Public Deliberation in Health Policy and Bioethics : Mapping an emerging , interdisciplinary field

For over two decades, the "deliberative turn" has rooted itself in the fields of health policy and bioethics, producing a growing body of deliberation in action and associated academic scholarship. With this growing use and study of citizen deliberation processes in the health sector, we set out to map this dynamic field to highlight its diversity, interdisciplinarity, stated and implicit goals and early contributions. More specifically, we explored how public deliberation (PD) is being experimented with in real-world health settings, with a view to assessing how well it is meeting current definitions and common features of PD. Our review provides an informative and up-to-date set of reflections on the relatively short but rich history of public deliberation in the health sector. This emerging, interdisciplinary field is characterized by an active community of scholars and practitioners working diligently to address a range of bioethics and health policy challenges, guided by a common but loosely interpreted set of core features. Current definitions and conceptualizations of public deliberation’s core features would benefit from expansion and refinement to both guide and respond to practice developments. Opportunities for more frequent cross-disciplinary and theory-practice exchange would also strengthen this field.


INTRODUCTION
For a little over two decades, the "deliberative turn" has rooted itself in the health field, producing a growing body of deliberation in action and associated academic scholarship.The growing popularity of citizen deliberation processes in the health sector should not be surprising.Their emphasis on informed, value-based reasoning and collective problem solving is appealing to a sector where some of society's most ethically controversial, fiscally challenging and politically charged decisions are made.Indeed, one of the earliest examples of the use of public deliberation to inform health policy is the now-famous Oregon Health Plan case, in which Oregonians were invited to provide guidance for prioritizing the medical conditions to be covered by the state's Medicaid program (Crawshaw et al., 1985).
Serious experimentation with public deliberation in the health sector began in the 1990s in the UK's National Health Service (NHS) where a variety of methods such as deliberative polling, citizens panels and citizens juries were used to inform local health care priority-setting decisions (Bowling, Jacobson & Southgate, 1993;Bowie, Richardson & Sykes, 1995;Lenaghan, New & Mitchell, 1996;Lenaghan, 1999;Coote & Lenaghan, 1997;Cookson & Dolan, 1999;Dolan, Cookson & Ferguson, 1999).This activity has been replicated to a lesser degree in other jurisdictions around the world, mostly through ad hoc researcherinitiated studies (Abelson et al., 1995;Bostwick, 1999;Stronks et al., 1997;Smith & Wales, 1999;Einsiedel, 2002;Abelson et al., 2003;Forest et al., 2004;Abelson et al., 2007;Menon and Stafinski, 2008;Secko et al., 2009;O'Doherty and Hawkins, 2010;DeVries et al., 2010).However, the establishment of several high-profile citizen deliberation processes and structures over the last decade suggests that public deliberation may be gaining some traction among policymakers as a mechanism for incorporating public values into health policy decision-making processes.In the early 2000s, for example, the principles of democratic deliberation inspired the design of twelve citizen dialogues held across Canada, which informed the recommendations tabled by a government-appointed royal commission, tasked to weigh in on the future of health care in Canada (Maxwell, Rosell & Forest, 2003).At approximately the same time, the UK body which oversees the assessment of health technologies -the National Institute for Health and Clinical Excellence (NICE) -created a 30-member Citizens Council to provide a source of social values input to inform their expertdriven, evidence-based health technology assessment process (Rawlins, 2009).The establishment of NICE's Citizens Council has, in turn, influenced the creation of two legislatively mandated public deliberation structures in Canada, one that provides advice on the Province of Ontario's pharmaceutical benefits program (Statutes of Ontario, 2006) and the other that provides advisory input on the performance of the Quebec health system (Government of Quebec, 2005).
Given the growing use and study of citizen deliberation processes in the health sector, we think there is merit in mapping this dynamic field to highlight its diversity, interdisciplinarity, stated and implicit goals and early contributions.More specifically, we are interested in exploring how public deliberation (PD) is being experimented with in real-world health settings, with a view to assessing how well it is meeting current definitions and common features of PD.Our motivation for this assignment was fuelled by our experiences as scholars and practitioners of public deliberation faced with the challenges of implementing and assessing PD within varied social and political contexts.To carry out this task, we drew on a selected body of empirical studies to examine the questions of how PD is operationalized in relation to its theorized goals.We hope that our results will offer some critical reflection on current definitions and how they might be expanded and strengthened.

METHODS
We carried out a targeted search of the published literature with a focus on original articles of empirical studies of public deliberation methods, practices and evaluations in the related fields of bioethics and health policy.

Sources, search strategy and article screening
Twenty databases of published English-language literature in the medical, health, social sciences and humanities were searched from 2000 to 2012.Search terms were developed through an iterative process to identify key component terms of interest such as: 1) public; 2) engagement, involvement, and deliberation; 3) bioethics and health (care) policy.Searches were carried out using different combinations of terms within each database to ensure that relevant material was not missed.The following two search strategies were undertaken: Search #1: ("public deliberat*") and (health* or ethics* or (health policy)) and (("public engage*") or ("public involve*")); search #2: ("public deliberat*") and KW=(health* or bio* or ethics).The collection of articles yielded through these searches was further screened to eliminate duplicates, and to include only empirical studies, which included "deliberation," "deliberate" or "deliberative" in the title or abstract.This search was complemented by a review of our respective literature databases constructed through previous work in the field and by our knowledge of scholarly and practice-based activity in the public deliberation field.

Review and analysis of selected articles
Our search yielded a set of 31 articles published between 2000 and 2012, which are described in Appendix 1 and discussed below.As a collection, this body of work helps us to describe the work being carried out in this arena and how it has evolved over the past decade.A second phase of our analysis involved the selection and in-depth exploration of a subset of the papers summarized in Appendix 1. Eleven papers were selected with the aim of covering a broad range of study aims, jurisdictions, decision contexts and approaches to design.A detailed review of these articles was carried out to extract relevant information corresponding to the core features of PD as described in a recent review of the PD literature (Blacksher et al., 2012).This review was used as the basis for our critical examination of current approaches to using PD to address health system challenges.

DESCRIBING THE FIELD
Our selective review of the public deliberation and health literature yielded a rich set of empirical studies that have examined the use and outputs of public deliberation or explored and assessed its design feature(s).The bulk of the studies reviewed were carried out at the local/regional, provincial/state or national level in Canada, the US, and the UK, with selected representation from Australia, Taiwan, and Israel (Rogers et al., 2009;Molster et al., 2011aMolster et al., , 2011b;;Guttman et al., 2007Guttman et al., , 2008;;Deng & Wu, 2010).
As Appendix 1 illustrates, the studies we reviewed cover a wide range of policy decisions, levels and content areas.Building on its early roots, there continues to be a strong emphasis on the use of public deliberation to inform local and state/provincial level planning and priority setting decisions such as planning for flu pandemics, setting local health goals, designing health services and locating health facilities (Abelson et al., 2003;Abelson et al., 2007;Rogers et al., 2009;Baum, Jacobsen & Goold, 2009).Public deliberation also continues to be used as a source of values input to inform various policy decisions related to the coverage of health services and more recently, health technologies (Bracci, 2001;Grogan & Gusmano, 2005;Paul et al., 2008;Bombard et al., 2011;Menon & Stafinski, 2008).The field is equally well represented by deliberations covering an expanding range of research ethics and regulatory policy issues including xenotransplantation, surrogate consent, biobanking and genetic testing (Einsiedel, 2000(Einsiedel, , 2002;;DeVries et al., 2010;Molster 2011Molster , 2012;;Maclean & Burgess, 2010;Longstaff & Burgess, 2010;O'Doherty & Hawkins, 2010).And while fewer in number, we also unearthed several noteworthy examples of deliberations used to inform large-scale health reform processes in Canada, Israel and Taiwan where major decisions about the structure, financing and organization of health systems were under review (Maxwell, Rosell & Forest, 2003;Guttman et al., 2007Guttman et al., , 2008;;Deng & Wu, 2010).
Despite a proliferation of methods now associated with the deliberative approach, the studies we reviewed continue to ascribe fairly generic labels to the form of deliberation being undertaken (Appendix 1, column 3).A notable exception is the citizen jury method, which is more consistently and accurately described across studies, likely due to the pioneering efforts of Ned Crosby and his founding of the Jefferson Center for New Democratic Processes in 1974(Jefferson Center, 2013).As the field has evolved, there has been greater specification of the core features of deliberation, and more consistency within research groups that have established track records in the field.However, method descriptions continue to be characterized by incompleteness or ambiguous terminology, challenging efforts to draw lessons from the field.
In contrast to the overall lack of method specificity, details about the recruitment and selection of participants is carefully tailored to the study context (Appendix 1, column 4).Common approaches include stratified random sampling from an available source (e.g., electoral rolls, community organization directories, etc.), and purposive sampling through various advertising methods (e.g., newsletters, websites, local media) followed by stratification on key demographic variables to ensure heterogeneity in the composition of deliberation participants.The field would benefit from more critical review of different recruitment methods and their rationales, which, to date, have been only superficially examined (Longstaff & Burgess, 2010).

How the field of PD in health has evolved
Public deliberation scholarship in the health field has evolved considerably over the last decade.Prior to 2003, the literature focused mostly on case studies documenting the outputs of public deliberation in various decision-making contexts with only limited attention to specific aspects of deliberation such as how it was implemented, with what success and to what extent its features aligned with the theory underpinning deliberative democracy.Since then, the field is characterized by more critical reflection and an emphasis on assessing its achievements.Bracci (2001) laid the early foundations for the development of evaluation and ethics frameworks through the evaluation of the patterns and ethical dimensions of state-level public deliberation and their influence on priority-setting and resource allocation.Similarly, Einsiedel and colleagues' (2002) work has inspired more rigorous evaluations of public deliberation through the use of a priori evaluation criteria and principles and comparative evaluation designs (Abelson et al., 2003;Grogan & Gusmano, 2005;Molster, 2011Molster, , 2012;;Abelson et al., 2007;Timotijevic & Raats, 2007).More recently, De Vries and colleagues (2010) have attended to these issues even further by developing and testing measures of "quality deliberation."While much of this activity has emphasized the evaluation of procedural elements, some attention is also being given to exploring the role of context in public deliberation, which involves unpacking the complex relationships between external environments, institutional settings and decision-making processes, and how these shape deliberation activities and their impacts (Abelson et al., 2007;O'Doherty & Hawkins, 2010;Jones & Einsiedel, 2011).Continued attention to these outcome-related dimensions of public deliberation is essential to answering central questions about its impacts on health policy decisions.

The interdisciplinarity and diversity of the field
Our selective review of published literature reveals that no single outlet for scholarly work exists in the field, although there appear to be preferred publication venues (Table 1).Of the 31 papers summarized in Appendix 1, two interdisciplinary health journals -Social Science and Medicine and Health Expectations -have published just over 40% of the papers.Beyond these two dominant outlets, however, the work is distributed fairly evenly across a diverse set of interdisciplinary health, social science and science journals.

EXAMINING ITS CORE FEATURES
As highlighted in the previous section, public deliberation and its core features have been translated into the health field in variety of ways and with different purposes.Drawing on an extensive body of theoretical work, Blacksher et al., (2012) propose the following minimum definition for public deliberation: (1) the provision of balanced, factual information that improves participants' knowledge of the issue; (2) the inclusion of diverse perspectives to counter the well-documented tendency of better educated and wealthier citizens to participate disproportionately in deliberative opportunities and to identify points of view and conflicting interests that might otherwise go untapped; and (3) the opportunity to reflect on and discuss freely a wide spectrum of viewpoints and to challenge and test competing moral claims.(p. 3) Each of these three elements emphasizes a different core feature or goal for deliberation.The first emphasizes the goal of ensuring that citizens are properly informed about the issue under deliberation; the second establishes the importance of paying careful attention to the representation of relevant and divergent perspectives; and the third focuses on the process created for free and open exchange to facilitate values-based reasoning.Blacksher and colleagues go on to stress the normative nature of public deliberation and its focus on reaching collective judgments about what "we ought to do," which bear on the more pragmatic goals of locating common ground and producing a set of deliberation "outputs" that can inform policy decisions.There has been much discussion in the literature about the merits and associated problems with each of these goals, but much less attention paid to how they are operationalized as unique or complementary features of deliberative practice (Blacksher et al., 2012).For example, what does it actually mean to include diverse perspectives in a deliberative process?What does a deliberation space look like that facilitates value-based reasoning?And to what extent have the architects of deliberation in the health field encouraged movement towards collective judgments and common ground, while ensuring that participants have the time and space to disagree and to air departures with dominant voices and interests?
Through our in-depth review of the subset of 11 papers, we examined how each element of the proposed minimum definition was articulated and put into practice through the design and implementation of various public deliberation initiatives in the health field.A detailed set of findings including verbatim excerpts from the papers we reviewed are presented in Appendix 2 and critically examined in the following sections.

Information provision and knowledge acquisition: Convergence of approaches
Within the subset of papers we reviewed, we found that information provision, as a core element of PD, was attended to quite carefully (Appendix 2, column 3).Most if not all efforts included as the principal information source a summary or briefing about the deliberation topic prepared by the research team responsible for running the deliberation.Presentations and/or Q&A sessions with content experts were offered as a complementary source in a number of initiatives, and in some cases, relevant "stakeholders" were invited to present a range of positions or interests on the issue.While not explicitly stated, this dual approach, with its primary emphasis on the synthesis of information by the research team, rather than invited experts, reinforces the commitment to providing "balanced, factual information."Indeed, some have identified criteria used to guide the preparation of information materials such as comprehensiveness, accessibility and objectivity (Einsiedel, 2002).While there appears to be convergence on the general approaches to informing participants, these have not been supported by rigorous evaluation to determine the most effective or efficient combination of information dissemination modalities.

Representation, representativeness and diversity: conflicting approaches and conflated ideas
The issue of which perspectives to "represent" in the deliberative process, and how to do this was the source of considerable ambiguity and divergence across the PD initiatives we reviewed (Appendix 2, column 2).Stated goals varied from seeking the statistical representation of community and population characteristics to more abstract notions of including "diverse perspectives" which may or may not include demographic diversity.Different emphases were also given to the goals of representing the interests of those "affected" by an issue or content area versus intentionally seeking a more "disinterested" set of perspectives.A notable gap in the examples reviewed were explicit references to representing "untapped" viewpoints as described in Blacksher et al.'s (2012) minimum definition, suggesting that this feature has not been adopted in practice.
The goals of representation and diversity were approached in other important ways in the PD initiatives reviewed, typically through efforts to ensure that the deliberation process itself allowed for a diversity of perspectives to be shared.This subtle yet critical distinction between representing a diversity of population characteristics vs. a diversity of viewpoints points to a weakness in our minimum definition and how the concepts of representation and diversity have been addressed in the PD literature to date.More specifically, notions of representing a diversity of perspectives through the recruitment process (e.g., by socio-demographics, interest characteristics, etc.) seemed to be conflated with the notion of designing a deliberative process that allows for a diversity of viewpoints to be shared (irrespective of the representativeness of participants).Some of the studies we examined were directly concerned with one concept or the other but rarely distinguished between the two or described how both might be achieved.

Value-based reasoning: specifying its key features
As one approaches the intricacies of deliberation, the feature of moral or valuebased reasoning is particularly challenging to specify, let alone decipher from accounts of public deliberation.What are its inherent features?How do we know when it is present?And what are the necessary conditions for supporting it?The studies we reviewed provided few clear depictions of how value-based reasoning was explained to deliberation participants and carried out (Appendix 2, column 4).For example, were values frameworks presented a priori or were participants expected to engage in value-based reasoning tabula rasa?That said, a number of supporting conditions appear to be associated with value-based reasoning.First, and most obviously, it occurs within structured task-oriented discussions.More specifically, discussions within smaller "break-out" groups is the typical setting used to identify the values underpinning various positions on an issue.The nature of these discussions is sometimes described as "collective problem solving," where individuals "with different backgrounds, interests, and values listen, understand, potentially persuade and ultimately come to more reasoned, informed and public-spirited decisions" (Abelson et al., 2007(Abelson et al., , p. 2117)).Many studies referred to two-step processes that involved an initial plenary discussion emphasizing common understanding of the deliberation topic, followed by smallgroup activity, reporting back and large-group discussion.
Moral or value-based reasoning also involves the complementary activities of knowledge acquisition, questioning and debate -all routinely cited in the studies we reviewed.More specifically, PD initiatives referred to participants "challenging," "questioning," "justifying" and "debating" each other and invited experts.While the act of identifying "ethical challenges," "tensions" and "tradeoffs" suggests the presence of free and open exchange, it still falls short of specifying the precise elements of value-based reasoning.Only a small number of studies we reviewed identified "normative criteria" to support a particular position or recommendation or used existing "moral frameworks" to guide deliberations (Bennett & Smith, 2007;Bombard et al., 2011).Building on these examples would be a fruitful avenue to explore in future work in this area.

Finding common ground: the need for greater transparency
Deliberative processes emphasize shifts from individual to collective stances.The act of reaching common ground and yielding collective outputs are its archetypal features.We found many clear statements of the type of collective output sought or obtained from deliberations (Appendix 2, column 5).For example, participants were described as "reaching a verdict," "producing recommendations," "developing majority opinions" and "reaching reasoned, informed and publicspirited decisions."Yet, as with our discussion of value-based reasoning, few details were provided about the steps taken to arrive at these outputs.For those approaches that were described more explicitly, they differed substantively, suggesting different epistemological approaches to the deliberative enterprise.For example, some PD initiatives sought "majority" and "dissenting" opinions as their collective output, acknowledging from the outset that consensus was never the goal.Others emphasized the process of identifying shared values and areas of agreement (qualitatively) and "collective preference statements" (quantitatively).A compromise position taken in a number of initiatives was one of "working towards consensus while allowing for disagreement."Regardless of the desired output for deliberation, the lack of clearly described methods for producing them needs to be addressed if they are to be given serious consideration (e.g., Who synthesizes the deliberation discussions and how?What is the process for vetting the outputs with deliberation participants?) Underlying these operational issues is a perplexing set of ambiguities.While the studies we reviewed included clearly identified outputs, the diversity of these outputs is striking.Moreover, their alignment with the related goal of yielding collective judgments on social issues (i.e., what we ought to do) is unclear and also likely to vary considerably.For example, does reaching a verdict, which is the goal of a citizens' jury, involve the same collective effort as developing recommendations or producing majority and dissenting opinions?And are these outputs, as well as efforts to identify shared values, moving citizen deliberators enthusiastically toward the goal of locating common ground without having thoroughly canvassed their views about what ought to be done?These issues have been fiercely debated in the broader PD literature but have not received the same level of attention as PD has been rolled out in the health field.

Informing policy through public deliberation
While the vast majority of public deliberation efforts in the health field have been initiated by researchers, they are motivated by a strong desire to address real-world policy problems and to inform relevant policy processes.We observed a variety of relationships between the initiators of public deliberation processes and relevant "policy receptors" (Appendix 2, column 6).Three types of relationships were generally observed: i) researcher-initiated PD activities which resulted in a report or recommendations that were presented or submitted to a relevant policy/decision-maker; ii) decision-maker-initiated activities designed to inform a particular policy process (less common); and iii) PD activities resulting from collaborations between researchers and decision-makers that yield public input on a particular topic while providing opportunities to study various aspects of deliberation.The establishment of formal links between PD and relevant policy decisions through the types of relationships described here would certainly appear to facilitate the uptake of PD outputs.However, efforts to document the impacts of PD on policy have begun only recently in the health field and have revealed the complexity of linking PD outputs to decision-making, the long-term horizons required to trace PD impacts, and the need for further conceptualizing about what constitutes "use" or "uptake" of PD outputs.
For example, Jones and Einsiedel (2011) traced the "institutional learning" impacts of a deliberative public consultation that employed a citizens' jury model in the early 2000s (Einsiedel, 2002).Their results revealed that despite the lack of substantive policy change, the exposure to the method of public deliberation facilitated a shift in institutional culture, which was reflected in a greater openness to including a broader set of actors in the policy process.And in their evaluation of the early years of the NICE Citizens Council, Davies et al., (2006) observed the Council's initial lack of "embeddedness" within the inner workings of NICE and the need to strike the right balance between maintaining the Council's independence from the sponsoring organization and ensuring that its outputs were on the organization's radar.Council members called for greater clarity about their role in relation to NICE decision-making and for greater transparency about how their input would be used.These findings were supported recently by Abelson et al., (2013) where a citizens panel established to inform the health technology advisory process in Ontario, Canada exerted various impacts on the HTA process that were mediated by direct exchange between citizens and experts, to clarify roles, promote accountability and build trust.As these examples illustrate, even when fully supported, public deliberation operates within a complex array of organizational settings and political structures that shape the degree and manner in which it is likely to exert tangible influences on policymaking and other aspects of collective problem-solving.

DISCUSSION
This mapping exercise is neither an exhaustive nor a comprehensive review of the full spectrum of public deliberation activity in the health field.By restricting our focus to peer-reviewed literature, we are aware that we have excluded learning from a rich field of practice, which has supported the study of public deliberation in the health field.Equally, our limited but deliberately narrow search parameters (e.g., empirical studies; deliberat* in the title or abstract) have excluded important theoretical contributions as well as empirical work that may have aspired to the principles of public deliberation without using its label (e.g., Ginsburg, Goold & Danis, 2006;Goold et al. 2005).Despite these omissions, we believe that our review provides an informative and up-to-date set of reflections on the relatively short but rich history of public deliberation in the health sector.We have explored the various ethical controversies and public policy problems that these types of processes have sought to address, and how the approaches used and features examined, have evolved over time.
Our review reveals a small but active interdisciplinary community of scholars and practitioners working diligently around the world to address a range of bioethics and health policy challenges using the principles of deliberative democratic theory.The desire to contribute to pressing health challenges has been a strong motivator within this community.Public deliberations have been undertaken to provide values input to national health reform initiatives and to local pandemic flu planning; they have contributed to the assessments of ethically controversial new technologies and have assisted communities to prioritize health services and benefits packages within constrained budgets.Despite a growing number of deliberative methods available (e.g., juries, panels, deliberative polling), most processes used in the health field ascribe to generic principles to allow for maximum tailoring to decision-making and issue contexts.
When it comes to operationalizing the principles of deliberative democracy theory, the subset of empirical studies we reviewed fared well, at first glance, in meeting a minimum definition provided for public deliberation.With few exceptions, clear efforts were made to inform participants, to represent a diversity of perspectives, to engage in free exchange and value-based reasoning and to locate common ground that would produce outputs that could potentially inform a set of decisions.In this sense, our minimum definition aligned quite well with what we found in practice.However, once we moved beyond this superficial level, to observe how each of these elements were implemented, we were struck by the degree of heterogeneity and ambiguity with which these core features have been operationalized, which are likely driven by the limitations of current definitions, the limited attention paid to them, or both.For example, the concepts of representation and diversity appear to be conflated in public deliberation studies, calling for greater specification of these goals and how they are to be achieved (i.e., through participant recruitment, through the deliberative process or both).Similarly, more attention needs to be given to clarifying, for practical purposes, what is required for value-based reasoning that produces collective judgments to occur.This could be done in two potentially complementary ways: i) inductively, through in-depth qualitative analyses of deliberative processes; and ii) by drawing on ethics and normative theory to provide models and frameworks.Finally, despite the strong motivations and formal links being established to ensure that the outputs of public deliberation inform decision-making, much more work is needed to understand what "use" and "uptake" of PD consists of and how to measure it.
In reflecting on this mapping exercise and the areas we have identified for further development, we see numerous opportunities for disciplinary and interdisciplinary scholars to contribute to this emerging, interdisciplinary field and to help expand, strengthen and more fully operationalize the proposed minimum definition.Communications and psychology scholars would provide valuable expertise to evaluate and improve the quality of the informational and social interaction features of PD.Philosophy and ethics expertise would bring needed rigor to the moral reasoning endeavour.Policy analysts are needed to map and assess the PD-policy linkages.Beyond these valuable disciplinary contributions, we see an even more critical role for more frequent and meaningful interactions between PD theorists and practitioners to learn from and enrich each others' work.

CONCLUSION
Our examination of citizen deliberation processes in the health sector has revealed a solid set of contributions covering a broad range of content areas, jurisdictions and aims.As experimentation and interest in citizen deliberation continues to grow, scholars and practitioners need to be rigorous in their approaches to design, implementation and evaluation.Current definitions and conceptualizations of PD's core features would benefit from expansion and refinement to both guide and respond to practice developments.Opportunities for cross-disciplinary exchange have facilitated this type of work already and should continue.We look forward to the next wave of innovations and experiments in this dynamic field.

Genetic testing and access to life insurance
To consider the strengths and limitations of citizens juries as a research method to elicit ideas about complex policy decisions; and to examine the outcomes of one citizens jury on genetic testing and life insurance Guttman (2007) See also Guttman et