Empty ethics: the problem with informed consent
Oonagh Corrigan, Centre for Family Research, Faculty of Social and Political Sciences, Free School Lane, University of Cambridge, Cambridge CB2 3RF e-mail: firstname.lastname@example.org
Abstract Informed consent is increasingly heralded as an ethical panacea, a tool to counter autocratic and paternalistic medical practices. Debate about the implementation of informed consent is constricted and polarised, centring on the right of individuals to be fully informed and to freely choose versus an autocratic, paternalistic practice that negates individual choice. A bioethical framework, based on a principle-led form of reductive/deductive reasoning, dominates the current model of informed consent. Such a model tends to abstract the process of consent from its clinical and social setting. By fleshing out the social process involved when patients and healthy volunteer subjects consent to take part in clinical drug trials, this paper attempts to address the problem arising from the current ‘empty ethics’ model. My arguments are substantiated by qualitative interview data drawn from a study I conducted on the process of consent as experienced by participants in clinical drug trials.
Informed consent has gained increasing salience within the health care field. The need to secure a patient's fully-informed consent prior to medical intervention for treatment or research purposes is increasingly heralded as an ethical panacea counteracting the potential danger of paternalistic and autocratic practices. Solutions to recent medical controversies, such as the retention of children's organs at Alder Hey in the UK, concentrate on the need for patients to be fully informed about procedures and their potential risks/hazards (Dyer 2000). In the USA too, discussions of cases such as the much publicised death of a patient in a gene therapy trial (FDA and NIH 2000) and the death of a healthy volunteer subject who died as a result of taking part in a trial involving the inhalation of a chemical compound as part of an asthma study (Steinbrook 2002), tend to focus on the lack of information given to subjects regarding inherent risks, and breaches of the informed consent process (FDA and NIH 2000). Policy responses to such cases are inclined to result either in a tightening of existing informed consent procedures, or in the introduction of informed consent procedures where hitherto none had existed. The assumption underpinning the implementation of informed consent is that doing so will protect the rights and welfare of individuals by offering them the opportunity to make free and informed choices. In general the informed consent process is depicted as an antidote to counter medical paternalism and as such, a polar opposition has been established with the empowered, informed, autonomous decision-making patient or research participant at one end of the divide and an all-powerful paternalistic authority at the other.
The issue of informed consent has been discussed and debated from a number of positions within the medical field. It is rather surprising then that mainstream medical sociology has, by and large, posed very few questions and has not contributed much in the way of theoretical or empirical insight to this issue. In a report of 377 international articles and publications of empirical studies on informed consent (Sugarman et al. 1999) the vast majority were from the fields of medicine, law, nursing, bioethics and psychology.
There are a number of contributions from medical sociology that draw attention to the complexities of decision-making, and render the medical encounter problematic by discussing the ways in which patients are often dependent on medical expertise and advice. For example, many studies examining new reproductive technologies have revealed the complexities of ‘choice’ and the extent to which informed decision-making is often highly constrained (Lupton 1997, Jallinoja 2001). Nevertheless, much of the medical sociology literature establishes an oppositional model of health knowledge and experiences, juxtaposing lay knowledge, perceptions and experiences of illness with expert and professional approaches (Williams and Calnan 1996). This challenge to expert, autocratic and paternalistic approaches to medicine foregrounds a more active and emancipated form of patienthood. Thus, the model of the patient as an autonomous health consumer is supported and the ‘active decision-making patient/paternalistic medical authority’ dualism is reinforced. Furthermore, beyond the realm of medical sociology such a position is connected to more general social theories that draw attention to the concept of the ‘reflexive actor’ (Giddens 1990, 1992). Giddens’ work reveals an understanding of individuals as social actors constantly seeking to reflect upon the practices involved in constituting the self and the body and maximising and making rational choices for the benefit of the self.
The current theoretical underpinnings of the principle of informed consent have largely been derived from bioethics and, especially, research ethics. In particular, the growth of principlism within this field has proved very attractive for governmental regulatory mechanisms (Evans 2000). The increasing centrality awarded to the moral dictate ‘respect for autonomy’ has been observed by sociologist and bioethicist Paul Wolpe (1998) who notes that in the USA autonomy has triumphed in relation to other bioethical principles, having become progressively more important over time. The dominance of autonomy in bioethics is also a reflection of the increasing centrality being awarded to individualism within Western liberalism more generally (D’Agostino 1998, Rose 1999). Inside the bioethical frame, autonomy is presented as the ability to act freely without constraint or coercion. To quote from moral philosophers Tom Beauchamp and James Childress's influential biomedical ethics text:
. . . the core idea of personal autonomy is an extension of political self-rule to self-governance by the individual: personal rule of the self while remaining free from both controlling interferences by others and personal limitations such as inadequate understanding, that prevent meaningful choice (Beauchamp and Childress 1989: 68).
As critics of this form of bioethics have argued, such an understanding of informed consent is premised largely on the autonomous individual and his or her rights, with little or no conception of the social aspects (Fox and Swazey 1984, Light and McGee 1998, Wolpe 1998). This approach tends to reify the process of consent by stripping it away from its context and reducing it to a rational-choice model of action. Such an ‘empty ethics’ model presupposes that autonomous individuals when presented with adequate information and given time to assess it will subsequently make a conscious decision whether or not to participate. There are further limitations with this model insofar as it is premised on a universal standard principle, which not only reduces the significance of other ethical principles but ignores the cultural context within which the process of consent takes place. Problems have arisen, for example, about the acceptability of research to collect genetic samples from the population on the South Pacific island of Tonga because of the islanders’ opposition to the individual informed consent procedures, which they criticise for ignoring the traditional Tongan role of the extended family in decision-making (Burton 2002). Although, in the case of the Human Genome Diversity Project, researchers attempted to avoid opposition to the research by using group consent, this has been rejected by many indigenous populations as it fails to address critical social issues of group identity and community rights (Reardon 2001). Also, as Renee Fox (1984) highlights, in China, medical ethics emphasises social relationships and views the individual as a social community. In Chinese culture ‘. . . thinking in an entirely abstract or speculative way about moral or social questions runs the risk of what Chinese scholars have historically called “playing with emptiness”’ (Fox and Swazey 1984: 339). However, even within Western countries where individual autonomy is more highly regarded, it is important to gain a richer more meaningful understanding of the clinical context and the social and political aspects of the consent process. This view is also beginning to be expressed in the bioethics domain:
. . . despite broad agreement about the need to obtain informed consent, there is some uncertainty about how or whether meaningful consent is achieved in practice, whether theoretical understandings of informed consent are useful or practical, and what practices help enhance the possibility that patients and subjects in fact meaningfully consent to treatment or participation in research (Sugarman et al. 1999: 2).
This article begins with the premise that there is a need for more socially nuanced concepts of freedom, autonomy and consent. By examining the process of informed consent as experienced by participants in clinical drug trials, I argue that this process is not situated outside the realm of power, but rather such decisions are made in contexts where prevailing discourses and norms shape the field of freedom and choice. As Nikolas Rose claims, autonomy and choice cannot be understood as based on politically innocent premises, but as products of systems ‘in which subjects are . . . obliged to be free’ (Rose 1996: 17).
The ascent of consent in biomedical research
Whilst the implementation of informed consent has only recently begun to emerge in the UK as a significant desirable aim in routine clinical practice, it has a more established history in the area of biomedical research where it is currently a prerequisite. The considerations prevailing in biomedical research may differ from those of clinical practice more generally. Nevertheless, as informed consent has become such an established standard practice in the field of biomedical research, this seems an appropriate arena to examine its implications.
During the past 50 years, various international and national ethical guidelines have proliferated that enshrine the concept of informed consent as the principal code to be adhered to so as to protect the individual patient or healthy volunteer subject from possible exploitation and harm. Such regulations are conceptually linked to the discourse of human rights and autonomy. The Nuremberg Code (1947), which is cited as the first major international ethical statement to stipulate these principles in biomedical research, was constituted in direct response to evidence that emerged during the Nuremberg War Crimes Trials (the Doctors’ Trials) about horrific medical experiments conducted on prisoners. The Code consists of 10 principles but the major principle, which was to become the primary ethical consideration in all biomedical research, stipulates that ‘the voluntary consent of the human subject is absolutely essential’ (Annas and Grodin 1992: 2). A year later, a more general response to the inhumanities of Nazism was established in the Universal Declaration of Human Rights. Thus, the opportunity to consent or refuse consent was declared as a necessary human right based on the dignity and worth of every individual and on respect for his/her freedom. In 1964 the World Medical Association (WMA) produced the Declaration of Helsinki, a more comprehensive set of guidelines which, while further emphasising the principle of consent, also included a distinction between therapeutic and non-therapeutic research1. Despite these regulations, however, informed consent did not initially become routine practice. It was not until two whistleblowers, Pappworth (1967) in Britain and Beecher (1966) in the US, gave details of unethical experiments routinely conducted, often on marginalised minority groups without their knowledge or consent, that procedures to ensure that informed consent was implemented emerged. During the early 1970s steps were taken to regulate ethical practices in research with the establishment of Local Research Ethics Committees (LRECs) in the UK, preceded by the formation of Institutional Review Boards (IRBs) in the US. Multi Research Ethics Committees (MRECS) were an addition, established in the UK in 1996 as centralised Committees overseeing multi-centred research. The principal mandate for these committees was, and continues to be, the review of proposals to carry out research on patients or healthy volunteer subjects within the medical environment. These bodies act as gate-keepers to safe-guard the welfare of subjects in trials and ensure that informed consent is obtained from patients involved in biomedical research.
The most recent set of international policy guidelines promoting ethical standards in the context of clinical drug research has been produced by the International Conference on Harmonisation (ICH 1998), a pharmaceutical industry-led initiative aimed at standardising Good Clinical Practice (GCP) to provide a unified standard for the EU, Japan and the US. These guidelines promote an international ethical and scientific quality standard for designing, conducting, recording and reporting trials that involve the participation of human subjects and claim consistency with the principles that have their origin in the Declaration of Helsinki. Great emphasis is again given to the concept of written ‘fully informed consent’ as well as specific stipulations about the nature of information that is mandatory for the researcher to disclose to the subject. Patient advocacy groups have also welcomed the growing recognition of the need for informed consent. Through their meetings and publications, the UK charity patient group, Consumers for Ethics in Research (CERES), actively promotes the right to informed consent and inclusion of research subjects as ‘partners’ or ‘participants’ (CERES 1999) in clinical research.
In this paper, I examine what ‘informed consent’ means in practice and highlight the necessity to open up debate about consent beyond the current polar oppositions of autonomous decision-making and autocratic paternalism. My arguments are substantiated with qualitative interview data drawn from a study I conducted on the process of consent as experienced by participants in clinical drug trials. This article focuses primarily on the process of informed consent as experienced by patients and healthy volunteer subjects.
Interviews were conducted with patients, ‘healthy volunteers’, nurses and doctors involved in seven different drug trials that took place in five different clinical settings. In total, 26 trial subjects and seven doctors and nurses involved in the consent process were interviewed (see Table 1). In addition, documentation given to subjects prior to consent, such as written information sheets and consent forms, were analysed.
Table 1. Sample of patients involved in seven different clinical drug trials
|No. of Subjects||6||2||6||5|
|No. of Subjects||4||1||2|
In this article I focus primarily on interviews conducted with patients and healthy volunteers. At one end of the spectrum, the subjects chosen were healthy volunteers, and at the other were patients whose conditions could be described as acute, chronic, and possibly terminal. I chose this sampling frame in order to gain insight into the process of consent when subjects were experiencing various degrees of health and illness in different clinical contexts. The healthy volunteer trials were conducted at a phase 1 clinical trials unit located within a pharmaceutical company's premises. The other trials took place at four different general hospitals in England. The interviews were conducted between September 1997 and September 1998. Interviews were semi-structured and questions were centred on the following broad themes:
- • A chronology of events leading up to the signing of the consent form.
- • Principal reasons for participation in the drug trial.
- • Factors influencing decisions to participate.
- • Anticipated benefits, if any, of trial participation.
- • Anticipated risks, if any, of trial participation.
- • Comprehension of what the trial involves and procedures to be carried out.
To gain as valid accounts as possible, interviews were arranged so that they could take place soon after consent had been given. I was very conscious that not only were my subjects potentially being over-burdened as research subjects, but that some of the problems my research might uncover regarding the social process of informed consent might also apply to my study. I took steps to ensure that the subjects realised that I was not part of the clinical research team and that their participation would not have an effect on their medical care. I arranged to interview the breast cancer patients in their own homes as I anticipated that they were potentially a more vulnerable group, insofar as most of these women were elderly and awaiting surgery with a relatively poor prognosis. The other interviews were conducted in the clinics where subjects were engaged in the drug trials. In the case of the two elderly male patients who had experienced minor myocardial infarctions, these interviews were conducted at the bedside and I decided not to tape record them as a precaution against causing these patients any undue anxiety or annoyance.
Ethical guidelines direct those involved in obtaining consent to ensure that participation should be reached after the subject has reviewed and considered the information given, and freely chosen whether or not to participate. Written information detailing the possible side-effects and risks that could be incurred as well as information about the randomisation process and other aspects relating to trial participation was given to trial subjects prior to consent. In some cases subjects knew in advance that they were going to be asked to take part in a clinical trial. The healthy volunteers, for example, had approached the trial unit of the pharmaceutical company, often in response to advertising campaigns, as volunteers for clinical drug trials. Also, the patients with hypertension and benign prostate enlargement had been initially approached through their community physician who had forwarded a letter from the hospital physician in charge of the trial briefly outlining the proposed research. Generally, the trial subjects I interviewed had some difficulty in fully understanding the information they were given. As Ray, one of the patients on the benign prostate study, told me in reply to my question ‘how easy was it to understand the patient information form?’:
Erm . . . it was a bit complicated at first really, I did not quite get the full gist of it – you know? And then when it got to the point when it talks about insurance, I thought what the heck is going on here?
The randomised double-blind process is standard, preferred practice for drug trials. For ‘adequate understanding’ to take place, however, patients need at the very least to be aware why they are being randomised and, if there is a placebo arm involved, that there is a chance that they will not be receiving an active treatment. My study revealed that patients were sometimes unclear as to what was being tested, and whether or not the trial involved active treatment. For example, one patient thought there was a possibility that he was not getting an active substance ‘. . . they may be using sugar for all I know’. Yet the study information made no mention of placebo and indeed both arms of the trial involved active treatments. Also a patient who was on the hypertension trial was unclear about whether or not he could be receiving placebo:
There are two sorts that one can take . . . but whether . . . which one I am on I am not sure. Which one I am on I haven’t a clue, and whether I am not on any I am not sure. It didn’t say in the thing [information sheet] that I might be given nothing but I don’t know.
In this case too, all arms of the drug trial were active so he was in fact receiving some form of drug therapy, and this was explained in the information forms he had been given. In contrast most patients in the benign prostate drug trial believed they would be receiving active treatment when in fact one in four would be receiving placebo throughout the entire trial period. The following extract is typical of the responses elicited from patients on this trial.
OC: Do you know what your chances are of receiving the trial drug?
Patient: I know for certain I am on a drug but I don’t know what it is because they haven’t told me.
OC: But are you definitely on some form of drug or medication?
Patient: Yes, yes.
OC: What do you know about the different drugs being tested?
Patient: Well I know there are two, but they don’t tell you which you are on, which is fair enough.
Some of the confusion surrounding this trial may be due to its complexity. As previously mentioned, there were four arms of the trial, including two different doses of one drug, one dose of another and a placebo arm. The following extract from the patient information form reveals how this was presented to the patients:
You are being asked to participate in a research study to determine the effectiveness and safety of two different drugs commonly used in this complaint, alfuzosin and tamsulosin. . . . The aim of this study is to look at two doses of alfuzosin (10mg and 15mg) in a new once daily formulation in comparison with tamsulosin (0.4mg once daily) and placebo (a non-therapeutic substance which has no active component and which is completely safe). It is hoped that the new once daily alfuzosin formulation will be as effective and better tolerated than the currently available formulations involving two or three doses per day. During the first part of the trial which will last 4 months, you will have an equal chance of receiving one of these four medications.
It is not difficult to see how the last sentence mentioned in this extract might lead patients to believe they would be receiving ‘one of these four medications’. However, other studies have also reported that trial subjects poorly understand the process of randomisation in clinical drug trials (Cassileth et al. 1980, Jan and DeMets 1981, Snowdon et al. 1997). Only one patient on the benign prostate trial clearly understood that there was a possibility he could be receiving placebo. His understanding was based on his previous experience on a drug where half the patients had been assigned to placebo and, by the end of the trial he believed he had been in the placebo group. His reason for consenting to this trial was that his chance of receiving an active substance this time was much better. This highlights the extent to which an individual's understanding of the consent information is sometimes based on prior experiences and personal biography.
The majority of the information forms that I examined presented information on risk under headings such as ‘what are the risks involved in taking part?’ and ‘possible side-effects’. As ethical guidelines require that subjects be given information on all known foreseeable risk, information about relatively less serious hazards were often listed alongside more potentially harmful effects. On one of the trials, for example, information about the relatively minor risk ‘for some bruising or inflammation at the needle site’ resulting from blood sampling immediately preceded information that the study drug might induce crystals in the urine. Although most of the information given included some indication of the likelihood of these risks occurring in most cases, these were referred to in general terms such as ‘a small number’, ‘few, if any’ and ‘common side-effects’. The following extract from one of the patient information sheets about the drug tamoxifen is typical of the kind of terminology deployed.
A small number of patients experience hot flushes. Far less common side effects are vaginal bleeding, vulval itching, nausea, vomiting, light headedness, visual disturbance, loss of hair and fluid retention.
Only one of the information sheets included a rough quantification of the risk. It was presented as follows:
Serious bleeding that is likely to require a blood transfusion in patients receiving either heparin or enoxaparin and aspirin is expected to occur in about 3–4% of patients.
As only one of the subjects mentioned that they would have liked more detailed statistical information, it is difficult to assess whether the conflation of more benign forms of risk with the potentially more acute forms, and the generalisations with regard to quantification, made meaningful interpretation of possible side-effects more difficult. Nevertheless, it is easy to see how definitions such as ‘a small number’ are open to a fair degree of interpretation and understanding.
Ethical guidelines reflect the concern among bioethicists and research advocacy groups that subjects be adequately informed and that they have sufficient understanding of their involvement in the trial prior to consent. There has been a number of recent initiatives addressing problems relating to the clarity of written information presented to subjects (ICH 1998). Nevertheless, focusing on these issues alone simply reinforces the model of the patient or healthy volunteer as an active, rational decision-making agent. It is important to recognise that such decisions do not take place in isolation. Rather, individuals draw upon wider cultural perceptions of science and medicine, and information relating to the trial is interpreted against background stocks of prevailing dominant cultural beliefs and norms.
Cultural perceptions of drugs and scientific expertise
During the decision-making process, patients and healthy volunteer subjects often deploy tacit knowledge based on personal experiences and public perceptions about the wider context of medicine and scientific development. Despite the pervasive nature of drugs in our culture, there is very little literature and empirical research about lay beliefs concerning drugs (Blaxter and Britten 1996, Britten 1996). Studies that have been carried out in this area suggest that the public's attitude to medicines is somewhat ambivalent. In the Health and Lifestyle Survey (Blaxter 1990), 35 per cent of the respondents cited medical advances generally, and drugs specifically, as the reason why people are healthier now than in the past. However, other studies (Calnan 1987) have found that while there is optimism among respondents about some aspects of modern medical technology, scepticism is often expressed about the value of drugs. With regard to the perception of new or experimental drugs, the public's ambivalence appears even more acute. A study by Slevin et al. (1995) on patients’ attitudes to participation in cancer drug trials revealed that 72 per cent of respondents thought that the greater chance of getting new treatments made the prospect of participation in research very appealing, while only 27 per cent of the same group thought that this was the case when the same question was posed substituting the word experimental for new. A further study conducted by the Advisory Committee on Human Radiation Experiments (ACHRA 1995) in the US found that patients’ perceptions of potential harms and benefits varied depending on whether a project was called a ‘medical study’, a ‘clinical investigation’, ‘medical research’ or a ‘medical experiment’. The term medical experiment evoked the most striking and negative associations. In contrast, when the term ‘medical study’ was used, such research was viewed as less risky, as less likely to involve unproven treatments, and as offering a greater chance of medical benefit (ACHRA 1995). This highlights the influence of the wording of the consent text. None of the information sheets I examined included the term experiment when referring to the drug trial. The most common term used was study, a term loaded with more positive connotations.
A few patients expressed concern about the possibility of experiencing side-effects. One woman who had consented to take part in a pre-operative breast cancer trial told me that initially she spent a lot of time worrying about and looking out for the side-effects: ‘I was looking for all those side-effects. That was terrible! I wouldn’t like to go through all that again’. Another patient said he had asked for reassurance from the doctor in charge of the trial that the drug was not linked to a class of drugs from which he had previously experienced very bad side-effects. Generally, however, subjects did not express a great deal of concern about the possibility of experiencing side-effects. This general lack of concern expressed by both patients and healthy volunteers about possible side-effects is supported with evidence from other studies which indicates that, after giving consent, patients were unable subsequently to recall any side-effects listed (Hassar and Weintraub 1976, Bergler et al. 1980, Cassileth et al. 1980, Estey et al. 1994).
The data from my study reveal that subjects tended to believe that the drugs being tested were safe, effective and likely to be an improvement on existing alternative drug treatments. For example, one of the breast cancer trial patients told me that ‘. . . the new drug is reckoned to be more effective than the old one, so there didn’t seem to be anything to lose’. In reality, most drugs being tested in clinical trials do not reach the market place, as during trials many drugs prove to be less effective than products already available or they show toxic development defects (Spink 1980).
Rebecca Dresser's (2001) study of patient advocacy groups in the US highlights the crucial role played by the media in contributing to therapeutic misconceptions about the likely benefit to patients from their participation in biomedical research. Journalists often use extravagant language deploying terms such as ‘breakthrough’ to dramatise a story, and complicated research indications are often over-simplified, creating unrealistic expectations of cures. Gabe and Bury (1996) also suggest that the media play a central role in shaping public discourse about health risks. However, in their study of the controversy surrounding the benzodiazepine drug, halcion, they conclude that the media-fuelled controversy over this drug exemplifies a new public critique of medicines, a challenge to medical scientific expert authority, and a decline of trust by the lay public about the infallibility of such expertise. In contrast, the findings from my study suggest that for some, trust in medical and scientific expertise remains largely intact. One of the patients on the hypertension trial described how, in choosing to take part in the trial, he felt safe placing his trust in the progress of medical science:
I am sure they are not going to give me anything that will do me any lasting damage, and somebody has to do it otherwise there would never be progress will there?
The work of Nikolas Rose (1989, 1993) draws attention to the role of institutions of expertise and the way that individuals are obliged to locate themselves in relation to the norms of truth and health as expounded by such institutions. The issue of trust in systems of expertise is rather more significant in the case of healthy volunteers because trial participation is not linked to therapy but, instead, subjects are paid for their participation in such trials. As one of the healthy volunteer subjects told me:
I thought if there was anything wrong or anything that could be allowed to happen that would cause long term injury or anything, then they wouldn’t be allowed to do these sorts of studies. I thought it has got to be pretty safe for them to be allowed to do it.
The specialisation of expertise means that there is no one individual body of experts and, as such, expert specialists are themselves reliant on other experts. Several of the healthy volunteers, for example, were scientists working for the pharmaceutical company conducting the trials. I asked one of them, who had consented to a trial that involved the ingestion of a radio-active material, about the potential harm from being exposed to radiation. He told me:
I do know that the levels are very low. All these come into the low categories that the authorities allow. I handle radioactivity fairly regularly in my work so I am familiar with the precautions and effects.
While his experience with radioactive material made him familiar with its use, he also trusted that the ‘authorities’ had set safe limits. Another healthy volunteer told me that he felt reassured because ‘all the animal data has been reviewed by the authorities, the ethics committees and the various review boards’. Despite public controversies and media attention surrounding drugs like halcion and thalidomide, my interviews reveal the extent to which subjects’ explicit or implicit trust in expert systems prevail, and there is very little evidence of a conscious challenge to expert opinion on drug safety. As one of the interviewees told me, ‘I shouldn’t think they would let you have some drug they don’t know anything about’. Even when subjects expressed misgivings or apprehension about potential risk, they were still willing to trust the system. As one of the healthy volunteer subjects said:
I remember him talking about crystals developing inside you which I wasn’t very clear about. It sounded like it could possibly be dangerous if it occurred but it didn’t seem to be very likely . . . I would have preferred if you had some statistical information.
Despite this he said he trusted that this was generally a safe thing to do and he had not heard of ‘any kind of danger involved in all this’. Also one of the patients on the hypertension trial told me:
While I understand that pharmaceutical companies are motivated by profit, nevertheless they do appear to be serving the public as well.
The following extract from an interview with another of the hypertension study patients also reveals the sense in which he knew that he was an object of research and yet at the same time he had an implicit trust that he would not be harmed:
OC: What do you think about this being a new drug?
Patient: I should think it has been fairly well tested before they give it to us guinea pigs.
The patient's use of the term ‘guinea pig’ illustrates that he understands his role as an object of research in the clinical trial.
In clinical drug trials the province of trust not only involves an investment in abstract medical and scientific systems but, perhaps more significantly, trust is placed in particular individuals – in doctors and nurses conducting such research. In order to appreciate the social aspects of the consent process it is important to explore the pre-existing norms and expectations of care that prevail in the clinical research setting.
Clinical norms and the expectation of care
The current model of informed consent necessitates an equitable doctor-patient relationship based on mutual participation. In other words, for valid informed consent to be operationalised the doctor cannot coerce, persuade or direct the patient or healthy volunteer to take part in the trial. My study, however, suggests that patients and doctors bring pre-existing norms and values to the clinical trial setting that shape their expectations and direct their behaviour. Patients generally have expectations about the doctor-patient relationship and assume that the doctor is acting exclusively in the patient's best interests. One of the physicians in my study, a consultant surgeon who routinely conducted trials on breast cancer patients under her care, acknowledged this and explained:
Sometimes patients view you differently once you have asked them to go into a research study because all patients like to feel that you have got their best interests, their personal best interests at heart, which you have. But they then see that I am actually looking at a broader picture, as well, of which they are only part of that picture, and that shakes some of them slightly because they feel that everything is focused on them, which it is to a certain extent, but I also have to be able to stand back and look at the bigger picture, and they have not seen that before and sometimes that's the first time they have seen that, and that just shakes their trust and faith in me a little bit, and in the doctor, that very individual doctor/patient relationship. And that is difficult because sometimes you see that little look of betrayal just flit across their eyes.
As the breast cancer consultant reveals, the request by doctors for patients to take part in research can disrupt the norms and values of trust that prevail during the clinical encounter to the extent that the doctor may perceive the patient's feeling of betrayal.
When and if the prospective clinical drug trial is fully explained and understood by the patient, then he/she will need to accept that the best treatment option is unknown. Indeed ‘clinical equipoise’ is a pre-requisite to the conduct of any randomised controlled trial (RCT). According to this concept of ‘clinical equipoise’, the requirement is satisfied if there is genuine uncertainty within the expert medical community about the preferred treatment. The concept of ‘equipoise’ is based on present or imminent controversy in the clinical community over the preferred treatment (Freedman 1987). Disclosure of uncertainty with regard to what might be the most appropriate treatment course for the patient, coupled with the additional uncertainty that accompanies RCTs, can be disconcerting for both doctors and patients (Taylor 1992). A participant observation study (Taylor 1988) undertaken in a breast clinic at the moment of cancer diagnosis found that the disclosure of uncertainty was used to initiate discussion about clinical trials. As the following extract from Taylor's study indicates, patients did not always welcome this approach:
Experimenter: We don’t really know which surgery is best. We do not have any real answers. We are collecting data to help us with these questions. Let me tell you about this clinical trial . . .
Patient: Doctor, I am asking YOU what you think is best for me. For God's sake you are a doctor . . . I don’t want my breast off . . . but then I want to live . . . (Taylor 1988: 118).
When a patient is offered participation in a clinical trial, he/she is often looking for advice about the best treatment option and reassurance from the doctor about her/his condition. In such a context, the request to consent can be interpreted as guidance to consent. As one of the patients who had consented to take part in the post-operative breast cancer trial informed me:
I saw the doctor and she said would I like to go in for this new drug, and I said ‘I don’t know anything about it, it's up to you, if you think it will do me good, all right I will go on it’.
Bamberg and Budwig (1992), who conducted a discourse analysis study, found that ‘the voice of research is most likely to be interpreted by the patient/research subject within the framework of curing’ (1992: 165). Furthermore, they argue that misconception cannot be explained in terms of the actual information disclosed but is based on prior conceptions or beliefs about the different roles of research and health care. In another study (Bevan et al. 1993), 38 per cent of patients who had consented to participation in clinical trials stated that their motivation for doing so was to comply with the doctor's request. Furthermore, assumptions are often made by those involved in clinical research – physicians, nurses and prospective patients alike – that the intervention being studied is the best treatment option (King 2000).
Healthy volunteers on the other hand had no prior expectations of therapy and were therefore not consenting for such reasons. One man, however, who had been taking part in trials for a number of years, said he had initially taken part in a study as a favour to a colleague who needed to recruit trial subjects in the pharmaceutical company where they both worked. All but one of the healthy volunteers informed me that their primary reason for consenting was financial reward. The exception was a young man who said that his father had recently died from cancer. This prompted him to respond to a local newspaper advertisement for trial volunteers as he wanted to play a part in furthering scientific progress in the hope that cures for diseases such as cancer might be found.
The two patients in my study who had consented to take part in a trial for patients having recently suffered a minor heart attack informed me that they were distressed at being asked to join the trial in such circumstances. The first patient I interviewed informed me that he was not very happy about being asked to make a decision about the trial so soon after experiencing his heart attack. He complained that he was in quite a lot of pain, felt unwell, and that being asked to make a decision with regard to the trial was ‘too much’ to ask from him. He said that it was difficult for him to make a decision in such circumstances, and that he would not have had a problem if they had just gone ahead and carried out the process without his consent. Because it was such a difficult decision to make at the time, he did not consent immediately but asked the doctor for more time to think about it and to discuss it with his wife. This was granted and he gave consent the following day. Similarly, the other patient on this trial stressed that, while he did not wish to complain, he really would have appreciated more time before he was approached about the study. He said that he had undergone a traumatic event and was in turmoil, and that to ask him to participate under such circumstances was ‘not right’. In these circumstances both patients felt that the request to consent was burdensome and an intrusion in their care.
Field of ‘choice’
International guidelines for clinical trials specify that trial subjects must be informed about ‘the alternative procedure(s) or course(s) of treatment that may be available to the subject, and about their important potential benefits and risks’ (ICH 1998). As such, meaningful consent relies upon awareness by the trial subject of alternative courses of action that can be pursued. Although none of the subjects in my study felt that they had in any way been coerced or forced into consenting to take part in the trials, some of the patients were unclear about what their alternative treatment would have been had they not been involved in the trial. When I asked patients what were their alternative treatment options their responses varied. Patients who had previously been receiving medication, such as those taking part in the hypertension study, were mostly aware of options that they had insofar as they could continue on their previous medication, and they had experience and knowledge of that particular option. However, when I asked one of the patients on the pre-operative breast cancer trial whether she knew what her treatment would have been had she decided not to go on the trial, she answered:
No, they didn’t say anything about that. They just said, ‘would you like to go on this trial and would you like to sign for it’ and I said, ‘yes’ . . . But if I had said, ‘no, I wouldn’t go on this drug’, what would have happened? Would they have taken me in? That is what I wonder.
There was also some confusion over alternative treatment choices for patients on the benign prostate trial. While the men on the trial were aware that their alternative treatment would have taken place in primary care, most were unclear about what that treatment would have entailed. When I asked one of the patients on this trial whether he had asked the research nurse any questions about the trial he told me:
Yes, I did ask her ‘what are the alternatives?’ Yes, (he looks through the information sheet) yes, ‘watchful waiting’ that is it, I wasn’t quite sure what that was. She did tell me, but I assume if I don’t have one of the alternatives I don’t have anything. I am not sure.
My study reveals, however, that for other patients, such as those suffering from established chronic conditions, the disclosure by the physician of therapeutic uncertainty opened up a field of choice enabling the patient to be more active in decisions with regard to treatment options. For these patients, the chance to be randomly, blindly allocated to one treatment or another had a more positive, empowering effect. For example, many of the patients on the trials for a new hypertension drug described their decision in much more active terms:
I have never been fully satisfied that my personal case had been well investigated. As to the cause of my high blood-pressure, I am not sure as to whether I do indeed have high blood pressure. And connected with that I was not totally happy with the treatment I was receiving anyway with some of the side effects. So I consented because I wanted better treatment and better diagnosis.
Another patient on the hypertension drug trial informed me:
I was interested and thought there were certain personal benefits to me, like a check up on my general health and blood pressure. That was why I came along in the first place.
All the patients on this trial mentioned the benefit of extra monitoring and some specifically welcomed the chance to try a new drug treatment that might potentially have fewer side-effects. Drugs for hypertension frequently cause unpleasant side-effects and often it is a process of ‘trial and error’ before a patient finds a drug which suits them. Furthermore, it is interesting to note that for both the hypertension patients and those with benign prostate conditions, trial participation involved being treated by hospital consultants/surgeons. Had they not agreed to consent to the trial, their condition was such that they would otherwise have been under the care of their local GP. Nevertheless, despite the tendency for these patients to view their participation in the trial in active terms some of them also cited a willingness to please the doctor as part of their decision to consent.
Undoubtedly, my study caused me to deliberate and reflect upon my own practices with regard to obtaining the informed consent of my informants. For the most part, I followed the same standards and procedures I have come to critique. I gave my informants written consent sheets explaining the reasons for my study, the proposed interview procedures, and what would happen to the interview data once the interviews were completed. I explained the main purpose of the study and asked them to sign consent forms and tick boxes to verify whether they were willing to have extracts of their interview transcripts published. I sought to differentiate my own research from the clinical drug trials, explaining that I was a university researcher and not in any way formally connected to the clinic. Nevertheless, I gained access to patients by first soliciting assistance from the doctors in charge of the trials who then passed on to their patients a standard letter that I had written, and they also asked the patients’ permission for me to telephone them. Furthermore, apart from the interviews with the breast cancer patients, the others all took place in the clinical setting. As such, my respondents may not have formed such a clear distinction between my research project and their treatment so that their consent to my study may also have been influenced by their willingness to please their doctor. In the case of the healthy volunteers, I was invited to be present when the doctor in charge of the study informed them about the trial. Interestingly, although the healthy volunteers were given individual consent forms, this was carried out in a group setting where they were also verbally informed about the trial by the doctor. Furthermore, my interviews revealed that for healthy volunteers the informed consent procedure for the trial was very much a formality in that having first volunteered to take part in paid research they were unlikely to have refused consent at that stage. Again, as I interviewed the healthy volunteer informants in the clinic, between tests during the time they were taking part in their trials, my presence may have been associated with the drug trial. Furthermore, as the patients and healthy volunteers were already research subjects, it is possible that some subjects could have felt bombarded by the research process. I have to concede that my research might occasionally have had negative consequences. For example, although I believe I acted sensitively, was polite and stressed the voluntary nature of their participation in the study, the mere act of approaching them, whether by telephone or face-to-face, could have been perceived as ‘bothersome’ or intrusive. I found that while most of the pre-operative breast cancer patients were willing to take part in an interview, asking them to commit to a ‘suitable’ time was not always easy for them as so much uncertainty and anxiety surrounded their condition. There were occasions where appointments had to be rearranged because of hospital treatment, and one of the patients decided to pull out because she had to go into hospital for surgery earlier than anticipated. Nevertheless, a few of the patients revealed that while they had not entirely welcomed the prospect of an interview with a stranger, they had found the interview process itself enjoyable, and one of the breast cancer patients has since written to thank me. The interviews with the breast cancer patients lasted much longer than did the others, and we usually chatted over a cup of tea or coffee before or after the official interview session. The narrative style of these interviews was different from those of the other trial subjects in that they spent a lot more time telling me of their fears and concerns about their condition in general, and how they and their families were coping with the knowledge that these women had a relatively advanced form of breast cancer. As I felt that the presence of a tape recorder might be unwelcome, in all but one trial I asked subjects whether they minded if I tape-recorded the interview and the majority voiced no objection. I did not, however, ask the patients taking part in the post myocardial infarction blood thinning trial. As these patients were in bed on the hospital ward, linked to a drip and coming to terms with the shock of having suffered from a recent heart attack, they were particularly vulnerable; I wanted to make the interviews as easy as possible for them. In this case, and in two other cases where subjects preferred not to be tape-recorded as they stated that it would make them feel nervous, abbreviated notes were taken during the interviews and were written up later the same day.
A number of sociological studies have examined moral decisions made by patients, doctors and nurses in the cultural context of their clinical settings (Anspach 1993, Mueller 1997). These authors draw attention to the way that the organisational and institutional settings impinge upon moral and ethical decisions and reveal that patients often become objects of the bureaucratic machinery. Allowing patients more time to consider their initial decision to consent, understanding consent as a process as opposed to a ‘once and for all’ act, and presenting information in a more accessible manner may bring some improvement to the consent process. Nevertheless, most clinical trials, and drug trials in particular, are subject to rigid study design and protocol requirements. That said, there are some new initiatives being tested (Donovan et al. 2002, Snowdon et al. 1998) that are attempting to improve the study design of randomised clinical trials so that patients better understand them, thereby making informed consent more valid according to its own terms of reference. Nevertheless, given that most clinical drug trials rely on standardised rigid protocol, there is very little current scope for a flexible methodology ethically sensitive to the needs of patients and healthy volunteers. Sociologist Renee Fox (Fox 1996) notes that in 45 years as a participant observer of patient-oriented clinical research she has witnessed a move away from ‘patient oriented’ research where the physician-researchers had more of a free rein with regard to the design of the experiment and where scientific research goals and standards were sometimes compromised for the sake of clinical care, to a more standardised research form characterised by an intellectual demise of patient-oriented clinical research.
Discussion and conclusion
It would be unfair to give the impression that the ascendancy of informed consent has been universally accepted within the medical and ethical realm. Indeed, the history of informed consent suggests that physicians often initially resisted its adoption. Debates about informed consent among ethicists and medical practitioners but these are narrowly focused on the relative merits of informed consent, which does not lead much beyond the paternalistic/autonomous decision-making opposition. Arguments that focus on informed consent as an absolute moral principle result in a reductionist abstraction and an empty ethics that strips the principle of consent away from its social context. On the other hand, arguments that plead for the recognition of the limits of consent in certain contexts argue for a more paternalistic approach. A summary of the viewpoints of three influential contributors to an on-going debate on informed consent that took place in the British Medical Journal illustrates my claim2.
First, Len Doyal, professor of medical ethics, argues that the moral importance of informed consent must be upheld and that ‘what is important here is our shared belief in the moral imperative of respecting human autonomy in almost all circumstances’ (Doyal 1998: 1000). As such, he adopts a fairly rigid, absolutist and universal perspective on the need for informed consent (Doyal 1998).
Mary Warnock, former chair of the Committee of Inquiry into Human Fertilisation, whilst generally wishing to see the moral principle of informed consent upheld, acknowledges that the concept is complex. Although Warnock is against its absolutist status and is in favour of allowing the principle to be breached in some cases, such as those interventions involving the ‘mentally incompetent’, she maintains that consent is the most important moral principle governing the use of human subjects in biomedical research (Warnock 1998).
Finally, from the perspective of Jeffrey Tobias, a consultant physician in radiotherapy and oncology, such rigid positions ignore the reality surrounding the difficulties of obtaining informed consent in practice. Tobias believes that in certain circumstances, asking a patient for his or her informed consent can be ‘needlessly cruel’. He claims that for the many patients who are ‘sophisticated’ (sic) and ‘well informed’, it is possible for an informed dialogue to take place between doctor and patient about the prospective trial. Tobias argues, however, that for the majority of patients who are ‘less educated, less well informed, and less able to marshall their arguments – a somewhat more directive or (without being pejorative) “paternalistic” approach will be far more appropriate and gratefully received’ (Tobias 1998: 1002).
The debate surrounding informed consent seems to be limited to arguments put forward by those who believe in its moral absolute status, those who believe it is a very important moral principle but that in some circumstances it would be ethical to breach the principle, and those who think it is largely unworkable and should be abandoned in favour of a more paternalistic form of medicine where doctors decide whether or not patients need to be informed. More recent discussions on informed consent have focused on the need to see informed consent as an on-going process rather than as a discrete act of choice that takes place in a given moment of time. Such an understanding, ‘when applied’ in practice, can open up the process of the trial itself, permitting the patient or healthy volunteer subjects to withdraw at any point during the study. My study also suggests that we need to look beyond consent, as one of the consequences the current focus is that it obfuscates other issues that arise in clinical trials. Many of the patients I interviewed, such as those with long-term chronic conditions, were unsure what would happen to them at the end of the trial. If the drug proved to be effective in treating their condition would they be able to carry on receiving it? Also, once the subject's participation in a trial has terminated, she is rarely informed about which drug treatment, if any, she had received on the trial, nor is this information subsequently routinely added to her medical notes. Subjects are not informed about the eventual results of the trial. While it is appreciated that clinical drug trials take years to complete, and this may mean that subjects entering into clinical trials at the beginning of the research may have to wait before the results are known, this does not seem to preclude the imparting of this information to the subject. These issues are coupled with the lack of transparency in general surrounding the licensing and testing of new drugs in the UK. Although the majority of drugs that enter into clinical trials do not reach the market, there is no compulsion to disclose trial results to the general public and it is therefore not known why these drugs fail.
While my study suggests we should be more cautious about the role of informed consent as an ethical panacea, I would not like to see a return to a more paternalistic approach where physicians in charge of the trials make decisions about trial participation. However, we need to broaden the debate on informed consent from its current tight and limited focus. Informed consent is an important ethical tool that protects subjects from overt coercion but we cannot ignore the often-dependent relationship between patient and doctor. Informed consent is premised on an equitable doctor/patient relationship that cannot always be fully realised. My study shows that for some patients, mainly those whose condition is of a mild and chronic nature, the informed consent decision-making process can open up the field of choice. For others, especially those who were seriously ill, the experience of being invited to take part in clinical drug trials was burdensome. Asking patients to take part in research can disrupt the doctor/patient relationship and challenge patients’ expectations of treatment and care. Patients frequently have over-riding emotional and physical needs. Depending on their condition and stage of ill-health, they often experience pain, shock and anxiety and it cannot simply be assumed that the imposition of the informed consent process will necessarily bring about an equitable doctor-patient relationship where patients make active choices. The trial subjects I interviewed all thought that the new study drug was likely to be an improvement on existing alternative drug treatment. Even in the cases of healthy volunteers, rather than patients hoping for a cure, issues of the subject's trust in scientific expertise are brought to the fore. My findings suggest that drug trial subjects place a high degree of trust in the doctor and the medical staff involved in administering the trial as well as the science behind it. This undermines claims made by Gabe and Bury (1996) of a new public critique of medicine, a challenge to medical scientific expert authority, as well as Giddens’ more general claim that expert knowledge is now chronically contestable and peoples’ trust in such expertise is in decline.
The implementation of a ‘valid’ informed consent process is more complex and difficult to attain than bioethics guidelines and policies imply. I use the term ‘valid’ here to mean valid within the terms of reference contained in bioethical guidelines. There needs to be a realisation that the type of illness a patient is suffering from, her anxiety about the likely trajectory of her illness, her expectations about treatment and, in general, her implicit trust in the doctor and medical science mean that ‘informed choices’ based on an adequate understanding of the information and on careful consideration of the potential benefits and risks, are difficult to achieve in practice. Furthermore, not only is the concept of informed consent problematic within its own terms of reference, but ideas of autonomy, freedom and choice belie the extent to which they are both limited and regulated. The dualistic opposition between liberal concepts of freedom and autonomy versus powerful autocratic medical practices fails to recognise that power is not just a phenomenon that is exercised as an external constraint, but that prevailing cultural norms, values and systems of expertise shape the field of choice.
This work is based on part of my PhD thesis, which was funded by the Fuller Foundation, Department of Sociology, University of Essex. I would like to thank my colleagues who have read earlier drafts of this paper and the three anonymous referees for their helpful comments.
The Declaration of Helsinki was updated and revised in 1975, 1983, 1989, 1996 and 2000.
The ‘education and debate’ feature on informed consent followed a year's ongoing correspondence in the BMJ about the topic, following an invitation by the journal for correspondence about the limits of informed consent.