JAMA  14 Mar 2012  Vol 307
1029   The Viewpoint pieces in JAMA this week are a strange mix of fact and fantasy. The first is a piece about industry payments to physicians and teaching hospitals in the USA. I am currently at Yale University alongside the authors of this piece, one of whom is a good friend. I am observing the US health system at first hand, so I can tell you what is coming to the NHS. You think it’s bad enough that private firms are already buying up NHS hospitals and child health services? You ain’t seen nothing yet. The whole of the American health system is saturated with the influence of the pharmaceutical and medical devices industry: huge payments are made directly to individuals or institutions to influence purchasing decisions. In the UK, we call this corruption, but only for the time being: look at those lists of payments to our MPs from private health providers published in the Daily Mail, of all places. No wonder our legislators are baffled by all the opposition they are facing from British doctors as they try to open up our antiquated socialist NHS to the healing influences of unbridled capitalism. Come on all you GPs who will be running NHS plc—there is money to be made by everybody! In America there is due to be public listing of all such payments in 2013. I wonder if it will make any difference. It is wonderful how much public disquiet people can handle when they have a few million in the bank.

1031    The next piece argues that everything will nonetheless be all right in the end, because genomics will come to the rescue. There is a technical fix for everything, and industry is smart enough to find it. Look, the cost of genomic profiling is falling all the time! Soon we will be able to treat fewer people because their DNA will tell us whether the drugs will work or not. And of course we will produce all sorts of better targeted drugs. It’s just a shame that they mostly cost $10K per month at present. But all this is bound to change, provided we look to the gleaming towers of lucrative science and ignore the ubiquitous realities of poverty and old age.

1033   And we must do more to prevent depression. The main reasons given in this paper are little to do with the misery of individuals but the misery that they may spread to others by being so miserable, and their effect on economic productivity. We must be on the look-out for people who do not smile enough, i.e. those with “subsyndromal symptoms”:  “Methods with proven effectiveness involve educational, psychotherapeutic, pharmacological, lifestyle, and nutritional interventions… The use of booster sessions and Internet technologies should be explored.” This sounds like a great idea. Pretty well everyone has pre-depression and if everybody gave each other encouragement, education, a good lifestyle and good nutrition we would all be happier. This used to be called the welfare state. But obviously it’s better to trap people in a hopeless cycle of debt and work (or lack of it) and then bring in private counsellors to address their pre-depression issues, and give them serotonin reuptake inhibitors.

1037   Every now and again, Canada is held up as an example of how to provide good healthcare. In fact, last week JAMA daringly had a Canadian advise the USA on how to set up a better health system. So, does the experience of Canada tell us that we should attempt to provide better care by reducing hospital staffing and costs? Well, actually, it tells us the very opposite. Mortality and readmission rates were lowest in hospitals with the highest costs and levels of spending. Canada works with a system of private providers. So if we want to improve the NHS, we should bring in private competition, and if we want good care we should purchase it from the providers with the highest costs. Or if we want rubbish care, we should buy if from the lowest bidder. This will be the responsibility of all GPs acting on behalf of their patients. You know it makes sense.

1072   Does This Patient Have a Severe Upper Gastrointestinal Bleed? My pre-test probability is high. I am an old man taking low dose aspirin, driven to dyspepsia by the thought of the undemocratic destruction of the NHS, seized with griping upper abdominal pain whenever I hear the words “Liberal Democrat.” I will spare you a description of my stools. Though this would be vital in the clinical situation, and might save me from a rectal examination in the emergency department. Here is a first-rate contribution to the first-rate Rational Clinical Examination series: “Melena, nasogastric lavage with blood or coffee grounds, or serum urea nitrogen:creatinine ratio of more than 30 increase the likelihood of a UGIB. (NB: Why don’t we have this in the UK?) The Blatchford clinical prediction score, which does not require nasogastric lavage, is very efficient for identifying patients who do not require urgent intervention.” The Blatchford score was devised in Glasgow (by a GP, amongst others) but the original Lancet paper from 2000 is still behind an Elsevier paywall. But you can get the score on Wikipedia. Everyone should work to make Wikipedia the foremost reliable source of medical information, and to make the Elsevier business model a thing of the past. This won’t help to save the NHS, but at least it will help to get knowledge out where it is most needed, i.e. the resource-poor world.

NEJM  15 March 2012  Vol 366
981   Here is the screening paradox summed up in a single study. The European Randomized Study of Screening for Prostate Cancer used prostate specific antigen—often called PSA because it has Perfectly Stupid Attributes for a screening test. Nonetheless, in this 11-year follow-up study, the risk of death from prostate cancer was lowered by 21-29%. But all-cause mortality did not differ between groups. Amazingly, I have seen this declared a success for PSA screening in some of the American medical press. Now clearly, if you don’t mind what goes on your death certificate, prostate screening is a waste of time. But to inform our advice to patients, let’s look at it from the perspective of somebody who had decided they would rather die of anything but prostate cancer. “To prevent one death from prostate cancer at 11 years of follow-up, 1055 men would need to be invited for screening and 37 cancers would need to be detected.” So if you had a “cancer” detected by screening, there is a one-in-37 chance that treatment would prevent your death within 11 years. And your odds of dying from anything in that period would be the same.

991    In my last year as a proper GP, I presented two cases from my practice of young women who had suffered cryptogenic stroke and had been found to have patent foramen ovale. OK, foramina ovales. I gathered together myself and a GP partner and a local stroke doctor and some interventional cardiologists from Oxford, and we had a high old time congratulating ourselves on our diagnostic acumen while the receptive audience admired real-patient videos showing the technical wizardry of percutaneous PFO closure. Clearly, we had saved these ladies from the dangers of recurrent stroke by the timely deployment of sophisticated interventions based on impeccable mechanistic insight. But even during the presentation, there was a slight feeling of unease when the evidence base for this presumption was interrogated: PFOs can be found in 25% of the population and there didn’t seem to be an adequately powered study to compare device closure of PFOs with medical therapy. Well, now there is, with a follow-up period of two years, and there is no difference in outcomes so far. So while I was trying for one last time to share some of the glamour of sophisticated sexy medicine, perhaps all that these patients needed was what I had tried to provide them with for the whole of my working life—good primary care.

1000   Another week, another study of a new oral treatment for relapsing-remitting multiple sclerosis. This one is called laquinimod, and as far as I can tell it is about the same as all the others—it produces a modest decrease in the rate of relapses and a 4.6% absolute decrease in progressive disability over two years. The study was placebo-controlled and conducted at 139 sites in 24 countries. It was paid for, designed, and analysed by Teva Pharmaceutical Industries, and the paper was written with help of an external writing agency. There seems no obvious reason why laquinimod should not now be licensed for use in MS at a cost similar to that of other novel agents, even though we know little about its long-term harms or benefits or how it compares head-on with these other treatments. We demand very little from our licensing agencies before drugs like these can be tested on patient populations. Patients are seldom told they are being used experimentally, doctors have no real means to judge their place in therapy, and meanwhile drug companies can get rich on what they charge for them. Maybe we are all fine with this as a business model: if we are so used to it, it must be right.

1010   Ingenol mebutate is a really neat treatment for actinic keratosis: four RCTs show that it works with 2-3 applications. It’s made from the sap of Euphorbia peplus, the petty spurge which is abundant in many parts of north America. It is doubtless very cheap for Leo pharmaceuticals to produce, and with such a tiny quantity needed for such a brief duration of treatment, perhaps they are thinking of giving the stuff away. Well, actually, one website quotes $699 for 141gm of 0.015% ingenol gel. But don’t be tempted to save money and find some petty spurge to put on your keratoses: the sap can burn quite badly.

Lancet  17 Mar 2012  Vol 379
1005    The National Patient Agency has surveyed mental health services in the UK from 1997-2006 to see if there is a relationship between suicide rates and implementation of recommendations to prevent suicide issued by the National Confidential Inquiry. These recommendations include such things as the formation of 24 hour crisis teams and assertive outreach groups for vulnerable patients who are poor attenders of conventional services. People who have both depression and alcohol problems, and are thus at highest risk for suicide, are no longer supposed to be pushed back and forth between services designed for one or the other problem but never both. By and large, the study finds a correlation between the degree of implementation of the recommendations and a fall in suicides. It isn’t the ultimate in scientific rigour, but it makes sense. Like all mental health provision in the NHS, it is a rather bleak minimum.

1045   Since serotonin reuptake inhibitors came on the scene, suicide rates have fallen in most countries. Is this cause and effect? There is no methodology which can possibly answer that question, but at any rate it makes it unlikely that they cause increases in suicidal behaviour at a population level. There was one large study of males in the USA which seemed to show that sertraline reduced suicidal behaviour compared to other antidepressants, or none. You can find this and other bits and pieces of information in this review of major depressive disorder: new clinical, neurobiological, and treatment perspectives. There is a convincing mechanistic biochemical explanation for every drug that works for depression, and for every one that has been tried and failed. All the time, extremely clever people are discovering more about the neurobiology of the depressed brain. This article discusses dozens of new pathways we could tinker with. Whether this would produce any results that are better than chance alone, nobody can tell.

BMJ  17 Mar 2012  Vol 344
I haven’t been able to access Latest Print Version on the BMJ website for several days, so I have picked a couple of research papers out of the Last 7 Days section.

There’s a general feeling that we don’t have enough good drugs for type 2 diabetes, and it’s hard to argue otherwise. So we should extend a welcome to any new drug that helps to lower blood sugar, especially if it doesn’t cause weight gain or hypoglycaemia, right? I’m ashamed to say that until about three years ago I would have nodded in assent. I may even have prescribed a few patients a dipeptidyl peptidase-4 inhibitor (or gliptin) just to show how up to date I could be. But what I was doing was performing an unwarranted human experiment, since we have no idea whether these drugs do more harm than good in the long term. A drop of one point in HbA1c is rarely worth the risk, especially if you don’t know what the risk is. This meta-analysis hedges its bets, as it must. But betting in the dark with the long term future of patients is not a good thing to do, ever.

But perhaps there is a subgroup of diabetic patients who are particularly likely to benefit from treatment with a gliptin, the drug rep or the paid lecturer might argue. Look, in this trial (he says, handing you the reprint), patients with red hair showed a much bigger drop in blood sugar, and moreover their LDL-cholesterol levels were halved. Here are some nice pens and do help yourself to the delicious sandwiches. With Wondagliptin you may be able to meet all your targets for red-headed diabetics, and we can help train your practice nurse to do it with the aid of daily blood glucose monitoring. Do I exaggerate? Not much, I fear. Anyway, back to subgroup analyses. Don’t believe them, especially in industry-funded studies. If you are an EBM nerd, or even if you are not, it is worth looking at this systematic review of the credibility of subgroup claims in randomised controlled trials.

Arch Intern Med  12 Mar 2012  Vol 172
397   About ten years ago, I advised you to train your larynx to pronounce the word ximelagatran, the name of the first direct thrombin inhibitor which seemed poised to replace warfarin for numerous clinical indications. If you followed my advice, you abraded your vocal cords and twisted your palate to no avail, since the drug was withdrawn in 2006 following reports of hepatotoxicity. Its successor dabigatran is easier to pronounce, works well at a fixed dose, and generally seems harmless to the liver. Boehringer Ingelheim seemed to have a winner on its hands. But alas, a cloud no bigger than a man’s hand has appeared on the horizon. Here is a meta-analysis which shows that dabigatran is associated with an increased risk of MI or ACS in a broad spectrum of patients when tested against different controls, and that is not something one likes to hear about a drug designed for long term thromboprophylaxis.

Plant of the Week: Euphorbia amygdaloides var robbiae

Like all members of the spurge family, this one seeps toxic sap, though whether it contains ingenol I have no idea. You could make your fortune by finding out. After all, it is a wood spurge that likes to grow in dark places, so its juices are bound to work for something that is caused by sun damage. Or does it work the other way round? I lack grounding in the science of herbalism.

This handsome plant bears the somewhat dismissive label of “excellent for ground cover.” All winter long, you have ignored its handsome rosettes of dark leaf, except perhaps on some magical sunny day when they have shone with a covering of frost. Now, as the ground warms, it is sprouting great flower-shoots of pale fresh green. Soon it will go back to its usual job of growing where nothing else will, and sometimes where you don’t want it to. Excellent for ground cover, indeed, and lovely in early spring.