Technology, Design, and Human Values in Healthcare
Communities struggle with finding ways for collaboratively exploring the value of healthcare technologies. Currently, a strong emphasis is being placed on the assessment of the costs associated with the health gains (expressed in quality-adjusted life years) that are achieved with these technologies. Following Hannah Arendt, we shall try to argue that such instrumental rationality is misplaced in discovering how technology can help to express human values. It typically reflects a society where processes of design and development, evaluation, and decision making involve separate trajectories and operate distinct from the realm of the lives of humans. We will present an alternative which is deliberative and transformative in nature. Its strengths and limitations will be explored, using the cochlear implant for deaf children as an example.
KeywordsTechnology assessment Values Dialectic Frame reconstruction Evaluation as learning
Designing and creating technology is a fascinating process. It is fascinating because it reveals something about our physical environment (what is physically possible, for instance, will we, in fact, be able to live on Mars?), about our social environment (what is socially possible, for instance, will we be able to join forces to conduct research on human (im)mortality?), and about ourselves (how clever and ingenuous are we, for instance, will we be able to discover the theory of everything, if there is one?). It is also fascinating because it unites considerations of what is possible with considerations of what is desirable. This is neatly expressed in questions like “Would we clone them if we could?” (suggesting that there are probably too few Albert Einsteins in this world and definitely too few Marilyn Monroes). It is also fascinating because it testifies of a basic trait of modern societies, which is being dissatisfied with the world as it is (Heller 1999). However, mankind has not an impressive track record when it comes to presaging how technology will enable the creation of value. The realm of healthcare is no exception. Over the past decades, accumulating evidence has made it abundantly clear that healthcare technologies need not always be merely beneficial. Healthcare technologies that were once considered of great value turned out to be largely ineffective or positively harmful (Dutton 1988). In response to such experiences, we have witnessed a growing demand for the critical evaluation of healthcare technology, preferably at an early stage of development. This has resulted in a veritable healthcare evaluation industry, with its idiosyncratic standards and logic (Klein 1982). In this process, Health Technology Assessment (HTA) commissioning organizations such as NICE (the National Institute for Health and Care Excellence) and HTA researchers have developed an evaluation methodology with a strong emphasis on cost-effectiveness. By translating health benefits into a single metric (quality-adjusted life years, or QALYs), league tables can be compiled, comparing different healthcare technologies in terms of costs per QALY gained. It is questionable, however, whether such analyses suffice to understand and communicate the value of healthcare technologies. Cost-effectiveness analysis is appropriate to explore whether a particular healthcare technology is a relatively efficient way of achieving certain preconceived goals. It is much less appropriate to collaboratively explore whether a technology results in practices that adequately reflect values that are considered important in the community concerned. The main objective of this chapter is to explore how existing methods of technology assessment may be adapted to better serve this role. The key features of this alternative approach are its interactive nature by engaging stakeholders, an emphasis on interpretation of evidence, and a focus on learning. We will refer to this approach as interactive technology assessment (iTA). In the next two sections, we will briefly discuss a number of key developments in the field of healthcare technology and the values that seem to be at work in this domain. We will then turn to the work of Hannah Arendt, who made a distinction between “work” (the construction of artifacts) and “action” (the public inquiry into the meaning of life) which is highly relevant to the question how technology bears on values and vice versa. We will then describe in more detail the method of iTA, followed by a case study, the application of this method to the evaluation of cochlear implants (CI) for deaf children. The chapter will be completed by a discussion of what we can learn from the case study: can iTA help stakeholders to discover how technologies should be designed and used in concrete situations so as to achieve maximal coherence among our multiple and varied value commitments, and, if so, what conditions should then be met?
Trends in Healthcare Technology
Many healthcare technologies have grown out of ideas that became more serious in the course of the nineteenth century, have picked up speed after the Second World War, and continue to develop at accelerated speed at present. An example is the left ventricular assist device or LVAD, which developed from the heart-lung machine and which is currently developing further because of the emergence of novel materials, energy-saving designs, and improved methods of control of coagulation and inflammation. In the context of this contribution to the volume, it is impossible to provide anything near an overview of the wide and fast technological developments in healthcare. Suffice, perhaps, to note the closely intertwined development of GRIN; Genomics, Robotics, Informatics, and Nanotechnology. To give an example: developments in informatics and genetics have created unprecedented opportunities for identifying the genetic origin of many diseases. Functional analysis of these genetic deviations will pave the way for gene therapy. Currently, such gene therapies are being developed for patients with serious eye disease and for patients with hearing disorders. Nanotechnology is part of so-called quadratherapeutics, where two types of nanoparticles are being combined with laser technology and radiotherapy in the experimental treatment of cancer. Another development that may have a significant impact on health is tissue engineering, allowing, for instance, the manufacturing of heart valves. The success of all these developments may be inferred from the estimated life expectancy of today’s newborns, two third of whom is expected to live 100 years or more. Clearly, this will pose novel problems, for which novel solutions will be sought. Such success comes at a price, though: an ever-increasing part of our national budgets is being spent on healthcare.
Design, Assessment, Evaluation
In the domain of healthcare, design may apply mainly to the development of devices, rather than to drug development. In drug development, large amounts of compounds are being tested in vitro and in animals for their therapeutic potential and safety profile. Design, here, may chiefly refer to the chemical alteration of such compounds in an attempt to improve their therapeutic-safety profile, to improvements in modes of administering drugs (as in the case of quadratherapeutics, mentioned above), or to attempts at targeting patients who are most likely to benefit and least likely to experience harm (personalized healthcare, such as restricting treatment with the antiretroviral drug Efavirenz to HIV patients with specific serotonergic polymorphisms). Not infrequently, drug discoveries are made serendipitously, as, for instance, in the case of Ritalin in the treatment of children with ADHD (Singh 2002). Device development usually extends over several decades, as, for instance, in the case of imaging devices (Blume 1992), the cardiac pacemaker, and the left ventricular assist device or LVAD (Sutton et al. 2007). The LVAD originated from the heart-lung machine, which enabled cardiac surgery. Since then, its development has been one of miniaturization, decreasing energy expenditure and improved control of thrombosis and infection, leading to various types of intracorporeal devices. Along the way, attempts at developing a total artificial heart were made but discontinued because of multiple problems that could not be resolved at the time. Until this day, unexpected problems have emerged with cardiac assist devices, demonstrating the strong “learning-by-doing” strategy in medical care (Starling et al. 2014). Cases like these point to the close link between designing, assessing, and evaluating healthcare technologies, with clinical experience giving rise to adjustments or technology redesign (e.g., in the case of the LVADs, continuous flow pumps as opposed to pulsatile flow pumps, development of new materials with different immunogenic properties, transcutaneous energy transfer, etc.). In this chapter, our focus will be on assessment and evaluation of healthcare technology, exploring what implications these may have for the design and redesign of such technologies.
Values in the Domain of Healthcare
The values underlying healthcare are multiple and varied. Although far from exhaustive, the following values seem to be at play in the current development and provision of healthcare. Firstly, there is the preservation of life itself. If someone has sustained a myocardial infarction, clinical management is aimed at revascularization of the heart in an attempt to rescue the life of the patient. Similarly, if a newborn child suffers from life-threatening cardiac or respiratory failure, the child will be supported by a heart-lung machine. The same holds for patients who sustained severe neurotrauma, for patients with an acute appendicitis, etc.: many healthcare activities are aimed at preserving life in patients with life-threatening conditions. Another value is the relief of suffering. The suffering may be mainly physical, as in the case of patients with inoperable colon cancer, who are prescribed opioids to relieve intractable pain. The suffering may also be mainly mental in nature, as in patients with severe refractory depression who receive electroconvulsive therapy. Yet another value is the restoration of functions, enabling people to conduct valued activities. These would include cataract extraction to restore vision, and hip or knee replacement to restore mobility, although such interventions will also relieve the suffering (pain, depression) associated with these conditions. Still other healthcare activities are directed at the preservation or restoration of human dignity. An example of this would be the treatment of drooling in children with cognitive disability. Such treatment will not help to preserve life and is unlikely to relieve suffering. And yet, many of us would feel that it would testify of a lack of respect if drooling in such children would be left untreated, assuming that an effective treatment is available (which is, indeed, the case). Many of the healthcare activities, mentioned so far, could be conceived as being directed at lifting, alleviating, or compensating for the restrictions, imposed by disease. Or, positively defined, directed at preserving, restoring, or augmenting capability, as defined by Amartya Sen (Venkatapuram 2011). Quite another value, related to the delivery of individual healthcare, is respect for the patient’s will. An elderly patient may, for instance, request deactivation of his or her cardiac pacemaker, believing that the pacemaker may prevent him or her from dying in a dignified way. Not granting such a request would testify of insufficient respect of a patient’s autonomy, provided that the request was not a rash decision or inspired by a temporary and uncharacteristic loss of stamina. Another deeply ingrained value in the healthcare domain is the avoidance of inflicting harm to the patient (“primum non nocere”, first do no harm). Perhaps, nowadays, the norm should read: avoid unnecessary harm. For, inflicting harm is unavoidably associated with any operative procedure and cancer treatment, be it surgery, radiotherapy, or chemotherapy. Still, inflicting avoidable or unnecessary harm to patients is considered a very grave offense, occasionally giving rise to the suspension or even prohibition of a professional’s continued practice. Finally, there is increasing awareness that healthcare should not be wasteful, that there should be proportionality between effort, burden to the patient, and expected benefit, and that all patients deserve equal concern for their suffering and anxiety. These may be considered as a commitment on the part of healthcare professionals to the value of justice or equity. To be sure, justice in healthcare has not been operationalized in a single, coherent fashion. On the one hand, there is a clear tendency toward maximizing the aggregate health gain by prioritizing the most cost-effective services. On the other hand, there is also the widely held conviction that healthcare should be used so as to mitigate gross inequalities in health experienced by humans during their life time. This conflict between a utilitarian and egalitarian conception of justice has been, and still is, one of the most vexing issues in healthcare policy (Daniels 2007).
Hannah Arendt: A Typology of Human Activities
How, then, do the development of technology and the reflection on the value of the practices that emerge from the use of these technologies relate to each other? We think that Hannah Arendt’s account of the human condition can be helpful in this respect. In The Human Condition, Arendt distinguishes three types of human activities: labor, work, and action (Arendt 1998). Labor involves activities (production and consumption) which are necessary for the maintenance of life. Work involves activities which are, literally, creative: the making of artifacts with which we equip the world and which serve to make our lives safer, more comfortable, more convenient, etc. Action, finally, involves communicative interaction between humans, aimed at collaboratively finding out the meaning and significance of their doings and beings. Each of these activities is, in its own way, indispensable to human life. In the absence of labor, we would not be able to continue living. In the absence of work, we would live, but in a very harsh way. In the absence of action, we might be able to live, and to live comfortably, but in a meaningless world.
Of the three types of human activities, action is the most elusive one, being not strictly necessary for the preservation of life, nor productive in the sense of resulting in tangible products that enrich our lives. And yet, it is quite clear that Arendt considers action the most crucially important activity, since it is through the collaborative construction of meaning that humans can achieve freedom and develop their identity. At the same time, however, she observes a tendency toward an ever greater emphasis on the preservation of life itself (labor) and the further improvement of the quality of our lives through the development of artifacts (work), at the expense of engaging in public deliberation about final ends (action). A clear indication of this is the tendency to use instrumental rationality, which is appropriate to work, as the sole guidance for assessing value. In our view, the strong focus on cost-utility analysis in the evaluation of healthcare technology is a striking example of this. However, as pointed out by Arendt, this is self-defeating. Having adopted instrumental rationality in the realm of work, thereby necessarily stripping everything of intrinsic value, humans cannot at the same time adopt this framework to ascribe – subjectively – meaning or value to artifacts, states of affairs, etc. Or, as Arendt puts it: Utility, when elevated to meaning, turns into futility. The ascription of value and meaning is, according to Arendt, the prerogative of action, which is inherently deliberative.1
The Meaning of Meaning
The key concept for action, then, is meaning, and in the context of assessing value of technology, the key question is not so much “What are the data?” but “What do the data mean, or what do they amount to?” This is not to say that data are unimportant. On the contrary. But the question of meaning goes beyond the data. In other words, when presented with data, we can intelligibly ask: What do these data mean? This question, about the correct interpretation of the data, refers to the practical implications: what, if anything, should we do and how should we act upon this information? The answer to that question cannot be obtained from the data themselves, but requires that the data are situated in a wider interpretative framework. Only then, we can make causal inferences and judge the relevance of the data, which is necessary to assess practical implications. To give an example, we may be presented with data on changes in global temperature. The question what these data mean goes beyond the data themselves. The question is about the possible and likely causes of these changes (causality), the likely future developments (prediction), and about the question whether we should be pleased about these changes, worried, or indifferent. Ultimately, it is about a judgment as to what an appropriate response would be. To take an example from the healthcare sector, we may be presented with data on the costs of healthcare over the past decades. Again, we can intelligibly ask: what do these data mean, what do they tell us, and how should we act upon them? Answers to these questions go beyond the data themselves and require an interpretative framework: What are the causes of the change in costs of our healthcare systems? Should we be pleased, worried, or indifferent? The change may be caused by the slow increase in productivity in healthcare as compared to other areas and interpreted as an indicator of economic growth. As such, at least from an economic perspective, the data should be a reason for joy, rather than concern, and give no cause to corrective measures (Baumol 1996). A final example to clarify the distinction between data and the interpretation of data might be changes in the burden of disease. Insel, for instance, reported that in spite of substantial efforts, the burden of disease associated with schizophrenia has not significantly decreased over the past decades (Insel 2010). The relevant questions are, again: what the reason for this might be, why it should be of our concern, and what our appropriate response to this finding might be? Only when we have answers to such questions, we can decide what we should do about it (if anything) and how we should act. “Action,” as defined by Arendt, refers to this complex of activities, namely, acting on the basis of the purposeful and collaborative collection and interpretation of data. Or, as Farrell e.a. formulated it: “Assessment processes are embedded in different sorts of institutional settings, within which scientists, decision-makers, and advocates communicate to define relevant questions for analysis, mobilize certain kinds of experts and expertise, and interpret findings in particular ways” (Farrell et al. 2001).
Plurality: It Ain’t Necessarily So
Another central feature of Arendt’s work, apart from the typology of human activities, is the essentially plural character of action. In other words, there are, and always will be, differences in how human beings interpret the world, including their own doings and beings. Our observations allow for multiple interpretations, that is, understandings of how things cohere and are brought about and judgments whether we should be concerned and how we should act upon them. This plurality follows from the fact that action requires that people are not forced to adopt a particular interpretation. Through action, people can achieve their freedom and, as such, reveal their identity to each other. This process results in, and at the same time requires, contrasting interpretations. Whereas in labor (activities that are necessary to maintain our lives) and work (activities that are necessary to make our lives comfortable, safe and efficient) the degrees of freedom are rather limited, this does not hold for action: “things ain’t necessarily so.” This plurality of interpretations gives rise to different views of what needs to be done and why: should we do something to counter global climate change, to bring the increase of public expenditures to healthcare to a halt, and to reduce the burden of disease associated with schizophrenia, and if so, what? Since these are clearly questions of collective, rather than merely individual action, some form of public deliberation is required, laying down procedural rules for who may be involved and in what way, and how, in the light of plurality, decisions are made as to what actions will be set in motion (Dryzek and Niemeyer 2010). In the next section, we will briefly discuss what this might look like in the context of the development and assessment of healthcare technology.
Implications for the Development and Assessment of Healthcare Technology: Interactive Technology Assessment
Many people would perhaps readily concede to the distinction between the generation of data itself and their interpretation, usually referred to as the assessment – appraisal distinction. It is suggested, then, that assessment is an essentially value-free kind of scientific research, producing data for the decision-making process (appraisal), where values are brought to bear on the available evidence. This distinction is, however, utterly misconceived, since it fails to acknowledge that interpretative frames are already operative at the time of data collection. At that stage, they serve to define which data are considered relevant, plausibly associated with the intervention, and amenable to scientific research. If we do not recognize this role of interpretative frames in the production of evidence, we fail to understand why the relevance of available evidence is not always endorsed by all stakeholders. To avoid this type of bias, then, involvement of the various stakeholders during the assessment process is vitally important (Reuzel et al. 1999). However, interpretative frames encroach even deeper in the process, since they also provide guidance to the design and development of healthcare technology itself (Schwarz and Thompson 1990). This would plead for the involvement of stakeholders at the stage of design and development of healthcare technology, in such a way that it can be qualified as a true instance of action. In the following sections, we will describe what such interactive technology assessment (iTA) and interactive technology design and development (iTDD) might look like. To do this, we will report our experience with the interactive evaluation of the cochlear implant for deaf children. This was an assessment of a technology which was more or less matured and which had to draw on the evidence that had been produced by then. The strengths and limitations of such an approach will be discussed, including a reflection on how the results of such deliberative practices may be integrated into processes where deliberation has been delegated to professionals (i.e., politicians).
Case Study: Interactive Evaluation of Cochlear Implants for Deaf Children
In this section, we will present our case study, the cochlear implant for deaf children. We will start by briefly describing the societal debate regarding this technology. This will be followed by a description of our interactive evaluation of the technology in terms of method and its key results.
The Technology and the Opposition from Deaf Communities
Deaf people are ill or incomplete individuals, are lonely and unhappy, cannot communicate effectively with others, and are all desperately searching for a cure for their condition. Such publicity is highly inaccurate. It also demeans Deaf people, belittles their culture and language, and makes no acknowledgment of the diversity of lives Deaf people lead, or their many achievements. It is stressful for hearing parents of deaf children who are already struggling to come to terms with their child’s deafness and are given a false impression that the implant will ‘cure’ their child.. (Deaf Australia, Policy on cochlear implants; www.deafau.org.au/info/policy_cochlear.php Accessed on January 29th, 2014)
The decision to implant such children is usually made by parents and guardians and Deaf Australia has long believed that parents and guardians do not have access to full and accurate information about the implication of deafness for their children’s lives. Parents are usually led to see their children as pathologically deficient and little information is available to them about the history, culture and language of Deaf people, or the possible lives of Deaf people in our society. Until such time as more complete information is available for parents, and more productive associations develop between parents of deaf children and adult Deaf people, Deaf Australia feels that decisions to implant young deaf children are questionable.. (Deaf Australia, Policy on cochlear implants; www.deafau.org.au/info/policy_cochlear.php Accessed on January 29th, 2014)
Interactive Evaluation of CI for Deaf Children: Methodology
The interactive evaluation which we conducted on CI for children was based on the method for joint evaluative design, described by Guba and Lincoln (Guba and Lincoln 1989). Briefly, it consists of identifying and engaging all stakeholders, where stakeholders are defined as those people or groups of people who are likely to experience the consequences of the evaluation and inquire into their constructions of problem and judgments of potential solutions, thus gradually developing a so-called joint construction of the problem and a preferred solution as the final result of the evaluation. Generally speaking, in the case of healthcare technology, stakeholders would typically involve patients and their relatives, (allied) healthcare professionals, healthcare insurers, manufacturers, hospital managers, representatives of advocacy groups, and policy makers. These may be approached directly by the members of the evaluation team, may come forward of their own accord, or become involved through snowballing. In the case of the evaluation of CI, 51 stakeholders participated, including ENT surgeons, audiologists, teachers from Institutes for deaf children, social workers, speech therapists, policy makers, hospital managers, and representatives from various advocacy groups (Reuzel 2002). Each of these participants was interviewed individually. The purpose of the interview is to reconstruct the interpretative frame of the stakeholder. An interpretative frame is the ensemble of a stakeholder’s judgment of various solutions to a specific problem, problem definition, background theory, and values (Schön and Rein 1994; Grin and van de Graaf 1996). Typically, an interview starts by asking what the interviewee considers as particularly problematic about the current situation (problem definition). From there, the interview may proceed by asking what strategies the interviewee considers most likely to resolve the problem (judgments of solutions). To better understand why a stakeholder considers a particular situation problematic and why he or she considers particular strategies most likely to be effective in resolving the problem, it is important to reconstruct background theory and values. The method of frame reconstruction (Grin et al. 1997) has been applied to a wide number of topics (Fischer et al. 1993, 2012) and has been found a valid and reliable method of qualitative data analysis (Moret-Hartman et al. 2007a). After checking and, if necessary, revising the reconstructed frame (respondent validation), a second interview round is held, where interviewees are confronted with the reconstructed frames of all participants (anonymously) and invited to respond to them. In particular, interviewees are asked whether, based on the reconstructed frames of the other participants, they wish to revise (parts of) their own interpretative frame and whether there are particular issues that they would like to have investigated or checked in the literature. In doing so, the material particularities of the situation are explicitly considered. The evaluation team, then, explores such issues by reviewing the literature and consulting experts. The findings are then fed back to the participants in a third interview round, where interviewees are also asked to review their own interpretative frame, taking into account the interpretative frames of the other participants and any findings from the literature or expert consultation. They are also invited to attend a meeting with all other participants, where the key findings will be presented and unresolved or contentious issues will be discussed, aiming at an agreement on how to proceed (or not) with the technology under investigation.
Interactive Evaluation of CI for Deaf Children: Key Findings
Here, we will report the key findings of the interactive evaluation of CI for deaf children that intended to design the socio-technical practices around CI so as to incorporate concerns from the stakeholders involved; a full account is presented elsewhere (Reuzel 2002). Firstly, it was found that those who were in favor of CI differed in their problem definition from those who were more critical of the technology. To the former, the key problem was simply that deaf children cannot hear, and that, for this reason, deaf children are unlikely to develop well in a hearing society. A CI is a logical remedy to this problem, although it is admitted that it is still an imperfect technology. Those who were more critical of CI framed the problem in a slightly – but importantly - different way. For them, the key problem is that deaf children, during the first months of their life, do not receive the sort of sensory input that is appropriate and necessary for their social, emotional, and particularly cognitive development. For them, the consistent use of sign language by the parents in the communication with their child is critical to the child’s development.
Interestingly, these positions turned out to be associated with differences in background theory and values. Proponents of the CI appeared to be reasoning from the assumption that sign language and spoken language are mutually competitive: acquisition in linguistic competence in sign language will interfere with acquisition in spoken language and vice versa. This explains their negative attitude toward the acquisition of sign language by deaf children. Those who were more critical of CI appeared to reason from the assumption that the two linguistic modalities could, in fact, be mutually reinforcing, as is known from bilingualism in spoken language. In terms of values , the views of proponents of CI could be largely resolved to the open future argument: do not choose options that are likely to constrain the range of future opportunities that are open to the developing child. Those who were more critical of CI were more likely to emphasize the importance of cultural diversity: there are multiple ways of organizing social life that can serve as a basis for the development of mutual and self-respect. Further key differences were observed in the appreciation of deaf culture: is it, or is it not, a fully fledged culture, and, relatedly, is sign language a fully fledged language? Finally, proponents and opponents of CI differed in their view of the significance of deaf culture to the social and emotional development of deaf children, notably with respect to the development of their self-esteem. It is important to note that most participants were not aware of the fact that they were reasoning from these premises, and how they influenced their problem definition and appreciation of the technology. Nor were they always able to provide compelling evidence or arguments in support of these premises. For this reason, we tried to retrieve literature that might shed more light on the issue of the antagonistic or synergistic nature of the relation between development in spoken language and sign language. This produced data from studies that had been conducted in Sweden, where deaf children had been followed over time, from before cochlear implantation to several years thereafter (Preisler et al. 2005). The researchers had used both, quantitative and qualitative methods, observing the children in their own environment, and investigated various aspects of their communicative development. Children who were most proficient in sign language at the time of cochlear implantation were more likely to benefit from the CI in terms of developing proficiency in spoken language. Although certainly not conclusive, these data seem to suggest that the two linguistic modalities can be synergistic, rather than antagonistic.
At the final meeting that was organized, it was agreed that early identification of deafness in newborn children is of crucial importance and that deaf children probably benefit from the consistent use of sign language by their parents, irrespective of whether they will receive a cochlear implant or not. Clearly, such recommendation has considerable implications for the parents, for how they are informed about sign language, deaf culture, and CI, and for creating sufficient training capacity in sign language.
In summary, then, the interactive evaluation had helped to overcome the impasse, following the intense debate on cochlear implants for deaf children, and it helped to develop a practical solution that was acceptable to all participants. It is important to note that the emphasis in an interactive approach is on learning between actors with different frames (Grin and van de Graaf 1996) and that it is both summative (what is the value of the technology, when used in this way?) and formative (how can the value of the technology be improved through the specific way in which the technology, and associate practices, is (re)designed?) (Scriven 1996).
Frame Reflective Analysis: A Reflection
The aim of our contribution to this handbook was to explore what can be done to gain a better understanding of how technologies relate to values that are considered important in our society and how we can find out what implications these value commitments might or perhaps should have for the choices that are made in the context of technology design and development. We have drawn on our experience with evaluating CI for deaf children, where we used a method of TA which is interactive, which comprises reconstruction of stakeholders’ interpretative frames, and which is directed toward learning. In the remainder of this chapter, we will discuss what might be learned from this case study. By relating it to a number of insights from various disciplines (policy sciences, philosophy, social science of technology), we will try to clarify its key characteristics and prerequisites.
The Dialectic Nature of iTA
As we have seen in the case of the cochlear implant, interactive evaluation is primarily concerned with the practical interpretation of facts: what implications should our findings have for the design of a social-technical innovation? In the case of the cochlear implant, we may consider evidence of how deaf people live in hearing societies. The question of interpretation goes beyond such data and asks: what are the underlying causes, how do we judge the situation from a normative perspective, and, consequently, what would be an appropriate response to the current situation? We have seen that in the case of CI, the available evidence was interpreted quite differently by different stakeholders, resulting in quite different views of what needs to be done. iTA may be thought of as an attempt to bring about a synthesis between actors with different interpretative frameworks. As such, it may be considered as truly dialectic: a discourse between multiple stakeholders who hold different views on a particular subject, who wish to reach agreement on the appropriate course of action on the basis of reasoned arguments. It will, therefore, be beneficial especially in situations like this, where actors’ problem understandings seem incompatible, i.e., when, at first sight at least, these problem definitions do not allow for a common solution. In such cases, it is quite likely that so-called Type III errors (solving the wrong problem) are made – more precisely, that the problem is defined in a way that privileges the viewpoint of particular actors. As Hoppe argues, this often is not an analytical mistake, but a matter of contextually prevailing rationality and network structures (Hoppe 2010). What we saw in the CI case was that stakeholder engagement and frame reconstruction resulted in the development and exploration of alternative frames, allowing for discovering what was “behind” proposed solutions (the CI). As such, the approach resulted in greater plurality in interpretation of the situation. On the basis of this feature, iTA may be conceived as a practical attempt to foster action, as defined by Arendt.
It is important to realize that deciding whether problem definitions are incompatible is not merely a matter of fact. Even where this seems to be the case, it may be a matter of creative judgment. Basically, this is because problems and conceivable solutions themselves, rather than being objectively given, are constructed through judgment. The implication is that the conclusion that no joint solution for a problem exists is contingent upon preceding judgment, or lack of prudent judgment, and may change through a process of further, creative judgment.
The Central Role of Judgment
The CI case also revealed the central role of judgment: judgment of the situation (what problems are experienced by deaf children growing up in hearing societies?), judgment of proposed explanations (do these problems originate from the lack of hearing in deaf children?), judgment of proposed solutions (e.g., the CI), judgment of available evidence (e.g., what can we infer from the available evidence concerning the clinical and cost-effectiveness of CI?), etc. Judgment here refers to the Aristotelian notion of phronèsis, which has been key to the development of interactive technology assessment. Phronèsis is practical wisdom: knowing what needs to be done in concrete situations in order to achieve maximal coherence among our multiple and varied value commitments (Richardson 1994). By its nature, a key feature of phronèsis is interpretation: it involves a double hermeneutic of interpreting both generic principles and the particularities of the situation. Following Arendt, it also essentially involves that different viewpoints are synthesized. This notion of phronèsis has been taken up since the mid-1980s in a then emerging approach within the policy sciences, sailing under the flag of the “argumentative turn” (Fischer et al. 1993, 2012) or “interpretative policy analysis” (Yanow 2000). Emphasizing the hermeneutic nature of this process, phronèsis has been conceived as a notion that may help understand the formation of interests, preferences, problem definitions and policy options, bringing back in the essence of politics, and human (inter)action more generally where it had been discarded by rational choice theory’s assumption that these are merely exogenously given and thus constant (Hindmoor 2006). Obviously, such understanding is especially useful when analyzing socio-technical change, in which preferences may evolve both as part of the (i.e., as a driver of or response to) innovation. Interestingly, this is exactly what happened in the case of the iTA of CI: stakeholders held different views on the technology; through frame reconstruction these could be resolved into different background theories; this occasioned partial revision (“learning”), resulting in agreement on how to proceed (“irrespective of whether a child will be notified for CI, start using sign language as early and consistently as possible”). In Richardson’s (1994) terms, it is an instance of the revisability of goals.
Should be organized interactively and participatory, enabling deliberative exchange of viewpoints
Should adopt an interpretative view on data gathering (which qualifies, but certainly not precludes, empirical-analytical analysis)
Should be “set up to induce transformative learning on the part of the participants as a result of the interactions with, (i) the problem situation, (ii) the way in which others define the problem and (iii) the particularities of the contexts in which these others operate. [It should enable] reflection on participants’ own interpretive frames, allows for a ‘fusion of horizons’ and potentially enables participants to act in line with the newly required understanding” (Loeber 2007, p. 399).
This summarizes, in a way which is equally precise as concise, the rationale for, and fundamental nature of, iTA, which thus transcends the mere idea of “participatory” policy analysis: it enriches the latter with elements of critical as well as hermeneutic-interpretative analysis, into what Hoppe has called “forensic policy analysis” (Hoppe 2010). iTA thus becomes a form of analysis which not only assesses but also shapes the design of socio-technical innovations and policies, which does so through considering them in the light of (different views on) the good life and individual identities, and which, in doing so, helps actors to articulate, and deliberate on, their different viewpoints on the predicament in which we find ourselves and actions that seem most appropriate to turn what is into what ought to be.
iTA: Contexts of Application
Thus understood, interactive technology assessment as a form of deliberative analysis and design may yield at least two contributions. First, in the context of health policy making, iTA as a form of deliberative policy analysis could inform policy making and thus help avoid Type III errors (the solving of the wrong problem). While this might help to embed a novel health intervention in health practices so as to do more justice to the life world of patients, opportunities to adapt the technology are likely to be limited in that stage. Thus, a second context of application where iTA might be useful are the practices in which medical innovations are being developed.
The rationale for using iTA to support formal policy making would be that it offers a departure from the often encountered practice that policy solutions reflect merely the problem definitions of those actors who have privileged access to the policy process. In healthcare these tend to be actors on the supply side, as the key policy mechanisms indicate “evidence-based admission” and “reimbursement” of drugs and devices, produced by an innovation system in the form of a tightly knit network of public knowledge institutions, the medical profession and industry. This intertwined policy and innovation network forms a typical case of a closed, institutionalized network, hardly accessible to other parties. Policy making is typically information driven, drawing on a community of experts with shared knowledge base and standards of rationality. In case of the health system, this system represents the dominant medical rationality (Schuitmaker 2013), understanding health and disease as located in a (universal) body and thus as mainly determined by healthcare interventions, ignoring other determinants like lifestyle, life circumstances, and agency of patients (Wilt 1995).
As Hoppe (Hoppe 2010) points out, such networks are well able to deal with structured problems, i.e., problems on which a fair degree of normative consensus and factual certainty are presumed to exist. Or, more accurately, such networks tend to treat all problems as structured ones – more specifically as problems structured in the same way as earlier ones – and shape solutions accordingly. This has many potential drawbacks. Health problems may be suboptimally addressed, especially when they differ in nature from the problems that used to dominate the agenda. One important example concerns the now increasingly important class of noncommunicable diseases (including diabetes, COPD, obesity, and cardiovascular problems) where precisely these other three determinants may play key roles, as we have shown for, e.g., irritable bowel syndrome (Moret-Hartman et al. 2007b) and low back pain (Gielen and Grin 2010). The cochlear implant example is another case in point, which moreover suggests that such one-sided rationality may also lead to innovations being rejected by those for (!) whom they were developed, generating more societal controversy and less health effects than desired.
Under such circumstances, iTA as a deliberative, participatory, transformative design practice may help restructure the policy problem, provided that the conditions, defined by Loeber, are met. The cochlear implant case shows that it may be difficult to meet all of these conditions. The first condition, organizing the analysis in an interactive, participatory and deliberative fashion, was largely met. Participants were selected to include technology developers (as a manufacturer), those providing or applying it (e.g., as a physician), users or receivers of the technology, decision makers, and those supporting or helping patients in using the technology (e.g., a nurse, a social worker, a teacher and so forth). Also, participants could suggest to invite others; newspaper advertisements were placed in national papers; a web page was issued; and it remained possible for additional participants to be included at later stages of the process.
Meeting the second criterion suggested by Loeber, adopting an interpretative, rather than a technical, approach to data gathering, proved to be more difficult. It was difficult for the evaluators not to reproduce – either by habit and training – or to accommodate institutional requirements or credibly respond to interventions by health professionals (Reuzel et al. 2007). At least as problematic was that the available literature was so strongly dominated by established medical rationality that it was difficult to test and/or underpin other viewpoints (an important exception being the longitudinal studies conducted by Preisler and her colleagues, as necessary for restructuring the problem) (Preisler et al. 2005).
The third condition mentioned by Loeber, inducing transformative learning (a learning process giving rise to novel, unanticipated interpretation of the issue and novel approaches to its resolution) was both promoted by the just mentioned rules for deliberation and participation, as well as by a project design favoring learning through successive interview rounds (Guba and Lincoln 1989) and hampered by the imbalances in terms of underlying knowledge base between various viewpoints. Other hampering factors included a lack of trust and power imbalances between participants (Reuzel et al. 2007).
This takes us to another set of conditions, pertaining to the relationship between the iTA practice as a locus for “problem restructuring” and the closed, institutionalized network with its tendency to deal with the problem through “normal” problem structuring. In the case of the cochlear implant, iTA was used to increase the acceptability of the technology by shaping its conditions of application. This was done in a context in which the decision to introduce (i.e., admit and reimburse) CI in the Netherlands had already been taken, while the fact that the technology already had been developed and shaped without much involvement of the deaf and their relatives had eroded acceptance and trust between them, the medical establishment, and government. This further added to power differences between participants that already were there due to differential access to decision making and the dominance of medical rationality. It would be crucially important if the policy community would learn from such cases so as to introduce institutionalized warrants to nurture such deliberative exercise and adequately absorb their details.
Finally, the agency of the evaluator seems of critical importance in establishing and maintaining the conditions for deliberative, participatory, and transformative judgment in the iTA. Thus, developing a critical mass of evaluators with the requisite competences is of key importance. In addition to process competence and strategic insights, sufficient mastering of, and a capacity to critically scrutinize, medical knowledge are quintessential in this respect.
Implications for Design for Values in Healthcare
What the case of pediatric cochlear implantation has shown is how the life world of intended users is easily obscured from sight by R&D and decision-making procedures in force that have a political rationality. To be sure, CI has been developed on the basis of a vision of this life world. For example, CI was originally designed as a safety tool that helped to pick up sound and hence warn for traffic and other dangers, rather than a communication device. When, soon after that, CI turned out to provide with a better hearing than had been anticipated, a new vision of deaf people being able to participate in a hearing society developed, giving direction to the process of further design. However, it is questionable whether the deaf themselves were in a position to contribute to shaping this vision, particularly when the vision on deafness as a handicap became the dominant paradigm in R&D circles. As soon as this was a political reality, a fierce debate ensued about the validity of the paradigm. As all stakeholders tried to “win” this debate, it appeared as if the life world was buried under a pile of arguments for or against CI.
What the case of pediatric cochlear implantation has also shown is that behind the roles stakeholders play in this political arena are humans who at a deeper, normative level have more in common than can surface in such an arena. Fights over funding CI at the expense of support for the deaf community and the acknowledgement of sign language, for example, had obscured the fact that all persons involved sought to achieve happiness for deaf children. In their life world, happiness involved: feeling safe in a trusted social environment, being able to communicate with persons in this environment, and having the opportunity to pursue what you value in life. Based on these shared values, cochlear implantation could be redesigned and accepted as a social construct, even though it did not change fundamentally as a device.
It is here that iTA demonstrates its value: its potential to have persons involved leave the political arena and share their life worlds instead. In these life worlds, there is nothing to win and much to gain.
iTA to Shape R&D Programmes
The example of the cochlear implant suggests that an interactive approach might also create a different relationship between R&D programmes and design priorities and patients’ needs and desires: witness the examples where much more congruency could be achieved than anticipated by those involved. Given that path dependencies (limitation of options as the result of previous decisions) are likely to be involved in technological development, there might have been more room for accommodating these needs and desires, had they been taken into account at an earlier stage, when development was still relatively open. It could, in other words, help pre-empt Type III errors (solving the wrong problem) in the process of designing the innovation that is particularly prone to such errors as the main players – technology developers and doctors – share by and large the same medical rationality.
By going through such a process at an early stage of development, the technology and the features of its application practices may be designed so as to reflect much better the life world, needs and demands of recipients, which in turn may yield timely recognition of important conditions for application, wider acceptance (and hence application), and more health effects. By repeating part of the process after some experiences have been gained, further development may be oriented toward improvement of these concerns.
In the past few years, various approaches to involving patients and other considerations early in the process have been developed. One example is the ELSI approach to elucidate ethical, legal, and social implications of a technology under development (see, for instance, www.who.int/genomics/elsi). While this approach may help to timely anticipate moral, social, and legal issues, it pays much less attention to the feedback of such considerations into the development and design process. The latter aspect gains much more attention in a second approach, constructive technology assessment, or CTA (Schot 1992), incorporating understanding of innovation dynamics in the form of such notions as “innovation journeys,” “development trajectories,” etc. While iTA, when used in innovation processes, could benefit from these insights, it could add to CTA the notion of phronèsis, yielding a better understanding of how novel considerations could enter the process through creative, communicative judgment, taking into the account the discursive and material particularities of the situation. We would suggest to coin such an approach interactive technology design and development or iTDD.
Now turning to the practice and methodology of TA in such a context, we may be brief on the rules that should prevail. Given the above rationale, an orientation toward the collective, a capacity for representative thinking, and a keen eye for the life world of the various stakeholders are clearly indicated. Thus the rules should, again, focus on warranting the deliberative, participatory, transformative nature of technology assessment.
One crucial issue would be to determine in what cases to undertake such an iTDD. While generic prescriptions seem hardly possible, one circumstance under which more often than not doing a constructive iTA could make a difference is in the early stage of a non-incremental innovation, like CI, or the more recent joint applications of genomics, robotics, informatics, and nanotechnology (GRIN). Such an approach may also help resolve controversies (see again the CI case) or help where compliance appears problematic. Crucial and particular examples where the latter is important concern interventions that presuppose, or seek to promote, agency by the patient and/or her informal caretakers. With the increasing recognition of lifestyle as a health determinant, the increasing interest in prevention, and the growing emphasis on more patient agency in healthcare, there seems to be a reason to develop a better capacity for such forms of evaluation.
A final issue for the practice of such constructive iTDD concerns how to actually achieve phronèsis against the dominance of one particular rationality. Practices of designing medical innovation are likely to be characterized by what Flyvbjerg (1998) has called the “rationality of power,” a notion that refers both to the fact that a practice tends to be dominated by the rationality of the most influential actors and the fact that those who “possess” rationality are likely to have the power to impose their problem definition. To be sure, we are not necessarily, and certainly not primarily, referring to attempts to deliberately control others. What we are hinting at are more subtle mechanisms: for instance, that it is through the minds and hands of innovators that concerns of others are translated into different design choices or that designers tend to orient themselves on the practices where the innovation is to be realized and where most likely established medical rationality is to prevail. One thing that may help here is the participation of intermediaries who are able to translate the concerns and life world realities of patients to medical innovation design practices. Nurses and other paramedics, as well as patients with a (para)medical training, may make important contributions here. Thus, stakeholders may be conceived as parties who are constantly trying to further elaborate and test their interpretative frames. It is not difficult to see that some parties have and have had more opportunities to do so than others. iTA and iTDD can be seen as means to redress such inequality, in the common interest of creating a greater plurality in interpretation of our life world and in our repertoire for responding to this predicament (Schwarz and Thompson 1990).
The design and development of technology is a fascinating process, fascinating because it reveals the impressive capacity and achievement of mankind on the one hand and his limitations in developing a proper understanding of these activities on the other hand. Specifically, we seem to have difficulty in reaching a public understanding of the value of technologies (in Arendt’s words: action vs. work). Technology assessment has been devised to fulfill this role. However, particularly in the realm of Health Technology Assessment, strong emphasis has been placed on various types of risk-cost-benefit assessment. These provide powerful tools to assess whether specific technologies represent efficient and safe ways of achieving particular ends. They are, however, insufficient in supporting the value inquiry, characteristic of action. We believe that stakeholder engagement in combination with reconstruction of their interpretative frames renders TA much more relevant to this task. This is not trivial since, following Arendt, it is in action that humans are capable of developing their identity and of exercising their capacity for freedom. Last but not least, as a spin-off, health professionals may gain a more satisfactory relationship with their clients in this way.
It is in this vein that we seek to reconcile labor, work, and action and thus retain the relation between the vita activa as a whole and the betterment of the human condition. For technologies like cochlear implantations are products of work that are meant to improve the human condition, both from an individual and a societal perspective. Yet, they also show that when work and action get out of sync and communication between various actors falls short, intentions for the good easily turn out to be courses to the bad. Through deliberative, participative, and creative procedures, a broader perspective on the human condition may be retained, embedding work and action as aspects of a life world that should not be separated.
It is to Richard Feynman that the observation is attributed that “Philosophy of science is about as useful to scientists as ornithology is to birds.” Not surprisingly, perhaps, we disagree. It is through the work of Hannah Arendt and many others that we can come to see more clearly the potential value of technology assessment and why the current practice of technology assessment does not live up to this expectation.
- Blume S (1992) Insight and industry. On the dynamics of technological change in medicine. MIT Press, Cambridge, MAGoogle Scholar
- Fischer F, Forester J (eds) (1993) The argumentative turn in policy analysis and planning. Duke University Press and University College London Press, DurhamGoogle Scholar
- Fischer F, Gottweis H (eds) (2012) The argumentative turn revisited: public policy as communicative practice. Duke University Press, DurhamGoogle Scholar
- Flyvbjerg B (1998) Rationality and power. Democracy in practice. University of Chicago Press, ChicagoGoogle Scholar
- Gielen AJ, Grin J (2010) De betekenissen van “evidence based handelen” en de aard van “evidence”. Lessen rond rugscholen en radicalisering. Hoofdstuk 2 (p. 59–78). In: Verlet D, Devos C (red.). Efficiëntie en effectiviteit van de publieke sector in de weegschaal. Studiedienst van de Vlaamse Regering, BrusselGoogle Scholar
- Grin J, van de Graaf H, Hoppe R (1997) Interactive technology assessment: a guide. Rathenau Institute Report W57. The Hague, SDU. (in Dutch)Google Scholar
- Guba EG, Lincoln YS (1989) Fourth generation evaluation. Sage, Newbury PakGoogle Scholar
- Heller A (1999) A theory of modernity. Wiley-Blackwell, HobokenGoogle Scholar
- Hindmoor A (2006) Rational choice. Palgrave, BasingstokeGoogle Scholar
- Hoppe R (2010) The governance of problems. The Policy Press, Bristol, p. 132 ffGoogle Scholar
- Reuzel RPB (2002) Health technology assessment and interactive evaluation: different perspectives. PhD Dissertation, Radboud UniversityGoogle Scholar
- Schön DA, Rein M (1994) Frame reflection. Towards the resolution of intractable policy controversies. Basic Books, New YorkGoogle Scholar
- Schuitmaker TJ (2013) Persistent problems in the Dutch health care system: learning from novel practices for a transition in health care with the UPP framework. PhD dissertation, University of AmsterdamGoogle Scholar
- Schwarz M, Thompson M (1990) Divided we stand. Re-defining politics, technology, and social choice. Univ Pennsylvania Press, PhiladelphiaGoogle Scholar
- van der Wilt GJ (1995) Alternative ways of framing Parkinson’s disease: implications for priorities for health care and biomedical research. Ind Environ Crisis Q 9(1):13–48Google Scholar
- Venkatapuram S (2011) Health justice. Polity Press, CambridgeGoogle Scholar