Giving children the best start in life by getting in early – even before conception – sounds like an unarguably good thing. Why on earth should we wait until the ‘damage’ has been done, and for children to be showing signs of distress before we offer any help?

After all, by the time children are say four, five or six, and displaying disruptive or aggressive behavior in schools, years of less than optimal parenting may have already shaped their personalities. But put in solid building blocks and sound attachments at the first opportunity and there will be far fewer difficulties down the line.

This is a lovely idea, and countless millions of pounds and dollars are poured into it. However, in the real world, getting these interventions to make a substantial difference is really very hard.

We need to separate three sets of outcome: the satisfaction of the parent, the quality of the parenting they deliver, and the outcome for the child. They’re progressively tougher to alter, but ultimately it is the next generation of children we wish to help. We must therefore resolutely focus on whether these interventions improve child outcomes, or are ever likely to.

Parents may find many of the interventions they are offered pleasing and recommend them to a friend. It could be anything from being cheered up by a friendly visitor, to being offered aromatherapy or even perhaps being given tickets to go to a concert. Whether this then leads to improved parenting, and improved child outcomes is another issue.

So, to take the examples given in this week’s articles - Newpin, Community Mothers and Home-Start – evaluation has shown that they improve a mother’s sense of satisfaction and sometimes their attitude towards parenting, but there is scant evidence that actual day-to-day parenting has improved, or that child outcomes are better.

Why so? Firstly, the model of support typically does not incorporate latest findings about what is most effective. The deeply-felt impulse of a volunteer to support somebody may be necessary but it will not be enough to produce change.

As Christine Puckering wrote [see Newpin: can it be as shiny as it sounds?], giving mothers support and even reducing depression will not automatically lead to better parenting, which we know to be essential for improving children’s outcomes. So intervention approaches must not only address the mother’s own needs, but also reckon in a clear, structured way with improving the parent-child relationship.

Doing this requires good organizational skills – to guarantee that a large number of parents offered the service take it up – and then well-honed intervention skills: not only an understanding manner, but also specific techniques for improving the parent-child relationship, for example through play and appreciation, setting clear limits and applying appropriate discipline. These skills are not easily acquired; they call for training and ongoing supervision.

And even with all these components in place, a number of evaluations have proved disappointing. If relatively advantaged parents are included, it is likely that their children will do well anyway; the additional input doesn’t make much difference. However well-intended, the supervision of staff delivering the intervention is often wanting, so that skills are not maintained. Thirdly, there is often a lack of knowledge, even a positive reluctance to use ‘evidence-based’ interventions.

The values attached to all this activity are changing. So, for example, if volunteers wish to provide a service where previously there was none, one might argue that whatever they do is OK, provided it is appreciated. However, if public money is being spent and there is an opportunity to provide something of a high standard which is likely to work, a whole different set of issues comes into play.

How well trained and supervised are the workers? Are they offering good value for money? Is it right to suggest an intervention to parents and children if it has been shown more than once not to benefit child outcomes?

These considerations are particularly important should there be other interventions around that seem more likely to improve child outcomes. The emphasis surely has to be on delivering what is right for children and their parents, rather than on permitting an organization to carry on doing the same old thing even if it doesn’t work.

These are painful matters to consider. People grow very attached to the interventions they are delivering. They want to feel justified in what they do.

The way forward seems to be to continue to have really good evaluations and to disseminate research knowledge into the wider discourse so that all workers can benefit.

With this in mind, last November UK Children’s Minister Beverley Hughes launched the National Academy for Parenting Practitioners with a ?30 million grant for a three-year investment. Initially quite a lot of the training will concentrate on evidence-based programs, but in time model curricula will be set up to raise the standard of what is taught to everyone delivering services to improve parenting.

There is also an ambitious research program to test new interventions or ones that have been shown to work in the US but not yet in the UK. There will be a program evaluation tool, to ensure that consistent criteria are applied.

There is a drive from the Department of Children, Schools and Families to expand the number of parenting practitioners, so we live in interesting times. But we must heed the findings of research evaluations and learn from them; otherwise we will be letting down the next generation of children.