Purpose The current wisdom in practice change is that there is no "magic bullet", and that a variety of practice change interventions should be employed. However, no theoretical framework exists for even understanding the failures of practice change interventions, let alone for predicting their effects or selecting from among the many interventions in any given situation. Our purpose is to develop a framework that will facilitate understanding of observed patterns of successes and failures and guide empirical research toward targeted choices of effective interventions.

Methods We began by examining the literature on practice change and knowledge utilization interventions, but found that theoretical foundations were either lacking or failed to incorporate relevant research from the cognitive, learning, and organizational sciences. Further, there was little integration between change at the individual, team, and organizational levels. Therefore, we followed a process analogous to that used by Donabedian in developing the structure/process/outcome framework for making sense of the medical quality literature. We developed a change process framework through an integrative review of the key findings of basic science research in cognition, the psychology of expertise, team functioning, and organizational behavior, combined with a wide range of apparently disparate or conflicting studies of practice change drawn from Cochrane and other systematic reviews.

Results The six-phase framework that emerged begins with awareness of new information, followed by understanding of the information and its implications for change, and subsequently acceptance of the need for change. The intention to change must then be formulated, the knowledge must be transitioned from declarative to procedural (i.e., into actual clinical behavior), and finally must achieve consistent performance. Transitions between each of these phases involve different cognitive or organizational mechanisms. The dynamics of transitions are not linear, and include such effects as sour grapes, anticipated regret, and inference of belief from behavior. The various practice change interventions reviewed operate at different transitions, but studies generally did not distinguish them. Many interventions relevant to some transitions, but very few relevant to others, have been studied.

Conclusions A framework for understanding practice change based upon theoretical foundations from several disciplines can be described. The next step is to extend the focused integrative review of the empirical literature into a comprehensive review, specifically seeking disconfirming findings.