Research methods unit I article review

Science and Engineering Ethics (2006) 12, 75-83 Science and Engineering Ethics, Volume 12, Issue 1, 2006 75 Keywords: risk, responsible conduct of research, research ethics, research integrity, placebo control, international guidelines of biomedical research ABSTRACT: Assessment of risk is one of the key issues in the field of responsible conduct of research which covers discourses of research ethics and research integrity.

The principle of minimizing risks and balancing of risks and benefits is one of the main requirements of research ethics. In addition, the content of informed consent that is another fundamental principle of research ethics derives from the assessment of risks and benefits related to a particular research project. Risk assessment also plays a crucial role in methodological design of the research project. This is an important point where research ethics and research integrity discourses overlap. Firstly, because the choice of a control group (e.g., placebo control) is a key ethical issue related to the protection of the research subjects’ interests. Secondly, because the quality of the research data, that is one of the key elements of research integrity, is closely connected to the choice of research methodology as well. The problem of biased interpretation or manipulation of risk related features of biomedical research should also be taken into account. Despite the importance of the concept of risk to the field of biomedical research, its relevance has not yet attracted a sufficient attention in the responsible conduct of research debate. Introduction: research ethics, research integrity and risk Modern biomedical research raises different normative issues and attributes complex responsibilities to researchers. The concept of responsible conduct of research (RCR) has recently been coined to cover all these different issues. It integrates among other topics such issues as scientific misconduct, questionable research practices, sloppy or careless research, conflict of interest, human subjects protection. 1 The topics listed above indicate that what is called RCR in fact covers what has been traditionally attributed to the fields of research ethics (RE) and research integrity (RI). The Concept of Risk and Responsible Conduct of Research Eugenijus Gefenas Medical Faculty of Vilnius University, Lithuania This paper was presented at the 6th International Bioethics Conference on the subject of ‘The Responsible Conduct of Basic and Clinical Research’, held in Warsaw, Poland, 3-4 June 2005.

Address for correspondence: Eugenijus Gefenas, MD, PhD, Medical Faculty of Vilnius University, PO Box 2484, 2051 Vilnius, Lithuania; email: [email protected].

1353-3452 © 2006 Opragen Publications, POB 54, Guildford GU1 2YF, UK. http://www.opragen.co.uk E. Gefenas 76 Science and Engineering Ethics, Volume 12, Issue 1, 2006 A division of normative issues related to RE and RI might seem to be rather unproblematic at first sight. While research ethics first of all concentrates on the rights and welfare of research participants, issues attributed to the scope of research integrity are mainly those dealing with the quality of research data and the relationship between researchers themselves. We might therefore think about RI in terms of the “ïnternal” part of the RCR because the issues dealt with do not overstep the borders of the research community (it is therefore sometimes called “ethics of researchers”). On the other hand, research ethics has traditionally covered the relationship between researchers and research participants targeting what we might call “external” responsibilities of the researchers. This seemingly clear watershed is however a bit more complicated. Even if we could separate the problem areas of RE and RI, there are some overlapping fields of concern. For example, the problem of conflict of interest should be dealt with from both perspectives because it influences the conduct of a researcher with regard to both the scientific community and research participants. A researcher who is entangled into the conflict of interest (e.g., due to the fact that he acts both as a principal investigator as well as a treating physician) might distort or manipulate the results of the study (the issue falling under the “mandate” of the RI discourse) as well as interfere with the autonomy or welfare of the research participant due to a biased interpretation of relevant features of the trial and/or a wish to enroll a sufficient number of research participants (that are key issues of RE). The process of designing a research proposal and choosing the methodology of the clinical trial might also bring both research ethics and research integrity perspectives together. It is very often stated, for example, that scientifically invalid or useless research is at the same time unethical. There is also a more complicated and subtle relationship between scientific and ethical aspects of research which leads to the overlap between research integrity and research ethics. We might refer here to those cases when considerations of risks and benefits to potential research participants influence the choice of methodology and study design as well as become a decisive factor in categorizing some practices as biomedical research to be reviewed by an ethics body. It is self-evident that risk-related criteria that are used in a process of ethical review of research projects make the concept of risk of paramount importance to research ethics. First of all, it concerns the principle of minimizing risks and balancing of risks and benefits. In addition, an accurate description of possible risks and benefits is a basis for informed consent that is another fundamental ethical principle to be followed in biomedical research: significant and probably the most important information to be conveyed to a research participant derives from a careful and systematic assessment of risks and benefits performed during the planning and preparation of the research project. It might, however, be claimed that the language and conceptual framework of risk-related features of ethical principles and corresponding regulatory/legal documents are much more ambiguous as compared with those related to the principle of informed consent. Even though there are many practical problems related to the implementation of the principle of informed consent (e.g., achieving a genuine informed consent in The Concept of Risk and Responsible Conduct of Research Science and Engineering Ethics, Volume 12, Issue 1, 2006 77 placebo controlled double blind clinical trials), its conceptual framework and content are rather well explicated. The importance of risk assessment in choosing a research methodology and study design should also make it one of the key concepts in the research integrity discourse where the quality of the research data is one of the main issues of concern. This is even more important taking into account the fact that assessment of risks and balancing of risks and benefits might easily become an object of manipulation and/or biased interpretation which brings no less dramatic consequences as other more often discussed research integrity issues related to research misconduct (e.g., fabrication, falsification, plagiarism) or questionable research practices. 1 That is why risk assessment related to biomedical research might be considered to be one of the important linking points between the fields of research ethics and research integrity as well as one of the central concepts of RCR. Despite this fundamental importance of risk/benefit assessment for the field of responsible conduct of research, the conceptual framework of defining risks in biomedical research is rather vague and ambiguous. Such concepts as “minimal risk”, “reasonable risk” or “acceptable risk” are not clearly defined in the international guidelines on biomedical research. Therefore, the purpose of this paper is to analyze the concept of risk and its role played for both research ethics and research integrity discourses.

The concept of risk and research ethics A standard definition of risk includes two main components: a quantitative and qualitative evaluation of possible harm. A magnitude of harm and probability of harm would constitute the quantitative component of risk. On the other hand, qualitative characteristics of risk allow us to distinguish physical, psychological, social and some other kinds of possible harm. 2 Let us first stress the importance of the concept of risk and its role played in defining the responsilities of researchers as well as the main guiding principles of ethical review. As has already been pointed out, the principle of minimizing risks and balancing risks and benefits is one of the most important components of the protection of research subjects and is included in all the major international research ethics guidelines, for example, the Declaration of Helsinki, 3 the Council of Europe Additional Protocol to the Convention on Human Rights and Biomedicine concerning Biomedical Research (hereafter AP to the CHRB), 4 International Ethical Guidelines for Biomedical Research Involving Human Subjects prepared by CIOMS (hereafter CIOMS Guidelines). 5 Together with the principles of respect for personal autonomy, protection of privacy and just distribution of harms and benefits it constitutes a core of research ethics.

The assessment of risk is structured into different levels of moral acceptability in all the major international guidelines on research ethics mentioned above. For example, such constructs as “minimal risk”, “reasonable risk” and “acceptable risk” are used to justify particular types of research projects. E. Gefenas 78 Science and Engineering Ethics, Volume 12, Issue 1, 2006 Risk related criteria are used both to indicate the type of research that falls under a more permissive rules of conduct or, in contrast, to restrict so-called “non-therapeutic” research on incapable or other vulnerable groups of persons. For example, the concept of minimal risk is used as a permissive criterion to justify an expedited procedure of ethical review for behavioural science research in the USA and Canada. 6,7 In a similar way, the absence of risk to psychological health is used to distinguish a type of research that does not require ethical review by research ethics committees according to the Council of Europe AP to the CHRB (see paragraph 17 of the Explanatory Report). 8 In this case, surveys, interviews, questionnaires and other types of non-interventional research that do not pose risk to the psychological health of research participants are exempt from the coverage of the mentioned instrument.

In contrast to previous examples, a criterion of risk is, however, much more often used as a restrictive one to strengthen the protection of research participants in non- therapeutic research that is defined as research without a direct benefit to the health of the research participant (see art 6 of the AD to the CHRB). CIOMS Guideline 8 says that “the risk presented by such intervention must be reasonable in relation to the knowledge to be gained”. The AD to the CHRB employs a term “acceptable” for the same purpose.

In addition, even more restrictive criteria of risk are supposed to be followed in research on incapable adults, children and other vulnerable groups. Different international guidelines differ regarding the level of risk allowed in research on vulnerable groups. For example, the article 17 of the AP to the CHRB strictly follows a “minimal risk” standard in research on persons not able to consent, which is defined as “a very slight and temporary negative impact on the health of a person concerned”.

The paragraph 100 of the the Explanatory Report to the AP to the CHRB provides examples of the interventions that might be considered as those not exceeding the minimal risk standard. These examples are among others: taking saliva, urine; taking small additional tissue samples during operation; taking a blood sample (capillary, peripheral vein); sonographic examinations, one X-ray exposure, or one exposure using magnetic imaging without a contrast medium”.

Similarly, the CIOMS Guideline 9 refers to the “low-risk standard”: that is the risk that should not exceed the risk attached to routine medical or psychological examination of incapable persons. However, the CIOMS Guideline 9 also provides a more liberal standard, a so-called “slight or minor increase above such risk” when there is a) overriding scientific and medical rationale for such an increase and b) research ethics committee’s (REC) approval. The Commentary to Guideline 9 explains that there is no agreed definition of what the “slight or minor increase” is; however, the Guideline says that its meaning is inferred from the REC’s reports that provide such examples as, additional lumbar punctures or bone-marrow aspirations in children with conditions in which such examinations are regularly indicated in clinical practice.

There might be problems related to the definition of minimal risk or “slight or minor increase over minimal risk”. For example, it is not easy to find an answer to the question as to what constitutes the “risk to psychological health” or whether such The Concept of Risk and Responsible Conduct of Research Science and Engineering Ethics, Volume 12, Issue 1, 2006 79 examples as “additional lumbar punctures or bone-marrow aspirations” could be regarded as justifiable cases of “minor increase over minimal risk”. We will, however, concentrate on two other important issues, namely, ambiguities that are related to the concept of acceptable risk as well as on the assessment of risk that is implicit in the process of designing biomedical research. These issues might have even broader implications on the responsible conduct of research debate. Defining acceptable risk Let us first concentrate on the possible impact of ambiguities in defining the concept of acceptable risk that is used to address the issue of risk in non-therapeutic research with capable research participants. In contrast to minimal risk which is defined as a threshold concept setting up an upper “absolute” limit to the risk that might be legitimately imposed on the research participants, a definition of “acceptable” or “reasonable” risk seems to be a “relative” and procedural one. For example, the Explanatory Report to the AP to the CHRB does not provide specific criteria to determine the level of what the acceptable risk is in non-therapeutic research. It just states that in case the “participants are able to consent to research, the level of risk and burden permitted (acceptable) is higher than that allowed for persons not able to consent (minimal risk and minimal burden)” (see paragraph 27). The relative nature of defining the term “acceptable” is further determined in relation to the benefits gained. It is important to note that the benefits that are referred to in this case are the benefits to society rather than to the individual participating in a biomedical research (see paragraphs 25 and 27). The procedural nature of defining the term ‘‘acceptable’’ is revealed by the following interpretations provided in paragraph 27 of the Explanatory Report to the AP to the CHRB: “Whether or not the risk and burden are acceptable will be considered carefully by the ethics committee and competent body that approves the research project. The final decision on whether or not the risk and burden are acceptable will be made by the persons concerned when they decide to give or withhold consent”.

One question to be asked in this context is the following one – how much higher the risk could be to be considered “acceptable” as compared to the minimal risk? The procedural and relative nature of defining acceptable/reasonable risk does not provide an answer to this question. It leaves the level of “acceptability” to be defined in every individual case. This might be regarded as a sign of flexibility to conduct clinical trials in different clinical as well as socio-cultural contexts. For example, a high level of risk associated with a new treatment/diagnostic procedure (very promising treatment with a very high probability of serious adverse events) might be justified in some special fields of medicine where chances of survival are very poor and new modalities of experimental treatment that are still in a very early stage of development promise some chance of success. Similar arguments have also been developed in relation to some experimental gene therapies.

However, at the same time the ambiguity of defining acceptable risk might be used to “import” risky research from the affluent to the developing countries. 9 It might also E. Gefenas 80 Science and Engineering Ethics, Volume 12, Issue 1, 2006 be used to justify placebo research where researchers might claim that withholding or withdrawal of a proven therapy does not present “unacceptable risk or burden to the participant” even in a case of rather risky research (see the case study presented below). As has already been pointed out, this might easily happen because the risks to the individual concerned are weighted not against individual benefits but against the generalizable knowledge gained as a result of the research.

Research methodology and evaluation of risk There are several contexts where evaluation of risk is implicit in the process of defining and designing biomedical research. We will briefly analyse the following two situations: a distinction between “therapeutic”/”non-therapeutic” research and the justification of a placebo control arm in clinical trials.

1. A distinction between “therapeutic”(also called research with direct benefit) and “non-therapeutic research” (or research without direct benefit) is the most straightforward example of risk assessment being an implicit factor in defining different types of research. This distinction is criticized as a misleading one by some authors. 10 However, it still plays a crucial role in all the mentioned international documents. As has already been shown in the previous section of the paper, therapeutic research is not restricted to specific thresholds of risk, such as “acceptable” or “minimal” risk, that are imposed on non-therapeutic research. What is particularly important to our discussion, defining a certain biomedical research project as a therapeutic one exempts it from the scrutiny of searching for the requirements of “acceptable”, “reasonable” or “minimal” risk (in research with incapable persons). It simply means that a researcher is much less constrained if he or she presents a research project as being a therapeutic one. This is also a powerful source of so-called therapeutic misconception so often distorting realistic expectations of research participants on risks and benefits related to a clinical trial. 2. Let us now consider how “the risk factor” might influence the design and methodology of biomedical research that brings both research ethics and research integrity perspectives particularly close to each other. An important example of this kind is an assessment of risk that is used to justify a placebo control group in the case of a clinical trial. This issue seems to be of particular interest to our discussion because it also shows the overlap between RI and RE discourses. This is because the choice of the control group influences the quality of the potential research data and at the same time is a crucial factor in protecting interests of the research subjects.

We are not going to present here a detailed analysis of this very complicated issue.

Our purpose is a limited one – using a recent amendment of the Helsinki Declaration, we will try to demonstrate the role the evaluation of risk plays in justifying the use of placebo in clinical trials. In its 2002 Note of clarification on paragraph 29 of the WMA Declaration of Helsinki, the WMA “reaffirms its position that extreme care must be taken in making use of a placebo-controlled trial and that in general this methodology should only be used in the absence of existing proven therapy.” The Concept of Risk and Responsible Conduct of Research Science and Engineering Ethics, Volume 12, Issue 1, 2006 81 However, at the same time it is stated that ‘‘placebo-controlled trial may be ethically acceptable, even if proven therapy is available, under the following circumstances:

• Where, for compelling and scientifically sound methodological reasons, its use is necessary to determine the efficacy or safety of a prophylactic, diagnostic or therapeutic method; OR • Where a prophylactic, diagnostic or therapeutic method is being investigated for a minor condition and the patients who receive placebo will not be subject to any additional risk of serious or irreversible harm.” Both of the conditions justifying the use of placebo are in fact based on the assessment of risks and benefits. The first one refers to the necessity to determine the efficacy or safety of the health care intervention. The terms efficacy and safety are conceptually linked to the risk/benefit analysis. According to the Dictionary on Clinical Trials the terms “Safe and effective “ usually implies a favourable risk/ benefit ratio. 11 The second justification of the use of placebo in spite of the existence of a proven therapy refers to the “minor condition”. This term also implies the evaluation of risk because what is defined as a “minor” or “serious” medical condition usually depends on whether or not the mentioned condition poses life-threatening consequences or long-term detrimental effects on the health of the individual concerned. The requirement to avoid “any additional risk of serious and irreversible harm” is an explicit risk assessment related statement. The above reasoning and “reduction” of the meaning of the mentioned conditions to the risk/benefit analysis might seem to convey a rather uncomplicated picture on the justification for placebo. However, it should be stressed that such a justification requires a rather specialized expertise and up-to-date knowledge of recent developments in the relevant areas of clinical pharmacology. Lay persons could hardly get to this specialized knowledge themselves and usually have to rely on the information provided by the experts in the field.

Case study In order to illustrate the importance and complexity of risk assessment for both ethical evaluation and methodological design of the research projects, let us analyze an example of a randomized, parallel-group, double-blind, placebo-controlled, phase III study, of the efficacy and safety of a new medication X and Lithium as monotherapy in 28 to 140 weeks’ maintenance treatment of Bipolar I Disorder in adult patients.

The justification of such a trial is supported by the EMEA guidance 7.2.2.2.

Prevention of manic/depressive episode. The Guidance states that “controlled studies against placebo and reference therapy (3-way studies) are first choice. In this way both the effect of the new compound can be shown and the relative benefit/risk can be assessed. If only an active comparator is used, assay sensitivity should be addressed adequately. Due to the changing course of bipolar disorder and the chronic character E. Gefenas 82 Science and Engineering Ethics, Volume 12, Issue 1, 2006 with recurrences of depressive and (hipo) manic symptoms with a long duration (at least one year) are necessary”. 12 It should be pointed out that Bipolar I Disorder is a life-threatening condition. For example, some authors argue that Lithium therapy might be regarded as a proven therapy to be applied in this case and that controlled clinical studies provide sufficient evidence that this therapy decreases the rate of suicides in the population of Bipolar I Disorder patients. 13 Therefore, following the Helsinki Declaration, justification of a clinical trial that includes a placebo control group could only be based on the assumption that existing clinical research data do not provide sufficient evidence to assure the efficiency and safety of the existing therapy. This is a highly specialized issue which requires comprehensive knowledge of the most recent research data in the field. We might just note that the above mentioned contradicting opinions show that at the moment there is a disagreement between the experts in the field about justification of the placebo use in this particular case.

Concluding remarks There are at least two key ways the assessment of risk is used in biomedical research.

First, risk assessment is a crucial component of national and international guidelines on research ethics. Second, it is also an integral feature of evaluating methodological aspects of biomedical research. The complexity and relevance of risk assessment to the choice of research methodology and study design makes it one of the key normative issues in the responsible conduct of research debate. This issue also reveals an interconnection between the fields of RE and RI. The field of research integrity seeks among other things to elucidate and prevent manipulations or biased interpretations of research data.

That is why the analysis of risk is of crucial importance for the RI discourse that is linked to the methodological and conceptual aspects of scientific research. The importance of evaluation of risk and balancing of risks and benefits is even more evident in the research ethics discourse. Minimizing risks and balancing of risks and benefits is regarded as one of the key principles of research ethics that directs the whole process of ethical review of biomedical research. What is more, the information provided to the research participants (which is a core of another fundamental ethical principle – informed consent) derives from the risk/benefit analysis discussed in the paper. Despite this fundamental importance of risk assessment, its role has not been fully recognized and implemented into the ethical review procedures. The process of ethical review is still too often mainly based on evaluating informed consent forms and does not always go deeply enough to reach the design and other methodological issues crucial to make a realistic picture of the research projects. Research integrity discourse could also be enriched by the insights from the analysis of risk in biomedical research.

The Concept of Risk and Responsible Conduct of Research Science and Engineering Ethics, Volume 12, Issue 1, 2006 83 Acknowledgements: I express my thanks to Professor Andrzej Gorski, organizer of the Warsaw conference, ‘The Responsible Conduct of Basic and Clinical Research’ (3-4 June 2005), where I presented the first draft of the paper as well as to the participants of the EU project EURECA* who commented on the later versions of my paper. * EURECA refers to the European project on delimiting the research concept and research activities; see http://www.eureca.manchester.ac.uk/.

REFERENCES 1. Rhodes, L.J. (2002) Beyond Conflict of Interest: The Responsible Conduct of Research, Science and Engineering Ethics 8(3): 459-468.

2. The Belmont Report: Ethical Principles and Guidelines for the Protection of Human Subjects of Research. (1979) Department of Health, Education, and Welfare, Washington, D.C.

3. WMA Declaration of Helsinki, Washington (2002), available on the Internet:

http://www.wma.net.

4. Additional Protocol to the Convention on Human Rights and Biomedicine concerning Biomedical Research (2005), Council of Europe, available on the Internet:

http://www.coe.int/bioethics.

5. International Ethical Guidelines for Biomedical Research Involving Human Subjects, CIOMS (2002), Geneva.

6. Sharp, H.M., Orr, R.D. (2004) “When “Minimal Risk “ Research Yields Clinically Significant Data, Maybe the Risks Aren’t So Minimal”. The American Journal of Bioethics 4(2): W32- W36.

7. Minimal Risk Review and Expedited Review (2005), Otawa University webpage, available on the Internet: http://web9.uottawa.ca/services/rgessrd/ethics/minimal_risk.asp.

8. Explanatory Report to the Additional Protocol to the Convention on Human Rights and Biomedicine concerning Biomedical Research [on line] (2005), Council of Europe, available on the Internet: http://www.coe.int/bioethics.

9. Gefenas, E. (2004) “Central and Eastern Europe: Research-related Problems for Transition Countries”, Ethical Eye: Biomedical Research, Council of Europe Publishing, pp. 121-132. 10. Levine, R.J. (1986) Ethics and Regulation of Clinical Research, 2 nd edition. Urban and Schwarzenberg, pp. 3-10.

11. Day, S. (2003) Dictionary for clinical trials. John Wiley & Sons, Chichester, p.172.

12. EMEA (1998) “Note for guidance on clinical investigation of medical products for the treatment and prevention and prevention of bipolar disorder’ (CPMP/EWP/567/98), available on the Internet: http://www.emea.eu.int/pdfs/human/ewp/056798en.pdf.

13. Goodwin, F.K. & Jamison, K.R. (1990) Manic-Depressive Illness. Oxford University Press, New York.