JAMA  1 May 2013  Vol 309
This week’s JAMA is devoted to child health. This was a mistake, because although children are generally interesting, health generally is not. A study from Quebec tries out various doses of vitamin D in babies and finds you can only get to a reliably high value by using doses which might cause hypercalcaemia. I’m not sure how many generalist readers need to know this. Likewise it’s faintly interesting that two doses of human papillomavirus vaccine given between the ages of 9 and 13 may give the same immunogenicity as three given between 16 and 26, but as we don’t definitely know how long this lasts from the trial, nobody is going to change practice. But among the skim-and-flick-past articles there is one important one, which brings good news for the parents of very premature babies.

1810   I am used to reporting depressing figures about premature babies, but this population-based prospective cohort of consecutive extremely preterm infants born before 27 weeks of gestation in Sweden between 2004 and 2007 is different. Admittedly 31% of them did not survive to the age of two and a half, but those who did were mostly in reasonable shape, which the investigators attribute to active perinatal care. “Of children born extremely preterm and receiving active perinatal care, 73% had mild or no disability and neurodevelopmental outcome improved with each week of gestational age. These results are relevant for clinicians counselling families facing extremely preterm birth.”

NEJM  2 May 2013  Vol 368
1675   To my mind, the words “meniscal tear” conjure up Monday morning at the surgery, with young men hobbling in to report crunching and swelling of their knee after a hard game of football on Saturday afternoon. The METEOR trial, however, was not interested in young amateur sportsmen: the mean age of the participants was 58 years, and 58% were female. Now at that age, about half of the knees you put through an MRI scanner will show meniscal tears of some sort, and in people with knee pain due to osteoarthritis, a judgment has to be made whether to interfere with the cartilage. If there is a financial gain to be made by interference, then you can expect arthroscopic trimming to follow swiftly. The alternative is physiotherapy, which was employed as a sophisticated-sounding package in the trial: participants were randomised to this or a prespecified range of arthroscopic procedures. A very worthy effort: undoubtedly there are far too many arthroscopic knee procedures carried out in the USA; and this trial showed similar outcomes in the two groups at 6 months. But it is a devil of a trial to generalize from. Of 14,430 patients assessed, 12,008 did not meet the trial criteria, and only 351 underwent randomization. Of these, 30% of the physio group had crossed over to surgery during the six months. So the intention-to-treat analysis does not really carry much weight, and I suspect orthopaedic surgeons will carry on doing whatever pays for the best and fastest car.

1695   Now for a trial which takes you straight to the heart of British general practice, though sadly no GPs seem to have been directly involved. The patients in PATCH-1 had been referred to UK dermatology departments for recurrent leg cellulitis, or else responded to direct advertising: they were randomised for receive either phenoxymethylpenicillin 250mg bd or a matching placebo. Those who got the penicillin suffered fewer recurrences, with a NNT of 5. When they stopped the penicillin, they got the same number of recurrences as the other group. This is useful knowledge, but it may not necessarily apply to your typical old lady visited by the district nurse. The average age of the participants was 58, and I am baffled at the failure to specify how many of them had diabetes. The choice of poorly absorbed straight penicillin as prophylaxis is also a bit odd: perhaps we need a PATCH-2 study using, dare I say it, twice weekly low-dose azithromycin.

1704   The reason I mention azithromycin is because it has already been shown to prevent respiratory infections in COPD, and overall, I think the evidence points to low doses being harmless, even in the presence of cardiovascular risk factors. The QT-prolongation/sudden death risks seem to be related to peak dosing, according to the discussion piece which opens this week’s NEJM. But it’s a tangled debate, with only observational evidence to go by. A Tennessee Medicaid study showed an extra cardiovascular death for every 21,000 patients prescribed azithromycin instead of amoxicillin. The latest study from Denmark, published here, shows no added risk from a five-day course in the whole of the population aged 18 to 64, but you could argue forever about comparators (it was Pen V again here), propensity scores and confounding by indication.

Lancet  4 May 2013  Vol 381
1532   The Lancet offers thin pickings this week. Here Zeke Emanuel, one of a famous trio of pugnacious brothers, criticizes the Helsinki Declaration, which will be fifty years old this year. He is right: it is a sprawly and incoherent document, aiming to set down the ethical principles of medical research, but straying into other areas, and badly neutered by political interference. I have actually read all the versions, and also their predecessor, the Nuremberg Code, while preparing a piece on the ethics of non-disclosure of human trials. We need a completely new version of the Declaration, with central insistence on the disclosure of all data from every human study, and penalties for failure to do so. Now the AllTrials petition has just reached 50,000 signatories—if you are not among them, we will forgive and embrace you if click here and sign at once.

1541   Gosh and golly: should I use tocilizumab monotherapy or adalimumab monotherapy for the treatment of rheumatoid arthritis? The answer, in most cases, is that you should use neither, but try methotrexate in everyone, on its own or with a “biological.” This head-to-head trial recruited patients with RA said to be intolerant to methotrexate or to have no response to it. You have to wonder how consistently these criteria were applied to the 326 patients recruited from 76 centres in 15 countries in North and South America, Australasia, and Europe. But anyway: Hoffmann la Roche declare a win for their product tocilizumab, an inhibitor of interleukin 6 receptor signalling, versus adalimumab, which is an antibody against TNFα. I defy anybody to be sure that this is not just an artefact of the doses used, or to know how to fit this in to the clinical management of RA: the editorial has a good try, but ties itself in knots. Gosh and golly: I’m glad I’m not a rheumatologist.

BMJ  4 May 2013  Vol 346
Orlistat, which interferes with fat absorption in the bowel, is the only drug treatment for obesity available in the UK and many other countries. The subtitle of the editorial which accompanies this article says that it is “still a useful option for some obese patients,” though that is not my experience. “All it did is make me fart for England” is one patient comment I remember; and since pétomania has not yet been adopted as an Olympic sport, this somewhat narrows its clinical usefulness. It may—or may not—sometimes cause liver inflammation too. The conclusion of this UK CPRD based study does a bit of bet-hedging: “The incidence of acute liver injury was higher in the periods both immediately before and immediately after the start of orlistat treatment. This suggests that the observed increased risks of liver injury linked to the start of treatment may reflect changes in health status associated with the decision to begin treatment rather than any causal effect of the drug.”

This next study was carried out in Australia using data from Massachusetts and Tokyo, and it shows that you don’t need to go remeasuring cardiovascular risk more than about every 10 years in low-risk groups. This is a really useful analysis. In higher-risk groups, measure more often if you have decided not to treat: above all, involve the patient in making any decisions, and help all smokers to stop.

I went to a course on systematic reviewing at about the time the Cochrane collaboration was set up, and haven’t had many refreshers since; but it strikes me as a difficult and in many ways subjective activity, despite everyone’s best efforts to make it a science. I don’t think it will ever come into its own until the raw individual patient data and all the meta-data for all trials are available; and then it will become such a massive and specialised effort that perhaps only two centres in the world will have the support and the expertise to do it properly. But I digress, and perhaps most readers don’t really care for these technical details: but they should, because this is the foundation for clinical decision-making, and when you have cancer you will be glad that somebody somewhere has done the job properly. Here’s a paper showing that it can be very hard to judge bias from the published versions of cancer trials: babies may be thrown out with bathwater, or the opposite, if that is possible. I am getting into deep problems with entropy, water, and babies here. Let the investigators speak for themselves:

“Overall, 23 trials (23%) were assessed as low risk of bias based on publications alone; however, with additional information, 66 trials (66%) were classified as low risk of bias. Had the 13 meta-analyses included only those trials at low (or lower) risk of bias (as recommended in the Cochrane Handbook), and if assessments were based on publications alone, five meta-analyses (38%) could not be undertaken, because none of the included trials were judged to be at low risk of bias.” Simples.

I’ve already made a joke at the expense of orthopaedic surgeons in this review, and indeed it’s pretty impossible to write a blog like this without some recourse to this rich tradition of medical scorn. So I commend to your attention this article on adolescent idiopathic scoliosis: just look at those diagrams: these spine surgeons deserve respect. Orthopaedics at its best does more direct good than almost any other medical specialty.

Plant of the Week: Osmanthus delavayi

I’ve left it slightly late to celebrate this excellent shrub, since the flowers on ours are already beginning to brown off a bit. “Osmanthus” means scented flower, and the rich fresh scent of this shrub and its commoner child O x burkwoodii are among the delights of early spring, when the sun comes out to warm them.

This delicious osmanthus is named for Jean Marie Delavay (1834-1895), who was sent as a missionary to western China in 1867 by the Missions Etrangères de Paris. His harvest of souls is not known, but cannot have equalled the harvest of 200,000 botanical specimens he sent back to Franchet in Paris. Unfortunately he did not send back many entire plants, but those that he did are generally magnificent. Anything with the specific name of delavayi is worth having if you can grow it.

Father Delavay’s osmanthus is usually grown as a medium-sized wall shrub and is said to be slightly susceptible to frosting. But while we have lost several nice little shrubs to frost in this protracted winter, O delavayi grown by a flimsy wooden fence has no visible damage at all, and has flowered abundantly. Left to its own devices, it affects a sprawling elegance, with arches of small dark evergreen leaves; but it is tougher than it looks, and if you want to train it into a tighter shape, it will respond to regular discipline from the secateurs. Definitely among the top ten evergreen garden plants.