It never fails to amaze me that still after all these years, the UK thyroid establishment still fails to recognise the changes in the TSH reference range of "normality" especially at the hypo end that nearly all other countries have adopted since the arrival of sensitive tests. The reference range with present day sensitive tests is close to 0.6-3.5. Anything above that is NOT subclinical, it is overt. That is it needs treatment (T4) to begin with at steadily increasing doses until health is restored. Our latest paper in Nuklearmedizin - on file at TUK - contact Lyn Mynott or Louise Warvill for a copy - shows the reference range to be as above and certainly not stretching to 10. Even at TSH = 2.8 I personally would put a patient on watch to see if there was deterioration over time. Frankly, I do not understand the wooden intransigence of the UK specialists to reconsider and realise how out of step they are with everyone else. It's not clever; it's sheer ostrich behaviour by people who know best but apparently know nothing and learn nothing. The sensitive tests have for TSH increased its sensitivity. This does not mean that this has increased its utility, but requires more careful analysis of what the results mean.