Monash Bioeth. Rev. (2015) 33:64–77 DOI 10.1007/s40592-015-0029-8 ON GAMETES AND GUIDELINES

Secondary use of empirical research data in medical ethics papers on gamete donation: forms of use and pitfalls Veerle Provoost1

Published online: 3 April 2015  Monash University 2015

Abstract This paper aims to provide a description of how authors publishing in medical ethics journals have made use of empirical research data in papers on the topic of gamete or embryo donation by means of references to studies conducted by others (secondary use). Rather than making a direct contribution to the theoretical methodological literature about the role empirical research data could play or should play in ethics studies, the focus is on the particular uses of these data and the problems that can be encountered with this use. In the selection of papers examined, apart from being used to describe the context, empirical evidence was mainly used to recount problems that needed solving. Few of the authors looked critically at the quality of the studies they quoted, and several instances were found of empirical data being used poorly or inappropriately. This study provides some initial baseline evidence that shows empirical data, in the form of references to studies, are sometimes being used in inappropriate ways. This suggests that medical ethicists should be more concerned about the quality of the empirical data selected, the appropriateness of the choice for a particular type of data (from a particular type of study) and the correct integration of this evidence in sound argumentation. Given that empirical data can be misused also when merely cited instead of reported, it may be worthwhile to explore good practice requirements for this type of use of empirical data in medical ethics. Keywords

Bioethics  Empirical ethics  Methods  Descriptive analysis

& Veerle Provoost [email protected]; [email protected] 1

Bioethics Institute Ghent, Department of Philosophy and Moral Sciences, Ghent University, Blandijnberg 2, 9000 Ghent, Belgium

123

Secondary use of empirical research data in medical ethics…

65

1 Introduction A decade ago, Borry et al. (2005), reported ‘the birth of the empirical turn in bioethics’. In their paper, the authors gave several reasons for ‘the clouded relations between the empirical and normative perspectives on bioethics’ in the past. Although, they stated, bioethics has increasingly accommodated empirical research, the difficulties regarding the role of empirical evidence in bioethics remain. Given the difference between the descriptive and the normative, numerous authors have discussed the methodological questions of integrating empirical research findings in normative ethics (for example: De vries and Van Leeuwen 2010; Dunn et al. 2012; Frith 2012; Hedgecoe 2004; Molewijk et al. 2004; Van der Scheer and Widdershoven 2004). Most of these methodological papers have focussed on empirical studies set up, conducted and reported by (bio)ethicists with the purpose of supporting their normative analysis. This paper doers not engage directly with the methodological literature, but instead focuses on the way medical ethicists make use of data collected by others. Rather than attempting to provide a typology or description of ways in which empirical data can inform bioethics (as others have done, for example: Borry et al. 2004; Hope 1999; Holm and Jonas 2004; Kon 2009; Solomon 2005; Sugarman and Sulmasy 2010), this study aims to provide an initial descriptive account, of the way in which authors publishing in medical ethics journals have made use of empirical data reported by others. This is done by undertaking a literature review of papers published in medical ethics journals on the topic of embryo or gamete donation, and exploring how they made secondary use of empirical data.

2 Method This review was initially undertaken for a presentation at the Brocher Foundation workshop ‘On Gametes and Guidelines: A symposium for exploring the use of interdisciplinary and empirical bioethics to inform the regulation of reproductive technologies’ (Ives and Provoost 2015), and as such it was limited to arguments about embryo and gamete donation. Aside from this purely pragmatic reason, the literature on the ethics of gamete and embryo donation provides a good case study for an analysis of the use of empirical data in ethics papers, given that the ethical debates around gamete donation have long been influenced by various kinds of data (including data about safety, efficacy and stakeholder views and preferences). A search was conducted for papers on this topic, written in English and published between 2001 and 2013 in the eighteen journals categorized under ‘medical ethics’ in the Journal Citation reports of the ISI Web of Knowledge in 2013. The aim of the search was to explore different ways in which empirical data are used and to identify examples to illustrate these uses. The search identified 47 papers that contained clear references to empirical research, including research manuscripts as well as commentaries, editorials and letters to the editor. Among this selection were

123

66

V. Provoost

two clusters of articles, relating to two target articles from the American Journal of Bioethics (Gurmankin 2001; Ravelingien and Pennings 2013) and their respective commentaries. One of the target articles reported an empirical study (Gurmankin 2001). That paper and its commentaries provided a particularly rich source of information as it offered an excellent opportunity to see how different authors responded to the same empirical study, interpreted the findings, and made use of them for their own normative arguments. References were studied in detail to determine whether they included empirical research (defined as systematic collection and analysis of data about a particular phenomenon, using methods with roots in the social sciences or medical research, with the aim of generating, analysing, and interpreting reliable and valid data). This broad definition is based on the definition of Borry et al. (2006). This definition allows us to include references to papers presenting data generated from many different types of studies (surveys, qualitative interview studies, clinical trials, laboratory science etc.).1 The 47 papers that made clear reference to empirical research were subjected to an iterative cycle of defining and refining types of usage and problems with that usage. The analysis focussed on the following: (1) the role or function of the empirical data referred to and how, if at all, they were used in the argument; (2) the problems or challenges identified in connection therewith and; (3) the authors’ general considerations or reflections about the need for, or the role of, empirical data in (bio)ethics and their calls for empirical research. Analytic codes and categories were derived inductively from the texts (for example, looking at different ways empirical data were used), constantly comparing emerging categories with both previous and subsequent texts. The emerging codes and categories were discussed with two other researchers, one of whom also confirmed the accuracy of the categories by means of an examination of a random selection of papers. Throughout the results section, quotes from the papers were used to illustrate the findings.

3 Results The studied papers’ references to empirical data varied from one citation in a single (foot)note in the introduction, to several citations in relation to central claims spread over the greater part of the paper. The types of studies referred to covered clinical or laboratory data (mostly to evidence medical risks) as well as the entire spectrum of social science studies, including surveys with stakeholders and the general public and qualitative (interview) studies. Not all authors used the data in support of a normative argument. Empirical studies were often introduced to help describe the background of a problem, a situation, current policies or recent policy changes. This was typically done in the introduction and often in (opening) paragraphs where 1

This is consistent with the way ‘empirical’ is used in the ongoing debates around empirical bioethics. Legal research (involving the use of case studies), as well as medical or fictional case studies were excluded from this definition, as were narrative accounts of medical techniques, descriptions of the history of medical treatments, or of events, policy, law or regulation, as this kind of research did not fit the definition of ‘empirical’ being used.

123

Secondary use of empirical research data in medical ethics…

67

reference was also made to accepted facts that established the current ‘state of the art’. Other authors used the data as a springboard for their normative analysis, but did not give it a central role. For the most part, arguments that relied on establishing facts by citing empirical data did so in order to establish the existence (and level) of risk or (potential) problems associated with certain practices or technologies. In some papers, empirical data were used to support proposals for solutions, thereby assuming or accepting the fact that there were problems. In all cases where problems were spelled out or presumed, the data were linked to normative assertions, although neither this link, nor the normative assertions, were always clearly spelled out. 3.1 Evidence used to recount problems that need solving Most authors used empirical data to note or describe a problem, often very clearly. Wilkinson (2012), for instance, referred to medical research on ovarian hyperstimulation syndrome (OHSS) when he argued that this is a risk that egg sharers are confronted with and are willing to take anyway (for their own treatment), while for donors this is an additional risk. When pointing to a problem, the authors typically made a normative claim, although often not explicitly. For instance, by pointing at the risks for oocyte donors, an author could implicitly endorse the normative claim that such risks should be avoided. In this sense, the mere stating of evidence of a risk has both an empirical and normative function. The risk is stated and evidenced, and the normative implication is that the risk ought to be avoided. 3.2 Evidence used to support a normative argument When used as part of a normative argument, empirical data were used to evidence factual premises—what Ashcroft (2003) referred to as ‘completing the hypothetical imperative’. In the case of Afshar and Bagheri (2013), for instance, the data played an important role in their argument. They referred to empirical data showing that prolonged embryo storage did not significantly decrease pregnancy or implantation rates, in order to challenge safety concerns related to such prolonged storage. This was an important element in their argument for the implementation of follow up tests of couples who donated surplus embryos 6 months after in vitro fertilization. In many of those cases, however, the argument (including the reference to empirical data) did not fit (well) into straightforward logical structures, such as the hypothetical imperative. Often, the data played a role in a secondary line of thought or as part of a larger, more complex normative argument. Ravelingien and Pennings (2013), for instance, provocatively argued that when the current shift toward open-identity policies in gamete donation is based on the normative claim that donor-conceived children have ‘a right to know their genetic forebears’, this right should be granted to all children, by means of a routine obligatory paternity test at birth. The aim of their paper was to show that, when one accepts the child’s right to knowledge about its genetic parentage as a fundamental right, the same concerns ought to apply in both donor conception and non-donor conception contexts. At one point, the authors

123

68

V. Provoost

acknowledged that the social discomfort for the child would be greater in the case of disclosure of misattributed paternity compared to disclosure of donor conception. This is brought up as a potential counterargument for their position. They continued by referring to empirical evidence about the effects of the timing of disclosure. This evidence is used to argue that in the same way as has been done in the context of disclosure of donor conception, one could rely on the role of counselling to address offspring’s difficulties with this information to minimise the negative impact of disclosure of non-genetic parenthood on the child’s self-esteem and adjustment. Likewise, evidence about the effect of the timing of the disclosure in the context of donor conception was referred to: ‘Offspring who were told during adolescence or early adulthood report significantly more negative feelings toward donor conception than those who were told at an earlier age…’ (Ravelingien and Pennings 2013, p. 38). The reasoning of the authors is that the analogy holds and a similar procedure could be applied to effectively reduce the negative impact of disclosure in both cases. Sometimes, conclusions were drawn that were not (or not clearly) linked to the data (see below, for the example of Fox 2008). In other papers, authors drew strong (and general) conclusions based on limited data. Based on the study of Gurmankin (2001) and her own qualitative follow-up study of 33 oocyte donors, Kalfoglou (2001, p. w2), for instance, writes: ‘The bottom line is that there is growing evidence that the conflict of interest between a donor and physician can lead to many questionable practices.’ The two studies referred to in support of this conclusion were based on small samples and the study design of both studies (focussing on preliminary phone calls of potential donors in Gurmankin’s study and on donors’ experiences 3 years retrospectively in her own study) did not allow for generalising conclusions. Also, neither study could provide a basis for inferences about a causal link between a (supposedly) proven conflict of interest and the ‘many questionable practices’ this would lead to. There was, for instance, no evidence to suggest that the issues raised by the interviewees—such as promises about aspects of care not being kept—were typical for the physician–donor relationship and would not occur in a physician–patient relationship. Also, the author chose to stress the negative instead of the positive findings, such as the fact that many participants were satisfied with the experience of being a donor and were willing to repeat it, or the fact that most women, 3 years after the donation, were able to recite many of the risks (as part of the information they received at the time). It is indeed a problem to find that a woman remembers being told that there were no risks by the staff of a clinic where she would donate her oocytes, but the question here is whether this finding, as part of a qualitative study, is sufficient evidence to conclude that there is a generalised conflict of interest potentially leading to questionable practices. 3.3 Evidence used to problematise a normative claim Another way of using empirical data was to make or assume a normative point (but not always argue for it) and then use data to problematize it. In the papers examined in this review, this was generally done in two ways: (1) by showing that practice

123

Secondary use of empirical research data in medical ethics…

69

was not in accordance with moral principles, rules or rights and (2) by showing that there was a problem with moral principles or rules. In the first type of reasoning, empirical data were used to show that a practice was not in accordance with specific established moral principles or rules (empirical claim), while it should be (normative claim). For example, Frith (2001) stated that ‘serious inconsistencies exist between the claim that donor offspring have a right to know the identity of their gamete donor and the way such a right is currently met in practice’ (Frith 2001, p. 473). To substantiate this, she discussed law and regulation recognising this right and referred to empirical studies that showed that large percentages of donor insemination parents planned not to tell their child about how it was conceived. In the second type of reasoning authors came to the conclusion that there were problems with the moral rules and principals themselves or conflicts between them that should be resolved differently. Also these statements combined an empirical claim with a normative claim. For example, based on the findings of the study of Gurmankin (2001), Berg (2001) pointed to a ‘limitation of the legal doctrine of informed consent’ that would preclude a true informed consent. In several of these extracts authors problematised moral principles or rules as well as the implementation of those rules and both categories of problematising were not clearly distinguishable. According to Adams (2013), for example, evidence from studies of adult offspring’s perceptions that showed that some offspring ‘are distressed by procreative choices that their parents were able to make’ (Adams 2013, p. 377) provided support for a ‘frame-shift’ of reproductive freedom: limiting this freedom in agreement with considerations regarding the welfare of the child. The empirical claim, for which several studies were referred to, was that ‘the majority of offspring believe that they should know the identity of their donors’ (Adams 2013, p. 373). According to the author, this demonstrated, that offspring themselves supported such a frame-shift. He stated: ‘Increasing reproductive freedom in these instances is diametrically opposed to an improvement in the conceptualised welfare of the offspring’ (Adams 2013, p. 378). 3.4 Vague empirical claims In several papers, vague empirical claims were made, in the style of ‘it is not uncommon….’ (Weinberg 2008, p. 175) or ‘it is very probably the case that’ (Price 2013, p. 531). Weinberg (2008) used several such claims, for example: ‘I suppose it is true that, generally speaking, parents who want children tend to make better parents than people who have children that they did not really want to have’ (Weinberg 2008, p. 176). These kinds of claims were not supported by references to empirical studies, but were empirical assumptions necessary to make an argument. That they are made without clear justification or evidential support may or may not be problematic, depending on their place in the argument. Although these kinds of claims are noteworthy, and potentially problematic, they are not the focus of this paper. Rather, I focus on instances where an author makes an empirical claim and supports it by citing empirical evidence.

123

70

V. Provoost

3.5 Critically assessing the evidence In the sample of papers used for this review, instances of critical assessment of empirical studies were mostly limited to commentaries to the Gurmankin (2001) paper, which between them criticised many different aspects of the empirical study. The study used preliminary phone calls made by a potential oocyte donor (who was really the researcher) to 19 centres to evaluate the risk information provided during that first phone call. It found that in the majority of these phone calls incomplete and/or inaccurate risk information was provided. These commentaries, as a very specific type of paper (where authors are invited to comment), are unlikely to be typical of medical ethics papers, but they nonetheless provide an excellent opportunity to explore the different ways in which medical ethics papers can critically engage with empirical data. The study in Gurmankin’s (2001) target article was criticized for being based on a small sample size. Authors who criticized this pointed to the limitations this brought along for statements about the significance of the problem, as identified by the author, and the generalizability of the findings (Brzyski 2001; Strong 2001; Stock 2001) specifically in relation to the author’s assumption of how centres operate (Brzyski 2001; Cohen 2001; Mastroianni 2001; Sauer 2001; White 2001). It was criticized for the way the concept of informed consent was operationalized (Brzyski 2001), for the choice of study variables (Brzyski 2001), and for its use of deception (Dresser 2001; Cohen 2001; Strong 2001). Critical remarks were also made about the presentation and the interpretation of the results (Cohen 2001), for instance for focussing too much on programs that fell short and not enough on programs that did a good job (Strong 2001). One author explicitly complained about such a lack of connection between the empirical data and the normative claim (recommendations based on the study): ‘The author’s final suggestion […] is so weakly connected to her data that it seems like a preconceived point she has grafted to her discussion’ (Stock 2001, p. 27). Overall, a study that was criticised was not necessarily condemned in full. Some authors still saw merits in these studies for drawing our attention to a problem. Others, however, thought that despite their strong critique of a study’s design, it still correctly identified shortcomings of a current practice (Sauer 2001). When this was the case, an empirical study was used for identifying particular problematic cases and providing interesting starting points for normative reasoning, rather than being used for numerical or theoretical generalisation. However, in several cases, there was still confusion about the type of conclusions that could be drawn from a particular empirical study, and there was a tendency to draw more general conclusions than the study (which had just been criticised) would allow: e.g. in reference to the Gurmankin study, ‘the size of her sample, though small, may be representative of the status quo or the state of the art, and that’s just the problem!’ (White 2001, p. 30). 3.6 Problems with the secondary use of empirical data Several things may go wrong when conclusions are drawn from empirical observations (in the studies cited) and an evaluation of these observations (a value premise). Below, some of the problems and pitfalls identified are described.

123

Secondary use of empirical research data in medical ethics…

71

One type of problematic use of empirical data is when an empirical claim is made that is not, and cannot, be supported by the study that is cited. Authors may, for example, simply generalise too far, or inaccurately report the findings, due to a lack of critical engagement with the original study. For instance, Berg (2001) argued that Gurmankin’s data (2001) pointed to a limitation of the legal doctrine of informed consent because potential donors could become significantly involved in the process of donation before receiving risk information. According to Berg, Gurmankin’s data provides ‘further evidence of the oddity of using an event model of informed consent (i.e. sign the consent just before undergoing the procedure), rather than a process model that promotes involvement at all stages’ (Berg 2001, p. 18). However, neither the level of the potential donor’s involvement in the process (at the time of the preliminary phone call) nor the type of model used for informed consent were measured by this study. Nothing was known about when or how long after the phone call, if at all, potential donors would have signed the consent form, nor about how much time there would be between signing the form and the start of the procedure. The study, which was the only empirical study cited by the author in relation to this argument, in fact examined only the preliminary telephone call of a potential candidate to the centre. In essence, the empirical claim made by Berg could not be supported by the Gurmankin study. Most authors in the sample did not appear to critically assess the empirical studies that generated the data they used. We might question to what extent this is a task of a bioethicist, but arguably, at least in some papers, looking more critically at the empirical studies could have improved the appropriateness of the conclusions. For instance, Adams (2013) cites Jadva et al. (2009) when he concludes: ‘‘When donor-conceived people have been asked whether they believe they should be told the truth of their conception, the majority do feel this way (Jadva et al. 2009), suggesting that not only should openness be practiced as a matter of principle but also that it is how donor-conceived people wish to be treated’’ (Adams 2013, p. 373). The study cited in support of this claim (Jadva et al. 2009), however, was an online survey that was completed only by people actively searching for their donor siblings [they were members of the donor sibling registry (DSR), a registry that helps donor conceived people search for and contact their donor and donor siblings]. In the paper about the study (Jadva et al. 2009), it was explicitly noted that ‘the sample may not be representative of all donor-conceived individuals’ and, moreover, ‘may not be representative of all members of the DSR’ (Jadva et al. 2009, p. 532). In Adam’s citing of the study, however, this limitation was lost, and empirical generalisations were made that the cited paper did not actually support. Another problem relating to the secondary use of empirical data is when these data are used in an ambiguous way. An example of this was found in a paper by Fox (2008), arguing against payment for gametes. Although empirical evidence did not play a clear role in the argumentation, the wording of the presentation of the data suggested a clear and essential role. For instance, when the author stated that ‘it is not obvious that’ X is the case and then started the first two sentences following that statement with the words ‘But experimental evidence suggests strongly that’ and ‘In one such prominent study’. In the second half of his paper, he stated: ‘The pursuit of profit in procreation, where no such opportunity existed before, will compete

123

72

V. Provoost

powerfully against non-pecuniary motives to give’ (Fox 2008, p. 165). Before citing the evidence to support his argument, he acknowledged that it was not obvious that abandoning an altruistic system would diminish the incidence of altruism (which could be read as a potential counterargument to his position). This acknowledgment was followed by a paragraph in which he used two empirical studies to suggest that subjects who observed altruistic behaviour in others, were significantly more likely to help others (and be driven to do so by magnanimous motives). The role of the evidence in this argumentation could be that by showing there is a reverse type of link, he suggests that it is reasonable that the link he described (and feared) would also exist because it can be reinforced by a tendency of individuals to copy behaviour. The author seemed to suggest that the evidence was related to the normative statement he made but the association was at least questionable. Because of the structure of the argument, the precise role of these references was ambiguous (and many roles seemed plausible). In any case, apart from the confusing structure of the argument, the evidence presented did not, and could not, refute the first statement although, to the reader, the wording (especially the word ‘but’) suggested otherwise. 3.7 Reflections on the role of empirical data Many authors acknowledged a lack of empirical data, especially about long-term effects or consequences of (new) treatments or techniques, using phrases such as ‘Based on what little we know’ (Wilkinson 2012, p. 180). Heng (2008), for instance, states: ‘Of even greater concern are the long-term health risks posed to women by egg donation (i.e. premature menopause and propensity to develop various gynecological cancers), even though current knowledge and available clinical data on this subject is lacking’ (Heng 2008, p. 414). Some authors called out for more research in general while others called out for a particular type of research, or for better quality research. In these accounts, obtaining good empirical data was seen as fundamental, and necessary prior to the ethical debate: ‘before any of these questions can be answered, however, the debate must begin with research’ (Sauer 2001, p. 2). Various reflections were made also on the role of empirical bioethics. For Mullen, besides drawing our attention to issues that otherwise would not receive much attention, empirical research was thought to make a debate more focused where it ‘has been degraded by rhetoric’ (Mullen 2001, p. W3). Others considered empirical research a tool for informing ethics studies by improving our understanding of some of the relevant issues, with its absence considered a difficulty: ‘‘Given the paucity of good quality evidence… it is possible only to speculate…’’ (Frith and Blyth 2013, p. 322). Empirical data were considered to be able to provide the empirical basis of an argument (Burr and Reynolds 2008) or refute or substantiate factual claims that formed a part of an argument (Mullen 2001). Also, critical questions were raised about the role of empirical data on specific aspects of gamete donation. Robertson (2010), for instance, raised such questions in his editorial of the Hastings Center Report in which Aaron Levine reported a survey

123

Secondary use of empirical research data in medical ethics…

73

of 2 months of advertisements for egg donors in 306 college newspapers (Levine 2010). While acknowledging the significance of empirical evidence in this debate, for example that nearly half of the ads exceeded the ASRM’s $10,000 recommended limit on compensation, Robertson did not think these findings could actually be used to criticise the current system of self-regulation which he described as flexible and uniform. He argued that our focus should be on the safety for donors and their free and informed consent rather than on what they are paid: ‘So while admiring Levine’s diligent efforts to bring facts to bear on a policy issue; I am left wondering whether there is an ethical ‘there’ there worth worrying about’ (Robertson 2010, p. 3).

4 Discussion The literature on the potential contribution of empirical research to bioethics is extensive and has grown substantially over the last decade. The purpose of this paper was to add to this discussion by providing a description of forms and pitfalls of the use of empirical data by means of references in papers published in medical ethics journals (albeit limited to one topic: the ethics of gamete donation). Wide-ranging uses of referenced empirical data were found, varying from simply describing the context of a problem and to ‘set the scene’, to being used explicitly to support a normative argument. When authors used the data in support of normative argumentation they often pointed to situations that were considered problematic because practice was not in accordance with moral principles, rules or rights or because the data purportedly highlighted a problem with a moral principle or rule. Based on the same empirical study, one author may reach the conclusion that we should focus on improving the application of a rule, while another may conclude that the rule in itself is shown to be problematic. This shows one of many forms of interplay between empirical findings and pre-existing ethical norms whereby the authors’ moral intuitions and reasoning are at the basis of determining whether and in what way our ideas about moral principles and rules should be modified (Sugarman et al. 2009). It reminds us that, not only in data collection itself but also in the way empirical data are integrated in a normative account, the step from data to conclusions is not straightforward. In this, it is important to adopt a reflexive practice including a critical interrogation of the relationship between the researcher and the world under study (Ives and Dunn 2010). Interestingly, other than in directed commentaries (where one should expect this), the papers rarely displayed critical assessments, which could have prevented improper use, of the studies they cited. This lack of critical assessment of studies used in argumentation ties into critiques about bioethics not importing the standards to which empirical researchers are held (Hurst 2010). In any case, although it is important to note that not all papers contained examples of problematic use, in several cases, authors could benefit from some reflection on the quality of the data they are using (as a basis of argumentation) in order to guarantee that they ‘abide by a high threshold of evidence’ (Sugarman et al. 2009). This necessitates the ability to recognise the quality of the data and the usefulness of the data for a particular

123

74

V. Provoost

argument, which is a major challenge for ethicists who usually do not have a background in empirical science methods. This is not only a clear challenge for authors conducting and presenting their own studies (Sulmasy 2010), but also for authors making secondary use of empirical data. Significant variations were found in the roles of the empirical data in particular arguments, including accounts with complicated and sometimes ambiguous logical structures where empirical data were used in a somewhat rhetorical way rather than playing a clear role in the argumentation. This was seen for instance in the paper by Fox (2008) where the wording presented the evidence as essential for the argument while the data dealt with another question and at most a side argument. This diverges from the idea that there might be a clear role for the empirical data, as in the form of a hypothetical imperative (‘If P then X’), described by Ashcroft (2003), where the data are cited as evidencing that P. Many of the accounts studied proved to be too multifarious to simply categorize into distinguishable structures with distinct roles for the data to play. To some extent this was due to an unclear argumentative structure, where empirical research was sometimes cited merely as part of a discourse. Such extracts suggest a form of ‘sciencing up a normative argument’ rather than actually using empirical evidence in a normative argument. Conversely, we should be aware that the complex structures of many of the (sound) arguments in which empirical data are used, are not straightforwardly translated into basic logical structures. This review cannot provide an exhaustive description of all possible practices regarding the secondary use of empirical data, nor can it be used to make claims about the accuracy of such use in medical ethics journals in general because the sample of papers studied was limited and not representative of medical ethics papers. Nonetheless, this specific sample, including papers that were commented on by many other authors, provided interesting material, at times showing how authors responded in a different way to the same empirical study. Given that empirical data can be misused also when merely cited instead of reported, it may be worthwhile to explore good practice requirements for the secondary use of empirical data. Such requirements could cover a wide range of aspects of use of empirical data, for instance the use of vague empirical claims, described by Hedgecoe (2004, p. 132) as ‘a relaxed use of empirical claims’ that ‘does no good for the theories being proposed, or for applied ethics as a whole’. As Sugarman et al. (2009, p. 67) state: ‘Looking forward, attention to how best to promote a more thorough integration of disciplinary perspectives may be more helpful than typologies of empirical research that are framed against the standpoint of some conception of non-empirical normative analysis’. Given the variety of usage of empirical data, it would seem an oversimplification to suggest there ought to be fixed quality criteria for the use of empirical evidence (regarding the selection and the incorporation in an argument) that are applicable to all. Different criteria may be needed depending on the purpose for which the data are used, as well as the theoretical framework for the empirical study, and the method used to integrate different forms of evidence in normative analyses (Sugarman and Sulmasy 2010). When the data play a central role in the argumentation, there is indeed a need for a description of the methods used for

123

Secondary use of empirical research data in medical ethics…

75

this interaction as well as a demonstration of how such interaction contributes to normative reasoning (Salloch et al. 2012). But even before that, we may want some reassurance about the pertinence and the quality of the selected data, and the clarity of their role in the argument. In conclusion, this paper suggests that it is important to study and pose critical questions about how we make secondary use of (by means of references to) empirical research in bioethics. The fact that many authors relied on studies uncritically, while at the same time various forms of problematic or hard to define usage were found, should give pause for thought. Critics of empirical ethics might be concerned that increasing reliance on empirical data may squeeze out space for the necessary justification and conceptual defence of value premises. However, for a sound integration of empirical evidence in normative argumentation, the quality and clarity of the structure of the argument remains essential and the use of (even excellent) empirical evidence will never make up for a poor quality argument. In that respect, empirical ethics will only be facing more quality demands because it needs both a sound argumentation (including justification of value premises) as well as a correct selection, interpretation and integration of the empirical data. This study provides some initial baseline evidence that shows that empirical data, in the form of references to studies, are sometimes being used in inappropriate ways, suggesting that we should also be concerned about the quality of the empirical data selected, the appropriateness of the choice for a particular type of data (from a particular type of study) and the correct integration of this evidence in sound argumentation. Acknowledgments The author gratefully acknowledges Elke Van Hevele and Guido Pennings for their input in assessing the categories generated in the analysis and Jonathan Ives (Guest Editor), Joke Meheus, Guido Pennings and the reviewers for their valuable comments and suggestions.

References Adams, D.H. 2013. Conceptualising a child-centric paradigm. Journal of Bioethical Inquiry 10: 369–381. Afshar, L., and A. Bagheri. 2013. Embryo donation in Iran: An ethical review. Developing World Bioethics 13: 119–124. Ashcroft, R. 2003. Constructing empirical bioethics: Foucauldian reflections on the empirical turn in bioethics research. Health Care Analysis 11: 3–13. Berg, J.W. 2001. Risky business: Evaluating oocyte donation. American Journal of Bioethics 1: 18–19. Borry, P., P. Schotsmans, and K. Dierickx. 2004. What is the role of empirical research in bioethical reflection and decision-making? An ethical analysis. Medicine, Health Care and Philosophy 7: 41–53. Borry, P., P. Schotsmans, and K. Dierickx. 2005. The birth of the empirical turn in bioethics. Bioethics 19: 49–71. Borry, P., P. Schotsmans, and K. Dierickx. 2006. Empirical research in bioethical journals. A quantitative analysis. Journal of Medical Ethics 32: 240–245. Brzyski, R.G. 2001. Putting risk in perspective. American Journal of Bioethics 1: 25–26. Burr, J., and P. Reynolds. 2008. Thinking ethically about genetic inheritance: Liberal rights, communitarianism and the right to privacy for parents of donor insemination children. Journal of Medical Ethics 34: 281–284. Cohen, C.B. 2001. The interests of egg donors: who is deceiving whom? American Journal of Bioethics 1: 20–21. De Vries, M., and E. Van Leeuwen. 2010. Reflective equilibrium and empirical data: third person moral experiences in empirical medical ethics. Bioethics 24: 490–498.

123

76

V. Provoost

Dunn, M., M. Sheehan, T. Hope, and M. Parker. 2012. Toward methodological innovation in empirical ethics research. Cambridge Quarterly of Healthcare Ethics 21: 466–480. Dresser, R. 2001. Donation, disclosure, and deception. American Journal of Bioethics 1: 15–16. Fox, D. 2008. Paying for particulars in people-to-be: Commercialisation, commodification and commensurability in human reproduction. Journal of Medical Ethics 34: 162–166. Frith, L. 2001. Beneath the rhetoric: The role of rights in the practice of non-anonymous gamete donation. Bioethics 15: 473–484. Frith, L. 2012. Symbiotic empirical ethics: A practical methodology. Bioethics 26(4): 198–206. Frith, L., and E. Blyth. 2013. They can’t have my embryo: The ethics of conditional embryo donation. Bioethics 27: 317–324. Gurmankin, A.D. 2001. Risk information provided to prospective oocyte donors in a preliminary phone call. American Journal of Bioethics 1: 3–13. Hedgecoe, A. 2004. Critical bioethics: Beyond the social science critique of applied ethics. Bioethics 18: 120–143. Heng, B.C. 2008. Should fertility doctors and clinical embryologists be involved in the recruitment, counselling and reimbursement of egg donors? Journal of Medical Ethics 34: 414. Hope, T. 1999. Empirical medical ethics. Journal of Medical Ethics 25: 219–220. Holm, S., and M.F. Jonas. 2004. Engaging the world: The use of empirical research in bioethics and the regulation of biotechnology. Amsterdam: IOS Press. Hurst, S. 2010. What ‘empirical turn in bioethics’? Bioethics 24(8): 439–444. Ives, J., and M. Dunn. 2010. Who’s arguing? A call for reflexivity in bioethics. Bioethics 24: 256–265. Ives, J., and V. Provoost. 2015. Monash editorial. Monash Bioethics Review. doi:10.1007/s40592-0150028-9. Jadva, V., T. Freeman, W. Kramer, and S. Golombok. 2009. The experiences of adolescents and adults conceived by sperm donation: Comparisons by age of disclosure and family type. Human Reproduction 24: 1909–1919. Kalfoglou, A.L. 2001. Navigating conflict of interest in oocyte donation. American Journal of Bioethics 1: W1. Kon, A. 2009. The role of empirical research in bioethics. American Journal of Bioethics 9: 59–65. Levine, A.D. 2010. Self-regulation, compensation, and the ethical recruitment of oocyte donors. Hastings Center Report 40: 25–36. Mastroianni, L. 2001. Risk evaluation and informed consent for ovum donation: A clinical perspective. American Journal of Bioethics 1: 28. Molewijk, A., A. Stiggelbout, W. Otten, H. Dupuis, and J. Kievit. 2004. Empirical data and moral theory, a plea for integrated empirical bioethics. Medicine, Health Care and Philosophy 7: 55–69. Mullen, M.A. 2001. What oocyte donors aren’t told? American Journal of Bioethics 1: W3. Price, D. 2013. Exploitation, akrasia, and Goldilocks: How many pounds for flesh for medical uses? Medical Law Review 21: 519–552. Ravelingien, A., and G. Pennings. 2013. The right to know your genetic parents: From open-identity gamete donation to routine paternity testing. American Journal of Bioethics 13: 33–41. Robertson, J.A. 2010. Is there an ethical problem here? Hastings Center Report 40: 3. Salloch, S., J. Schildmann, and J. Vollmann. 2012. Empirical research in medical ethics: How conceptual accounts on normative-empirical collaboration may improve research practice. BMC Medical Ethics 13: 5. Sauer, M.V. 2001. Egg donor solicitation: Problems exist, but do abuses? American Journal of Bioethics 1: 1–2. Solomon, M.Z. 2005. Realizing bioethics’ goals in practice: ten ways ‘‘is’’ can help ‘‘ought’’. Hastings Center Report 35: 40–47. Stock, G. 2001. Eggs for sale: How much is too much? American Journal of Bioethics 1: 26–27. Strong, C. 2001. How should IVF programs handle initial disclosure of information to prospective ovum donors? American Journal of Bioethics 4: 23–25. Sugarman, J., N. Kass, and R. Faden. 2009. Categorizing empirical research in bioethics: Why count the ways? The American Journal of Bioethics 9: 66–67. Sugarman, J., and D. P. Sulmasy (Eds.). 2010. Methods in medical ethics. Georgetown University Press.

123

Secondary use of empirical research data in medical ethics…

77

Sulmasy, D. P. 2010. Reading the medical ethics literature: A discourse on method. In: Methods in medical ethics, ed. J. Sugarman and D. Sulmasy, 315–328. Washington, DC: Georgetown University Press. van der Scheer, L., and G. Widdershoven. 2004. Integrated empirical ethics: Loss of normativity. Medicine, Health Care and Philosophy 7: 71–79. Weinberg, R. 2008. The moral complexity of sperm donation. Bioethics 22: 166–178. White, G. 2001. No news here. American Journal of Bioethics 1: 29–30. Wilkinson, S. 2012. Is the HFEA’s policy on compensating egg donors and egg sharers defensible? Medical Law Review 21: 173–212.

123

Secondary use of empirical research data in medical ethics papers on gamete donation: forms of use and pitfalls.

This paper aims to provide a description of how authors publishing in medical ethics journals have made use of empirical research data in papers on th...
174KB Sizes 0 Downloads 13 Views