It’s October and that means the excitement of a new term and new students. I sit them down on their first day and tell them several things. Everyone will be telling them to aim for a first-class degree but that would be gross hypocrisy from me – I only ever got a II-i! There are 38,000 bottles of wine in the cellars at St John’s, but they should remember that they have six years to drink them, while everybody else has only three, so they can afford to drink at half the rate of the other students (though this advice is rarely taken; medics and vets seem to drink at twice that rate!).
They also have six years to learn enough jargon to be able to persuade clients to pay them even when they don’t know the cause of their pet’s disease – idiopathic will be their favourite word. And finally, only 80 percent of what I will tell them will be true. The rest I’ve either got wrong at the beginning, have forgotten or… it’s a joke. But they are Cambridge students, so it’s up to them to work out which belongs in the 80 percent and which in the 20.
After their first week, we’ll go over the new words they’ve learnt in those first seven days: dorsal and ventral, medial and lateral, proximal and caudal in anatomy; hyperglycaemia, polydipsia and ketoacidosis in biochemistry; action potential, depolarisation, voltage-gated sodium channel in physiology – the list goes on and on…. And that’s just in the first week – they’ve got 240 more to go before they graduate!
But saying that 20 percent of what students get taught is wrong is probably a wild underestimate, according to a great book I’m currently reading. The Half Life of Facts by Sab Arbesman (get it for 5p on Amazon “new and used” if you’re fast, but it’s well worth the full 15 quid if you want to buy it new).
When I was a child, we saw dinosaurs as slow lumbering cold blood creatures, all scaly and green. Now they are portrayed as brightly coloured, often with feathers, fast and voracious. When I was a student, cats with kidney failure should be fed cottage cheese and given anabolic steroids. Now look at the plethora of kidney diets available and medical therapies from phosphate binders to amlodipine.
Things change. Back in 2002, Thierry Poynard and his colleagues from the Service d’Hepato-Gastroenterologie, in the Groupe Hospitalier Pitie-Salpetriere in Paris, produced a great paper entitled “Truth survival in clinical research: an evidence-based requiem?”
They looked at research papers published from 1945 to 1999 about cirrhosis or hepatitis in adults. In 2000, 285 of 474 conclusions (60 percent) were still considered to be true, 91 (19 percent) were considered to be obsolete, and 98 (21 percent) were considered to be false. The half-life of truth from those research papers was 45 years.
The number of conclusions still considered correct after 20 years – their 20-year survival – was lower in meta-analyses (at 57 plus or minus 10 percent) than that from non-randomised studies (87 plus or minus 2 percent) or from randomised trials (85 plus or minus 3 percent). Interestingly, the survival of conclusions from studies of high methodologic quality was not greater compared with those which were of low quality.
Truth be told, I’m not quite sure what this is telling me and I’m going to have to read some more papers in this area. When I typed “truth survival in clinical research” into Google Scholar, it gave me 146,000 results in 0.09 seconds. I’m afraid it’s going to take a good deal longer than that to work through even the first page’s worth of papers – I’ll let you know how I’ve got on this time next month.