JAMA  28 Sep 2011  Vol 306
1329    Intra-aortic balloon counterpulsation reduces left ventricular load and improves outcomes in animal models of myocardial infarction. But in previous small human studies of MI without shock, it hasn’t been shown to do anything much, and this trial confirms that it does not reduce infarct size significantly. But such is the success of primary PCI for MI these days that it would need something bigger than this 337-patient, nine-country trial to settle the issue with hard end-points: here there were just 3 deaths at six months in the PCI-plus-IABC group versus 9 deaths in the PCI-alone group.

1338   Reading surgical case-series reports from more than a hundred years ago, I’m struck by how many of the themes of modern outcomes research are already present – adjustment of case-mix, the need for proper follow-up, individual operator versus institutional outcomes, the relationship between operator volume and outcomes, and the learning curve effect. Here Brahmajee Nallamothu and colleagues take a look at the results of carotid stenting according to the annual volume and experience at the time of the procedure among new operators who first performed carotid stenting. It’s a nice study design, and confirms the intuitive expectation that the more of these procedures you do, the better your outcomes tend to be. It raises important issues about how we can best protect patients while taking new operators through their inevitable learning curves. And also what we do with those operators whose learning curves aren’t climbing enough. As for the more general issue of whether anybody should do carotid stenting in the first place, I will leave that for more learned persons to decide. Medicare decided to give it the go-ahead, but that does not make it a good procedure.

1344   Saw palmetto extract achieved fame about fifteen years ago when the balance of evidence at the time seemed to show that it was a safe, cheap, and effective remedy for the symptoms of benign prostatic hypertrophy. Kind Nature was thought to have placed relief for nocturic old males in the berries of Seroana repens, and even sceptical doctors like me would scrawl the annoying common name of this plant on a scrap of paper for patients to take down to the local alternative remedy shop. Hokum, according to this double-blinded RCT, using increasing doses of saw palmetto extract versus placebo. The stuff makes no difference at all.

N.B. Could any kind reader point me to the origin of the saying “Use new remedies early, before their effect wears off?”

NEJM  29 Sep 2011  Vol 365
1173   Could it be that the gene gnomes have pulled it off at last? A discovery that will genuinely guide treatment in a common condition – as common as inhaled glucocorticoid treatment in asthma? Well, there is much rejoicing in their ranks over the discovery that people with a certain GLCCI1 variant show only one third of the response to inhaled steroids as other asthmatics. For the second week running, the NEJM runs an editorial with “personalised treatment” in its title, this time by the great editor himself, Jeffrey Drazen. But without wishing to dampen the celebrations, I would point out the last phrase in the results: “with genotype accounting for about 6.6% of overall inhaled glucocorticoid response variability.” So more than nine times out of ten, genome profiling would not help you in this aspect of clinical management. Even if you could ever access it and afford it.

1184   Many of you will know the famous Fletcher-Peto curve of the progression of chronic obstructive pulmonary disease with and without smoking cessation – it dates back to 1976 and is a classic for several reasons, one being that it nicely illustrates the use of a graph to represent the natural history of a condition. I am surprised that this kind of depiction has never caught on as a teaching aid for other conditions: still more surprised that there is no similar graph in this paper on the variation in trajectories of COPD as measured by FEV1 after a bronchodilator in 2163 people over 3 years. Not surprisingly, the course of decline varies a lot between individuals and is normally distributed. For a longer term delineation of the natural history of COPD, nicely illustrated, you need to look elsewhere.

1193   Nicotine is one of many neurotoxic substances that plants have evolved to poison insect predators: the one in laburnum plants (Cytisus spp.) is called cytisine. It actually hits the same α4β2 nicotinic acetylcholine human brain receptors as nicotine, and laburnum extracts have been used as a smoking cessation aid since the 1930s, when Hitler encouraged the use of herbal remedies against the filthy habit of smoking which was sapping the fighting power of the German people. The Soviet people followed suit, and cytisine still remains available for next to no roubles in Russian pharmacies. In the capitalist West, we buy it chemically modified at great expense in the form of varenicline. Why not instead spend thruppence and use cytisine? This study shows that it definitely helps – 8.4% abstinence at 12 months after a 25-day course compared with 2.4% on placebo. A shame this wasn’t a head-on trial with Champix/Chantix/varenicline.

Lancet   1 Oct 2011  Vol 378
If there are important, believable randomised controlled trials out there waiting to be published, once again The Lancet has decided against letting us have them. The original papers this week have all been on the website for weeks, and I let you have a couple during the lean weeks of August and September. Here is one I didn’t:

1219   The circulation is circular, and so is a lot of the thinking that goes on around it. I read this modelling paper about the cost-effectiveness of options for the diagnosis of high blood pressure in primary care with particular care because the topic is important and I’ve been interested in it ever since I became involved in the first primary care ambulatory blood pressure studies which were done in Oxford 20 years ago. Another couple of Oxford colleagues are involved in this paper: but I am afraid I can’t agree with some of their modelling assumptions and hence their conclusion. The data to compare home monitoring with ABPM in relation to hard outcomes are simply lacking, as they say; Framingham is a less accurate predictive tool than QRISK for the UK population; a practice can buy 22 home monitors for the price of one ABPM machine, and patients much prefer these and often buy their own. The jury remains out and a wholesale (or even retail) move to ABPM is premature.

BMJ  1 Oct 2011  Vol 343
There are very few journals where a jobbing general practitioner could find a chance to publish these days, and the BMJ is not one of them. I got my chance back in 2000, when they still did Brief Reports, but these have now disappeared from the main journals. General practice is the most fertile of all research settings, potentially, and very much the best generator of patient-important questions; but funding arrangements are such that it is virtually impossible to realise these advantages, and even academic departments of primary care usually offer little help. Perhaps this may be about to change, according to this editorial on a new European strategic report. But some of us remember the UK Mant report from the 1990s, and will not be holding our breath.

I usually read editorials with personalised medicine in the title expecting to find a lot of genome-babble to disbelieve, but this strange piece is from a renin enthusiast hoping to reform the treatment of hypertension. Those with exceptionally long memories will remember the vogue for renin-based classifications in hypertension in the 1960s, but Morris Brown argues from first principles that now we could apply better assays more logically to the individualised treatment of blood pressure using specific drug classes which did not exist then. Perhaps: but some evidence would be nice.

Medical papers in praise of chocolate are rather common these days. They used to guarantee media attention, but on that I cannot possibly comment as I live a hermetic existence 3,000 miles from all British organs of public intelligence. I am also on a chocolate-free diet since I need to lose weight. Nevertheless the ideas of chocolate-driven cardioprevention set forth in this systematic review of observational evidence are very attractive: all we need is some way of making cacao palatable without the addition of sugar.

Arch Intern Med  26 Sep 2011  Vol 171
1542    Here is a paper which shows that patients who visit their doctor more frequently get more treatment. And if the condition is diabetes and the treatment is to the wrong targets, then that is not a good thing. The targets here were HbA1c below 7; BP below 130/85; LDL-cholesterol below 100mg/dl. The evidence? Harm to most patients in the first case; no benefit in the second; and no evidence in the third. And how often should one monitor each? Six monthly seems logical for HbA1c which is a reflection of glycaemic control over 3 months. Monthly seems the maximum for BP if seriously out of control, to allow for multiple measurements and regression to the mean. And LDL-C? Why measure it at all when we know that all these people will benefit from a high-dose statin. But this study conducted in two American primary care facilities linked to a teaching hospital shows that such targets are best reached by seeing diabetic patients every two weeks. So now you know another way to harm people with diabetes and waste health service money: see them too often. But in the USA, Merck is supporting an organization (ACE) which is publicizing this paper under the headline: Frequent Doctor Visits Benefits Patients With Diabetes (sic).

1571    So far the plant neurotoxins we have mentioned are nicotine, cytisine, and theobromine (in chocolate): now it’s time for caffeine. The stimulant properties of tea and coffee were discovered by Europeans at approximately the same time, and were followed in both cases by ultimately successful attempts to discover and steal the source plants and grow them in European colonies. Anything pleasurable and mildly addictive has always attracted medical censure in certain quarters; but as with chocolate, the evidence of benefit from caffeine outweighs any evidence of harm. Not that this is too wonderful a study: out of 50,000 US women, those who drink caffeinated coffee report less depression than those who don’t. Not exactly a randomised trial.

Fungus of the Week: Grifola frondosa

The alleged English name for this fungus is “Hen of the Woods,” which seems a bit odd as there is another quite different fungus called “Chicken of the Woods.” Why stick to the poultry yard when naming fungi that grow on trees? Why not “Pig of the Woods?” Or “Porpoise of the Forest?” I think I will stick with the Latin.

The Grifola grows at the base of trees, usually oaks. There is a gigantic version (G gigantea) which produces enormous brown fronds and kills its host: this one is sensibly proportioned and grey in colour and lives in a state of mild parasitism. It appears as a little cluster of fronds growing up through the soil, growing steadily larger until it has had enough, usually at about 50-70cm. It is simply there to shed spores, after all.

To harvest this fungus for the table, you have two options. You can pick a biggish specimen and then pull off the fronds when you get home, discarding the woody central mass. Or, if you are lucky, you can pick the whole thing while it is less than 20cm across and cook the lot. But the great challenge is to render it clean enough to be edible. Bear in mind that it has grown through the soil, and if the soil is tenacious, then you must resign yourself to a bit of mild pica.

Unfortunately the soil round the red oak from which I get my supplies is sandy, so the pleasure of eating this fungus is distinctly modified by loud crunching noises as grains of silica encounter dental enamel. The structure of the fronds is actually very beautiful but almost impossible to clean. Look carefully at the underside of many fronds and you will see the Mandelbrot set in all its wonder: a multiplicity of similar shapes at every scale, none predictable. This may give you enough satisfaction, without you feeling the need to proceed to the cooking stage.

However, if you are prepared to brave the grimaces of your loved ones and the inner doubts which I have sown in your mind, proceed as follows:

Recipe A: for fronds. Tear the fronds from a young but well-developed specimen. Under running water, attempt to remove all soil, insects, grit etc using your fingernails, a shaving brush and/or a toothbrush. Do not expect to succeed. Heat olive oil moderately in a frying pan and add the fronds while they are still wet. Reduce the heat so that they stew for 10 minutes. At this point add some scraps of chopped shallot and season with salt and pepper. Simmer further until the mushroom liquor is nearly evaporated, but do not let them brown. Serve with finely chopped parsley, eat jauntily and claim that they have a mild and delicate taste and that everybody should try some.

Recipe B: for a whole infant cluster. Cut the base from your cluster and attempt to remove dirt etc. from under the little fronds under running water. Do not expect to succeed. Cut the cluster into slices about 0.75cm thick. Cut some very fatty bacon or pancetta in cubes or pieces and fry until the fat has been largely melted into the pan. Now add the sliced grifola, followed by a few scraps of shallot. This time you can allow a little browning to occur, but beware of drying out. Serve with or without parsley, and murmur “delicious” as you eat the bacon with as little of the fungus as you can get away with.