As a public health consultant, committed both to the use of the best evidence, and to utilising available resources to achieve the best outcomes for all children and in particular to reduce inequalities this challenge is significant. At a time of reduced financial resources it is also immediate.

How should we respond? On the one hand this kind of trial is what those committed to evidence based practice repeatedly ask for. It has scale, rigour and longitudinal follow up. Ian Roberts, the eminent professor at the London school of hygiene and Tropical Medicine calls RCTs “the truth test”; they’re only real way to tell the differential impact between a new intervention and usual treatment. However strong are our instincts that an intervention works, those instincts often fail us; we know from extensive experience that this is true. The RCT is the best mechanism we have getting to the nub.

Taken from  this perspective the RCT has spoken; and we should listen. Now, one response may be that this is just one trial out of a number that have previously developed a robust evidence base for FNP internationally; as with swallows and summers, one non-significant trial, a failed intervention doesn’t make. However one of the main tests of an intervention is its replicability to different contexts. As has been said often, FNP in a US context cannot be realistically compared to the UK where there is universal health visiting with it’s particularly strong existing focus on vulnerability.

These arguments are compelling, but they don’t close the book. Thinking about the RCT some questions emerge. Firstly there is the focus of the primary outcomes. None of the outcomes classified as primary (birth weight, maternal smoking in pregnancy, repeat births and hospital attendances) strike me as the main reasons why I commission FNP. All are, I think, problematic in different ways as primary outcomes.  Conversely, the secondary outcomes which showed small positive impacts; intention-to-breastfeed, child cognitive development (at 24 months only), language development (at 12 and 18 months), levels of social support, partner-relationship quality, and general self-efficacy are absolutely core to my sense of the purpose of FNP.

Secondly I have real concerns about the challenges of implementing and evaluating an intervention of this nature. The FNP evaluation began not long after introducing the intervention to the UK.  Unlike a pill which does not evolve over the life of a trial, people do and those people and the relationships they form are central to this intervention. FNP is about the selection of the right people to be nurses; about the training they receive; and the supervision they experience. All of these contribute to the nature of the core relationship between nurse and client. All of these core elements are likely to evolve and develop over time, contributing to effectiveness in recruiting participants, maintaining their involvement, connecting with them and challenging them. Yet these elements were put under the RCT microscope when realistically they were in an early stage of their development.

I’ve never previously been either a committed believer or firm sceptic about FNP. I’ve always seen a strong theoretical and scientific justification for much of its approach while being concerned about some aspects; are really all the ingredients equally active; do we lose something from selecting on demography rather than vulnerability being two key ones. I’ve got no doubt that this evaluation will prove very useful in putting the model under further scrutiny, and I look forward to much of the discussion that will no doubt follow. Whether though it gives as a clear a signal to commissioners as the authors suggest, for me it’s too early to say.  I think the debate that follows is going to be key.