Thankfully the American's shifted St. Patrick's Day to their convenience - the carnage happened last weekend so last night did in fact turn out to be relatively civilized. Shame.

But thankfully this left me in a good state of body and mind to fully appreciate the keynote address by Del Elliott - the Exec Director and driving force behind the Blueprints agenda.

Three main themes resonated through the cavernous breakfast hall filled with wafts of Starbucks coffee and yogurt-bathed melon. First, the standards by which researchers, policy makers and practitioners evaluate evidence varies enormously. Elliott tells us that between 30 and 40 per cent of programs on Federal lists of 'evidenced-based practice' have insufficient evidence or are even ineffective when measured against robust standards of evidence. No wonder policy makers and funders make such a mess of prevention.

Second, that fidelity of program implementation is directly related to effect size. And that this is not linear. Instead Elliott suggests a step-wise relationship: there is a particular threshold of fidelity that must be obtained before we start seeing even marginal impacts on child outcomes. It's no good only implementing just 50 per cent of a model program's components (you’ll probably get zero impact)- you need to be at least delivering 70 or 80 per cent before you start reaping the benefits. The greater the fidelity the greater the impact on children's lives.

Finally, we now know a lot more about what it takes to scale up effective programs. We have all the key ingredients: dissemination capacity, fit to population, site preparation, training etc. We now need to get people following the recipe; if you get the proportions wrong the cake will collapse.

Talking of which, I might have another perfect banana bread slice…