NHS isn't be all and end all
While I understand Neil Robinson's point about square root transformations of relative prevalence (Letters, July 26), the essence of the problem is that the practice population multiplier is not subject to the same transformation – or conversely that the relative prevalence is transformed at all.
If you consider the same number of diabetic patients in two practices, one of 5,000 patients and one of 10,000, then because these multipliers are treated differently, the square root applied to one only, the practice with a larger list gets 30 per cent more per patient!
Which genius of numeracy at the BMA let that one slip through?
Dr Anthony Pickering