JAMA  14 Nov 2012  Vol 308
1916   Last week we learned that male doctors who were randomised to take a daily multivitamin preparation had cardiovascular events and died from them at exactly the same rate as those who took placebo for a median of 11.2 years. This week we learn that they also died of cancer at the same rate, but by aggregating all the cancers you can just squeeze a statistically significant reduction in total cancer incidence out of the data. So come on, fellow male physicians: if 83 of us take a daily multivitamin for 11 years, one of us may avoid a cancer, though all of us will die at the same rate. Golden docs in trials all must/ As chimney-sweepers, come to dust.

1889   Migraine is a stupid condition which we don’t understand: as a lifelong migraineur I’ve read any number of review and research articles about it over 40 years and I am only slightly the wiser. But here is a nice clear paper which tells us something definite about those worrying white matter hyperintensities that often turn up when MRI scans are performed on people with migraine. The sensible Dutch set up a prospective population-based observational study of adult participants (aged 43 to 72) with migraine and an age- and sex-matched control group, in the year 2000, and here they report on the incidence of MRI brain lesions 9 years later. There is a striking gender difference: in women, MRI lesions are much commoner in those with migraine than those without, while in men there is no difference. These lesions look like tiny areas of ischaemia, but when you follow up the women there is no difference in progression of these hyperintensities between the migraine and the control group (and there is none in the men either). So whatever is going on does not appear to be an active process: nor does it seem to be reflected in any significant change in cognitive scores. So when I get yet another visitation of zig-zags and defocussed vision at least I can be fairly sure that it doesn’t mean I am losing a bit more brain.

NEJM  15 Nov 2012  Vol 367
1883   Lyme disease is a popular diagnosis on the eastern seaboard of America, but an uncommon curiosity in most parts of old England. I’ve only seen it twice in 35 years of UK practice, and one was just a few days ago. The lady concerned lifted her trouser leg to reveal a classic large erythema migrans lesion which appeared two weeks or so after the initial bite had settled. But she was adamant that she had not been in any woody or ferny areas, until I reminded her of a poolside walk in her village which fitted that description. We shall do some serology and see. We looked it up together on Google and the Wikipedia picture could have been her very leg. Now if she were to get the same thing again, long after adequate treatment, might that mean she was having a relapse of Lyme disease? This study of 22 paired episodes of erythema migrans from which Borrelia burgdorferi had been isolated shows that in every instance the genotype of the organism was different in the two cases. Treated Lyme disease probably never recurs.

1891   After a while, it gets a bit boring restating the fact that the only lipid lowering agents which have been shown to reduce clinical events are the statins. I believe that this is true of people with primary hypercholesterolaemia as well as of the general population, but if I am wrong, I know I will quickly be corrected. The most potent dose of statin that most people can tolerate is atorvastatin 80mg, but in some individuals even this is not enough to reduce LDL-C sufficiently. In this phase 2 trial, Sanofi and Regeneron Pharmaceuticals tried out their new monoclonal antibody targeted at serum proprotein convertase subtilisin/kexin 9 (PCSK9), which has the effect of reducing degradation of LDL receptors and so gets more LDL removed from the circulation. This sounds very neat and ingenious, and it does actually work: SAR236553 lowers LDL-C more in primary hypercholesterolaemia than does atorvastatin 80mg, and can be added to it for extra effect. Time to give it a catchy name and start marketing it? I would say not. Just because monoclonal antibodies sound like magic bullets, that doesn’t mean that they are either harm-free or as effective as surrogate end-point reduction would suggest. In the real world, there is no substitute for real time and real results: we need quite lengthy phase 3 studies.

1930   Sitting where I am 15,000 years ago, I would be surrounded by tundra vegetation and probably within easy walking distance of the nearest glacier. I always marvel at how people manage to survive and multiply in those regions of the world where the default setting is death from hypothermia. It’s bad enough when the oil runs out and the central heating packs up in middle England in 2012. Here’s an informative update on the treatment of accidental hypothermia from a US perspective. This is not a richly evidence-based area: “In a single tertiary care center, 14 different rewarming methods were used to treat 84 cases of accidental hypothermia.” Wow: leading chefs have some things to learn from these guys. None of the fancier invasive rewarming methods has been shown to be superior to gentle external warming. In centres where it is available, cardiopulmonary bypass or extracorporeal membrane oxygenation can be used. Success can follow hours of cardiopulmonary resuscitation and repeated asystole—I know that from experience in the distant past. This applies particularly after cold water submersion: “The longest period of submersion that a person has survived without neurologic impairment was 66 minutes in a child who was 2.5 years old (the child’s core temperature was 19°C).”

Lancet  17 Nov 2012  Vol 380
1741   Here is one of the most important papers to have come out of UK general practice for some years. It reports a cluster randomised trial with 9.6 years of follow-up which shows that screening for diabetes in high risk individuals has no benefit. It used a validated risk score to identify 15,000 individuals at risk who were offered screening, with a control group of 4,137 who were not. Individuals identified with diabetes were then randomized to usual care or intensive multifactorial treatment. I know that at least one of the investigators was expecting the trial to demonstrate benefit, but there was simply no reduction in all-cause, cardiovascular, or diabetes-related mortality within 10 years. And this trial not only shows a lack of benefit from the early detection of diabetes, but also casts grave doubt on the doctrine of a “legacy effect” from early tight control, as claimed by the UKPDS investigators.

1759   Metal-on-metal hip resurfacing was supposed to lead to better implant survival than conventional metal-on-polythene total hip replacement, but it became clear about a year ago that this had been a massive mistake, involving tens of thousands of patients (many in younger age groups) in premature prosthesis failure and the need for reoperation. This sobering trawl through the data in the National Joint Registry for England and Wales (NJR) for primary THRs undertaken between 2003 and 2011 concludes “Before further new implant technology is introduced we need to learn the lessons from resurfacing and metal-on-metal bearings.” It is shameful that so many scandals have had to happen before anybody started taking the regulation of medical devices seriously: and we are still a long way off a solution.

1778   The issue of screening mammography in the UK has been mired in controversy for some years, in part thanks to wildly misleading promotional material distributed to women by the screening bodies. In a situation like this, the British answer is to gather together the great and the good and reach a nice sensible compromise based on the interpretation of the evidence which most closely fits with current practice. But the Independent UK Panel on Breast Cancer Screening has come up with the most startling claim I have read about any population intervention ever: “Breast screening extends lives. The Panel’s review of the evidence on benefit suggests a 20% reduction in mortality in women invited to participate in a 20-year screening programme.” Gosh, that’s amazing! Given that far fewer than 20% of women die of breast cancer, having your breasts squashed into an X-ray machine must have extra life-extending benefits yet undreamt of. I think the Panel mean a 20% relative reduction in breast cancer mortality, and some editor should have spotted this mistake, which occurs twice in the paper. But then, this is The Lancet.

BMJ  17 Nov 2012  Vol 345
Varenicline is known as Champix in this country and Chantix in the USA, and it is probably the most effective drug for helping people to stop smoking. But one meta-analysis of placebo-randomised trials showed an increase in cardiovascular events while taking varenicline; another did not. This is a Danish database study comparing the cardiovascular event rates in people prescribed varenicline for smoking cessation compared with those prescribed bupropion: there is no difference. Though it does have mood-lowering effects in some people, this is a drug that does more good than harm in most.

The newer anticoagulants are at least as good as warfarin for the primary and secondary prevention of stroke in atrial fibrillation, but it is hard to know which is best among them—the direct thrombin inhibitors (dabigatran) and the factor Xa inhibitors (for example, rivaroxaban, apixaban). Several cardiovascular academics here look at the differences between them in the trials and reach the conclusion that only head-to-head new trials will tell us anything substantive.

Years of selection and education make us believe that we are special and that, when the chips are down, doctor knows best. And even if we wanted to share decision making, there just isn’t time. So we guess what the patient wants, decide what is best for them, tell them our decision, and move on. This is how medicine operates in every country and context, and it is going to take decades to change. But change it must, because this is no way to treat people, and it generates conflict and waste on a scale which is simply unsustainable. Patients whose views are actively sought, and who really understand their options, frequently choose less investigation and less treatment, so there may be some immediate savings to be made, particularly in extravagantly wasteful systems like the USA’s. All this is dealt with at length in a recent King’s Fund paper on the silent misdiagnosis; here is a great summary of the same argument by the same authors.

Arch Int Med  12 Nov 2012  Vol 172
1537   The Archives are beginning to annoy me, by printing too many interesting papers which I have to report on. Here is a piece with the cumbrous title “Use of New-Generation Oral Anticoagulant Agents in Patients Receiving Antiplatelet Therapy After an Acute Coronary Syndrome: Systematic Review and Meta-analysis of Randomized Controlled Trials.” Those of you who have been following these reviews for some years (or actually reading the primary literature—heaven forbid) will know what a keenly fought-over field this is. But for pharma companies hoping for their next blockbuster drug, a very disappointing one. “The use of anti-Xa or direct thrombin inhibitors is associated with a dramatic increase in major bleeding events, which might offset all ischemic benefits in patients receiving antiplatelet therapy after an ACS” is the conclusion of these Hungarian meta-analysts. How dramatic? Well, at least three-fold: not quite all the rooms in Bluebeard’s castle, but enough to warn you off opening any more doors.

1548   Now here is a hen’s tooth: a clinical prediction rule that doctors may actually use. It is called the PLAN Score and it is a bedside prediction rule for death and severe disability following acute ischaemic stroke. Not the first ever, of course, but perhaps the least cumbersome, factoring in just 9 easily available clinical variables. It had a C statistic (look this up if you don’t know what it is) in the mid 0.8 range for death by 30 days, severe dependence, and death within one year.

1557   Here is another hen’s tooth: a trial of a weight-reducing agent which has no commercial sponsor. The drug is zonisamide, a curious sulphonamide agent which is licensed for use in intractable epilepsy but has been promoted (illegally) for other indications. Under the United States Food, Drug, and Cosmetic Act, Elan Pharmaceutical Inc was criminally fined $203 million and Eisai was civilly fined $11 million for misbranding Zonegran for improper promotional activities. This trial was conducted at Duke University and involved 225 subjects with a mean BMI of 37. Zonisamide taken at a dose of 400mg for 12 months resulted in weight loss of 7 kilos, at the expense of many adverse effects. This is a pretty hefty dose, but anything less failed to have a significant effect.

1582  Angio-oedema is a common side effect of angiotensin-converting enzyme inhibitors (ACEIs) and this American database study seeks to compare these with  angiotensin receptor blockers (ARBs), and much less-used the direct renin inhibitor aliskiren, while adjusting for a range of possible confounding factors. It’s a reasonable study which reaches the conclusion you’d expect: ARBs are less likely to cause your face to swell up while aliskiren is too seldom used for us to be sure, but it’s probably worse than ACEIs. Researchers using the vastly bigger UK Clinical Research Database should be churning out this kind of thing by the shedload every week.

Plant of the Week: Mahonia gracilipes
November has reached its halfway point and everything is gloom and dampness. The very ground seems to rot, and even the fungi give up and decay. But a few precious shrubs refuse to succumb to seasonal affective disorder and try to cheer us at this dismal start of winter. Certain mahonias are among them, and I think this is my very favourite, for its small habit and its lovely little flowers of purple and white.

For some reason this plant is very seldom seen, though it could fit into any garden and thrives on neglect, and even tolerates dry shade, like most mahonias. It has attractive sprays of evergreen leaf and is named for the elegance of its stem, though this can become a bit lanky with time. The solution is quite simple: just cut the whole plant back in late winter, and it will shoot up again with a shorter stem or stems. The curious and beautiful flowers can appear at any time between August and December.