Introduction

In 2008, the WHO Commission on Social Determinants of Health set forth a global agenda to eliminate health inequities in a generation, calling action on health inequities an ethical imperative. 1 To even partially respond to this target, better policy uptake of what is by now a very broad body of evidence about the social determinants of health and their negative consequences for health equity is badly needed. Yet, there is little evidence to suggest that peer review journals, whose functions are to screen research for rigor and originality and to consolidate a foundation for future research are the right vehicles for advancing the use of research (including health equity research) in policy. Knowledge Translation (KT) is a growing movement in clinical and health services research, aimed to help make research more relevant2 and to move research into practice and policy.3 KT encourages innovative knowledge sharing strategies that reach far beyond the traditional practices of scholarly publication and dissemination to create better opportunities for evidence-informed decision-making. On first glance, KT appears to offer promising strategies to help close the “know–do” gap and increase the application of research to reduce urban health inequities. On closer inspection, however, the promise of KT for urban health and health equity research falters. In this paper, we explore why this is the case, and we offer some ideas from policy studies that may enrich KT and extend options for researchers who seek to contribute to progressive, evidence-based social change in cities.

Health Inequities Research and the Need for Evidence-Based Change in Cities

Urban health research (and research concerning social determinants of health inequalities more generally) often looks at relationships between health and other social policy domains: how often do homeless shelter users visit hospital emergency departments, and what could reduce their need for urgent care? What is the prevalence of diabetes in urban Aboriginals — and who has the right to own or use this information to develop culturally specific interventions? Does mental health status improve among residents in revitalized social housing developments — and why? Do safe consumption facilities reduce overdose rates among drug users? And are they associated with increased crime in the neighbourhoods where they are located? These are vital issues to know about, if a society aims to provide relevant services and care to its most disenfranchised populations. But who needs to know, and what actions should the research community take to make this knowledge available, accessible, and useful to the appropriate stakeholders? These practical questions about problem-solving and mobilizing research evidence to influence urban health policy move us well beyond the activities of research itself, toward the domain of KT. Yet, when we pursue these questions, we find that they also reveal fundamental blind spots in the conventional KT framework about how, when, and under what circumstances, research evidence may influence policy-making. As we show below, these blind spots limit how useful KT strategies may be for urban health equity research and policy.

In what follows, we start by defining what has come be known as KT and trace its roots in evidence-based medicine. We then focus in on three major gaps, or what we hold to be misrepresentations of policy-making in the KT literature, to demonstrate why strategies based on a conventional KT framework are often ineffective for helping health equity research to have policy impact. Specifically, KT conceptualizes research utilization in terms of the technical implementation of scientific findings, on the part of individual decision-makers (or individual decision-maker groups, such as a particular type of professional) who can be “targeted” for a KT intervention, in a context that appears to be absent of political interests. The focus of KT analysis then turns on what is the most effective intervention to promote implementation.

A typical example, published recently in Implementation Science, compares two KT strategies to encourage adherence to hand hygiene guidelines among hospital nurses. The control strategy includes staff education, reminders, feedback, and targeting adequate products and facilities. The test strategy includes all of the above elements in the control strategy, supplemented with managers’ engagement and leadership, and setting team norms and targets.4 In our experience, urban health problems and what is needed to prompt urban health policy change do not resemble this clinical behaviour change model. In contrast, drawing upon concepts from policy studies, we show, first, that policy-making to promote urban health equity is likely to involve far more steps than the simple implementation of evidence-based solutions. Research evidence may or may not play roles in influencing change at any of these steps. Second, the complexity of the policy-making process requires abandoning the notion of a single decision or single decision-maker and acknowledging instead that policy change (or the maintenance of the status quo) tends to result from diverse, nonlinear negotiations among multiple actors operating within a policy network. Any number of these actors, or combinations thereof, may be potential “users” of research-based knowledge, for different purposes, at different stages in the policy change process. Finally, and most importantly, policy decisions that affect population health inequities occur, by definition, within a field of unequally distributed social, economic, and political power. Power relationships have direct and indirect effects on the ways that research evidence is received, perceived, and mobilized in the policy development process, and these need to be accounted for when planning and implementing a KT strategy for urban health. By supplementing the conventional KT model of policy development with the concepts of policy stages, policy networks, and a discourse analysis approach for theorizing power in policy-making, our goal is to clarify the conditions under which urban health research is likely or not to have an influence on public policy.

Making Research Relevant: The Emergence of Knowledge Translation in the Health Sciences

The most influential definition of KT, developed by the Canadian Institutes of Health Research, and recently adapted by the World Health Organization, is, “the synthesis, exchange, and application of knowledge by relevant stakeholders to accelerate the benefits of global and local innovation in strengthening health systems and improving people’s health”.5 KT models for influencing health policy have emerged through extrapolations of the tradition of Evidence-Based Medicine (EBM) to other domains of decision-making (EBDM).6 Sackett et al.7 describe EBM as “the conscientious, explicit, and judicious use of current best evidence in making decisions about the care of individual patients.” The implicit assumption underlying EBM is that exposure to relevant health care information will trigger research-accountable decisions and research-accountable decision-making behaviors among physicians who have been trained to follow research-based recommendations. KT strategies to extrapolate EBM to non-clinical decision-making contexts have focused on addressing: (1) characteristics of the research process and research product; and, (2) the research literacy of the “decision-maker.” Lavis has aptly framed such strategies as “researcher push” and “user pull,” respectively.8 Alongside these initiatives, complementary work has also been pursued, particularly in Canada, to advance a more robust account of social interaction in KT. In particular, Jonathan Lomas and the Canadian Health Services Research Foundation have made a major contribution by emphasizing the importance of effective communication and mutual understanding between researcher and decision-maker in order to promote better alignment between evidence that is needed and evidence that is made available. Some strategies include research uptake training for policy staff, to create greater demand for research, designing calls for more relevant research proposals in consultation with policy-makers, and involving policy staff directly in the research process of question formation, data collection, and data analysis.9 Likewise, the CIHR has endorsed the recursive Knowledge-Action Cycle which incorporates linkage and exchange approaches with proactive efforts to identify and overcome barriers to implementation, plus evaluation activities to assess implementation/application of research results.10

What has not changed, in the adaptation of KT from evidence-based decision-making, is the focus on evidence in evidence-informed decision-making, to the relative exclusion of attention to the process of decision-making itself. Researchers are increasingly encouraged to modify the topics of research and the presentation of research findings (e.g., improve accessibility, clarity of messaging, focusing on target audiences’ interests) and to educate decision-makers to appraise research evidence.11 This emphasis on evidence and relative inattention to decision-making itself has important implications for the relevance of KT in health equity policy-making.

Thus, it would appear at first glance that KT has something to offer those doing health equity research, particularly as it applies to urban settings. After all, the ultimate goal of such research is often to improve the health status of marginalized urban populations and to do that policy and program change is required. However, as the balance of this paper will show, there are at least three reasons to suggest that the KT model of policy making presented in the conventional literature needs to be substantially amended in order to be useful for urban health policy and research.

Policy Making for Urban Health Equity Is Much More than Implementation

As KT moves from clinical settings to the policy process, it retains a framework that characterizes the influence of research on practice (i.e., on “decision-making”) in terms of implementation of evidence-based solutions to clearly defined problems. For example, in clinical settings the theory is that effective KT will make it more likely that clinicians will follow clinical practice guidelines about patient care. As a result, one of the metrics for evaluating KT “success” is whether or not research-proven health care interventions are increasingly executed outside the research setting and that better health outcomes for patients are achieved. While the impact of KT on clinical decision-making has been mixed12 the complex policy implications of urban health research defy the reduction of evidence use to implementation. For example, a comprehensive policy response to research showing correlations between housing affordability and health status in a given jurisdiction13 could necessitate coordination across multiple departments, agencies and levels of government and would likely take place over a number of years.14 A problem-solving or solution-implementation model, based on clinical decision making for an individual patient, is an obviously inadequate analogy for such wide-sweeping, long-term policy change. A variety of conceptual frameworks are available from policy studies that can generate more nuanced accounts of policy change related to social determinants of health compared to models based on EBM. We have shown elsewhere that introducing the stages framework of policy making can be a useful heuristic model for advancing thinking about KT for policy change, even if the stages framework is itself but an approximation of reality.15

If nothing else, the policy stages framework emphasizes the simple but profound fact that policy making is often a long-term process. First, an issue or problem has to become part of the government’s agenda. Before research evidence can ever influence a government to solve a problem, there is the prior step of deciding that the problem exists, that it matters and that it can be addressed. Faced with literally thousands of issues that it could focus on, a government selects a platform, or portfolio of issues, which it believes it has the power, resources and political support to change.16,17 Of course, research evidence is but one input, but it can be — and often is — an important influence on agenda-setting and identifying that a problem exists that needs to be solved. Indeed, by presenting an empirical basis for some normative arguments (e.g., by modelling the broad scale and long-term outcomes of poverty reduction policies, or by raising awareness of the dire health consequences [and health care costs] of homelessness and the potential benefits that could be obtained through investments in social housing), urban health research can play a particularly important role in agenda setting.

Additional stages of the policy development process include assessing the range of possible responses to the problem against a diverse range of criteria (of which scientific evidence is but one such criterion); coordinating and negotiating a series of decisions by a diverse set of actors over a period of time; eventually implementing the programs authorized by the policy; and, finally, in a well-performing policy-making system, evaluating and modifying the policy decisions.18 Research evidence may (or may not) be taken into account during any of these stages. Enriching conventional KT with the stages framework helps to show a spectrum of possible entry points for research in the policy-making process, far beyond simple implementation. It can guide urban health researchers to become more strategic analysts of the policy process, and better equipped to contribute their evidence at different moments in the process.

KT for Policy: Many Decisions by Many Actors

In the KT account of policy development, which as we have seen is modeled after clinical decision-making situations, attention focuses on the behavior of an individual decision-maker organization/sector to whom the research project and findings should be narrowly “targeted” in order to trigger individual behavior change and action.19 In the KT literature it is routinely asserted that evidence that is not tailored or packaged for a target audience is less likely to be reviewed or adopted into use.20 Another commonplace assertion is that “linkage and exchange” activities to engage a specific decision-maker partner directly in the research process will produce research evidence that is more relevant and more likely to be applied in practice.21,22

Our experience of urban health problems and interventions does not resemble this single decision, single decision-maker model posited by KT. Instead, urban health challenges typically involve multiple sectors and levels of government and can be relevant to diverse stakeholders of varying degrees of influence, only some of whom may actually make decisions, for example, clients, advocacy groups, journalists, the judiciary and police, and private sector stakeholders, in addition to elected officials and government bureaucrats. For example, which unique sector, or level of government, for example, would be the ideal participant and target audience for research related to social housing and health?13 Should researchers emphasize linkages and exchanges with state or provincial representatives in the health care sector (who have limited responsibility for housing); with municipal representatives responsible for administering (but unable to expand investments in) social housing; or with legislators in national government who are promoting (perhaps at the behest of advocacy groups) a national housing strategy? Surely a better answer than to target just one key audience for this research would be to forge connections and collaborations with all these decision-makers, and especially with other important stakeholders as well.

In the political science literature, the concept of the policy network has emerged to counter “ideal type” depictions of decision-making in institutional hierarchies and to better represent policy processes as they actually occur.23 From the policy network perspective, focusing on formal and macro-level decision-making bodies like Congress or Parliament or, in an urban context, City Council or state or provincial government, ignores critical aspects of the policy-making process which involve the circulation of resources of information, support, and authority across state and non-state actors in relation to a particular policy problem or set of problems. It is in these decentralized and more or less coordinated interactions between state and societal actors that policy making — particularly complex policy-making — unfolds. The policy network concept builds on the view, described above, of policy change as a staged process. Using a policy network lens enables us to “map” the presence of influential parties in the policy development process, the interrelationships between actors, and the types of power they can mobilize.24 Understanding the array of policy stakeholders who are involved at different stages of policy change, researchers and KT staff can make more strategic and thoughtful decisions about whom to transfer and exchange their knowledge with, when, and for what purposes. For example, the policy network emphasizes the importance of the interpersonal relationships between individuals and structural linkages between a wide range of actors in government, in the private sector and among nongovernmental organizations. With respect to social housing and health, a policy network perspective draws our attention away from an exclusive focus on governments and broadens the analysis to include private developers and land owners, the construction industry, and the myriad of social advocacy organizations seeking to expand the stock of social housing.

Generally speaking, an effective response to social determinants of health will almost inevitably require a mode of collective action, negotiation and coordination among an array of diverse stakeholders25 — a view of policy change that move us quickly into the realm of politics. Yet, conventional KT, which focuses on changing technical and practical decisions made by individual actors, has relatively little to say about collective processes, or about cultivating relationships with diverse, not to mention, competing, stakeholder groups. Introducing a policy network approach to KT shifts our attention to how a policy issue is understood and framed by a much wider range of influential actors, and encourages us to undertake a more nuanced analysis of how diverse groups could use — or ignore or abuse26 — research evidence.

KT For Policy: Taking Power and Politics Seriously

The inherent complexity of urban health policy making, in terms of both the multiplicity of steps and the multiplicity of actors and interests involved, cannot meaningfully be grappled with absent a clear analysis of the specific power relations at play with respect to the particular policy issue in question. Yet, the KT literature is all but silent on the role of power in determining which ideas rise to the top of policy agendas, which survive to the stage of implementation, who has the authority and resources to negotiate policy solutions, and how particular courses of action are perceived as acceptable options or not. The third, and for urban health research, the most limiting blind spot in the KT model of policy development is this lack of a theory of power. Here we use a discourse analysis approach to explore just two related dimensions of this challenge. First, how power relations shape the ways that policy problems and solutions are defined, and second, the potentially damaging consequences for researchers who describe policy options in terms that are oppositional to prevailing definitions and norms.

Central to the conventional KT literature is the tacit assumption that researchers and policy-makers will agree on the meaning of a health policy problem; that is, they accept the same framing of the problem.27,28 This discursive agreement is required for researchers to generate evidence-based recommendations that are salient to, and executable within, a decision maker’s range of action and priorities. A further step is to include government and health system officials directly in the research process either by means (deemed “linkage and exchange” or what some have called “integrated KT”).29 When an issue is noncontroversial, researchers, policy makers and the wider communities that are affected may share a common, uncontested definition of the problem. This situation is often the case for some (but certainly not all) clinical or health services issues (e.g., there is relatively little disagreement that infection control in health care settings is in part due to inconsistent hand hygiene practices among health care providers, as described above). However, many urban health problems decidedly not uncontroversial and their definitions are regularly disputed.

For example, there is a multitude of competing ways of framing illicit substance use — a pervasive urban health challenge, including addiction and pathology, criminality, mental illness and self-medication, and cultural deprivation. These different meaning frames will lead participants to different accounts of what matters in relation to substance use, what needs to be done and who is responsible for doing it. Importantly, these diverse discourses are not equally authoritative or persuasive. Rather, the authority to name or frame a problem and make it stick is a marker of power, and the struggle to challenge, refute and redefine meaning frames through discourse is the stuff of politics.30 The discursive strategies that shape how a problem is defined in turn influences the types of solutions that may be sought by different stakeholders. For example, determining whether injection drug use should be approached by using a criminal justice lens or a public health lens has dogged policy related to Vancouver’s Insite supervised injection site for several years.31 In cases such as these, there may be limited or no political willingness for decision-makers who use a criminal justice lens to coordinate with harm reduction researchers in problem solving. This simple reality can make the KT goal of close cooperation between researchers and decision makers an awkward proposition.

On the contrary, one of the goals of health equity research may be to resist how powerful constituencies define issues affecting marginalized populations and to persuade decision-makers to understand problems and solutions in a new light. However, conventional KT accords relatively little value to research impacts that are reflected in changed attitudes rather than changed actions and it provides minimal guidance about how to advance controversial research evidence among decision makers. Instead, when researchers propose problems or solutions that do not align with prevailing meaning frames, they become vulnerable to having their KT efforts conflated with “advocacy,” which is said to undermine their claim to scientific credibility.32 The reference to advocacy is not in itself inaccurate. Indeed, it can be argued that all KT is a mode of advocacy — on behalf of the use of evidence in decision-making. However, when the charge is invoked to discipline researchers and to silence research that runs counter to prevailing discourses, we see clearly that what determines whether policy ideas survive or not is always shaped, at least in part, by the interests of powerful social groups, regardless of the strengths of scientific evidence.

Conclusion

How can urban health research (and researchers) make a stronger contribution to public policy making and help reduce population health inequities in cities? KT frameworks have been developed to help move research evidence into practice and realize better health outcomes for patients and populations; therefore, it seems natural to look to these frameworks as a way of advancing the impact of urban health evidence. However, few urban health problems bear a strong resemblance to the clinical or health system problems posited in the KT literature. There are serious challenges associated with applying conventional KT principles to support the uptake of health equity research. First, KT posits (and measures) research uptake in terms of discrete, implementation events. This conception fits very poorly with the realities of public policy making and program design where even the simplest models assume that action is the result not of one decision but many, taken over a more or less long period of time. Measuring the impact of research on policy making in general, and on policy that deals with urban health challenges in particular, cannot, therefore, be reduced to a simple implementation model of research impact. Second, conventional KT is concerned with individual actors and actions. But addressing urban health problems requires choices and change on the part of a large number of actors who come together in policy networks and communities. Third, and most important, conventional KT is silent on the role of power and politics in shaping public policies. As a consequence, KT strategies can be difficult to execute in relation to urban health issues, and they may not be the optimal approaches for increasing the societal value and impact of urban health research. In this paper, we have shown that enriching KT with the concepts from policy studies can help generate a more adequate account of the conditions under which urban health research may be relevant for urban health policy. Importantly, these new concepts reveal not only more potential barriers but also a broader array of potential opportunities for research uptake than the conventional KT model permits.

To strengthen the link between research evidence and on-the-ground advancements in urban health, much more work to expand the conceptual limits of KT is badly needed. As Greenhalgh and Wieringa33 have recently observed of the KT canon generally, we need a “wider range of metaphors and models” that “allow us to research the link between knowledge and practice in more creative and critical ways” and “move beyond a narrow focus on the ‘know-do gap’ to cover a richer agenda.” In urban health, at the very minimum, these models will need to account for urban policy making as a process, not a discrete event, and will not shrink from the idea that policy, especially equity policy, is informed by politics.