Taking snapshots of preventive interventions: On the effectiveness of preventive interventions for youth and how it relates to implementation and conflict of interest

Intervention studies This dissertation describes three trials in which the effectiveness of three preventive interventions for youth were tested in the Netherlands. The interventions aim to improve the social and emotional development of children in elementary school (PATHS), reduce alcohol use and mental health problems in students in secondary education (Preventure), and empower adolescent second generation migrants (POWER). The results revealed no effectiveness of the PATHS intervention, which was probably due to the implementation strategy that was tested in this study. Intervention effects of Preventure were also not found regarding the (secondary) mental health outcomes that were examined in this dissertation. The intervention effects of POWER were limited, but greater in groups in which the intervention was implemented with more fidelity. These studies teach us that replication studies are needed within the field of prevention; it cannot be assumed that what works within one context or specific situation will also works in other contexts or situations. Measuring the implementation during a trial has to be seen as equally important as to measuring the actual intervention outcomes. It is the product of these two factors that ultimately produces the intervention effects. The author additionally argues that more effort is needed to search for interventions that not only work in ideal settings, but also still work when implemented on a large scale in real life settings. Conflict of interest This dissertation also searches for an answer to the question how common conflicts of interests in intervention trials are. To what extent are intervention researchers also the developers of the intervention or working for the licensee, and how does this relates to trial outcomes? It was tested whether this effect was mediated either by the cynical view or the high fidelity view. The cynical view implies that researchers with a conflict of interest manipulate the research results, or (un)consciously make choices during the study that increase the chance of finding positive intervention effects. The high fidelity view holds that licensees and program developers, who are also the researchers, implement the intervention in the most faithful manner during the trial, and thereby achieve better trial outcomes. The results reveal that having a conflict of interest is very common in Dutch intervention trials of youth interventions; in 84 percent of the articles at least one of the authors had a conflict of interest. Analyses regarding an international dataset of intervention trials of school-based substance use prevention programs revealed that a conflict of interest was prevalent in 75 percent of the cases. Using meta-regression analyses, it was demonstrated that having a conflict of interest does indeed affect the study outcomes significantly. It could, however, not be confirmed whether the high fidelity or the cynical view moderated this effect. Based on these studies, the author argues that more independent intervention research is needed, as well as more transparency about conflicts of interest in intervention research