Some years ago, just after starting as a consultant, a patient with an established diagnosis of multiple sclerosis (MS) asked me to prescribe Diet Coca-Cola for her. I was taken aback by the unusual nature of the request, and had to ask why. It turned out that she had watched a Channel 4 documentary programme on television the previous evening, in which a lady called Cari Loder had presented a proposal for a regime of treatment that appeared to benefit her MS. The combination was vitamin B12, a somewhat out-of-fashion antidepressant called Lofepramine, and L-phenylalanine in generous quantity. The latter, the TV commentary had advised, could readily be obtained by drinking a litre or two of Diet Coke every day.

To a new consultant, attempting to fashion my clinical practice around a sound scientific evidence base, this sounded like nonsense. I obtained a recording of the relevant TV programme and made enquiries to find out whether clinical trials of this astonishing therapy had been done. How had subjects been recruited? How many? How was blinding achieved? Which independent body had analysed the results? There were, of course, no answers to these questions. Indeed there were no trials. The TV programme had just presented one woman's experience: at best an unblinded ‘n=1’ trial. Yet the documentary included not just film of Cari Loder herself, but also recordings and testimonies of many other patients who had …