Several prevention programmes in the substance use field developed in the United States have been transported to other countries, including Sweden. A number of questions have emerged regarding the transportability of these programmes, with concerns about cultural adaptation and fidelity [1,2]. Whether these programmes have been successful in their international application still remains to be shown. Reports from both Swedish and Norwegian attempts to implement a set of evidence-based programmes have demonstrated no effects in trial communities compared to controls [3,4]. What has been largely overlooked so far is the extent to which these programmes have demonstrated effects at the community level in the United States, even though a recent paper claims exactly that [5]. It is in this context that Harold Holder's paper [6], with its seven themes, fills a critical gap.

Some of Holder's points are of a policy nature, calling for more transparency regarding funding and personal investments. Some are of a more technical nature, e.g. single versus two-tailed t-tests. Some finally address issues of broader generalizability, e.g. whether the results of these academic efficacy trials, with selected participants, can be applied in real communities. Overall, the paper raises the question of whether policy decisions based on the presumed effectiveness of these programmes have been premature.

Since the turn of the millennium the Swedish government has invested serious money into the alcohol and drug prevention field. Sweden, with a history of heavy drinking throughout the 19th century, developed comprehensive control policies, initially in the alcohol field, and subsequently also in the drug field. When Sweden joined the European Union, several components in its alcohol policy had to be discarded. Researchers and politicians alike recognized that this would probably result in increased drinking and increased problem levels. In response, the Swedish parliament adopted a comprehensive national alcohol action plan where the importance of national control policy was still acknowledged, but the emphasis was shifted to local community prevention. This shift was, to an important extent, inspired by reports from successful prevention trials in the United States.

A first observation is that gradually, over the years, smaller effect sizes are reported as more studies are published. This is not surprising or strange; initial efficacy studies, whether in the treatment field or prevention field, are characterized by unique and favourable conditions. The mistake is to draw conclusions based on these early results.

A second observation is the complexity of many of these programmes, calling for elaborate manuals, training, technical support, supervision, etc. Most local communities lack the resources for this and most countries lack the prevention infrastructure that could provide it. One observation in the Swedish six-community trial was that, on average, even communities with dedicated leadership and full-time prevention coordinators succeeded in implementing selected programmes in 30–40% of the target schools [4]. There is also a concern about the fragility of these programmes, which the repeated calls for fidelity illustrate: it seems that even minor deviations from the original, often laborious construction would lead to loss of effectiveness. This, then, is a call for simplification, to strive for the identification of core concepts and their dissemination [7].

Returning to the Swedish situation it is notable that, following an initial surge in drinking following the liberalization of policies, consumption peaked in 2004 and has since decreased somewhat, contrary to earlier predictions. The question is why—the answer is not the dissemination of manualized prevention programmes, because whether effective or not they have reached only a small fraction of the target groups. One possible effect of the national action plan was that it contributed to problem recognition and popular mobilization. Studies have found that a clear shift in popular opinion occurred around 2004, following which support grew for restrictions on availability (e.g. the alcohol retail monopoly) [4,8]. This, then, may be the main effect of many prevention programmes, whether evidence-based or not, that they can increase popular acceptance of restrictive policy by raising awareness. Whether this actually is the case should be the object of future research.

Declaration of interest