Journal Watch - September 2007

JAMA 5 Sep 2007 Vol 298

993 There is nothing like teaching medical students and trainees to bring home the variety of human personality. Or to show up one’s own limitations. The objective structured clinical examination (OSCE) has been used in Canada for more than 15 years, and in the UK it is now used at various stages of clinical student training and beyond to test competence in history taking, physical examination, communication and problem solving. Between 1993 and 1996 the Medical Council of Canada used OSCEs to test doctors in the years immediately after graduation, and this study analyzes aspects of performance in these tests with later complaints to medical regulatory authorities. Sure enough, the doctors with the worst marks for communication skills attract a significantly higher number of complaints.

1002 I would rate the preceding paper as amongst the best written I have read in the generally dismal literature of medical education research; from this next study I learn that there is actually a grading instrument for such papers, known as MERSQI (medical education research study quality instrument). The authors demonstrate that there is a link between MERSQI scores and the level of study funding. Have MERSQI upon us, O Lord, and grant our funding requests.

1010 Now for showing up one’s own limitations. There is a nice little biostatistics quiz at the end of this paper about medicine residents’ understanding of biostatistics. It takes about fifteen minutes. Read the questions carefully and you’ll get a better score than I did – I lost more marks from insufficient attention than I did from my areas of gross ignorance (analysis of homogeneity and logistic regression – like most doctors). The average for residents with higher degrees and more advanced training is 50%, which is a bit dire.

NEJM 6 Sep 2007 Vol 357

977 As I said last week, we are going to have to get used to the sight of studies about the analysis of single-nucleotide polymorphisms relating to common diseases using powerful SNP analysing chips and associated statistical software, though whether we will actually read them is another matter. This one uses exemplary methodology, comparing SNPs in large cohorts with rheumatoid arthritis or systemic lupus erythematosus with matched controls in two widely separate localities – the USA and Sweden. The results are, to the ignorant reader, decidedly underwhelming in relation to the effort expended. Various combinations of alleles around STAT4 are associated with various increases in risk for RA or SLE or both, but not to any dramatic degree.

987 Dronedarone has been waiting in the wings for several years, hoping to supplant amiodarone in various contexts, including atrial fibrillation. Unfortunately one of the first studies, in heart failure, had to be stopped because of a mortality increase. So the studies reported here of patients with a history of episodic AF excluded those with heart failure, and the results are modestly good. Dronedarone prolonged the median time to relapse by a factor of more than two and reduced the heart rate at relapse by 15 bpm compared to placebo. And we know that it has far fewer side-effects, and a much shorter half-life, than amiodarone.

1000 Interventional cardiologists are also very keen on trying to cure atrial fibrillation, or prevent it in patients with sinus node disease. The pacing mode likely to become fashionable is called dual chamber minimal ventricular pacing, which is somewhat better at preventing the development of persistent AF than conventional dual-chamber pacing in these patients.

1009 We recently learnt that non-fasting hypertriglyceridaemia is more predictive of cardiovascular risk than fasting measurements, but this reviewer doesn’t let that get in the way of his rather nebulous discussion of fasting hypertriglyceridaemia, about which we know a lot more. Not much of it is very important, however, and it would be hard to spot a take-home message anywhere here. Mine would be to look out for diabetes and too much alcohol.

BMJ 8 Sep 2007 Vol 335

481 There is a group of patients at increased risk of coronary heart disease who we may be missing – those with a family history of premature CHD. In our zeal to record smoking, BMI, cholesterol and so forth we tend to miss the family history – but it would be easy to get the data from a patient questionnaire, looking out specifically for CHD in a male relative under the age of 55, or a female relative under 65. They are at added risk even if their other risk factors are normal.

486 The psychological impact of screening relates to the preparedness of the screened population – I think this is the take-home message of this Cambridge trial of screening the population quartile at highest risk of type 2 diabetes. The investigators manage to get two papers out of this observation (see also p.490) without giving us their substantive results on diabetes detection. When these appear, they might form a greater ADDITION to the sum of human knowledge.

497 When the thiazolidinediones first appeared, I was enthusiastic about the potential of new agents which might work by insulin sensitisation rather than by flogging the dying islet cells. A couple of patients promptly went into heart failure and I’ve been more cautious since. The jury is still out over the exact balance between harm and benefit in these agents, and indeed whether this is a class effect (see Annals of Internal Medicine website), but I think we can agree with this paper that the safest antidiabetic agent in patients with cardiac impairment is metformin – the risk of lactic acidosis is tiny (see p.508).

502 Nearly a quarter of us go around with permanently blocked or runny noses – dignified with the name of perennial rhinitis. The authors of this review state that most of it is due to Dermatophagoides pteronyssinus (the European house-dust mite – Americans can save their curses for their own brand, D. farinae) and that skin tests reliably detect this. There isn’t, however, any reliable way of eliminating house-dust mites so we’re back to steroid sprays or just living with it.

513 The SMART studies demonstrated a while ago that a budesonide/formoterol inhaler not only provides good control of chronic asthma but also effectively treats exacerbations usually with no more than a single extra puff per day. This message is relayed to us on the Change Page, which still adheres to its silly policy of only citing the key studies on the website rather than the printed version.

Lancet 8 Sep 2007 Vol 370

829 If you treat 80 diabetics for 5 years with a fixed combination of perindopril and indapamide you might avert one death. Lots of terribly distinguished people supervised and wrote up this immense trial, which is presumably why it features in The Lancet. The funding source was the George Institute, a non-commercial institution, though this is nowhere mentioned in the text. So this is not a marketing trial, but as the editorial (p.804) states, it’s unlikely that it constitutes any ADVANCE beyond the known fact that blood pressure lowering in diabetes is a good thing.

841 Once again, The Lancet provides us with some awesome international studies of prevalent disease – in this case, mental illness in its widest sense. The boundaries are unclear, of course – there is a review on p.819 of a book which accuses psychiatry of turning normal sorrow into depressive disorder. And this comes during a week when in a rare loss of temper I found myself telling a community psychiatric nurse that such distinctions are meaningless – at least as regards the “caseness” of a woman who had been an in-patient for over a year due to her perceived risk of self-harm, and whose husband is now dying from motor neurone disease. Her needs are unmet and so are those of the majority of people with disorders of mood, anxiety and substance abuse throughout the world, and particularly in developing countries.

851 “For there is nothing either good or bad, but thinking makes it so” is what we hear Hamlet say in standard modern editions (2.ii. 248), but these words only occur in the Folio version, not the earlier first or second Quartos, and probably represent Shakespeare’s third thoughts as he added the whole section on Hamlet’s profoundly depressive view of all Denmark as a prison. Depression is itself the worst of prisons, leading to greater impairment of function than any other chronic disease in the World Health Surveys – which do not, however, include Denmark.

859 The theme is developed in two masterly surveys of the global burden of mental illness in relation to physical illness – no health without mental health (p859), exploring the single truth that lies within the Indo-European root for “whole”, “holistic”, “holy” and “health”.; and an exploration of the entirely inadequate resources addressed to this basic human need (p.878).

891 We have been hearing an awful lot about human papillomavirus and cervical cancer in vaccine studies lately, and if you want to learn more about the exact mechanisms, here is a nicely illustrated seminar on the topic. The conclusion cheerfully asserts, “Much of the cervical cancer problem can be solved with existing or soon-to-be available technology, sufficient will, and modest resources.”

Ann Intern Med 4 Sep 2007 Vol 147

294 The arrival of a new diagnostic test which simplifies the diagnosis of a common and important disease should be the occasion of great excitement and research in general practice – that was what I told my work partner Harold Hin about endomysial antibody testing for coeliac disease ten years ago. He took me at my word and did a landmark study, while I continued to fumble about with BNP and “heart failure”. The monkey oesophagus bioassay he used had a specificity of 100% in that study, which was one of several that pushed the prevalence estimate of coeliac disease up by one order of magnitude. But even now, judging by the agreed modern prevalence figure of 1%, we are missing half of the patients with coeliac disease in our practice. Most of you are missing three-quarters. Now this Dutch secondary-care study reveals that everyone with coeliac disease carries a specific HLA-DQ type and that HLA testing had a specificity of 100% compared with small-bowel biopsy. So surely if you combine EMA with HLA-DQ testing you’ll have a completely reliable non-invasive way of diagnosing CD? Alas, it is not quite that simple – but to find out why you’ll need to put on your best Bayesian thinking-cap and read the editorial on p.339.

311 We’ve all probably had a bit of left ventricular hypertrophy at some time or other, because our cardiac myocytes are fantastically adaptable in their response to loading and they grow and shrink within weeks. As they grow in size, they produce a bigger electric charge – measured by ECG – and of course this growth also leads to an increase in cardiac mass and volume, best measured by MRI (CMR), but usually estimated by ultrasonography (echocardiography). The correlation between ECG and echo is very poor – see the review on In fact the whole business of measuring ECG-LVH in hypertension is a bit passé. If you really wanted to know how much strain the ventricle was experiencing, you would have to do serial measurement of cardiac mass and BNP over decades. But in the real world, this study (LIFE) shows that regression of ECG-LVH from its pre-treatment voltage in hypertensive patients results in less hospitalisation for heart failure. For other outcomes, look at the Framingham database, where ECG voltage data are available for nearly 60 years.

330 Hyperkyphosis is traditionally known as the “dowager’s hump” – a dowager being a widow who was allowed to enjoy the use of her husband’s assets while she lived. These noble ladies presumably led lives of genteel inactivity and lived much longer than most of their female contemporaries, giving them time to develop advanced osteoporosis of the spine. But it is wrong to assume that this is the main cause of hyperkyphosis, according to this narrative review. Although it is present in up to 40% of elderly people, only a minority of even the most severe cases have demonstrable vertebral fractures. There is a lot we don’t know about why the elderly stoop.

Play of the Week: the three versions of Shakespeare’s Hamlet

Hamlet is perhaps the best example of Northrop Frye’s definition of a great literary work as one whose circumference cannot be determined. Every attempt to describe or discuss it stands condemned by its own inadequacy. Perhaps even Shakespeare stood in awe of his own creation: there is certainly good evidence that he returned to it more than once.

The play was an immediate success, and the first version of it to appear in response to customer demand was the Quarto of 1603, an unauthorised garbled transcription by someone who tried to take notes during performances. Shakespeare replied with his own version, the Second Quarto of 1604/5, “newly imprinted and enlarged to almost as much againe as it was, according to the true and perfect Coppie” The play was now exceeding the bounds of performance time, so Shakespeare had probably augmented the text for this printed version. After his death, his works were collected in the Folio edition (1623), which contains a lot of new plays and some old ones in direct transcriptions from the printed Quartos: but in the case of Hamlet, there are further considerable additions and changes from the Quarto of 1605. These clearly come from the hand of Shakespeare himself, and include the section on Denmark as a prison to which I referred.

I had better say no more. People who project their own vanities and neuroses on Shakespeare always make fools of themselves – Frank Harris, TS Eliot, AL Rowse, and in the last few weeks, Germaine Greer. Sufficient that we have three versions of Hamlet to read and wonder at. Why, some day I may even go to a performance.


JAMA 12 Sep 2007 Vol 298

1163 Just a few years ago, it seemed almost certain that giving people group B vitamins including folic acid would lower homocysteine and improve vascular health. But all the trials begun with high hopes have ended in failure: B-group vitamins do indeed reduce homocysteine but they have no effect on vascular disease in any risk group, including patients with advanced renal disease. This study randomised such patients to a good strong cocktail of folic acid, pyridoxine and cyanocobalamin, but achieved no reduction in total mortality or vascular events over median follow-up of 3.2 years.

1180 The abbreviation TZD comes in useful when discussing a thiazolidinedione – the word itself sticks trickilly in the throat-zone if we thrillingly try to trill it. The TZDs have been under a cloud lately mainly as a result of the meta-analysis of rosiglitazone and cardiac outcomes published months ago in the New England Journal, even though this was immediately questioned by a re-analysis of the same trials on the Annals of Internal Medicine website. And here come two studies which appear to show that not all TZDs are the same. in this meta-analysis of the RCTs using pioglitazone, it does not show the same tendency to increase myocardial infarction as rosiglitazone; in fact it decreases the risk of MI and all-cause mortality. Both drugs cause an increase in symptomatic heart failure, but not death from heart failure.

1189 Here is another meta-analysis of the rosiglitazone trials; once again, the increase in myocardial infarction is found to be significant (RR1.42, 95%CI 1.06-1.91), but there was nonetheless no increase in cardiovascular mortality. If you take these two papers together, the prospects for this glitazone in competition to pioglitazone do not look rosy, especially since the latter is the cheaper drug (at least in the UK). Pio is certainly the TZD I shall be prescribing in the future.

1196 Good palliative care depends on close assessment and constant attention to symptom control, and after pain, intractable nausea and vomiting must count as the most distressing symptoms. There are plenty of good manuals of palliative medicine but for a really good update on the latest options for nausea/vomiting, this paper is hard to beat. Keep it handy.

1209 If you have the slightest interest in evidence-based medicine, then you must read this short paper (and if you don’t have any such interest, why are you practising medicine?). The bane of every conscientious doctor’s life is trying to overcome the limitations of applying summary results of clinical trials to individual patients. Whole books have been written about this by EBM sceptics and chaos theorists, but the authors of this commentary (David Kent and Rodney Hayward) outline the solution in three pages of clear prose and a single figure. The present system of subgroup analysis can hide huge variations in risks and benefits, so instead every major interventional trial should report multivariate risk-stratified analyses based on easily obtainable clinical variables. That is after all what we doctors try to do with our brains for a considerable portion of each day. This landmark essay brings Sackett heaven a step nearer.

NEJM 13 Sep 2007 Vol 357

1072 For some years the town of Norwalk, Ohio, has had to live with the infamy of association with a virus causing epidemics of severe gastroenteritis, but its citizens can return to respectability now that the “Norwalk-like virus” has been renamed the norovirus. Here is a little piece describing the beast, which is a particular bane to those who consume raw molluscs. Oyster-eaters of the world should unite to develop a vaccine.

1075 When my father developed coronary heart disease, he became intolerant of motor exhausts, and I remember him having an angina attack near a bus station. At the time I was inclined to put everything down to his rich Polish emotional life, but this study shows that the effect may well have been causal. In this study of male volunteers with stable coronary disease, brief exposure to dilute diesel exhaust fumes really did promote myocardial ischaemia and inhibited endogenous fibrinolytic activity.

1094 As every medical student knows, the clotting factor cascade is long and complex and full of opportunities for lucrative interference by pharmaceutical companies. For some reason many of the drugs they develop are given names resembling Astérix characters, such as idraparinux, a long-acting inhibitor of activated factor X. It is given by weekly injection at a standard dose, and in this trial it proved as good as heparin followed by warfarin for preventing venous thrombotic events after deep vein thrombosis but not as good following pulmonary embolism. Extended over a six-month period, it was associated with an increased risk of major haemorrhage (p.1105).

1121 The ischaemic myocardium is subject to all manner of insults, crude or subtle. Crude ischaemia causes necrosis, while subtle ischaemia causes slow cell death (apoptosis) or a state of suspended animation called hibernation. Then along comes the interventional cardiologist and submits the myocardium to new insults, subtle in the form of drugs, or crude in the form of reperfusion. Reperfusing the dormant myocytes can cause cellular damage and electrical instability, described at length in this review article. The current best method of prevention, as mentioned a couple of weeks ago, is to induce ischaemia elsewhere in skeletal muscle – ischaemic preconditioning. That’s crude too, but effective.

BMJ 15 Sep 2007 Vol 335

542 Searching for something original to review in this week’s BMJ, I turned to this short paper describing the reality of blood pressure measurement in British general practice in the era of incentive payments. Spurred on by cash and the reminders which flash up on our computers, almost every GP in the land hits the targets for recording BP, and also for treating it, since the targets are unambitious. The authors argue for greater concentration on achieving control in those at greatest risk, and observe that high BP readings often result in “therapeutic inertia”, i.e. no change in treatment – sometimes justifiably, often not. The editorial by David Wald (p.523) takes a somewhat different line. It attacks the very notion of payment for recording risk factors, since they don’t account for the bulk of cardiovascular disease. He comes close to recommending age as the only criterion for treatment, quoting the works of NJ Wald. Are they by any chance related? Might he be hinting at the desirability of a Polypill for all over-55s? Should this be known as the Waldensian heresy?

558 This clinical review of adult coeliac disease is not exactly original either, but is a serviceable summary of where we stand at present. It does not however suggest how we might use genetic tests, and promises “many alternatives to a gluten free diet “ in its Ongoing Research box, but mentions only one.

Lancet 15 Sep 2007 Vol 370

937 For years I referred to the Stent Wars in these columns, as if they were a great intergalactic battle, but in the end they have come to resemble a few square metres in the Somme, fought over till they are nothing more than a pulp of mud and human remains. So is General Paclitaxel’s drinks cabinet a few feet nearer Berlin? No, actually, von Sirolimus has moved his one a few feet nearer Paris. But overall, according to this collaborative network meta-analysis of outcomes associated with drug-eluting and bare-metal stents, it is complete stalemate.

949 Some years ago, too, I advised readers to practice pronouncing “ximelagatran” because this direct thrombin inhibitor was bound to supplant warfarin, and INR measurement would become a thing of the past. Alas, ximelagatran turned out to have substantial and lasting liver toxicity in some patients, so we may instead have to learn to pronounce dabigatran etexilate. I don’t know why the inventors of anticoagulants have such a grudge against the human vocal tract. Let’s hope that this orally available thrombin inhibitor is kinder to the liver, because in this study dabigatran proved the equal of enoxaparin for preventing venous thromboembolism following hip replacement.

Arch Intern Med 10 Sep 2007 Vol 167

1730 As I instructed you some weeks ago, start taking vitamin D supplements now, or move south and take your clothes off. This meta-analysis of the RCTs of vitamin D supplementation shows a total mortality benefit and, as you’ll remember from the NEJM review article, there are a lot of qualitative health benefits too. I’m off to buy a bottle of 800 IU cholecalciferol tablets this morning, though more may be better (worried about hypercalcaemia? This doesn’t occur even at doses of 10,000 IU/day). The editorial on p.1709 sums things up nicely.

1745 As we get older, we get little white matter lesions in our cerebral cortices which show up on MRI. This Italian study explores the association between these and incident hip fracture in people over 65. Sure enough, there is one, but it’s fairly weak.

1752 A few decades ago, someone said that we had entered the age of safe surgery and dangerous medicine. At the beginning of the twenty-first century, that’s truer than ever: this report on serious adverse drug events reported to the US Food and Drug Administration between 1998-2005 highlights a veritable pandemic of iatrogenic harm. In this short period, reports of serious reactions went up 2.6-fold and fatal reactions 2.7-fold. Opioids top the US death scale and oestrogens cause the most non-fatal harm. Anti-TNF drugs are climbing up both scales. The “atypical” antipsychotics clozapine, risperidone and olanzapine also feature increasingly.

1791 Helping patients to stop smoking is often the biggest favour we can do for their health, and until recently, bupropion was the best drug we had to help them (in fact nortriptyline was just as good, but nobody was hyping it). But since the drug was quite expensive, and the trials involved supportive counselling, we were instructed to follow a complex protocol involving named smoking cessation advisers and a restricted prescribing schedule. I can’t say I took much notice, and in fact ended up following a pragmatic schedule of GP visits very similar to those in this Italian trial, which achieved a 25% quit rate at a year – about the same as the formal trials.

1798 If you are an obedient soul, your heart will benefit. This is the message of this study of self-reported medication adherence from the Heart and Soul study, which asked 1015 outpatients with coronary heart disease whether they took their drugs as the doctor prescribed. Those who did had half the event rate of those who didn’t – an astounding effect size!

Plant of the Week: Fuchsia magellanica var. molinae “Sharpitor”

Outdoor hardy fuschias are among the best things to be seen in our gardens at this time of year, and “Sharpitor” is a personal favourite; though I have, as usual, managed to kill it. It has pale variegated leaves and pink flowers and looks rather lovely all summer. But if you want it to live through a dry spell in its early days, it is best to water it. The more ordinary magellanic fuschias with green leaves and red flowers also deserve a place in the garden, and these are quite indestructible.

There are several thousand different, less hardy, fuchsias which give great joy to their propagators throughout the greenhouse-owning classes of southern Britain. They are of course pronounced “fewshers” rather than anything resembling, well, you know, an improper word. Patients sometimes make gifts of particularly gaudy examples to their doctors: these should be treasured, and any evidence of their demise carefully hidden.


JAMA 19 Sep 2007 Vol 298

1312 The only item of interest I could find in this week’s JAMA was this review entitled Prescribing Amiodarone, something most of us do reluctantly at the behest of a cardiologist. A nasty drug, causing corneal microdeposits in almost everybody, and photosensitivity in many; with liver and pulmonary toxicity in a substantial minority. There are better drugs for most indications, the most common exception probably being atrial fibrillation with systolic heart failure.

NEJM 20 Sep 2007 Vol 357

The New England Journal also offers thin pickings this week, and the reason is similar – half of its space for original papers is taken up with genomic studies of zero interest to most clinicians. Such studies are no doubt very demanding, cutting-edge work and may - who knows? – one day translate into something of practical importance. But why publish them in the core journals where generalists look for the best evidence to inform day-to-day practice? Those interested in genomics could be provided with a brief update page once a month. Perhaps it could be presented in a cumulative way, so that instead of being told that TRAF1-CF is a risk locus for rheumatoid arthritis (p.1199) we would be shown all the known loci. Eventually it might possibly start to make sense.

1179 Antenatal corticosteroids are widely given to women at high risk of premature labour because they have been shown to accelerate lung maturation and so reduce the risk of respiratory distress syndrome and serious neonatal morbidity. Here is a follow-up study of two year old children born to mothers who had been randomised to repeated steroid injections or placebo (saline) injections in the Australasian ACTORDS trial. There was no difference between the groups in body size or neurosensory disability. A similar American follow-up study (p.1190) shows a similar result, with a tendency (not statistically significant) to more cerebral palsy in the treatment group.

1229 A low-key review of gynaecomastia makes the point that it is very common in adolescence and then again as men age. Teenage gynaecomastia usually lasts about six months, whereas in older men it gets steadily commoner, especially if they are prescribed drugs like bicalutamide or spironolactone. But once or twice in a professional lifetime you may see gynaecomastia as the presenting sign of testicular cancer, so make sure you do the appropriate examination (after putting together an appropriate explanation).

BMJ 22 Sep 2007 Vol 335

599 In last week’s BMJ David Wald wrote an editorial questioning the assumptions of a paper; this week he writes the paper and somebody else questions its assumptions. It’s all about screening for familial hypercholesterolaemia, based on a meta-analysis performed by the authors, from which they derive a screening strategy which fulfils 8 out of 10 criteria established by Wald senior. It involves measuring cholesterol in all children aged 15 months and if it’s high, working back to their parents (it’s a dominantly inherited disease, so one parent will generally have it).The editorial picks holes in this argument, and I don’t foresee any screening programme until we know more about the potential adverse effects and cost-effectiveness of statins given to children and adolescents.

608 The authors of this clinical review of the management of infertility make it clear that this is no more than a brief synopsis. It reads as quite a struggle for brevity, skimming along from ovarian reserve tests to the latest dilemmas in assisted conception.

612 Evidence based case report: Do all fractures need full immobilisation? This question obviously invites the answer no, and is followed by a nice little vignette of how evidence can be used in the real conditions of general practice. Paul Glasziou is even allowed some references. But at the start of the journal (This Week) this article is billed as “do all fractures need full mobilisation”, and on p.585 the drug which causes most deaths in the USA is twice referred to as oxycondone. Evidence based question: does the BMJ need more careful editing?

Lancet 22 Sep 2007 Vol 370

1010 Helicobacter pylori is fighting back, and the standard 7-day eradication regimes no longer provide clearance rates above 80%. For better success rates we need to use a two-week course of triple therapy, or perhaps add in a bismuth salt. Or perhaps use sequential therapy as discussed in this editorial. More trials are needed, outside the small group of Italian centres which have provided most of the current data.

1035 Some months ago, a reader who works in Cambodia asked if I could cover more papers about the developing world; not in detail, but just so that she could look them up. It didn’t seem much to ask, and I promise that I will try soon to append a brief list for anyone who might find it useful, unqualified though I am to comment on them. One strategy is simply to read The Lancet. Never a week goes by without an article of global interest: here is a study showing that simply handing out free insecticide-treated bednets can reduce children’s total mortality in Kenya by 44%. I’m not sure one should seriously quote this figure, given the loose study design and the wide confidence intervals, but the message is nonetheless clear: malaria is an important and preventable cause of death, and simple measures widely applied can make a big difference.

1040 Mortality in children under 5 is the lowest ever, but it is not nearly low enough, and the decrease by 2015 is likely to be 27% rather than the Millennium Development Goal 4 of 67%. I recently saw a letter from Bruce Charlton in the BMJ rapid responses claiming that the reduction in child mortality was proof of the success of global capitalism. Strange then that the biggest improvements have been in Cuba and Nicaragua.

1089 Sudden death due to stress is common in nineteenth century fiction, and can even be found in the novels of EM Forster. But what is the real cardiovascular toll of stress? According to this interesting and comprehensive review, it is considerable, and certainly includes sudden death, as well as other interesting phenomena such as apical ventricular ballooning and acute heart failure. Public speaking is a recognised standard stressor, so next time someone tries to persuade you to give a lecture, adding that it won’t kill you, you can reply that it might. As for chronic stress and the effects of depression and personality type, there are abundant associational data and many plausible physiological mechanisms, but to date the universal panaceas of serotonin reuptake inhibitors and cognitive therapy have failed to show any effect on cardiac outcomes in depressed people with heart disease.

Ann Intern Med 18 Sep 2007 Vol 147

357 We try to see all our patients with type 2 diabetes at least twice a year: most keep acquiring comorbidities and get seen much more frequently. We measure their HbA1c, weight and blood pressure, tweak their medications, make sure they have had an eye check, and their ten minutes are up. But what about exercise? OK, we may go as far as some parting mumbles about joining a gym, but this study shows that we should probably be putting it first on the list. And prescribing it. And doing it ourselves. Aerobic (an exercise bike) is better than resistance (a weight machine); and the two together are best of all at reducing HbA1c. But of course the benefits are also seen in weight reduction and cardiovascular fitness.

386 It’s no good looking to current pharmacology to address the problem of type 2 diabetes. This systematic review confirms what we feel from experience: the only really beneficial drug is metformin, which some patients find hard to tolerate, and all the more recent, expensive drug classes (thiazolidinediones, a-glucosidase inhibitors, and meglitinides) are no improvement on the older, cheaper agents, i.e. sulfonylureas and metformin.

Plant of the Week: Rosa “Dublin Bay”

The gardens of England have yet to turn truly disconsolate, but they are starting on their way towards the muddy bleakness of November. However, shimmering among ours is the reddest rose of all, blooming even more freely than it has done for the last three months. If you possess some large book of rose photographs, you will find that this one looks as if the reds on the film or the print have gone wrong. They glow in improbable ways. But that is exactly what this plant does in the real garden. That is why we decided that we must have it, even though it lacks what we had hitherto regarded as the most essential requisite of every rose – scent. On very warm summer evenings, you might catch a hint of apples.

Just be grateful for its astonishing redness and its astonishing generosity of flower. As you make your way into the September garden, mournfully whistling a last song by Strauss, you will suddenly find yourself attempting Ernst’s fiendish variations on “The Last Rose of Summer”.


JAMA 26 Sep 2007 Vol 298

1401 In the classic British wartime film In Which We Serve, Noel Coward assumes captaincy of the warship HMS Torrin and takes every opportunity to surround himself with young men in shorts. Addressing his crew, he declares that “a happy ship is an efficient ship, and an efficient ship is a happy ship.” That is the message of this American study of detecting and treating depression in the workplace. Depression is the commonest cause of prolonged sickness absence, but employers have shown little enthusiasm for screening and treatment programmes. This study attempts to prove that such a programme using telephone screening, outreach and care management reduces absence from work and is cost effective. It’s nice to see that they don’t regard this as the only, or necessarily the most important, criterion of success. As Michael Marmot writes in this week’s Lancet, “(it is) not just because better health makes economic sense, but because it is right and just”.

1412 Hong Kong sounds like a dangerous place for the bowel: more than a fifth of the older population has colorectal neoplasia, rising to more than a third of those with newly diagnosed coronary artery disease. I’m not sure whether this is just an epidemiological oddity or whether it might turn out to shed some important light on the pathogenesis of colon neoplasms. The conclusion as usual speculates on a variety of possible mechanisms by which the conditions could be connected.

1420 Can oily fish prevent type 1 diabetes? We won’t know for some time, but this study does suggest that a high intake of omega-3 polyunsaturated fatty acids offers some protection in high-risk children. So did a Norwegian case/control study which found that children with diabetes were less likely to have taken cod-liver oil, but the DAISY study tried to assess dietary intake of omega-3 fats from all sources and even to measure fatty acids in erythrocyte cell membrane. The end-point is islet autoimmunity and the hazard ratio just reaches statistical significance (CI 0.21-0.96) for a protective effect. An interventional study is under way.

1429 In a long list I once compiled for a proposed BMJ series called “Commoner Than You Think”, pituitary dysfunction following brain injury or subarachnoid haemorrhage was a late entry which slipped in due to the interest of a Spanish endocrinologist. Here is a systematic review of the studies to date, 19 in number. In fact nearly half of patients who survive an aneurysmal SAH have evidence of hypopituitarism . A lot commoner than you think, then, and worth looking for.

NEJM 27 Sep 2007 Vol 357

1281 Thimerosal, thiomersal, let’s call the whole thing off. It was all rubbish to start with. It started with a worry that the mercury in this vaccine preservative might cause neuropsychological damage in children as they received ever more vaccinations in early childhood, and on a precautionary basis in 1999 the US Food and Drug Administration asked vaccine manufacturers to use a different preservative (see p.1278). A wave of anxiety and litigation followed, but in fact the whole scare was baseless, as this study of outcomes at 7-10 years clearly demonstrates. It also demonstrates the statistical truth that if you look at enough outcomes and enough subgroups, there are bound to be some “significant” associations: but in the case of thiomersal, the marginal goods and bads cancel each other out. Thimerosal (US), by the way, is the same compound; it’s just an example of the linguistic phenomenon of metathesis.

1301 There is an association between high levels of LDL-cholesterol and cardiovascular events which a lot of people think is causal, because statins lower LDL-C and also prevent CV disease. On the other hand, other drugs lower LDL-C and don’t prevent CV disease. Other drugs raise HDL-C, which ought to prevent CVD, but these fail every time. Expect lots more bickering (and don’t even mention homocysteine), and more studies like this one, which shows that even if you lower LDL-cholesterol markedly with statins, low HDL cholesterol levels remain associated with more cardiovascular events. But probably don’t cause them, in my opinion.

1316 In these days when every patient is scoured for evidence of renal impairment using the laughably blunt tool of estimated GFR, it’s refreshing to come across a review that reminds us of the days when renal failure was diagnosed by smell, and called uraemia (uremia in the USA, due to the linguistic phenomenon of Websterism, after the perverse decision by Noah Webster in 1806 to differentiate American spelling from British spelling). Here in the UK, renal physicians live in a world apart, surrounded by equipment most of us don’t understand. This review questions whether they understand it either, so many are the unresolved questions which surround renal dialysis.

BMJ 29 Sep 2007 Vol 335

625 The legalisation of physician-assisted death will be the beginning of a slippery slope towards legalised murder, put the vulnerable under duress to accept euthanasia rather than be a burden to their relatives, be seen as a cheap alternative to good palliative care, and put doctors under impossible strain from the volume of demand. Those who shout these arguments so loudly every time the subject is raised may care to read the evidence in this editorial. In the Netherlands and Oregon where it has been legal for a decade, demand is steady or falling (at about 1 death per 1000); it is usually requested by well-educated people; and it may be associated with a reduction in involuntary death-hastening acts by doctors such as terminal sedation.

651 The Royal College of General Practitioners does not have a research budget as such, but somehow managed to start an observational study of oral contraception 40 years ago, when The Pill was a new thing and ethics committees had not even been invented. Anything might have happened to the millions of women who opted for this revolutionary form of sexual emancipation; many religious leaders of the time (and since) hoped it would all end in disaster. Perhaps, from their point of view, it has. Shocking numbers of young people are indulging in sexual intercourse for nothing but pleasure. But the effect on cancer has been beneficial for most women.

655 Neonatal infection with group B streptococcus remains rare but often devastating, and in theory it is preventable – currently by the use of prophylactic antibiotics around the time of birth, and soon perhaps by the use of a vaccine. This paper is a virtuoso display of modelling; the 8 authors started by doing 32 systematic reviews and ended up comparing 341 models for cost-effectiveness (see editorial,p.622). Despite this, all the contributors seem to have remained alive; and their conclusion is nice and simple: give antibiotics to all preterm and high-risk groups.

659 Chest pain presenting to an accident and emergency department usually guarantees immediate attention, and in big UK departments it seemed like a good idea to set up chest pain units to streamline assessment and perhaps reduce admissions. But this study of 14 British hospitals can’t find any evidence of improvement in process end-points such as reattendance and admission; in fact a trend towards an increase.

663 The other journals have carried several reviews of polycystic ovary syndrome in the last couple of years, but his one is by far the clearest and most practical, as well as the shortest; and it manages to squeeze in anovulatory infertility as well. Sensibly, it points out that the best treatment for PCOS is usually weight loss; lose 5-10% of your body mass and you lose 30% of your fat and most of your insulin resistance.

Lancet 29 Sep 2007 Vol 370

1129 Meta-analyses of thiazolidinedione trials lithely try tying down risk-zones reliably. Say it again, faster. One outcome they look at is “heart failure”, a term which I once argued in a BMJ editorial is so vague and alarming as to be clinically meaningless and unhelpful to both doctors and patients. John Cleland, though a longstanding friend and fellow-maverick, would not go quite so far. But he is fairly outspoken in the editorial which accompanies this study (p.1103). The glitazones do tip patients into a syndrome of fluid retention and breathlessness but they quickly recover. This is not the same “heart failure” about which John has written so many books, and I one. Nor is it really among the most patient-important outcomes in diabetes (editorial p.1104). Let me suggest “The UKPD Study has done more harm than good for diabetes management”: now there’s a theme for a for-and-against feature in the BMJ.

1137 The avian influenza virus H5N1 has dropped from the headlines for the time being, and let’s hope it never comes back. A molecular pathology study of two victims shows that the virus can travel widely around the body, affecting the brain and invading the placenta.

1146 Politicians of every party regularly affirm their determination to move people from chronic sickness benefits to paid employment, but nothing ever seems to happen. If they were motivated by good will towards the ill, they could almost certainly help them; whereas tightening the entitlement rules just leads to greater effort in circumventing them. This study shows that with individual placement and support even people with severe mental illness can succeed in finding competitive employment.

1153 Last week, I rather misrepresented Bruce Charlton’s letter to the BMJ, which appears in print this week: he didn’t claim that capitalism was the best way to reduce child mortality, but to eliminate world-wide poverty. Maybe. But there seems no current check on the enormous inequalities in wealth which capitalism generates, and which in turn generate enormous inequalities in health. The Whitehall Studies conducted by Michael Marmot are classics of the British literature on this subject, and he has subsequently chaired an international group called the Commission on Social Determinants of Health to look at the wider picture on behalf of WHO.

He raiseth up the poor out of the dust, and lifteth the needy out of the dunghill;

That he may set him with princes, even with the princes of his people.

Psalm 113:7-8.

Arch Intern Med 24 Sep 2007 Vol 167

1837 The ideal cancer biomarker would appear at an early, treatable stage of growth and have a predictive value of 100% in all populations, so that everyone would be cured and nobody would be worried. Some hope. But while optimists would claim that we are already near to finding some useful cancer biomarkers, it takes a really extreme extension of thought to credit the white cell count with any predictive value (see editorial, p.1822). This paper reports that if you look at the superabundant data of the Women’s Heath Initiative study (144,000 women) and take the highest quartile of wbcs and compare it to the lowest, you can find a slightly higher rate of common cancers (breast, colorectal, endometrial and lung) in the latter

1884 One of the commonest problems in elderly people is systolic hypertension which is usually the result of arterial stiffening as evidenced by a wide pulse pressure. As you strive to lower the systolic, down goes the diastolic, and somewhere around 65mmHg you begin to wonder if you might be overdoing it. The conclusion of this analysis of the Syst-Eur Trial is that you can carry on down to 55mg if you like, but the data are paradoxical and there is little evidence of benefit.

1892 I have spent a working lifetime (well, not quite: I am not dead, and retirement is some time off) hoping that someone might come up with an effective treatment for chronic low back pain. And here it is: sham acupuncture. Traditional Chinese acupuncture is just as good, but needs to be learnt; while sham acupuncture, as I’ve said before, simply requires a set of needles and the plausible manner which many of us have sought to acquire over the years. This widely publicised German trial confirms what other acupuncture trials on different parts of the body have shown: put the needles where you like, and you will achieve a useful effect in many people. There is no need to follow the ancient Chinese pricks.

Plant of the Week: Vitis vinifera var “Purpurea”

People have attempted to grow grapes in Southern Britain for two thousand years, and a small trickle of sour wine has marked the warmer periods of English history. Now, we are told, the climate of Britain’s wine-making areas closely resembles that which first gave rise to Champagne, Chablis and Muscadet. Over the last twenty years, it has become impossible to buy a really dry Champagne, or a really flinty Chablis, or a really biting Muscadet, because they all now taste of the warm south. So come on British viticulturalists – bring on a lovely sparkling Kent, a Hampshire full of grass and minerals, a nice dry Dorset.

But living in North Oxfordshire, I think I will stick with the ornamental purple vine. It sprawls all over a fence, competing with rampant ivy, and has leaves that turn from dark purple to bright orange at this time of year. Hidden beneath them are clusters of small black grapes. You can put these in bowls as house ornaments, though they quickly shrivel and become covered with clouds of fruit flies. Or you can use them as the Romans did their British grapes – to make verjuice, a sharp bitter condiment, often mixed with bird droppings for a better flavour. Excellent with dormice fried in honey.

Developing World Checklist

Challenges in the Prevention, Diagnosis, and Treatment of Malaria in Human Immunodeficiency Virus–Infected Adults in Sub-Saharan Africa

Barriers to improvement of mental health services in low-income and middle-income countries





Page last edited: 13 March 2009