You have got to do something. Young mothers and their babies living in socially deprived areas do poorly. Isolated, unprepared, and hard to reach; the obvious way to help is through their peers. Why bother with research. It’s obvious. Just implement it. 

But, it didn’t work, and worse. At the NAPCRG annual meeting in Montreal, Margaret Cupples (The Queen’s University, Belfast) described their high quality randomised controlled trial where young mothers in deprived areas were mentored by local trained volunteers. A great study, well conceived, carefully designed and embedded in the community, it was adequately powered and recruitment was successful.  When they looked at the results, however, there was no change in the primary care and secondary care outcomes for either mothers or babies. No benefit. No improvement in well being. No health gain. But this was just the first surprise. When they looked more closely at their findings, searching for an explanation, they noted that some of the mentors found it difficult linking up with their mothers. Perhaps this variability in the number of contacts had affected the outcomes. It had but, when they looked at those who had most mentor contacts, (more than six), these mothers were worse off. No one knows why. It may have been that this group needed greater input anyway, or more worrying, maybe they were harmed by their involvement.

We might leave it at that. But, the qualitative component of the work revealed further surprises. The mentors themselves had been affected by their involvement. They began full of enthusiasm and good intentions but, when they found it difficult to make contact with the mothers, were frustrated when the mothers did not turn up, had their phones turns off or, did not return calls, they become increasingly irritated and ultimately disillusioned. They stopped being interested and lost heart. They too had been damaged by the experience.

You have got to do something. It was a similar message on a larger scale from Richard Wender (Thomas Jefferson University) in his keynote address: A call to arms for primary care researchers. Since 1971 the US has spent 500m on research in cancer but, for the big four cancers, long term survival has barely altered. We gained immense understanding of the biological models but little improvement in survival. This translational gap is unsurprising when you see such disproportionate investment – where primary care gets 0.41% of NIH funding.  Soul searching among the research funders has brought a realisation of the need to look more closely at applied clinical research. US research has failed to meet satisfactory grades. This, perversely, is an opportunity for primary care. He called for a research revolution.

But, there are too few researchers and, as he listed the possible barriers to developing primary care research, he had no easy solutions. The picture was familiar to most European researchers where, there have been considerable although tortuous advances. He made a telling observation, however, that it is easier to find research critical of primary care than publications reporting achievements and advances. I am still not sure where he will find the researchers in a country where there is already a shortage of primary care doctors and many undersubscribed residency programme. Doing something is going to take some time.

Superstitious? Spare a thought for our New Zealand colleagues who set off to fly to Montreal early on Friday the 13th. Time zone changes gave them an extended Friday the 13th that lasted 34 hours, and they landed even before they had left.

Domhnall MacAuley is primary care editor, BMJ