JAMA  3 Aug 2011  Vol 306
493    This issue of JAMA is devoted to war and violence, things that most of us have not experienced. Those who do experience them, like most of the population of Europe in the generation before mine, are never unscathed. The study here discovers that military veterans with post-traumatic stress unresponsive to antidepressants do not respond to the addition of risperidone. No surprise there: they are just the extreme end of a spectrum of people who seldom respond to treatment but mostly respond to time. Looking at table 1, it seems that three quarters of these veterans had not seen battle for over 40 years: and an unspecified number are listed as having fought in WW1, which would make this the first published trial of a posthumous intervention. More typically, it took men who fought in that war about ten years to recover any feeling of normal life; such at any rate was the case with the Great War’s greatest writer, Ford Madox Ford. You don’t believe me? Read Parade’s End and No Enemy.
NEJM  4 Aug 2011  Vol 365
395    Reducing mortality from lung cancer is a worthy aim, which we all pursue as best we can while governments draw billions of pounds from tobacco revenues. What is the place of screening in all this? Here are the results of the National Lung Screening Trial (NLST) which shows that compared with annual chest X-ray, low dose helical computed tomography done for 3 consecutive years in people at high risk can reduce lung cancer mortality by 20% at six years. An editorial by Harold C Sox, attempts to place this in the context of previous trials, and the authors themselves give a very cautious and thoughtful appraisal of their success in the discussion section of their paper. The false-positive rate for CT in this group was very high (23%), but led to surprisingly few invasive tests and very little mortality in contrast to earlier trials. The benefit in overall mortality seems significant and not entirely due to lead-time bias and other kinds of confounding. But the analysis of screening studies is a complex art: if you want a highly readable introduction to it, go to Overdiagnosed (Beacon Press, 2011) by Gilbert Welch, Lisa Schwartz and Steven Woloshin. Then pore over the figures and charts in this paper and see if you can form your own opinion. And then think of the whole picture: in the context of a cost-effective health service, is this something we should start rolling out?

439    Chronic hypertension in pregnancy gets a rather generalised review here – the “chronic” is there to distinguish it from acute pre-eclamptic toxaemia, though of course it is itself a risk factor for PET. There is a useful table of blood pressure lowering drugs which are safe to use in pregnancy. Hypertension in women of child-bearing age is not common and needs investigation, in my view preferably with measurement of renin and aldosterone under controlled conditions: but the article does not go into this but simply suggests following the JCN 7 guidelines in those on multiple drug therapy. If you want to know what these guidelines are, a reference is given. For an old man like me, the rate at which Americans throw out initials and acronyms is a cause of despair. Did I hear that right? Am I cognitively impaired? There is no need to answer these questions.

Lancet  6 Aug 2011  Vol 378
487   For the third week running, we get a negative trial of a new intervention for type 1 diabetes. It is a gloomy sight: most of the beta-cells lie slain or dying on the autoimmune battlefield. Researchers seem to be seized with the desire to fight back and protect the few remaining cells, even though the war is effectively over. The manufacturers of teplizumab are no exception, and the Lancet allows them to report this trial as full of promise, though it was a complete flop. Teplizumab when given within 12 weeks of the diagnosis of T1DM made absolutely no difference to sugar levels or insulin requirements. It interferes with some aspects of T cell function and complement binding, and keeps beta-cells alive longer, as shown by maintained C-peptide levels at 12 months. In an editorial called “Anti-CD3 antibodies for type 1 diabetes: beyond expectations,” Jean-François Bach’s imagination takes flight: “Ideally, type 1 diabetes should be regarded as a medical emergency and treatment with teplizumab could be started within a few days after diagnosis, as compared with several weeks or months as is done now.” The basis of this statement is a post-hoc analysis of a small subgroup at the highest dosage who showed the highest C-peptide response. Is there no such thing in diabetes as waiting for some evidence, preferably of benefit to patients?

507   In a review some years ago, I read that every mammal has its own species of Helicobacter and that in most cases these bacteria behave more like commensals than pathogens. The implication is that before the antibiotic era, every human carried H pylori, and that by trying to eradicate it, we are carrying out a huge (un)natural experiment. Of course, these ancient and resourceful spirochaetes have not taken this lying down. Deprived of their luxurious bath of hydrochloric acid and assaulted with antibiotics, they have formed a global resistance movement, and in some places they shrug off the normal triple therapy of proton pump inhibitor, amoxicillin and clarithromycin. This large trial was set up in several Central American countries to see if a quadruple regime including metronidazole would result in better eradication than triple therapy given for two weeks, but it did not. Here the old stuff still works, and those wishing to prevent ulcers do not have to give up their beer and tequila for two weeks while enjoying their enchiladas.

BMJ  6 Aug 2011  Vol 343
A little study (called HONEST) from a Dutch renal clinic concludes “The findings support the combined endeavours of patients and health professionals to reduce sodium intake.” They took 52 outpatients with mild hypertension and proteinuria with an estimated glomerular filtration rate above 30 and gave them all a massive daily dose of lisinopril (40mg). They were then randomised to 6 week crossovers between a normal or low sodium diet (unblinded) plus placebo or valsartan 320mg. The low sodium diet reduced BP by 1mm Hg more than the valsartan. To be HONEST, I can’t see why this trial was even done, let alone published.

Screening seems to be the topic of the week in these journals and the hottest topic in screening is always mammography. Over here in the United States, it is an article of faith that annual mammography saves lives and that denying it to women under 50 smacks of death-panel socialism. In the UK, on the other hand, it is becoming impossible to find expert speakers willing to defend screening mammography in public: and that’s mammography every 3 years from 50-65. It’s becoming clear that for every life saved, 15 women undergo unnecessary surgery and hundreds undergo the anxiety of further investigation. And here is a study looking at the impact of screening in several European countries, comparing adjacent areas where it was introduced many years apart. “Our study adds further population data to the evidence of studies that have used various designs and found that mammography screening by itself has little detectable impact on mortality due to breast cancer.”

Ann Intern Med  2 Aug 2011  Vol 155
137    In the National Lung Screening Trial reported in the New England Journal, the rate of transthoracic needle lung biopsy following a positive CT or X-ray was only 2-3%. On the other hand, if such a screening programme were scaled up, this would still mean an awful lot of people. Here the authors of the book Overdiagnosed which I recommended above, plus their Dartmouth colleague Ronda Wiener, look at the risk of complications following such a procedure. They studied the outcomes of nearly 16 000 transthoracic biopsies, mainly performed for the confirmation of peripheral lung cancer. Haemorrhage is a rare complication, but can be life-threatening; whereas pneumothorax, at 15%, is far from rare. With more helical chest CTs being done for all sorts of reasons, and a quarter of them revealing pulmonary nodules, order in some extra chest drains.

171   I recently shocked an audience of young doctors by stating that my chance at 61 of having localised prostate cancer stood at about 25%. Since then I have looked at various sources (including Overdiagnosed) and found that according to one study of prostates examined histologically after traumatic death, the true figure may be 60%. So if a urologist decided to do some transrectal biopsies on me and hit the right/wrong spot, I might well have a cancer diagnosis. So what would then be the best nonsurgical management strategy – leave alone, local radiotherapy (brachytherapy), or external beam radiotherapy? The conclusion of this systematic review of different modes of radiotherapy for localised prostate cancer is that nobody actually knows. A most reassuring fact to share with your patients.

Plant of the Week: Campsis radicans

This gaudy native of southern North America made it across from Virginia to England very quickly – perhaps before 1650. South of New York it is an unstoppably vigorous climber which will clamber over and suffocate or pull down anything in its path. But it seldom gets that rampant in our cloudy isles, and worse still, it often doesn’t bother to flower at all, unless carefully grown against a sunny wall. This wall must necessarily be of stone rather than red brick, so as to set off its loud and cheerful trumpets of scarlet, yellow, or orange. Pale south Oxfordshire oolitic limestone is ideal, if only the campsis would flower more often. When it does, its flamboyance is a wonderful corrective to the tasteful, hydrangea-and-rose dominated August gardens of that genteel part of England.

This plant does not mind cold – it is apparently hardy to minus 30C, and grows well in southern Canada: it just needs plenty of summer sunshine. Its commonest name is the trumpet vine or trumpet creeper, though its other American name, “cow itch vine,” hints at earlier, earthier times in ol’ Virginny.