This is a formal write up of the same phenomenon, which is a good example of how we should do the epidemiology on a new disease. But it almost seems that as we rid the world of polio, another enterovirus takes its place. Let's all very much hope not.
Here is the report of a large scale trial of a new ante natal screening modality - analysing cell free DNA in maternal plasma for trisomy 21. Note the use of Area Under the Curve for the primary comparison.
The results are pretty impressive but don't let that blind you to some practical problems (of which more later). cfDNA detected all 38 of the 38 pregnancies affected by trisomy 21 compared to only 30 out of 38 detected by standard methods (nuchal translucency plus biochemistry). Could not be better. But...
Of 18,955 women who were enrolled, results from 15,841 were available for analysis. So some technical problems. And this was in a study population, not the real world of human error, not-very-good laboratories and systems etc.
Also I was surprised to see that there were some false positives - how could that possible happen? How could you detect trisomy 21 when it wasn't there? The answer lies in things like mosaicism. and other tricks of mother Nature.
So cell free DNA is very, very good but it's not perfect.
You're the manager of a supermarket. Do you get paid for the excellence of your displays, or the number of people who enter your store, or the number who ask what something costs? No - you get paid for how much they buy. BUT THIS ISN'T SOMETHING UNDER YOUR CONTROL.
Ok. So now you're a Director of Public Heath. Should you be rewarded for the excellence of your programmes, how many people sign up, how many complete - or for how much death rates drop? I was never brave enough to ask for my annual objective to be the drop in premature deaths, or teenage pregnancy, or road deaths. I suppose that was partly because I didn't there was any hope of being given resources commensurate with the task. Mostly, though, cowardice.
And so to the Quality outcomes framework (QOF), routinely described in the international literature as 'the world's largest pay-for-performance scheme'. A piece of massive number crunching has compared death rates in small areas with the QoF scores of local practices. Answer: as far as we can tell no relationship between the two.
That's not entirely surprising. The scheme covered the family doctors of an entire country, and accounted for a significant slice of their income. So it had to be introduced gently - with process targets rather than hard outcomes.