The blind spot of contemporary science is experience Maturana and Varela, 1987, as cited in Senge, Scharmer, Jaworski, & Flowers, 2005, p. 196

The purpose of this special supplement is to present a set of research papers by Carole Estabrooks and her team on the topic of research use in practice. Significantly, in addition to the original research papers, a number of leading researchers, scholars, and policy makers in the field of knowledge translation (KT) have been invited to address key elements that impact upon research developments in the area and to critique the work of Estabrooks et al.

Through this process of inquiry, we hope to demonstrate the importance of scholarly critique and how the development and refinement of the research approach itself is enhanced through continuous dialogue and debate. Everyone who has been involved in the generation of the supplement-Carole and her team, research colleagues, and the many reviewers-has contributed both to the process of knowledge generation and to the outcomes presented in this supplement. In this respect, it is important to acknowledge the impact of the process we chose to follow as much as the results, that is, the papers, themselves.

Interest in KT and implementation science is escalating. From the original studies of Horsley, Craine, Crabtree, and Wood (1983) in nursing, there have been many iterations, and much has been learnt. Bibliometric analyses of the research utilization literature have recently been published in Nursing Research (Estabrooks, Winther, & Derksen, 2004), and a growing number of reviews (many of which are referenced in the supplement) are available to help identify our current state of knowledge of what works.

From a policy perspective, health administrations across the world are looking to understand how best to improve the quality, effectiveness, and safety of the healthcare they deliver. They have been influenced by two significant movements in health policy, namely, the quality and safety movement and the evidence-based practice movement. Together, these approaches have helped to influence research investment and set an international health policy agenda that has been conducive to supporting and deepening our understanding of how research is used in practice. In Canada, the Canadian Institutes of Health Research and the Canadian Health Services Research Foundation have both been instrumental in supporting leading edge work in this area. Indeed, we are very privileged to have the Vice President for Knowledge Translation of Canadian Institutes of Health Research, Ian Graham, contributing to the supplement. Similarly, in the United States, federal agencies such as the Institute of Medicine, in its work around patient safety, and the Agency for Healthcare Research and Quality, in its programme of work around evidence-based practice implementation, have moved the agenda forward. Some of these significant developments are described by Titler and colleagues. Examples from Australia and the United Kingdom (Bucknall, Rycroft-Malone, and Dopson, respectively) also demonstrate health administrations' investments in research into KT and implementation. It is within this context that Nursing Research has decided to publish this special supplement on advances in theory and methods in KT research.

The Propositions

Trained in approaches to theory development and testing that some would characterize as positivist, and expert in advanced research design techniques, Estabrooks and colleagues have utilized data from two large nursing data sets to look at knowledge utilization and those individual, unit, and organizational factors that influence the use of research in practice. Two significant methodological approaches have been presented, and concomitant challenges are discussed: First, there is the issue of having to construct a set of variables to make proxy or phantom measures for research utilization using a technique called structural equation modeling. The paper of Midodzi et al. provides the rationale for doing this, whereas the papers of Estabrooks et al. and Cummings etal. describe the assumptions, methods, and processes involved. Commentaries on the appropriateness and risks involved in using structural equation modeling are found in the critiques of Titler, Sales, and Dopson.

The second challenge has been trying to understand and measure the impact that different organizational variables have at different levels of the system. Thus, for example, in the paper of Estabrooks et al., the authors put forward the case for using a technique called multilevel or hierarchical level modeling to analyze the different impacts that certain variables have on research utilization. Interestingly, the findings of Estabrooks et al. are different from the conclusions drawn by Cummings et al. in their article. This outcome is discussed in the commentaries and is, in particular, picked up by Dopson, Titler, and Sales. For those unfamiliar with advanced techniques, the article of Adewale et al. provides an explanation of how this method could be used to acknowledge the kind of natural aggregations that are present in nursing and healthcare environments. Thus, understanding the impact of certain variables in a unit or in the wider hospital system would be accommodated by hierarchical level modeling. Estabrooks argues in her prologue that we need to move away from a rather simplistic bivariate design in KT research in nursing and embrace more sophisticated techniques that reflect the natural data structures we find in organizational research and the natural aggregations of clinical practice.

Estabrooks and colleagues have used the Promoting Action on Research Implementation in Health Services (PARIHS) Framework (Kitson, Harvey, & McCormack, 1998) as a heuristic device to help make sense of the many variables and interactions that occur in practice. In addition to the critique of advanced methodologies, the papers offer an objective assessment of the strengths and limitations of this conceptual framework to date. Like all heuristic devices, the PARIHS Framework has its strengths and its limitations, but the important thing is to see how the conceptualization aids in our understanding of the complex world we inhabit. However, what is particularly encouraging in the papers is how discussions around the strengths and limitations of the PARIHS conceptual framework are based on emerging evidence from the investigations of Estabrooks et al. (see also Wallin, Estabrooks, Midodzi, & Cummings, 2006).

Some Wider Issues

In addition to the primary propositions, a number of other important issues are raised. Bucknall elegantly outlines the important contribution that decision science makes to our overall understanding of KT and reminds us of the need to hold onto the importance of the individual and his or her experience within the wider organizational system. Interestingly, despite the argument that organizational factors are important, the evidence presented here still attributes most impact to individual behavior and characteristics. Indeed, for research use to occur, research evidence needs to be screened, interpreted, and applied to the specific situation by individuals who may belong to a group or in an organization. The processual factors (as Dopson calls them) that impact upon the transmission of such decisions from one set of clinicians to others is indeed a central issue-one that is discussed in terms of the concept of facilitation of research within the PARIHS framework (Harvey et al., 2002). However, Dopson quite rightly argues that we ignore at our peril the interactions, social networks, power, politics, and cultural variations that we experience in our own organizations and how these contexts enhance or detract from successful research use in practice.

From a theoretical and methodological perspective, we are faced with a series of paradoxes. Rycroft-Malone and Dopson raise legitimate questions about how the deductive, positivistic approach to theory development and research design fits with and complements more interpretative, naturalistic inquiry approaches. In essence, is it more appropriate to think of getting research into practice as a logical, predictable process with its own rules, like a chess game (Kitson, 2005; Kitson et al., in press), where the elements that we have uncovered through research can be manipulated, controlled, and influenced? Or are we dealing with a dynamic situation that, although we have roughly similar elements, each time we look at the picture, it changes in its own unique, dynamic way? Thus, instead of the metaphor of a chess game, are we more likely to understand KT by thinking of it as looking through a kaleidoscope: the elements are all there, suspended in a space in time, but depending on the "movement," the pictures created will all be different? These are the philosophical and theoretical questions that we need to address. We also must utilize all our senses and intellectual capacity to begin to construct appropriate conceptual frameworks and theories that will help our own knowledge generation. And from these constructions, we need then to refine our use of multiple approaches to advance the science of KT (Green, Caracelli, & Graham, 1989; Ivankova, Cresswell, & Stick, 2006).

Our biggest challenge in this field of research is to avoid rushing to solutions and certainty and to resist the belief that there will be straightforward replicable explanations. As researchers, policy makers, and clinicians, we have to learn to tolerate ambiguity, paradox, and uncertainty to an extent that makes our rational and intuitive minds hurt. Embracing our own quest for knowledge with honesty, integrity, and respect for multiple perspectives and experiences means that we are more likely to build a strong evidence base for KT. We would do well to embrace the suggestion put forward by Graham and Tetroe to establish an international research agenda on KT and more importantly create an international research community of committed researchers, policy makers, and clinicians from a broad range of relevant disciplines.

There are some famous examples of such endeavors, one example being the influential work of Van de Ven and colleagues in the 1980s on research on the management of innovation (Van de Ven, Angle, & Poole, 2000; Van de Ven, Polley, Garud, & Venkataraman, 1999).

Carole and her team have shifted the focus of the debate and have tried, through the design of this special supplement, to demonstrate the importance and benefit of debating and refining our understanding of this research agenda around KT. We are entering an exciting, if not challenging, time and one that will reap reward for patient care if we diligently and collectively move forward together, remembering that the most important thing in contemporary science is direct and personal experience of the phenomena that we are attempting to understand.

Alison L. Kitson, RN, PhD, FRCN

Guest Editor

Royal College of Nursing, London

[email protected]

References