Article publié dans Oxford Bibliographies, le 7 novembre 2014.
Auteurs: Émilie Robert, Valéry Ridde.
Introduction:
Public health interventions are by nature complex and are implemented in contexts and with social actors who have a mutual influence on each other. It thus becomes nearly impossible, and more or less irrelevant, to control the interventions or to find direct causal relationships. It follows that most public health interventions are natural experiments for which appropriate methodological fit should be the gold standard. Therefore, researchers should adopt a pragmatic, non dogmatic position; be responsive; and adapt to the natural environment of the real world (as there is no such thing as a laboratory setting for public health interventions). Because of their complexity, we need to go beyond just knowing the interventions’ effects; we also want to know how those effects are produced, for whom, and under what circumstances. Methods should be selected that allow us to understand the inherent complexity of public health interventions. The social sciences have contributed significant methodological suggestions to the field of evaluation of complex interventions. In this context, the war of paradigms is unproductive. However, taking into account real-life situations does not mean abandoning the use of theories, conceptual frameworks, or the intervention theories identified in the preevaluation phase (evaluability assessment). Moreover, the issues related to an evaluation often extend beyond just methods. Budget constraints need to be considered, as well as the usefulness of the responses to the questions and concerns in terms of effecting change in practices and interventions (see the Oxford Bibliographies article “Knowledge Translation and Exchange”). Beyond studying effects, it also becomes essential to examine interventions’ implementation (fidelity and adaptation) and operationalization, as well as their relevance, appropriateness, sustainability, and acceptability. Thus the evaluation questions and concerns should guide the methods, and not the reverse. To the extent possible, we will present all these consubstantial aspects of a pragmatic and real-world evaluation approach from a conceptual perspective and then illustrate their application. (The authors wish to thank Donna Riley for translation and editing support.)
Télchargez l’article complet ici: