that consent be obtained only after the person is satisfied that they have received the
information they want, have understood what is involved and have been satisfied that
their concerns have been addressed and dealt with to their satisfaction. Usually, the
discussion and the obtaining of consent can be dealt with on the same occasion. However,
it is good clinical practice to allow potential participants the time and circumstances to
consider their decision and they should be allowed to think over the issues if they wish to
do so before deciding whether or not to consent.
Ordinarily, there should be a signed and dated consent form that is stored with the
individual research record. When video- or audiotapes are used the recording should
include a question and the respondent’s answer about the fact that the interview/test/
observation is being recorded, and that the participant is willing for that to be done. The
consent form need not repeat the information in the information sheet, but it should
include a statement that the information sheet has been provided and that the participant
has been able to have his or her queries answered. If a written consent is not to be
obtained, the ethics committee should be satisfied that verbal consent is sufficient and
that the reasons for not obtaining written consent are valid and ethically acceptable. If
oral consent is used, a record must be kept of the relevant circumstances (including the
information given, how consent was obtained and who was present).
Ordinarily, a written consent should be obtained; this should specify
that the information sheet has been given, read and understood and
that what is involved has been explained to the satisfaction of the
participant.
5.5 Inducements and undue influence to participate
The GMC (1998) has stressed that there must be no coercion placed upon potential
participants to agree to take part in research. There are three main situations in which
there may be particular concerns. First, whereas it is appropriate to reimburse participants
(see paragraph 4.3), the level of reimbursement must not be such that it could act as an
inducement to agree to participate. Second, hierarchical relationships within organisations
can place undue pressure on people to participate. This may arise if patients wish to please
their doctor by participating and if they fear that non-participation may jeopardise their
medical care or the time they have to wait for it. It may also arise with non-patient
23
ROYAL COLLEGE
OF PSYCHIATRISTS
RESEARCH
GUIDELINES:
samples if the respondents are students or employees, who may feel that their careers will
be fostered by participation in the research (see paragraph 5.6(g)). Two safeguards are
needed:
(a) an explicit statement in the information sheet (see paragraph 5.3);
ETHICS
(b) discussion with potential participants on whether or not they wish to
participate, which should be undertaken by someone other than
the person in a hierarchical relationship with the respondent.
Third, detained persons (either patients or prisoners) may feel under pressure to
consent to participate or may think that participation will afford them more favourable
treatment or will make early release more likely. These groups are considered more fully
in paragraphs 5.6(e) and 5.6(f), but the safeguards to be employed comprise:
(a) relevant explicit statements in the information sheet;
(b) discussions on whether to participate by a researcher who is not in a
position (real or perceived) to influence detention;
(c) an independent opinion on participation from someone with relevant
experience.
Scrupulous care must be taken by researchers to avoid undue influence on potential participants to agree to take part in research.
Attention is particularly necessary with respect to payments to participants, the dangers of pressure implicit in hierarchical relationships and the pressures (and opportunities) that may be perceived
by detained persons.
5.6 Competence and capacity to consent
Of all the ethical issues regarding research with psychiatric patients, those surrounding
competence and capacity to consent are the most problematic for both practical and
legal reasons7.
Competence is a legal concept that, by its nature, has to be categorical. That is, in
order to decide whether a person can or cannot provide consent, it is necessary to
determine whether a particular individual is or is not competent to consent to a given
procedure at a specific time.
Capacity is a clinical term that is generally taken to mean that a person has the
cognitive and other psychological skills/understanding necessary to be capable of coming
to an informed decision on whether or not to consent. At one time, this used to be
thought of in ‘status’ or categorical terms, meaning that an individual either did or did
not have the capacity to consent. Research findings and clinical experience have made
it clear that that is a mistaken approach because skills/understanding constitute a graded
dimensional phenomenon (rather than something that is categorically present or absent).
Moreover, it is likely to be influenced by training and by how the issues are expressed.
Also, it may fluctuate over time according to mental state and circumstances. In addition,
the level of capacity will vary according to the complexity of particular decisions and the
person’s experience of the related features, and according to context (see Appendices 2
and 3).
The need to consider capacity in dimensional functional terms, together with the
recognition that individuals can be helped to have the capacity necessary for specific
decisions in specific contexts, mean that researchers are under an obligation to seek to
enable potential participants to achieve the capacity required. This requires that the
issues be put in practical concrete (rather than abstract) terms; that steps should be taken
(if necessary, over several sessions) to help the person understand what is involved in the
24
7. The Medical Defence Union (1997) guidance and the British Medical Association and the Law
Society publication (1995) to which it refers, are reasonably clear and helpful, but the therapeutic/
non-therapeutic distinction, which we reject, is used. The Law Commission (1995) report adopts a
broadly functional approach. The legal situation remains somewhat ambiguous because the key
issue of the true meaning of capacity and how to measure it remains unresolved (see Kennedy, 1997).
decision and in its consequences; when appropriate, reference should be made to parallel
situations that the person has already experienced; and the discussion should use language
appropriate to the person’s understanding.
Irrespective of the categorical decision on competence that is legally required, it is
always clinically desirable for respondents to be helped to understand what is involved
and to participate in decision-making at whatever level is possible8.
INFORMED
CONSENT,
CAPACITY AND
COMPETENCE
The capacity to give consent is task- and time-specific, it constitutes
a graded dimension of understanding, and it is something that can
be influenced to some degree. Researchers should seek to help respondents achieve the capacity needed for the specific decision
needed. Although, legally, a categorical decision on whether a person is competent to give consent is required, individuals whose capacity falls below that level should be helped to understand what is
involved and to participate in decision-making. Ethics committees
should satisfy themselves that the materials and process used to
facilitate understanding are adequate.
5.6(A) CHILDREN
The issues already outlined (paragraph 5.6) with respect to the need to aid understanding
of the research participation issues, and to include in decision-making respondents whose
capacity falls short of that required for competence, apply to children (Smyth & Weindling,
1999). Researchers should start with the assumption that, until proved otherwise with a
particular child, all children are able to achieve some degree of understanding (however
partial). Even when the child lacks the capacity to give consent, it is good research
practice to engage the child in the decision-making process and to seek to achieve the
child’s assent to whatever is required.
At one time the criterion for including children in research was that it was therapeutic
and, therefore, in the child’s personal best interests. That is no longer regarded as an
ethically acceptable criterion because it would serve to exclude children from research
that could lead to major advances in the treatment or prevention of diseases that are a
substantial source of morbidity or mortality in early life. Rather, the criterion is that the
research needs to be done with children because it concerns disorders or diseases of
childhood and because answers to the scientific questions could not equally well be
obtained using adult participants who are competent to give consent for themselves.
Also, it is necessary that the research should not be against the child’s best interests.
In this circumstance, parents are able to authorise children’s participation in a study
(and should sign an appropriate consent form making that explicit) provided it presents
no more than minimal risks9. On that criterion, it would be permissible, for example, to
agree to venesection or a low radiation dose X-ray or scans (such as magnetic resonance
imaging) that do not involve a radiation risk. Steps should always be taken through
adequate preparation and specific interventions (such as anaesthetic cream for
venesection) to reduce the distress of the child. Videotapes and demonstrations have
been found to be very helpful in gaining children’s participation in procedures such as
scans.
The situation is not quite as straightforward when the research (therapeutic or nontherapeutic) involves risks that are more than minimal. Provided that the scientific need
is sufficiently great, specifically applies to children and could not be met with research on
competent adults, the research may still be ethically acceptable10. We emphasise that the
scientific need must be judged strictly in terms of the potential benefits with respect to
8. There is a growing literature on semi-standardised approaches to the assessment of capacity/
competence, but there are no widely accepted instruments even for use in the clinical context and
the validity of attempts to seek to develop a standard tool are highly dubious, given that capacity is
task-specific.
9. In the USA, minimal risks have been defined as those no greater than the risks normally
encountered during routine physical or psychological examinations or tests (Federal Policy for the
Protection of Human Subjects, 1991). Although this has no legal standing in the UK, this seems a
sensible, workable definition to use.
10. The legal situation in the UK regarding such research on children remains unclear.
25
ROYAL COLLEGE
OF PSYCHIATRISTS
RESEARCH
GUIDELINES:
ETHICS
the disorders from which the children suffer (or to which they are at risk) and not in
terms of the scientific interest as viewed from the perspective of the researcher. However,
we recommend that a strong case would need to be made and, in addition to obtaining
parental assent, it would be essential to obtain an independent professional opinion on
both the scientific needs and ethical acceptability. The research should be approved only
if that opinion is unambiguously supportive. Parents are ordinarily the best judges of what
is acceptable for their children but there may be occasions when, in their fully reasonable
concern to achieve a solution to the disease/disorder from which their child suffers, they
may be too ready to give assent. Parental assent is a necessary but not a sufficient criterion
for approval for a child to be included in research involving more than minimal risk.
Two situations require specific consideration. First, there is the question of what is
acceptable when a child expresses (verbal or non-verbal) refusal to participate, despite
parental assent having been obtained. We advise that, when the procedures are more
intrusive than those required for ordinary clinical care, the child’s refusal should be
accepted as a sufficient reason not to proceed. The child’s perception of what is intrusive
should always be considered. Second, there is the question of whether parental refusal
can be overridden. Clearly, there are circumstances (such as when there has been child
abuse or neglect) in which it cannot be assumed that parents will act in the child’s best
interests. The legal situation is clear that when the child has been removed from parental
care, the parental responsibilities are then transferred to some other person or body.
When that has not happened, it would be unethical to proceed without the assent of the
parent. Ordinarily, the expectation is that both parents should assent, but if the child is
living with just one parent who is exercising decision-making in other aspects of the
child’s life, the assent of that one parent may be sufficient. Given the complications of the
many permutations on child care that are possible, each case will need to be decided on
its own merits. Great caution should be exercised before proceeding when the parents
disagree over the child’s participation in a study.
It is good research practice to engage children in the decision-making process even when they lack the capacity to give consent. The
issues with respect to assessing capacity outlined in paragraph 5.6
apply to children. Parents should be able to authorise children’s
participation in research provided it presents no more than minimal
risks. If the research involves risks that are greater than minimal, it
could be ethically acceptable if the scientific need is sufficiently great,
specifically applies to children and could not be met with research
on competent adults. A strong case would need to be made in this
circumstance and, in addition to parental assent, it would be essential for there to be unambiguous support from an independent professional with respect to both scientific needs and ethical acceptability. With procedures that are more intrusive than required for
ordinary clinical care, a child’s refusal should be accepted as a sufficient reason not to proceed, irrespective of parental consent.
5.6(B) ADULTS INCOMPETENT TO CONSENT TO RESEARCH
26
Despite extensive legal and ethical discussion, it remains the case that, in the UK, the
law does not recognise proxy consent for adults. The issue of obtaining informal consent
with adults suffering from a psychiatric disorder is not a problem in most instances. Many
patients do not have an impaired capacity (Appelbaum & Grisso, 1995; Grisso &
Appelbaum, 1996; Wong et al, 1999, 2000) and, even when initially they appear to lack
capacity, it is often possible to improve understanding through the provision of appropriate
information, discussion and education; the improvement may often be sufficient for
informed consent to be obtained (Appelbaum, 1998). This may be the case, for example,
in the early stages of dementia – a common condition in the psychiatric care of older
people.
Nevertheless, there are important common situations requiring research in which
potential participants are not competent to consent. This applies, for example, to the
more severe forms of dementia or learning difficulties and to some cases of schizophrenia
or bipolar affective disorder, to mention a few examples.
In these circumstances it is important to have a way of proceeding in order to ensure
that individuals from these groups are not prevented from profiting from the benefits of
research simply because of their membership of a group likely to have impaired capacity.
We recommend, following the same principles as those discussed in relation to children,
that it is ethically acceptable to include in research (therapeutic or non-therapeutic),
without informed consent, adults who are incompetent to consent provided that:
(a) the study is concerned with the disease, disorder or disability from
which the patient suffers;
INFORMED
CONSENT,
CAPACITY AND
COMPETENCE
(b) the research cannot be undertaken with validity in less vulnerable
groups with the same disorder who are capable of giving consent;
(c) adequate steps have been taken to help the individual gain sufficient
understanding so that informed consent could be obtained;
(d) the individual’s closest relative or cohabiting partner has given assent
on the basis of adequate information and discussion;
(e) the patient’s professional carer (medical or non-medical, as
appropriate) is of the firm opinion that the risks and benefits for the
individual, taken in consideration with the relevance of the research,
warrant the person being included in the study;
(f) this opinion is reviewed and ratified by an appropriately experienced
clinician who is independent from the study;
(g) the risks are minimal (using the criteria as we have defined them
above);
(h) the research is not against the individual’s best interests;
(i) the study as a whole has been reviewed and approved by the
appropriate ethics committee.
No individual should be disqualified, by virtue of his or her group
membership, from participating in research (as a result of incompetence to give consent or other reason) that could be of benefit in
relation to the disease, disorder or disability from which he or she
suffers. It is ethically acceptable to proceed without personal informed consent provided that a specified set of conditions has been
met. These include the relevance of the research, the fact that the
research cannot be undertaken with validity in less vulnerable groups
with the same disorder, the assent of the individual’s closest relative
or cohabiting partner, the support of both the person’s professional
carer and an independent clinician, minimal risks and approval by
the appropriate ethics committee. However, as with children, the
patient’s refusal should be accepted as sufficient reason not to proceed, irrespective of other consents.
5.6(C) EMERGENCY RESEARCH WITH ADULTS INCOMPETENT TO CONSENT
There are some further considerations that apply to circumstances in which research
decisions need to be taken very quickly. This would be the case, for example, with
patients who are in a state of severely impaired consciousness following head injury or
stroke (see Appendix 4). This is a life-threatening situation requiring emergency action
but in which existing treatments are unsatisfactory. In this circumstance, the research is
in the patient’s best interest because the emergency necessitates intervention, and the
research risk may be reasonable compared with the risks of existing therapy, or the
expected outcome without therapy. The patient is incompetent to give consent at the
time the research intervention needs to be introduced and, in most instances, it would
not be feasible to contact a close relative (or they may be in too distressed a state for their
consent to be truly informed and dispassionate). In keeping with the decisions of a local
or multi-centre research ethics committee with respect to a specific randomised controlled
trial of treatment in such an emergency situation, we recommend that it is ethical to
proceed without consent if eight conditions are met:
(a) the existing treatment is of dubious or limited effectiveness;
(b) the research treatment is promising but unproven;
27
ROYAL COLLEGE
OF PSYCHIATRISTS
RESEARCH
GUIDELINES:
ETHICS
(c) the research risk is reasonable compared with the risks of the existing
treatment or natural outcome (but not necessarily minimal in the
ordinary sense);
(d) the doctor dealing with the emergency treatment of the patient
considers that the balance of risks and benefits for the individual
patient warrant going ahead with inclusion in the study;
(e) the decision to include the patient in the randomised trial (or other
type of study) is reviewed and ratified by an appropriately experienced
clinician who is independent of the research;
(f) there is no objection from any available relative or cohabiting partner;
(g) the study as a whole has been reviewed and approved by the
appropriate ethics committee;
(h) in keeping with Food and Drug Administration (1999) recommendations, we advise that, in addition and whenever possible,
waived consent should be followed by deferred consent. That is, as
soon as possible, the patient, family or guardians should be informed
about the research (British Paediatric Association, 1992) and given
the option of continuing in the trial or withdrawing from it.
It is ethical to waive consent in the study of emergency treatments
for life-threatening situations affecting individuals who are incompetent to give consent, provided the principles outlined in paragraph 5.6(b) have been followed. The specifics in such emergency
situations differ, however, in two respects. It is acceptable to proceed without the assent of a relative if that cannot be obtained in
time (but not ethical to proceed in the face of a relative’s objection)
and the risks and benefits should be judged in relation to those
associated with existing treatments or outcomes (rather than minimal risks in an absolute sense). In planning research that deals with
these circumstances, there should be appropriate consultation with
the relevant user groups (see paragraph 4.16).
5.6(D) PEOPLE WITH LEARNING DIFFICULTIES
28
The same general principles outlined in paragraphs 5.6 to 5.6(g) apply to individuals with
learning difficulties (see Appendix 3). It should not be assumed that people with learning
difficulties are incompetent to give consent; active steps should be taken to enable them
to achieve sufficient understanding to give informed consent; and the level of decisionmaking ability required should be less for research without significant risks but with
potential benefits. Research involving people with learning difficulties is ethically
acceptable provided that: the research is of major relevance to learning difficulties; the
risks are minimal; the individual does not object or appear to object either verbally or nonverbally; the research does not intrude unreasonably on the person’s privacy or freedom
of action; appropriate assent is obtained (see below); and the study has been considered
and given approval by the relevant ethics committee.
Two specific issues need consideration. First, the minimal risk criterion, if regarded as
an absolute, would rule out certain types of research that might be crucial for an
understanding of the causes of severe learning difficulties and that could not be
undertaken validly in less vulnerable groups. Thus, for example, some investigations
might require the person to be heavily sedated (which might also be viewed as above a
minimal risk). We advise that such research would be ethical provided that its quality
was high, its potential value in relation to learning difficulties was great, the absolute level
of risk for the individual was low and that the assent of a close relative and approval of an
independent professional was obtained (as in paragraphs 5.6(a) and 5.6(b)). In the case
of an individual without close relatives, it would be appropriate to obtain assent from a
primary non-professional carer who is well known to the individual and apparently trusted
by them. It would also be important to obtain the views of relevant user group
representatives. However, the risk–benefit ratio of such research would have to be clearly
favourable for ethics approval to be justified.
The second point derives from the first. UK law at present does not recognise the
parents of adults with learning difficulties as having any rights in relation to their care.
This is widely resented by the families and is seen as inappropriate, and very insensitive,
by most professionals. We recommend that consideration needs to be given to a change in
the law on this issue. In most ordinary circumstances parents are the people most likely to
look after the benefits of the individual in relation to treatment or research. We recognise
that there are circumstances in which this might not be the case (see paragraph 5.6(a)
above) and, hence, the procedures we propose require the support of an independent
professional (approved by the ethics committee), as well as the assent of the closest
relative.
INFORMED
CONSENT,
CAPACITY AND
COMPETENCE
Active steps should be taken to help individuals with learning difficulties to achieve sufficient understanding for them to give informed
consent. Even when this is not possible, the individuals should be
helped to be involved in the decision-making process. When the competence to give consent is lacking, research may be ethically acceptable if it is relevant to learning difficulties, it is not against the individual’s best interests, it does not intrude unreasonably on the person’s privacy or freedom of action, appropriate assent has been obtained from a close relative, the procedures have been approved by
an independent professional, the study has the support of relevant
user group representatives and the study as a whole has been approved by the relevant ethics committee.
5.6(E) STAFF AND STUDENTS
There is a good tradition of researchers trying out procedures on themselves before
asking participants in studies to undergo them. Also, there are situations in which control
data may be obtained most appropriately from the researchers themselves (e.g. in scanning
studies), although attention will always need to be paid to comparability in features that
could affect findings. Researchers, however, need to be very sensitive to the problems
inherent in asking others (such as staff and students) to volunteer for research. When
there is an implicit or explicit hierarchical relationship with those individuals who are
being recruited as participants, there is a substantial risk of a perception that covert
coercion is present (because there may be a concern that agreement to take part in a
study will be favorable to career progression and a refusal will put progression at risk). We
advise that these risks must always be carefully considered whenever staff or students are
to be used as research participants. It is usually undesirable for researchers to use staff or
students with whom they have a hierarchical relationship, and if this is intended, the
application to the ethics committee must provide a reasoned explanation of why it is
necessary and how the coercion risk will be dealt with. The risks are somewhat less, but
still present, for staff and students in the same institution but outside the responsibility of
the research group. The risks are least with staff and students at other institutions. In all
cases, however, scrupulous attention must be paid to the risks of perceived covert coercion,
and thought needs to be given to the possibility of using other sources of volunteers. In
addition to the issue of possibly perceived coercion in a hierarchical situation, there are
ethical issues involved in asking colleagues (i.e. staff or students) highly personal questions.
In research where this is involved, we recommend that staff or students should not be
asked to participate.
There are risks of perceived covert coercion if staff and students
who are in a hierarchical relationship with the researcher are approached to volunteer to participate in research. It is usually desirable to avoid the use of such groups in research and, if their use is
needed, there must be a reasoned case included in the ethics committee application; this must specify how the coercion concern will
be dealt with.
5.6(F) DETAINED PATIENTS
The obtaining of fully informed consent is an especially sensitive issue in relation to
patients detained under the Mental Health Act, mainly because of the actual or perceived
problem of coercion, but also because they are likely to be seriously ill (with the consequent
29
ROYAL COLLEGE
OF PSYCHIATRISTS
RESEARCH
GUIDELINES:
ETHICS
uncertainties over their ability to give informed consent). The general principles already
outlined in earlier paragraphs of section 5.6 apply, but particular care should be taken to
ensure that the process of gaining informed consent should be as transparent as possible
to the patient and that it should be obtained by someone other than the medical officer
who is seen by the patient as having a decision-making role in relation to the detention.
As advised in earlier paragraphs, it should be essential to obtain the opinion of an
independent professional; an approved social worker or legal representative might well
be the most appropriate person.
When the patient is incompetent to give consent, the same procedures as outlined in
paragraph 5.6(b) apply. When the patient is competent, the procedures should be those
that apply to any patient except that especial care is needed to ensure that there is no
possibility of perceived coercion and, because of that, an independent professional opinion
should always be obtained. The minimal risk criterion applies in the case of incompetent
patients but not detained patients capable of giving informed consent.
For detained patients the general principles of paragraph 5.6 apply,
but especial care is needed to ensure that there is no possibility of
perceived coercion and that an independent professional opinion
has been obtained. For research on detained patients to be ethically
acceptable, the focus of the research must be relevant for their disorders, it must not be possible for the research to be undertaken in
a valid and generalisable fashion with less vulnerable groups, it must
not be against their individual best interests and particular care
must be taken to ensure that the informed consent is real. Ordinarily,
too, it should be expected that the research procedures should be
non-invasive. Invasive research may be ethically acceptable but particularly strong justification with respect to risks and benefits would
be essential (see paragraph 5.6(a)).
5.6(G)
PRISONERS
Prisoners, or offenders under community detention orders, may be approached to take
part in two main sorts of research:
(a) that which pertains to their status as offenders, in order to further
an understanding of the nature of antisocial behaviour or its effective
prevention or management;
(b) that designed to further the understanding or treatment of some
disorder to which they may be particularly vulnerable.
In the latter case they should not be approached to take part in any research that
would not be appropriate for their peers in the wider community. As with other groups, it
is important that they not be excluded from research that is relevant in relation to the
problems they experience or the disorders from which they suffer. Equally, it is crucial
that researchers are aware of the many opportunities for perception that there is covert
coercion, or the opportunity for gain, both of which could invalidate consent procedures.
Particular care should be taken to ensure that prisoners and supervised offenders
understand that their decision whether or not to participate in research will not affect
their sentence or conditions. This should be made explicit in the information sheet.
For prisoners the same general principles as those outlined for detained patients apply with respect to possibly perceived coercion,
but in most cases there is not the same issue with respect to competence. In the few instances where that is relevant, a comparable set
of procedures should be followed at the individual level. When competence is not in doubt, the main concern is to ensure that the
research, and the procedures of obtaining consent, provide adequate
protection to the individual. Ethics committees should consider
whether, given the nature of a particular study, independent advice
should be sought regarding the ethical acceptability of the study.
30
6. OPERATION OF ETHICS COMMITTEES
INFORMED
CONSENT,
CAPACITY AND
COMPETENCE
Ethics committees vary greatly in the types of research that they cover. Some serve wide
areas of NHS research, with very few projects concerned with mental health issues.
Some serve teaching hospitals, medical schools or universities, with many proposals dealing
with pilot projects by students, but also many that deal with complex high technology
research, much of which is inter-disciplinary. A few are concerned with postgraduate
psychiatric research and teaching centres with a very high throughput of psychiatric,
psychological, nursing and social work research concerned with psychopathological
questions, dealing with both patient and general population samples. It is inevitable,
therefore, that such committees will vary in their mix of expertise and experience, as well
as in the applications that they have to assess. This has unavoidable implications for the
details of the ways in which they operate. In this report, we concentrate, therefore, on the
basic principles that we wish to see applied.
6.1 Legal framework
Ethics committees review protocols, make recommendations on the ethical
acceptability of specific studies and advise on broader issues. Currently, however, they do
not, even vicariously, have responsibility for the implementation and conduct of any
study, or for the behaviour of any individual researchers. Those responsibilities remain
firmly with the principal investigators and their employers.
We recommend that clearer lines of responsibility need to be established. At present,
institutions vary greatly in their rules regarding ethical review of studies carried out by
their employees and/or carried out on their patients or staff. We advise that it is desirable
that there be uniformity in expectations and that these follow the lines of paragraph 3.1.
It is also not always clear whether individual researchers have the right to disregard the
recommendations of ethics committees. We suggest that employing authorities make it
explicit that research undertaken without ethical approval will not be supported by them
and that undertaking unethical research is automatically a justification for disciplinary
action. If an individual believes that an ethics committee decision is mistaken, there
should be a right to appeal against the decision, and ethics committees need to decide
how they operate such an appeals system. We recognise the unacceptability of an
arrangement whereby the body that came to the initial decision also assesses the appeal.
We advise that in the event of such an appeal, the ethics committee should as a matter of
course have the appeal assessed by a different ethics committee, the choice of committee
being agreed with the person making the appeal. The decision should lie with that
second committee and not with any individual (such as a dean or clinical director) who
has been outside the group with experience in considering ethical issues.
Uniformity should be sought on expectations regarding which studies need to be submitted to ethics committees. The legal implications
of disregarding ethics committee decisions should also be clarified.
6.2 Composition of ethics committees
Ethics committees face complex issues, and it is essential that their membership provides
the expertise and experience required for the task (see National Bioethics Advisory
Commission, 1998). This has to include knowledge on the relevant scientific issues, and
it is essential that membership includes mental health specialists if research in this field is
to be considered. At present, that is sometimes not the case, and it needs to be remedied.
Expertise should also be available on those aspects of research that are relevant to many
studies in the psychiatric field such as: epidemiology; statistics; experimental methods;
randomised control trials; pharmacology; imaging methods; genetics; the design and use
of research interviews and questionnaires; and the clinical issues involved in caring for
31
ROYAL COLLEGE
OF PSYCHIATRISTS
RESEARCH
GUIDELINES:
ETHICS
32
vulnerable groups whose ability to give informed consent is an issue to be considered.
Experience in these areas may well be provided by researchers who do not themselves
work in the mental health field. We recognise that often it may not be practicable to have
a membership with all the relevant research skills; when this is the case it should be a
matter of routine to seek expert advice on the relevant clinical or research issues (such as
drug effects or radiation hazards). Much research into mental disorders is inter-disciplinary,
and this should be reflected in the ethics committee membership; clinical psychology
expertise is particularly relevant.
There is a similar need for membership to cover the relevant areas of clinical
experience. Clearly, this has to include psychiatry if applications are likely to deal with
research into mental disorders. In addition, however, the coverage should provide for
experience with age groups presenting special issues that need consideration; childhood
and old age are obvious examples. If applications are likely to include these groups, the
committee should have clinicians whose practice is relevant to them. The same applies to
learning disabilities if they are included in the applications covered by the committee.
Clinical research will often impinge on care outside hospital, and it should be routine to
have a general practitioner on each committee. With psychiatry, care frequently involves
other disciplines, and there is a comparable need to have a nurse and/or a social worker
included in the membership.
Ethical issues involve far more than scientific and clinical considerations, and it is
essential that the membership includes several lay (meaning non-clinical) people who
are not employees of the trust or university served by the committee. However, if lay
people are to have an effective role, it is important that they have some relevant experience
such as through holding (or having recently held) a non-clinical role in hospital or
community services or in a medical charity. It is also important that patients, their families
and non-statutory bodies that represent them should also be able to provide input to
ethics committees. Ordinarily, however, it is preferable for such members not to be patients
of the trust served by the committee (to avoid their having too personal a view – either
positive or negative). A further consideration is that the committee should be sensitive to
possible special considerations arising from differences in gender or variations in ethnicity,
and this should be reflected in its membership.
Although it is necessary for the committee membership to cover a wide range of
interests and experience, it is also essential that it works as a cohesive group and that
members respect the complementary expertise of others. Accordingly, we advise that
members should be chosen for their personal qualities as effective committee members
with an interest in ethical issues, and not as advocates or representatives of particular
interests. It should also be explicit that they accept that all discussions in the committee
are confidential, although the decisions of the committee should be made available in
appropriate reports. In order to ensure that the relevant expertise is available at all committee
meetings, it should be acceptable for members to have previously agreed alternates.
It is necessary that ethics committees are trusted for their dispassionate consideration
of ethical issues and not be open to the perception that they are there to defend research
interests. For this reason, we recommend (in line with Department of Health advice, see
NHS Management Executive, 1991) that it is highly desirable for either the chair or vicechair of ethics committees to be a lay person and we advise that, in time, it might be
desirable for this to become a mandatory requirement. We do not, however, recommend
that now because of the practical difficulties some ethics committees have experienced
in getting lay people with both sufficient knowledge of the clinical and research issues
and the time and commitment to consider many varied applications.
As evidenced by the need for the present working party, ethical issues in the mental
health field are becoming increasingly complex. Training in research ethics is becomingly
increasingly widely available and committee members should be strongly encouraged
(and financially supported) to obtain such training. Ethics committees need to consider
whether they should specify a minimum level of training that is expected of each
committee member.
When ethics committees only rarely receive applications dealing with mental health
issues, it may be reasonable not to include in their membership either professionals or user
group representatives with experience in this area of work. However, when they do
receive an application on mental health issues it should be a standard expectation that
either they obtain expert advice through coopted membership or they should pass the
application on to be dealt with by another ethics committee that does have the necessary
expertise.
OPERATION OF
ETHICS
COMMITTEES
Committee membership should reflect the range of research and
clinical skills relevant to the applications that they consider and
the range of disciplines involved in the case of patients with disorders in the field covered, and should include both lay members and
people who reflect the interests of patients and their families. It is
highly desirable that the chair or vice-chair be a lay person. Sufficient experts in clinical and research aspects of mental health
should be included in the membership of committees to cover the
range of research topics they are likely to have to deal with.
6.3 Resourcing of ethics committees
Experience has shown that the time needed by committees to deal with applications can
be very greatly reduced if applications have been processed efficiently first by an
experienced and skilled administrator with the necessary training in ethical issues and
the committed time to do what is required. The administrator should be able to vet all
applications and seek clarification or further details from applicants whenever this seems
to be needed.
Ethics committees should have a properly resourced professional
administration.
6.4 Application forms
The ease and thoroughness of ethical scrutiny of studies is greatly influenced by the
quality of the design of application forms. It should be mandatory for the purpose and
procedures of the study to be described succinctly and accurately, using language that
can be easily understood by committee members without either scientific or clinical
training. The details should always include the nature, size and method of recruitment
of the sample, funding source, payments to be made, research design (how the study will
achieve its objectives), procedures or measures to be used (if non-standard they should
be included with or described in the application, as appropriate) and an explicit statement
of the risks and the ethical considerations.
The information sheet and consent form should always be provided. The name,
professional position and employing authority of the senior researcher (who must be of
consultant/senior lecturer level or their equivalent) must be specified, as well as the same
details for the researcher concerned with the day to day running of the study. The
source of funding must be specified and details of payments to researchers and participants
provided. The application must indicate how informed consent and data protection
issues will be dealt with, and whether (and in which form) research findings will be fed
back to participants.
Attention should be paid to the design of application forms to
ensure that all necessary details of the study and ethical issues are
provided in a standard way.
6.5 Scientific review
As noted above, ethical issues require consideration of the scientific quality of the proposed
research. That need is reflected in our recommendations on committee membership
(paragraph 6.2). The degree of scientific scrutiny should always be proportional to the
potential risks to participants or researchers. When the scientific merits of studies have
already been considered by peer review prior to seeking ethical approval, ethics committees
should not normally make further suggestions on the science unless it is clear that a major
issue has not been addressed. Many funding bodies, however, do not undertake peer
review, so scientific quality cannot be assumed on the basis that funding is provided.
33
ROYAL COLLEGE
OF PSYCHIATRISTS
RESEARCH
GUIDELINES:
ETHICS
In addition, many of the major funding bodies (such as the MRC and the Wellcome
Trust) require ethical review before they consider funding. As, currently, such a small
proportion of applications get funded, clearly this means that ethics committees have to
review many studies that never get undertaken. The rationale of funding bodies is that
they would be unwilling to fund unethical research, so this has to be a prior consideration.
Also, if the ethical review leads to a change in design, the scientific appraisal will be based
on the wrong design. Equally, however, if peer review leads to a change in research
strategy or tactics, the ethical review will have been based on an inappropriate design.
There is no entirely satisfactory way out of this impasse but we accept that, unless the
funding system changes, ethics committees will need to undertake some form of scientific
scrutiny prior to peer review. We recommend that ethical approval should always be
contingent on the science receiving peer review approval, if that is being obtained, and
that it should be mandatory for applicants to inform the ethics committee of any changes
in research design or procedures (including sampling), as well as any changes in the
source of funding or payments to be made to researchers or participants.
It is usual for studies to undergo some changes during their course, as a result of
practical or scientific difficulties encountered. We recommend that it should be a routine
for researchers to be asked once per year to provide details of any changes in the research,
and in the research personnel, as well as of any ethical difficulties experienced.
Continuation of ethical approval should be contingent on this information being provided
and the changes in protocol agreed by the ethics committee.
Ethical review of studies must include an appropriate degree of
scientific scrutiny and continuation of ethical approval should be
made contingent on the researchers providing information on any
changes in design, funding or personnel involved.
6.6 Multi-centre studies
Increasingly, research is based on samples from more than one hospital or clinic and on
large epidemiological samples that are widely spread geographically. In the past this has
posed a major problem to researchers who have had to seek separate ethical approval
from multiple ethics committees, each of which uses a different style of application form,
and may raise contradictory scientific or ethical issues. Multi-centre research ethics
committees have been set up in the UK to obviate these difficulties (NHS Executive,
1997). There are no issues specific to research in the mental health field that apply to
their operation, and we make no recommendations about their operation. It should be
noted that studies involving five or more recruitment centres must go to the multicentred research ethics committees before going to the local committee. However, these
were primarily set up to deal with large multi-centre randomised control trials and other
clinical research presenting similar coordination issues. They appear well designed for
that purpose but their operation does not map well onto large scale epidemiological
studies based on non-patient samples or on samples of patients obtained through voluntary
registers or membership of support groups. In these circumstances it is appropriate for the
study to be considered by the local ethics committee serving the employing institution of
the lead researcher. It is imperative, however, that the committee has the necessary
expertise to review the study. If it does not, they should seek advice from one that does.
In addition to large scale multi-centre studies, there are many more studies based on just
a few collaborating clinical facilities. According to present arrangements, these still need to be
considered by several separate local ethics committees. In these circumstances, we recommend
that the lead review should be undertaken by the committee with the most appropriate
expertise to undertake the assessment of that particular type of study. The other committees
must consider any ethical issues that are specific to their clinical facility, and must have the
right to raise serious ethical concerns if they consider that they were overlooked by the lead
committee. Nevertheless, ordinarily, it should be expected that the same application form
should be accepted by all committees, and that the recommendation of the local committee
should be followed. It should also be a requirement, however, that the separate committees
share and consider any expert external advice that any of the committees have obtained
regarding the study or its procedures.
34
Studies of non-patient samples or samples of patients obtained from
registers separate from local centres should ordinarily be considered by the appropriate local ethics committee, rather than being
passed on to a multi-centre ethics committee. When research involves several centres (but below the number requiring referral to a
multi-centre committee), the relevant local ethics committees should
decide on which one should take the lead, with the expectation that
the other committees would accept their decision unless there were
some special additional considerations to be taken into account.
OPERATION OF
ETHICS
COMMITTEES
6.7 Conflicts of interest
Increasingly it is becoming standard practice to require members of all committees to
make a formal declaration of potential sources of conflict at the time of joining the
committee and to update these once a year. We recommend that this should be expected
of all members of ethics committees. It is accepted that it is entirely proper for members to
have affiliations with organisations that have interests relevant to some applications coming
to the committee; what is improper is for this not to be known to other members of the
committee. It should also be obligatory for members to make overt any such possible
conflicts of interests in relation to individual applications, and to withdraw from the room
if the conflict is felt by the member or the rest of the committee to make it inappropriate
to remain present during discussions. The chairman’s ruling on this matter should be
absolute. Potential conflicts concern any links (including paid or unpaid consultancies,
membership, office holding and other connections) that have possible financial, lobbying
or hierarchical implications.
Formal declaration of potential conflicts of interest should be expected of all members of ethics committees, and if these create a
possible problem in relation to individual applications, affected members should withdraw during their discussion.
35
ROYAL COLLEGE
REFERENCES
OF PSYCHIATRISTS
RESEARCH
GUIDELINES:
ETHICS
Appelbaum, P. S. (1998) Missing the boat: competence and consent in psychiatric research.
American Journal of Psychiatry, 155, 1486–1488.
––– & Grisso, T. (1995) The MacArthur Treatment Competence Study: 1. Mental-illness and
competence to consent to treatment. Law & Human Behavior, 19, 105–126.
Barnes, D. E., Hanauer, P., Slade, J., et al (1995) Environmental tobacco smoke. The Brown and
Williamson documents. Journal of the American Medical Association, 274, 248–253.
Beecher, J. K. (1966) Ethics and clinical research. New England Journal of Medicine, 274, 1354–
1360.
Bero, L., Barnes, D. E., Hanauer, P., et al (1995) Lawyer control of the tobacco industry’s external
research program. The Brown and Williamson documents. Journal of the American Medical
Association, 274, 241–247.
British Medical Association & The Law Society (1995) Assessment of Mental Capacity: Guidance
for Doctors and Lawyers. London: BMA.
British Paediatric Association (1992) Guidelines for the Ethical Conduct of Medical Research
Involving Children. London: British Paediatric Association.
Caldicott Committee (1998) Report on the Review of Patient-Identifiable Information. London:
Department of Health.
Choo, V. (1998) The line between research and audit. Lancet, 352, 337–338.
DeMets, D. L., Pocock, S. J. & Julian, D. G. (1999) The agonising negative trend in monitoring
of clinical trails. The Lancet, 354, 1983–1988.
Dickert, N. & Grady, C. (1999) What’s the price of a research subject? Approaches to payment
for research participation. New England Journal of Medicine, 341, 198–202.
Food and Drug Administration (1999) Informed Consent: Exception for Emergency Research.
http://www.fda.gov/cder/present/dia698/diafda4
Federal Policy for Protection of Human Subjects (1991) Federal Register, June 18; 56: 28012-8.
Fienberg, S. E. & Resnick, D. P. (1997) Reexaming the bell curve. In Intelligence, Genes, and
Success: Scientists Respond to The Bell Curve (eds B. Devlin, S. E. Fienberg, D. P. Resnick et
al), pp. 3–18. New York: Copernicus.
Freeman, T. B., Vawter, D. E., Leaverton, P. E., et al (1999) Use of placebo surgery in controlled
trials of a cellular-based therapy for Parkinson’s disease. New England Journal of Medicine,
341, 988–992.
General Medical Council (1998) Seeking Patients’ Consent: The Ethical Considerations. London:
GMC.
Glantz, S. A., Barnes, D. E., Bero, L., et al (1995) Looking through a keyhole at the tobacco
industry: the Brown and Williamson documents. Journal of the American Medical Association,
274, 219–224.
–––, Slade, J., Bero, L. A., et al (1996) The Cigarette Papers. Berkeley, California: University of
California Press.
Grisso, T. & Appelbaum, P. S. (1996) Values and limits of the MacArthur Treatment Competence
Study. Psychology, Public Policy and Law, 2, 167–181.
Hanauer, P., Slade, J., Barnes, D. E., et al (1995) Lawyer control of internal scientific research to protect
against products liability lawsuits. The Brown and Williamson documents. Journal of the American
Medical Association, 274, 234–240.
Hilts, P. J. (1996) Smokescreen: The Truth Behind the Tobacco Industry Cover-up. Reading,
Massachusetts: Addison-Wesley.
Kennedy, I. (1997) Consent: Adult, Refusal of Consent, Capacity. Commentary on ReMB[1997]2
F.L.R. 426. Medical Law Review, 5, 317–325.
Law Commission (1995) Mental Incapacity (Law Commission Report No. 231). London: HMSO.
Macklin, R. (1999) The ethical problems with sham surgery in clinical research. New England
Journal of Medicine, 341, 992–996.
Medical Defence Union (1997) Consent to Treatment. London: Medical Defence Union.
Medical Research Council (1992) Responsibility in Investigations of Human Participants and Material
and on Personal Information. MRC Ethics Series. London: MRC.
––– (2000) Personal Information in Medical Research. MRC Ethics Series. London: MRC.
––– (2001) Human Tissue and Biological Samples for Use in Research. Operational and Ethical
Guidelines. MRC Ethics Series. London: MRC.
Melton, I. J. (1997) The threat to medical-records based research. New England Journal of
Medicine, 337, 1466–1470.
36
National Bioethics Advisory Commission (1998) Research Involving Persons with Mental Disorders
that may Affect Decisionmaking Capacity: Vol. I. Report and Recommendations of the National
Bioethics Advisory Commission. Rockville, Maryland: National Bioethics Advisory
Commission.
NHS Executive (1997) Ethics Committee Review of Multi-centre Research: Establishment of
Multi-centre Research Ethics Comittees, HSG(97)23.
NHS Management Executive (1991) Local Research Ethics Committees. Health Service Guidelines.
London: Department of Health.
Nuffield Council on Bioethics (1993) Genetic Screening: Ethical Issues. London: Nuffield Council
on Bioethics.
––– (1999) The Ethics of Clinical Research in Developing Countries. London: Nuffield Council
on Bioethics.
Ong, E. K. & Glantz, S. A. (2000) Tobacco industry efforts subverting International Agency for
Research on Cancer’s second-hand smoke study. Lancet, 355, 1253–1259.
Pappworth, M. (1967) Human Guinea Pigs: Experimentation on Man. London: Routledge.
Patel, N. (1999) Getting the Evidence: Guidelines for Ethical Mental Health Research Involving
Issues of ‘Race’, Ethnicity and Culture. London: Mind.
Royal College of Physicians (1994) Independent ethical review of studies involving personal
medical records. Report of a working group. Journal of the Royal College of Physicians of
London, 28, 439–443.
––– (1996) Guidelines on the Practice of Ethics Committees in Medical Research Involving Human
Subjects (3rd edn). London: The Royal College of Physicians.
––– (1999) Research based on archived information and samples. Journal of the Royal College of
Physicians of London, 33, 264–266.
Royal College of Psychiatrists (1990) Guidelines for Research Ethics Committees on psychiatric
research involving human subjects. Psychiatric Bulletin, 14, 48–61.
Slade, J., Bero, L. A., Hanauer, P., et al (1995) Nicotine and addiction. The Brown and Williamson
documents. Journal of the American Medical Association, 274, 225–233.
Smyth, R. L. & Weindling, A. M. (1999) Research in children: ethical and scientific aspects.
Lancet, 354 (suppl. 2), 21–24.
Wald, N. J., Law, M., Meade, T. M., et al (1994) Use of personal medical records for confidential
research purposes. British Medical Journal, 309, 1422–1424.
Walton of Detchant, Lord, Doll, R., Asscher, W., et al (1999) Consequences for research if use
of anonymised patient data breaches confidentiality. British Medical Journal, 319, 1366.
Wong, J. G., Clare, I. C. H., Gunn, M., et al (1999) Capacity to make health care decisions. Its
importance in clinical procedures. Psychological Medicine, 29, 437–446.
–––, –––, Holland, A. J., et al (2000) The capacity of people with a ‘mental disability’ to make a
health care decision. Psychological Medicine, 30, 295–306.
World Medical Association (1989) The Declaration of Helsinki (1964, revised 1989) The World
Medical Association Handbook of Declarations. Ferney-Voltaire, France: World Medical Association.
REFERENCES
37
ROYAL COLLEGE
APPENDIX 1 – EXAMPLES OF POSSIBLE STANDARD
OF PSYCHIATRISTS
RESEARCH
DESCRIPTIONS
GUIDELINES:
ETHICS
Example of medical term
Suggested alternative
Analgesia
Animal studies
Assigned
Auditory hallucinations
(Brain) scan
Claustrophobia
Combined treatment arm
Pain killer
In the laboratory
Put/allocated/placed
Hearing voices
Picture of your (brain)
Fear of enclosed spaces
Part of the study in which patients will
receive both/all treatments
Genetic material
Difficulty in breathing
Involve/mean
Every 3 months
The first treatment you had/would have
Studying how common
Affecting the way
Treatment/care
Psychiatric problems/mental ill health
Feeling sick
Tablets
The way your body handles the drug
An inactive substance
Before you come into hospital
Allocated by chance
Treatment plan/programme
Make you
Stop taking part in the study
DNA
Dyspnoea
Entail
Every 3/12
First line treatment
Investigating the incidence
Jeopardising
Management
Mental illness
Nausea
Oral medication
Pharmacokinetics
Placebo
Prior to your hospitalisation
Randomisation
Regime/Regimen
Render you
Terminate your participation
38
APPENDIX 2 – CHILDREN’S LEVEL OF UNDERSTANDING OF
MEDICAL DECISIONS
By M. Rutter
The issue of children’s capacity to understand what is involved in research closely parallels
the queries regarding their understanding of decisions about their own treatment.
Accordingly, in this appendix, the topic is considered first in the latter context because
decisions on how research consent should be dealt with ought to follow good clinical
practice. Discussions about children’s consent to medical treatment tend to be dominated
by the twin issues of whether children are competent to make informed decisions and
whether, even if they are competent, it is right to allow them to make serious and difficult
choices (Alderson & Montgomery, 1996). The questions are normally posed in this way
because the really tricky legal, medical and ethical problems arise when there is a clash
between the expressed views of parents and of children, and where one or other of those
also clashes with professional opinion. The evidence on these points is considered further
below but they do not constitute the most appropriate initial basis for discussion. Very few
decisions about treatment are of the life and death variety, very few involve such stark
clashes and very few have to be decided in a hurry. Accordingly, the starting point needs
to be the more everyday question of whether or not, in the course of ordinary medical
practice, young children should be involved in discussions about their medical problems
and about how such problems should be dealt with.
INDIVIDUALS’ RIGHTS TO PARTICIPATE IN HEALTH CARE DECISIONS
The basic principle is an assumption that all people, of whatever age and whatever
degree of handicap, should be able to be properly informed about medical matters that
pertain to them and be involved in decision-making about their care and treatment. This
means that the onus needs to be placed on the justification for not doing this in individual
cases. That could arise either because the individual’s level of understanding was so
extremely limited that it would be obviously pointless to do so, or because of evidence
that it would be damaging. So far as the first possibility is concerned, everyone would
accept that there is a point when it would be absurd to try to have a discussion. Thus, it
would be ridiculous to try to engage a neonate in discussions about treatment, to take an
extreme example. Equally, however, the same would apply to a profoundly retarded
individual whose level of understanding of language was, say, below the age of 6 months.
The question then becomes one of asking at what chronological or mental age children
begin to have sufficient understanding to engage in a meaningful conversation about
medical matters. We consider the evidence on the growth of understanding further
below but, at this point, it is sufficient to note that detailed naturalistic studies of young
children’s conversations in the home make it quite clear that even 3- and 4-year-old
children can and do discuss matters of moment and engage in decision-making (Dunn,
1988). Obviously, the ways in which they do so, and their capacity to deal with complex
matters, is very much less than that of adults, but it is far from zero. The implication is that
health care professionals should be expected to talk with children, even during the preschool period, about medical matters that involve them. This needs to be done in ways
that the children can understand but that are no different in principle from the exactly
39
ROYAL COLLEGE
OF PSYCHIATRISTS
RESEARCH
GUIDELINES:
ETHICS
comparable need to do so with respect to adults. Thus, most adults can scarcely be
expected to understand all the complex considerations that go into decisions about
different forms of medical treatment, or different surgical procedures, for specific
conditions. It would make no sense to present the patient with the sort of technical
evidence that would be taken into account by an expert in the field. On the other hand,
it would be generally accepted that it is the duty of health professionals to provide the
patient with as much understanding of these issues as possible and to engage them in the
decision-making. Exactly the same applies in childhood.
The other reason for withholding information or leaving people out of decisionmaking is that it would be harmful to them to do otherwise. Once more, it seems appropriate
to start with the adult case. There was a time when some doctors would ‘protect’ their
patients from the worry of knowing that they have cancer or some other life-threatening
disease by concealing this information from them or giving them information that was
misleadingly optimistic. The assumption seems to have been that if you did not tell people
about terrible things, they would not think about them themselves. It is now obvious that
is not the case. Indeed, often, people’s fantasies about what might be wrong with them
and about the chances of successful treatment are worse than the reality. Rather, it has
become clear that the appropriate approach is to discuss these matters honestly with
patients but with a sensitivity to individual wishes on the extent to which they want to
discuss matters in detail. That is, the opportunity for discussion should always be made
available, and deliberate misinformation is not acceptable, but there needs to be an
appreciation that there are valid individual differences between people in the extent to
which they wish to be involved in detailed decision-making and in learning about all the
specifics of what is entailed. The maintenance of hope and of a positive attitude is
important, particularly because there is good evidence that attitudes of mind may influence
the course of somatic disease. There is a parallel with parents’ desire to protect children
in the occasional requests from families that their loved one (an adult) should not be told
that they have cancer. It would now be accepted that although health professionals need
to pay attention to what families say about patients’ sensitivity, feelings and wishes,
nevertheless, the starting point has to be the rights of individuals to be properly informed
and to be able to participate in decision-making. It follows that there should be a similar
starting point with respect to children. The question is whether there are circumstances
in which the needs of children, and the potential damage brought about by engaging
them in decision-making, is of a degree that an exception should be made to this general
principle that people should be informed and engaged in decision-making. Such exceptions
should have a clear, well justified, basis.
GROWTH IN CHILDREN’S LEVEL OF UNDERSTANDING
In considering children’s level of understanding, we need to be guided by well established
empirical research findings rather than by theories, either outmoded or contemporary
(Rushforth, 1999). The evidence is clear-cut that people’s mental abilities, social
understanding and emotional appreciation do indeed increase greatly over the course of
development (Keating, 1990; Rutter & Rutter, 1993; Ceci et al, 1994a,b; Walden &
Garber, 1994). Thus, at about the age of 2 years, children first develop a capacity to make
inferences about the causes of events and to appreciate the psychological state of other
people (Kagan, 1981). Over the next couple of years or so their ability to understand what
other people are thinking increases markedly (Baron-Cohen et al, 1993). However, strategic
thinking of a planful goal-oriented nature does not become well established until later
childhood or early adolescence (Kail, 1990). At all ages there are marked individual
differences in the extent to which children are suggestible but, on the whole, younger
children are more prone to be misled by leading questions and by a wish to please adults
(Ceci & Bruck, 1993; Bruck & Ceci, 1999).
The growth in mental abilities continues through the teenage years (Justice, 1996).
Thus, in early adolescence, young people’s thinking tends to become more abstract,
multi-dimensional, self-reflective and self-aware, with a better understanding of relative
concepts. They are better able to hold in mind several different dimensions of a topic at
the same time, and so generate more alternatives in their decision-making. In addition,
40
they become better able to monitor their own thinking for inconsistency, for its gaps in
information and for the accuracy of its logic. This greater mental sophistication that
comes during the teenage years is accompanied by related developments in the ways in
which young people think about themselves. During adolescence, there is a marked
increase in emotional introspection, together with a growing tendency to look back with
regret and to look ahead with apprehension. That is, not only do young people become
increasingly able to consider the long-term consequences of their own actions, and of
what happens to them, but also they tend to think about such consequences more in
terms of their own sense of responsibility and better awareness of the effects of what they
do on other people.
These developments in intellectual capacities and emotions derive, in part, from
continuing brain development (which extends into the teenage years and beyond), and,
in part, from life experiences. However, it would be a mistake to view the changes as just
a growth in the power of information-processing. The increase in mental abilities reflects
several different features including the knowledge base, an ability to use more of the
information available, more sophisticated mental planning capacities in order to decide
how best to tackle decision-making and an increased flexibility in the use of mental
strategies (Sternberg & Powell, 1983).
The evidence that children’s understanding, and their ability to make difficult
decisions, increases greatly over the age period between infancy and adulthood is
overwhelming. At first sight, it would seem a straightforward matter to determine when
young people’s cognitive capacity has reached a point when they are competent to take
decisions on health care and medical treatment themselves without the need for parents
to decide things on their behalf. Unfortunately, it is evident that several crucial features
much complicate that assessment.
To begin with, decision-making is influenced by both cognitive capacity and
emotional understanding and these two aspects of functioning do not necessarily
develop with the same timing. Also, however, there is huge individual variation both
in the time taken to reach particular mental levels and in the ultimate levels attained.
This is, as it were, the mental equivalent of the enormous individual variation in the
age at which children acquire their permanent teeth or reach puberty. It would be
misleading to make a general claim that girls reach sexual maturity at, say, 12 years,
when some have their menarche at 9 years but others not until 15. The same applies
to cognitive maturity. Equally, not all adults have the same level of understanding.
This is apparent, for example, in the spread of IQ scores, some 2.5% have an IQ below
70 but an equal proportion have one above 130.
That analogy makes it seem tempting to search for a good test for the skills involved in
decision-making. Of course, a test of general intelligence does provide a rough and ready
guide to the likely level of relevant mental skills, but it is highly fallible for three different
reasons. First, standardised tests of children’s understanding may underestimate what children
can and do achieve in more ordinary day to day situations, as well illustrated by the findings
on children’s ability to ‘mentalise’ or appreciate what other people are likely to be thinking
(Dunn, 1988). Second, this variation in mental performance is even more marked when
people’s everyday experiences are unusual. Thus, there is much evidence of the contextspecificity of knowledge (Ceci et al, 1994a). People tend to be very much better at solving
problems when they concern issues well familiar to them than when exactly the same problem
is presented in an unfamiliar fashion. Third, failure on a mental task does not necessarily
mean that the relevant cognitive skill is missing. If children can be helped with remembering
different facets of a decision-making task, or if they can be shown how to adopt a different
mental strategy, they may solve problems that are not ordinarily possible for children until a
much older age (Bryant, 1974; Goswami, 1991).
Much the same applies in the field of emotional understanding. Thus, quite young
children express mixed feelings, but it is not until middle childhood that they gain much
insight into their own emotional lives and realise when there is emotional ambivalence.
Also, over the same age period, children learn a range of thinking techniques to control
their emotions. But this capacity is influenced by circumstances. Thus, Harris (1989)
found that children experiencing distress in hospital were less able than other children of
the same age to appreciate mixed feelings.
APPENDIX 2
41
ROYAL COLLEGE
OF PSYCHIATRISTS
RESEARCH
GUIDELINES:
ETHICS
ASSESSMENT OF COMPETENCE TO TAKE DECISIONS
It follows from these considerations that any assessment of a particular child’s competence
to take a specific decision regarding health care or treatment needs to be based on how
they deal with that specific decision rather than on any standardised tests (although
these may be useful in alerting professionals to the possibility that the child is unusually
mentally advanced or delayed in comparison with others of the same age). Furthermore,
the assessment needs to be based on how the child deals with the matter after discussion
and help with understanding and with ways of thinking about the issues. First impressions
can be misleading.
But, even with the most detailed and skilled of assessments, it will still be the case that
capacity needs to be considered as a graded dimension and not as something that is either
present or absent. Even pre-schoolers are likely to have some appreciation of what is involved
in their health care but they will have a quite limited capacity to appreciate long-term
consequences and will be restricted on their capacity to balance competing considerations.
WHEN, IF AT ALL, SHOULD CHILDREN BE PROTECTED FROM DECISION-MAKING?
The final issue concerns the question of whether there are circumstances in which it is
damaging for children to be expected to make difficult decisions. The query needs to be
split into two parts. First, at least after the early pre-school years, it is never acceptable for
children to be excluded totally from participating in decision-making on matters of health
care that concern them. The weight given to their views should properly be influenced
by the level of their understanding, but their wishes should not be disregarded as of no
consequence. Second, that does not necessarily mean that children, or even young
adolescents, should be expected to be the final arbiter on a serious decision. They are
used to turning to parents for advice and guidance and most young people continue to do
so on serious matters right through adolescence, despite rebellion on more everyday
concerns such as hair style, dress and time in at night (Rutter, 1979; Rutter & Rutter,
1993). Shared decision-making will be the preferred option in most instances. Perhaps
the most obvious circumstance when it is particularly damaging to force (or expect)
children to take the final decision is when one parent argues for one course and the other
parent presses for the opposite. Inevitably, such decisions will be felt by the child as having
an element of conflict over loyalty. The closest parallel is provided by decisions over
which parent should look after the child when there is a parental divorce or separation.
It would be usual nowadays to seek to understand the child’s feelings and wishes and to
give them considerable weight, and that ought to apply equally to health care decisions.
Ordinarily, children are not, and should not be, expected to choose between their parents,
but, equally, there ought to be grave hesitation before taking a decision that runs counter
to a child’s firmly and consistently expressed opinion when that is judged to be based on
an adequate (not necessarily complete) understanding of the issues, and when it does
not appear to be the result of undue influence exerted by one or other parent. The same
should apply to health care decisions.
EXTRAPOLATIONS TO CAPACITY TO CONSENT TO RESEARCH
42
Several key principles that are pertinent to the capacity to consent to research participation
derive from the findings on children’s development and their application to clinical
decision-making. First, a child’s ability to understand constitutes a graded phenomenon
and not something that appears in categorical fashion at some particular age. Even quite
young children can understand some of the relevant issues and it should be a standard
expectation that every effort is made to help children appreciate what is asked of them
and to participate in the decision-making process in so far as they are able. Second, even
with an individual child, the level of understanding is not a fixed feature. It will be
influenced by the socio-emotional context, the extent to which past experiences are
relevant to the particular decision, it can be increased by discussion and education and
its evaluation will be much influenced by whether questions are put in an abstract or reallife familiar way. Third, although legal considerations require that categorical decisions
on capacity have to be taken, it needs to be appreciated that this requirement derives
from practical considerations (in situations in which people may differ on whether there
should be research participation, someone’s view must prevail) and not from the underlying
psychological reality. The situation parallels that in clinical decision-making. The severity
of psychopathology varies in a continuous fashion but at some point a categorical decision
has to be taken on whether or not to treat or whether or not to admit to hospital. Fourth,
a child’s ability to understand may change over time according to alterations in mental
state, increased biological maturation or as a result of new experiences. This possibility
will need to influence the way in which children are dealt with over time in research
participation. Fifth, because children can understand to some degree, it should require
special justification to override a child’s clear and sustained objection to any research
procedure that carries significant risk. Finally, children are much influenced by the
views of those whom they love and respect. This means that they are likely to be influenced
by their parents’ attitudes. In most circumstances, this is likely to be protective, but care
needs to be taken in the unusual situation in which parents seem to be seeking to
influence children against the children’s own best interests.
APPENDIX 2
REFERENCES
Alderson, P. & Montgomery, J. (1996) Health Care Choices: Making Decisions with Children.
London: Institute for Public Policy Research.
Baron-Cohen, S., Tager-Flusberg, H. & Cohen, D. (1993) Understanding Other Minds: Perspectives
from Autism. Oxford: Oxford University Press.
Bruck, M. & Ceci, S. J. (1999) The suggestibility of children’s memory. Annual Review of
Psychology, 50, 419–439.
Bryant, P. E. (1974) Perception and Understanding in Young Children. London: Methuen.
Ceci, S. J. & Bruck, M. (1993) Suggestibility of the child witness: a historical review and
synthesis. Psychological Bulletin, 113, 403–439.
–––, Bronfenbrenner, U. & Baker-Sennett, J. G. (1994a) Cognition in and out of context: a tale
of two paradigms. In Development Through Life: A Handbook for Clinicians (eds M. Rutter &
D. Hay), pp. 239–259. Oxford: Blackwell Scientific.
–––, Baker-Sennett, J. G. & Bronfenbrenner, U. (1994b) Psychometric and everyday intelligence:
synonyms, antonyms and anonyms. In Development Through Life: A Handbook for Clinicians
(eds M. Rutter & D. Hay), pp. 260–283. Oxford: Blackwell Scientific.
Dunn, J. (1988) The Beginnings of Social Understanding. Oxford: Blackwell.
Goswami, U. (1991) Analogical reasoning: what develops? A review of research and theory.
Child Development, 62, 1–22.
Harris, P. L. (1989) Children and Emotion: The Development of Psychological Understanding.
Oxford: Basil Blackwell.
Justice (1996) Children and Homicide: Appropriate Procedures for Juveniles in Murder and
Manslaughter Cases. London: Justice.
Kagan, J. (1981) The Second Year: The Emergence of Self-Awareness. Cambridge, Massachusetts:
Harvard University Press.
Kail, R. V. (1990) Memory Development in Children (3rd edn). San Francisco: Freeman.
Keating, D. P. (1990) Adolescent thinking. In At the Threshold: The Developing Adolescent (eds
S. S. Feldman & G. R. Elliott), pp. 54–89. Cambridge, Massachusetts/London: Harvard
University Press.
Rushforth, H. (1999) Communicating with hospitalised children: review and application of
research pertaining to children’s understanding of health and illness. Journal of Child Psychology
and Psychiatry, 40, 683–691.
Rutter, M. (1979) Changing Youth in a Changing Society: Patterns of Adolescent Development and
Disorder. London: Nuffield Provincial Hospitals Trust (1980, Cambridge, Massachusetts:
Harvard University Press).
––– & Rutter, M. (1993) Developing Minds: Challenge and Continuity Across the Lifespan.
Harmondsworth, Middlesex: Penguin and New York: Basic Books.
Sternberg, R. J. & Powell, J. S. (1983) The development of intelligence. In Cognitive Development,
Vol.3, Mussen’s Handbook of Child Psychology (eds J. H. Flavell & E. M. Markman),4th edn,
pp. 341–-419. New York: Wiley.
Walden, T. & Garber, J. (1994) Emotional development. In Development Through Life: A
Handbook for Clinicians (eds M. Rutter & D. Hay), pp. 403–455. Oxford: Blackwell Scientific.
43
ROYAL COLLEGE
APPENDIX 3 – DECISION-MAKING AND ASSESSMENT OF
OF PSYCHIATRISTS
RESEARCH
COMPETENCE IN INDIVIDUALS WITH A LEARNING DISABILITY
GUIDELINES:
ETHICS
By W. I. Fraser
BACKGROUND CONSIDERATIONS
Although the distinction between therapeutic and non-therapeutic research is becoming
less tenable, this distinction has often guided ethics committees in relation to research on
learning disability. In the past, people with learning disability have frequently been
automatically excluded from participation in non-therapeutic research. Supporters of
normalisation now generally consider this devalues people with learning disabilities. Also,
researchers know this detracts from the representativeness of research when samples do
not include the ‘lowest’ range of IQ distribution. A good example is provided by current
research into the neurodevelopmental origins of schizophrenia, in which it is only recently
that schizophrenia in individuals with learning disability have begun to be included.
Ethics committees have to weigh the same benefits, discomforts and risks for individuals
with learning disability as for any other group of participants, and this need applies
equally to therapeutic and non-therapeutic research.
People with learning disability have been excluded unthinkingly and automatically
from many non-intrusive surveys that do not affect their privacy, simply because there
were uncertainties about their ability to give or withhold consent.
Research with persons with learning disability can be justified when:
(a) it attempts to provide knowledge about the care and treatment of people
with such clinical conditions;
(b) it cannot be carried out without their involvement;
(c) it does not expose them to more than negligible risk, is not unnecessarily
invasive or restrictive and does not unduly interfere with their freedom
and privacy.
Minimal risk can be defined as the situation in which the probability and magnitude
of harm or discomfort are no greater than those encountered in ordinary life or during the
performance of routine physical or psychological examinations or tests (Council for
International Organisations of Medical Sciences & World Health Organization, 1993).
Research journals in learning disability now use the above as an ethics framework for
publications. Researchers in learning disability, like researchers in other fields, are faced
with guidelines listing questions about every project involving people with learning
disability. These include: How much information is to be given to the subjects? What
could prevent the subjects from understanding the information? Will the subjects have
acquired appropriate information to decide freely to take part? What detail of explanation
is required? (Informed consent in clinical practice has not required detailed
understanding.) Information may not be sufficiently clear. It may be given too quickly or
under conditions where the subject is unable to concentrate. The individual may have
difficulties in understanding because of anxiety, depression and dementia on top of low
intelligence. Although serious mental illness may impair understanding, it does not
necessarily have this effect.
A recent study by Wong et al (2000) of capacity in groups with mental illness, learning
disabilities and dementia showed that capacity was significantly impaired in the learning
disabled and dementia groups, but not in the mental illness group. With the exception of
the dementia group, the other groups benefited from more information. In the learning
disabled group, simplification of information increased the capacity to make decisions.
Those with learning disabilities are particularly likely to be suggestible in their
decision-making. Most people with learning disability will fail suggestibility tests
and even when they pass on oral vignette tests of vulnerability, the validity of such
44
tests in real situations, where wishes to please override common sense, may be very
low. They may say what they believe is expected of them. In individuals of all ages
with learning disability, explanations should be in language and terms that are
appropriate to the person’s stage of development. It may be helpful to use pictorial
information to aid communication and, where appropriate, use should be made of
augmented communication symbolics (such as BLISS) or of signing (such as by
Makaton).
Learning disability does not necessarily imply incompetence, and an inability to
make one kind of decision does not necessarily mean a general incompetence. It is good
practice to get the views of close relatives and an independent person who knows the
subject (such as an advocate) but, whatever their views, the decision of the subject is
paramount in most circumstances.
The various approaches to the assessment of capacity in learning disability were
reviewed by Wong et al (1999). Traditionally, reliance used to be placed on either the
outcome of individual decision-making (capacity being determined on whether the
decision met some external criterion) or on the status of the person as belonging to a class
of low IQ that automatically meant a lack of competence. Both approaches have been
rejected on empirical and conceptual grounds.
Instead, a functional approach is now preferred (Hoggett, 1994; Arscott, 1997).
This focuses on the person’s relative abilities and the extent to which these meet the
demands for a particular decision-making task. As in other groups, so also for those
with learning disabilities, the following are necessary: retaining and understanding
information about the procedure; appreciating that the information is of personal
relevance; weighing up the information to reach a decision; and communicating the
decision (Appelbaum & Grisso, 1995; Grisso & Appelbaum, 1996; Law Commission,
1995). This functional approach (President’s Commission for the Study of Ethical
Problems in Medicine and Biomedical and Behavioural Research, 1983; Law
Commission, 1995; Scottish Law Commission, 1995), is based on the extent of the
person’s understanding, knowledge, skills and abilities to meet the demand for the
task involved in making a particular decision. It is decision-making and time specific.
While evidence from standardised assessments of mental states is essential background,
this should not replace a specific evaluation of decision-making abilities in relation to
the particular procedure.
The way in which relevant information is presented to the subject is crucial. Regardless
of whether or not the person has a ‘mental disability’, material seems easier to understand
when it is presented in simple language and separate elements, rather than in an
uninterrupted text (Grisso & Appelbaum, 1998). Consideration should be given to the
possibility of improving relevant functional abilities by education, treatment or simplifying
the task by presenting it in simpler language or by use of pictures or by augmented or
alternative communication systems (Tymchuk et al, 1986). There is an obligation to take
all practical steps to improve communication in learning disability. The status and
functional approaches are usually combined so that a functional assessment is only carried
out when status places the presumption of capacity in doubt.
APPENDIX 3
ABILITIES RELEVANT TO DECISION-MAKING
Wong et al (1999) listed the following as relevant abilities: (a) communicating a
choice; (b) understanding relevant information; (c) retaining the information; (d)
appreciating the purpose and significance of the information; and (e) reasoning, and
rational manipulation of information to arrive at a decision (Appelbaum & Grisso,
1988; Berg et al, 1996). With respect to the fourth requirement, dealing with
appreciation of the situation and its consequences, it is necessary for the person to
recognise that he or she has a disorder. Todd and Shearn (1997) pointed out that
carers are sometimes reluctant, for political correctness, to inform these people that
they have a learning disability. The fifth requirement of handling information
rationally centres on the process by which the individual reaches a decision, not its
outcome. It focuses on whether the person’s decision was a rational product of his or
her underlying beliefs and, in the case of people with learning disability, not simply
an intention to preserve face or to please.
45
ROYAL COLLEGE
OF PSYCHIATRISTS
RESEARCH
GUIDELINES:
ETHICS
46
LEARNING DISABILITY ADULTS WITHOUT CAPACITY
Substitute judgement (requiring the proxy to decide which is the best estimate of a person’s
choice) has been much employed in the past by practitioners in learning disability to make
treatment decisions. This was in accordance with the practice accepted at the time by a
responsible body of appropriately skilled medical opinion. This was most recently challenged
following the admission to hospital, without using the Mental Health Act 1983, of a person
who lacked capacity but was not objecting. The Court of Appeal found against Bournewood
Community and Mental Health NHS Trust (L v Bournewood Community and Mental
Health NHS Trust decision by the House of Lords in the appeal, 1998). The House of Lords
in June 1998 overturned the Court of Appeal’s decision but the issue will undoubtedly arise
again. It is being considered by the European Court of Human Rights.
In Scotland (although very rarely employed) there is a court provision to appoint
someone to have the power to give consent on behalf of an adult – a tutor dative. Enquiry
should, therefore, be made as to whether there is a tutor dative and if there is, what is the
extent of power to consent given by the appointing court. Presumably, the law would
permit this to apply to research but no examples of this use are known.
For any non-therapeutic procedure involving people with learning disability, more
systematic assessments of what constitutes ‘best interest’ will be required. A general rule
is that the capacity required should increase as the risk–benefit ratio goes down (American
Psychiatric Association, 1998; Hansson, 1998; Clegg, 1999). Potential proxies show lack
of agreement on health care decisions for people who lack capacity (Ouslander et al,
1989). Large differences exist among and within professional groups and carers as to the
legal capacity of clients (Biesaart & Hubben, 1999). Conflicts of interest make parental
caregivers and independent advocates of limited use as proxies. The most serious ethical
concerns are raised by higher risk studies offering no prospect of direct benefit to incapable
subjects. The Nuremberg Codes and the Declaration of Helsinki do not authorise this
form of research: “It should be permitted only after vigorous ethical review or public
notice and opportunity for comment”.
The criteria for capacity now widely accepted in learning disability are those
suggested by the Law Commission (1995) for England and Wales, adopted in the proposed
Mental Incapacity Bill (Wong et al, 2000). The best example of decision-making assessment
is that of Wong et al (2000), drawing on the methodology used in the MacArthur Treatment
Competence Study (Grisso & Appelbaum, 1995). This comprises an information sheet
providing the necessary elements of information, including: (a) the purpose of the test;
(b) the nature of the procedure; (c) the risks of having the test; (d) the risk of not having
the test; and (e) voluntariness – the principle of free choice. These should be written in
simple language, employing a readability measure to check this, and printed in large font.
Wong et al (1999) provided a useful flow chart for consent to health care interventions,
indicating what is ethically and legally defensible for people with learning disability. This
has value for research consent.
This decision assessment to elicit valid responses of capacity to make a decision
tested by Wong et al (2000) has a satisfactory inter-rater reliability. It comprises a
spontaneous account, involving participants relating what in their view is a blood test,
before any information disclosure, to establish a baseline. In their study, an information
sheet was then presented in its entirety under two conditions: uninterrupted disclosure
and elements disclosure. The sheet was read out to the individuals and given to them in
writing. Participants were then asked for paraphrased recall and a recognition task was
used to identify whether a series of statements matched the information. For learning
disabled people, as for people with mental illness, the best understood item was the
procedure, followed by the purpose, voluntariness and then risk. People with dementia,
in contrast, found voluntariness less difficult than purpose. As Wong et al (1999) pointed
out, participants who need support in their everyday lives are not used to considering the
consequences of their decisions. Wong and colleagues are currently investigating a strategy
that may make these more abstract elements more concrete, such as by using pictures.
The simplification of decision-making tasks significantly increased capacity for the learning
disabled by presenting information in this uninterrupted form and then by constituent
elements and by amending the response to include recognition and non-verbal
demonstration.
Most adults with moderate disabilities who have a research project explained twice
understand what the research is about and how often the researcher wishes to meet
them, but less than half know they can withhold consent at any time (Arscott et al,
1998).
APPENDIX 3
ASCERTAINING THE LEARNING DISABILITY PERSON’S PERSPECTIVE
Advocacy organisations in learning disability, such as People First, are already guided by the
Green Paper, Who Decides (Lord Chancellor’s Department, 1997). Best interest is decided by
the following factors: (a) ascertainable past and present wishes and feelings of the person
concerned and the factors that the person would consider, if able to do so; (b) the need to
permit a person to participate as fully as possible in any decision affecting him or her; (c) the
views of other persons, if appropriate and practical to consult about the person’s wishes and
feelings, and what would be in his or her best interests; and (d) the purpose for which any
action or decision required can be as effectively achieved in a manner least restrictive of a
person’s freedom of action (i.e. an attempt has to be made to make some assessment of the
preferences, hopes and values). Clearly this is a complex task, which is generally not carried
out in the treatment of learning disabled persons and may well have a significant impact and
different perspective for patients with backgrounds that differ from those of the researcher.
The speciality of learning disability has a tradition of individuals with that condition being
thought of as belonging to a dependent category, an attitude that is not value-neutral. They
may even be regarded as not necessarily entitled to life (e.g. use of the severe handicap as a
justification for termination of pregnancy in the Abortion Act) and less entitled to contribute
to society (Edwards, 1997). Sometimes, too, they may be thought not to have a degree of
ethical or moral development sufficient to enable them to take part for altruistic reasons and,
therefore, not to be able to take part in non-therapeutic research. The lack of opportunity to
take part is disenfranchising them and disproportionately excludes them from sharing the
burdens and benefits of ordinary people (Edwards, 1997; Arscott et al, 1998; Clegg, 1998,
1999).
Acknowledgements
This Appendix draws heavily on the papers by Wong et al, provided by Dr. A. J. Holland,
one of the authors, and I am much indebted to him for making them available at a time
that was prior to publication.
REFERENCES
American Psychiatric Association (1998) Guidelines for assessing the decision-making capacities
of potential research subjects with cognitive impairment. American Journal of Psychiatry,
155, 1649–1650.
Appelbaum, P. S. & Grisso, T. (1988) Assessing patients’ capacities to consent to treatment. New
England Journal of Medicine, 319, 1635–1638.
––– & ––– (1995) The MacArthur Treatment Competence Study: 1. Mental-illness and
competence to consent to treatment. Law & Human Behavior, 19, 105–126.
Arscott, K. (1997) Assessing the capacity of people with learning disabilities to make decisions
about treatment. Tizard Learning Disability Review, 2, 17–28.
–––, Dagnan, D. & Kroese, B. (1998) Consent to psychological research by people with an
intellectual disability. Journal of Applied Research in Intellectual Disabilities, 11, 77–83.
Berg, J.W., Appelbaum, P. S. & Grisso, T. (1996) Constructing competence: formulating standards
of legal competence to make medical decisions. Rutgers Law Review, 48, 345–396.
Biesaart, M. C. I. & Hubben, J. H. (1999) Incompetence in practice in health care in the
Netherlands: Report of a study. Journal of Intellectual Disability Research, 43, 454–460.
Clegg, J. A. (1998) Critical Issues in Clinical Practice. London: Sage.
––– (1999) Ethics in intellectual disability. Current Opinion in Psychiatry, 12, 537–541.
47
ROYAL COLLEGE
OF PSYCHIATRISTS
RESEARCH
GUIDELINES:
ETHICS
48
Council for International Organisations of Medical Sciences (CIOMS) & World Health
Organization (1993) International Ethical Guidelines for Biomedical Research Involving Human
Subjects. Geneva: WHO.
Edwards, S. D. (1997) The moral status of intellectually disabled individuals. Journal of Medicine
& Philosophy, 22, 29–42.
Grisso, T. & Appelbaum, P. S. (1995) Comparison of standards for assessing patients’ capacities
to make treatment decisions. American Journal of Psychiatry, 152, 1033–1036.
––– & ––– (1996) Values and limits of the MacArthur Treatment Competence Study. Psychology,
Public Policy and Law, 2, 167–181.
––– & ––– (1998) Assessing Competence to Consent to Treatment. Oxford: Oxford University
Press.
Hansson, M. O. (1998) Balancing the quality of consent. Journal of Medical Ethics, 24, 182–187.
Hoggett, B. (1994) Mentally incapacitated adults and decision-making. The Law Commission’s
project. In Decision-Making and Problems of Incompetence (ed. A. Grubb), pp. 27–40.
Chichester: John Wiley & Sons.
Law Commission (1995) Mental Incapacity (Law Commission Report No. 231). London: HMSO.
Lord Chancellor’s Department (1997) Who Decides: Making Decisions on Behalf of Mentally
Incapacitated Adults. A Consultation Paper issued by the Lord Chancellor’s Department.
London: The Stationary Office.
L v Bournewood Community and Mental Health NHS Trust decision of the House of Lords in
the appeal (1998) Health Service Circular, series number 122. London: Department of Health.
Ouslander, J., Tymchuk, A. & Rahbar, B. (1989) Health care decisions among elderly long-term
care residents and their potential proxies. Archives of Internal Medicine, 149, 1367–1372.
President’s Commission for the Study of Ethical Problems in Medicine and Biomedical and
Behavioral Research (1983) Making Health Care Decisions. Washington DC: US Government
Printing Office.
Scottish Law Commission (1995) Report on Incapable Adults. Scottish Law Commission Report
No. 151. Edinburgh: HMSO.
Todd, S. & Shearn, J. (1997) Family dilemmas and secrets: parents’ disclosure of information to
their adult offspring with learning disabilities. Disability & Society, 12, 341–366.
Tymchuk, A.J., Ouslander, J.G. & Rader, N. (1986) Informing the elderly: a comparison of four
methods. Journal of the American Geriatrics Society, 34, 818–822.
Wong, J.G., Clare, I.C.H., Gunn, M., et al (1999) Capacity to make health care decisions. Its
importance in clinical procedures. Psychological Medicine, 29, 437–446.
–––, –––, Holland, A.J., et al (2000) The capacity of people with a ‘mental disability’ to make a
health care decision. Psychological Medicine, 30, 295–306.
APPENDIX 4 – WAIVED CONSENT FOR RESEARCH
INTO ACUTE GRAVE ILLNESS, INCLUDING UNCONSCIOUS
PATIENTS, UNDER EMERGENCY CONDITIONS
APPENDIX 3
By J. Pickard
1.
CLINICAL CARE
It is generally accepted that when a patient presents acutely with a depressed level of
consciousness because of injury or illness the clinician may undertake whatever treatment is
necessary to ensure the patient’s life or health, without waiting to obtain consent. A note
should be made in the clinical records to explain the absence of formal consent. It is good
practice to involve relatives or other carers in the decision-making, but in an emergency, the
desirability of this should not delay the clinician taking such actions as the best interest of the
patient demands. It is important that the clinician confine himself to providing only necessary
treatment (Council of Europe, 1997; Medical Defence Union, 1997).
2.
NEED FOR RESEARCH INTO ACUTE GRAVE ILLNESS AND UNCONSCIOUS PATIENTS
There is general agreement that research is necessary if emergency medical, surgical and
neonatal care are to be improved. Some currently accepted treatments may prove to be
counterproductive and new treatments require critical evaluation. This is an international
view. For example, the proposed Maryland Law on research involving decisionally
incapacitated individuals emphasises that research may be essential if science is to
understand and ultimately combat disease of the brain, including Alzheimer’s disease,
severe psychiatric disorders, severe trauma, stroke, other causes of decisional incapacity
and the medical problems that are associated with these conditions and disorders (British
Paediatric Association, 1992; Council of Europe, 1997; Anonymous, 1998).
3.
CONSENT TO RESEARCH
The General Medical Council (GMC, 1998) recommends that a practitioner should seek
further advice where research involves children or adults who are not able to make a
decision for themselves. The GMC emphasises that in these cases the legal position is
complex or unclear and there is currently no general consensus on how to balance the
possible risk and benefits for such vulnerable individuals against the public interest in
conducting research. However, the GMC goes on to note that a number of public
consultation exercises are underway and guidance should be sought from bodies such as
the Medical Research Council (MRC) and Medical Royal Colleges, as well as the relevant
research ethics committees.
Such bodies generally agree that exceptions are required to the need for informed
consent for research provision equivalent to those governing that of medical care in an
emergency. On many, but not all, such occasions it may be impractical, or meaningless,
to attempt immediately to obtain consent for a proposed research study from relatives,
parents or guardians before proceeding. To require such an attempt always to be made
would also inhibit much potentially valuable research. Provided, therefore, that the specific
approval of a local research ethics committee has been obtained for the project overall, it
would be ethical to carry out research on unconscious and decisionally incapacitated
individuals including children without obtaining consent on such occasions of extreme
urgency.
The World Medical Association , as part of its revision of the World Medical
Association (1999) Declaration of Helsinki, notes in its paragraph on waiver of consent:
“When permitted by applicable law, the requirement for informed
consent may be waived by the independent research ethics
49
ROYAL COLLEGE
OF PSYCHIATRISTS
RESEARCH
GUIDELINES:
ETHICS
committee. It may be justified in research and emergency situations
in which patient-subjects of temporary or enduring loss of decisional
capacity and interventions or procedures must be initiated before
informed consent can be obtained from patient-subjects through
legally authorised representatives. In the latter case the research
ethics committee may require special procedures to protect the rights
and welfare of the research subjects.”
The proposed revision of the Declaration of Helsinki has attracted some controversy,
but not with regard to this section (paragraph 25) (Brennan, 1999; Levine, 1999).
The MRC (1998) guidelines for good clinical practice and good clinical trials state
that situations do exist in which fully informed consent may not be possible (e.g. emergency
settings). In these cases procedures agreed in the existing guidelines should be followed,
provided a favourable opinion has been given by the appropriate independent ethics
committee. Similar guidelines are provided by the Council of Europe (1997).
The Food and Drug Administration (FDA, 1997) notes that, in emergency situations,
when prior consent by the subject is not possible, the consent of the subject’s legally
acceptable representative, if present, should be requested. When prior consent of the
subject is not possible, and the subject’s legally acceptable representative is not available,
enrolment of the subject should require measures described in the protocol and/or
elsewhere with documents of approval/favourable opinion by the ethics committee to
protect the right, safety and well-being of the subjects and ensure compliance with
applicable regulatory requirements. Previously the FDA had specified that such waived
consent applied only in the very restrictive circumstances of testing investigational
therapies for life threatening emergency medical conditions and for whom available
treatments were unproven or unsatisfactory. It added four additional safeguards or
additional protection:
(a) Consultation with representatives of the communities from which the
subjects will be drawn.
(b) Public disclosure prior to the commencement of the study sufficient to
describe the study and its risks and benefits.
(c) Public disclosure of sufficient information following completion of the
study to apprise a community in the research to the study and its results.
(d) The establishment of an independent data and safety monitoring board.
However, it has been pointed out that some of the special precautions may be counterproductive (Mariner, 1995). There is a problem with defining the appropriate community
for consultation. Lindley (1998) suggested that the appropriate community may be the
patients and relatives who have personal experience of the particular illness. Communities
should be informed that entry into placebo controlled randomised control trials offers
patients, who are in the randomised placebo group, some benefits. First, they are not
exposed to the risks of novel treatment and second, they have the benefit of being
treated by a better-educated and interested team with better records, follow-up and
ancillary treatments.
4.
WHAT SHOULD BE THE MECHANISM OF WAIVED CONSENT?
(a) A local or multi-centre research ethics committee needs to be
convinced that the prospective subjects have a life threatening
situation, available treatments are unproven or unsatisfactory and
clinical research is required to determine which intervention is most
beneficial. In addition, obtaining informed consent is not feasible.
The research is in the patient’s best interest because the emergency
necessitates intervention and the research risk is reasonable
compared to the risks of existing therapy. Clearly in the context of
potential treatments for head injury and stroke, for example, it is
not reasonable to expect that treatment risks are minimal in relation
to the standard of available treatment. Existing treatments may be
50
dangerous and have indeed proven to be so. Some treatments may
be beneficial but certainly carry a risk, for example, thrombolysis for
stroke and myocardial infarction. Hence it has been recommended
that a better guideline would be that such a trial is ethical if the
treatment is promising but unproved, provided that the potential
risks are considered acceptable by the public.
APPENDIX 4
(b) Rather than vague review by the community, which is already explicit
in the composition of ethical committees that include lay members, the
opinions of patients and their families who have personal experience of
a particular condition should be sought.
(c) Four tiers of decision making have been discussed by various law
commissions including advanced statements, family members, patientselected proxy decision-makers and court-appointed guardians. In the
absence of any of these the doctor must make the decision (Luttrell,
1997). The problem with advanced statements is that the incompetent
patient has become a different person and hence there is a need for
both physicians and surrogates to focus more on the incompetent patient’s
current condition rather than on previous preferences (Fletcher &
Spencer, 1995).
(d) Much discussion centres on the best interest test and the substituted
judgement test. The two tests may not be mutually exclusive and the
Law Commission in 1993 recommended the use of the test based on
best interest with a strong element of substituted judgement (see also
Lord Chancellor’s Department, 1997).
(e) The FDA (1999) have led the way in auditing the use of exceptions
from informed consent for emergency research and have published the
results on the internet.
Patients with head injury and impaired consciousness may be unable to give properly
informed consent, and in this emergency situation it may not be medically appropriate to
delay the start of treatment until proxy consent can be obtained. Hence, the doctor in
charge should take responsibility for entering such patients, just as he or she would take
responsibility for choosing other treatments. However, the requirements of the relevant
research ethics committee must be adhered to at all times. An information leaflet on the
study for patients and their friends and relatives will be available in all drug packs
(http://www.CRASH.ucl.ac.uk).
5.
DEFERRED CONSENT
It is generally agreed that the patient, the family or the guardian(s) should be informed
about the research as soon as possible both for reasons of ethics and courtesy (British
Paediatric Association, 1992). However, there remains discussion about the notion of
‘deferred consent’ (Mariner, 1995). The Royal College of Physicians (1996) considers
that the local research ethics committee should advise on what, if anything, the
patient or family should be told later. Informing the patients that they are participants
in a trial, after entry, even if offering the possibility of withdrawal, may not prevent
their inclusion as contributors to the data in an ‘intention to treat’ design. This does
not appear to be a major ethical issue, whereas the right to withdraw from an act of
involvement clearly is. Somberg (1996) noted that in his experience patients and
their families were thankful to have the option to be told, and to go along with the
research project as it continued or to withdraw from the project once they could be
presented with the information and make a choice of their own volition. The FDA
(1997) noted “The subject or the subjects legal representative should be informed
about the trial as soon as possible and consent to continue and other consents as
appropriate should be requested.”
Deferred consent should not stand alone as a mechanism but it is simply the final
step in a well-regulated mechanism of waived consent.
51
ROYAL COLLEGE
REFERENCES
OF PSYCHIATRISTS
RESEARCH
GUIDELINES:
ETHICS
Brennan, T. A. (1999) Proposed revisions for Declaration of Helsinki: Will they weaken the
ethical principles underlying human research? New England Journal of Medicine, 341, 527–
531.
British Paediatric Association (1992) Guidelines for the Ethical Conduct of Medical Research
Involving Children. London: British Paediatric Association.
Anonymous (1998) Research and the incompetent. Bulletin of Medical Ethics, 141, 22–24.
Council of Europe (1997) Convention of Human Rights in Biomedicine. Articles 1–22.
Food and Drug Administration (1997) International Conference on Harmonisation: Good Clinical
Practice: Consolidated Guidelines. www.fda.gov/cder/guidance/iche6.htm.
––– (1999) Informed Consent: Exception for Emergency Research (22 June 1998). www.fda.gov/
cder/present/dia698/diafda4.
Fletcher, J. C., & Spencer, E. M. (1995) Incompetent patient on the slippery slope. Lancet, 345,
271–272.
General Medical Council (1998) Seeking Patients’ Consent: The Ethical Considerations. London:
GMC.
Levine, R. J. (1999) The need to revise the Declaration of Helsinki. New England Journal of
Medicine, 341, 531–534.
Lindley, R. I. (1998) Thrombolytic treatment for acute ischaemic stroke: consent can be ethical.
British Medical Journal, 316, 1005–1007.
Lord Chancellor’s Department (1997) Who Decides: Making Decisions on Behalf of Mentally
Incapacitated Adults. A Consultation Paper issued by the Lord Chancellor’s Department.
London: The Stationary Office.
Law Commission (1993) Mentally Incapacitated Adults and Decision-Making: A New Jurisdiction.
Consultation Paper No. 128. London: Stationery Office.
Luttrell, S. (1997) Making decisions about medical treatment for mentally incapable adults in
the UK. Lancet, 350, 950–953.
Mariner, W. K. (1995) Research and emergency care without consent: new proposed FDA rules.
Lancet, 346, 1505–1506.
Medical Defence Union (1997) Consent to Treatment. London: Medical Defence Union.
Medical Research Council (1998) Guidelines for Good Clinical Practice in Clinical Trials. Clinical
Trials Series. London: MRC.
Royal College of Physicians (1996) Guidelines on the Practice of Ethics Committees in Medical
Research Involving Human Subjects (3rd edn). London: The Royal College of Physicians.
Somberg, J. C. (1996) Research and emergency care without consent: New proposed FDA rules.
Lancet, 347, 475–476.
World Medical Association (1999) Proposed revision of the World Medical Association Declaration of
Helsinki (WMA document:17.C/Rev 1/98). Ferney-Voltaire, France: World Medical Association.
52