Those interested in the hospital volume/quality debate may like to take a look at this study from the USA which shows that admission to the highest-volume centres there is strongly influenced by social class, race and insurance status. Once upon a time, our National Health Service would have been the ideal place to carry out a contrasting study, showing how it is possible to allocate specialist services on the basis of need, optimal volume and accessibility. But I’m sure the Government are right in thinking we are better served with a haphazard network of competing bankrupt hospital trusts and overpaid, underused private treatment centres.

This study looks at children born with intellectual disability and traces their mental health over a couple of decades. By the criteria of the study, boys start off worse but end up better than girls; in both sexes, the burden of mental disturbance remains very high (30%+) and, in the USA at any rate, largely untreated.

Should we follow the American lead and give influenza vaccination to all babies and toddlers aged 6 to 23 months? The answer may well be no – see BMJ 912 below, which cites a systematic review with 95% confidence intervals stretching from 0.18 to 1.69. But if we ever do, the huge US experiment at least teaches us that it is very safe.

Although I always pounce eagerly on any paper in the Rational Clinical Examination series, this one will go into my file for the sake of completeness rather than because I expect ever to perform another lumbar puncture. But for those who take or interpret samples of CSF, it is essential reading. Use a small-gauge atraumatic needle and reinsert the stylet before removal; patients do not need bed rest afterwards. For clinical decision rules about who needs an LP, see Arch Dis Child 2006;91:647 or J Infec 2006: Aug 1 e-pub.