Journal Watch - October 2008

JAMA 1 Oct 2008 Vol 300

1551 One of the people who taught me psychiatry was Anthony Storr, who was quite famous in his day as the author of various books about creativity, sexual deviation, music, Jung, and other subjects redolent of student life in the sixties. The last one I read was about psychotherapy, at around the time I first started in general practice. The only bit I remember is his statement that to be of any use, psychotherapy has to involve some hard work on the part of the therapist as well as the subject, and that if the therapist doesn’t go home feeling tired, the chances are that he has done little good that day. It would often comfort me as I staggered home feeling knackered. But that of course doesn’t of itself mean that I (or he) had done any good: and I am still agnostic about the effectiveness of long-term psychodynamic psychotherapy. Stuff happens; you get over it, or you don’t. People are deeply grateful for any kindness and insight shown to them along the way. And if you are their GP, then it is nice if you can keep up the kindness for as long as it takes, which may be more than thirty years. But whether this is “effective” I have little idea, and despite the methodological pains of this meta-analysis (and I mean “pains” both ways), I still can’t tell.

NEJM 2 Oct 2008 Vol 359

1464 The continuous monitoring of blood glucose is now possible, and if it were soon to be linked with the continuous infusion of insulin, then we would have a replacement for the islet cells and a sort of “cure” for type 1 diabetes. This trial used a variety of subcutaneous monitors which had to be re-sited about weekly, and relayed the levels to clinical centres which then advised the patient about insulin dose adjustments. A pretty cumbersome feedback loop – and it didn’t achieve any improvement in HbA1c control compared with conventional home monitoring in children and adolescents. The only improvement was seen in adults aged 25 or over, and we will need a lot more evidence than this 6-month study before adopting continuous monitoring as a practical strategy.

1477 This clinical review of autosomal dominant polycystic kidney disease begins with a vignette of a healthy young man who has severe haematuria following a knock during basketball: not an untypical presentation, according to the text. About half of these patients will need renal replacement therapy later in life; and in women, extensive hepatic cystic disease can also occur, especially if we feed them exogenous oestrogens. Although this disease is written in the DNA, it is possible that we may be able to arrest its progression with drug treatment in the future: arginine-vasopressin inhibitors may at last find a use here.

1486 According to the world map in this article, Britain is a hotbed of hepatitis B infection, but because ours is mainly transmitted in adolescence and adulthood, it carries little risk of persistence; whereas if it is transmitted perinatally, as in Asia, it is immunologically tolerated and hangs about for decades, leading to fibrosis and hepatocellular carcinoma. Unfortunately the only usefulness of drug therapy – which is the subject of this review – is in youths and adults with recent acute infection. As for the virus antigen types and the actual drugs, you can leave the detail to infectious disease buffs and hepatologists.


Lancet 4 Oct 2008 Vol 372

1223 “Supplementation with N-3 polyunsaturated fatty acids should join the short list of evidence-based life-prolonging therapies for heart failure” declares the front cover of The Lancet. Well, just about: but it’s hardly front cover stuff. If one patient out of nearly 2,000 in this Italian trial had been knocked over by a Vespa or eaten a bad mushroom, the hazard ratio might have crossed the 95% confidence upper limit from 0.998 to 1.000 and the result would have been declared a draw. So “evidence-based” here means based on a borderline result in a single trial. Fifty-six heart failure patients would need to take these fatty acids for 4 years to prevent one death, perhaps.

1231 I love statins and I believe they have done more than any other drugs (alcohol always excepted) to prevent heart failure. But once your heart is packing up – whether the systolic ejection fraction is normal or impaired – statins will cease to do any further good. This finding from the GISSI-HF trial (randomising patients to rosuvastatin or placebo) may seem odd but it is consistent with the rest of the evidence we have about statins in established HF. Incidentally, the evidence about alcohol in established HF is more promising, though not definitive.

1240 It seems that every problem type 2 diabetic I send to our all-singing, all-dancing regional diabetes centre comes back on exenatide, an incretin mimetic which does seem to help many of them reduce their insulin and lose weight. Now for the really good news: exenatide works even better if it is given as a single weekly injection rather than as a twice daily shot. Overall glycaemic control is improved and there is no added risk of hypoglycaemia, and there are similar reductions in body weight. This was an open-label 30 week study (DURATION-1 from USA and Canada, sponsored by the manufacturers) which set out looking for non-inferiority and ended up finding superiority.

1251 I is important. But it is an exaggeration to say that a single atom of iodine weighs 126.9 g, as claimed in the first sentence of this review of iodine-deficiency disorders. It would be a bit unlikely for such atoms to evaporate from seawater and then be carried over land in rain clouds, there to nourish the function of our thyroid glands. We would be too busy taking shelter from falling iodine atoms and our thyroid glands would be so heavy we would be bent double. Which is what can almost happen, but due to too little rather than too much iodine, causing massive goitre. Still worse, iodine deficiency remains the commonest cause of preventable mental impairment worldwide. Iodine supplementation in areas of endemic cretinism needs to be maintained assiduously because even a few months of iodine deficiency can affect brain development. Watch out, Derbyshire.


BMJ 4 Oct 2008 Vol 337

792 One of the reasons I enjoy out-of-hours medicine (apart from the pay) is that you get to be a real doctor again. Much of the workload consists of sick children, the vast majority of whom will get better without any action on anyone’s part. Your job is to assess them in such a way that you will (a) reassure the worried parent and (b) never miss a kid who is seriously ill. And this is what you’ve been trained to do and nurses haven’t. There is a cumbersome system – the Manchester triage protocol – which allows nurses to do the sifting in hospital contexts. Not surprisingly, it errs greatly on the side of safety, so a high proportion of kids and parents are kept waiting anxiously until they can see a doctor, who by this stage may also tend to err on the side of caution. This Dutch validation study just convinces me that we need more experienced GPs working in tandem with hospital emergency services for children.

795 In fact we need more GPs triaging adults out of hours as well. Provided they realise that a transient ischaemic attack or minor stroke is not simply a diagnosis without a treatment, as it used to be for most of my working life, but a medical emergency requiring urgent investigation to prevent a major stroke. This Oxford study shows that confusion about how and when to seek help for TIA can worsen outcomes. The responsibility should lie with the Department of Health, which agreed to pass the organisation of out-of-hours cover to primary care trusts and NHS Direct. Because they wanted to encourage a plurality of providers, they sowed this confusion. It is silly to argue that an hour or two more of local GP surgery opening, or alternative provision by Darzi centres is going to make any real difference.

798 The cervix of the uterus is oddly important considering it’s usually just a small hole we look for down a speculum. You have to remember that it’s the entrance through which most of us come into the world, the only alternative being a large abdominal incision. So does removing a section of it for cervical neoplasia have any adverse effect on the outcome of subsequent pregnancies? This meta-analysis (plus another study on p.803) shows that most forms of conisation, especially cold knife, carry a risk of severe adverse pregnancy events including perinatal mortality, though this is lower for laser procedures.

Fungus of the Week: Boletus edulis

This is the edible fungus which we all shout about when we find it. It looks so good and solid: in Germany it is known as Steinpilz, and it is impossible to tell on a market stall whether these fungi are real or indeed carved out of stone. There was a stern warning of “Non Toccare!” where we saw them on sale in Italy, in case you were tempted to find out. In fact the felted cap of these mushrooms cries out to be touched and stroked. The Italian for them is porcini, meaning little pigs, to my way of thinking; though the books say that it is because pigs like eating them. The alleged English common name is Penny Bun, though everyone I know uses the French name cep.

In Italy they were 48 € per kilo for good specimens without maggots (“No vermi”), and dishes containing fresh ones tended to be expensive. Dried ones, however, are ubiquitous in pizza and pasta dishes. There must be a lot of them in the Italian woods, whereas in England I have never found a spot where they can be found reliably or in quantity. If I did, I wouldn’t tell you anyway.

If you are lucky enough to find a nice fresh young specimen with no vermi, be sure to slice it finely and eat it raw. It has a lovely nutty taste with a hint of sharpness. In Bologna, we had a salad of raw porcino slices with shavings of parmesan, fennel and black truffle with a drizzle of olive oil and a few spots of balsamic vinegar. But since fresh black truffles are harder to come by in England than in Emilia-Romagna, try interleaved slices of fresh bolete (Boletus badius works equally well) and bresaola, again with shavings of parmesan, some good oil, and the famous vinegar of Modena. And the best wine for these dishes – if you can find it - is a good earthy dry Lambrusco, which of course bears no relation to the sweet fizz sold under this name outside Italy.

JAMA 8 Oct 2008 Vol 300

1653 I don’t know how typical our local deep vein thrombosis ultrasound service is, but it often takes two working days to look for a clot in someone who has a raised D-dimer, which can mean three or more days if they turn up on a Friday. Days during which they have to get daily injections of low molecular weight heparin, whether they need them or not. According to this Italian study, two-point ultrasonography can be done by almost anyone using almost any machine after less than two hours’ training, and by examining just the femoral canal and the popliteal fossa, you can accurately rule out proximal DVT. The study randomised over two thousand patients to this simple examination or full-leg colour-coded Doppler ultrasonography and found no difference in outcomes at three months.

1665 Communication at the end of life is a subject much talked about by palliative care professionals, but what evidence is there that it actually makes a difference? As far as I know, there are bits here and there, but this pragmatic study provides quite a lot. It was necessarily observational (you weren’t going to randomise half a group of dying patients to be without any communication), and prospectively analysed the effect of end-of-life discussion on three outcomes – the mental health of the patients, how much interventional medicine they underwent near death, and how their caregivers coped with bereavement afterwards. Just over a third of this cohort of 332 dying patients reported explicit end-of-life planning discussions before the start of the study. These discussions were neutral in terms of causing depression or anxiety to patients, but significantly reduced invasive treatment before death and showed evidence of improving caregiver adjustment to bereavement. By contrast, aggressive medical treatment up to the time of death trebled the rate of major depressive illness in bereaved caregivers.

1674 On the basis of observational evidence and two randomised trials, circumcision has been advocated as a public health measure against HIV infection in Africa, where most transmission in heterosexual. But does it work for men who have sex with men – the main risk group in the Western world? A meta-analysis finds that there is no conclusive evidence either way. Perhaps, inspired by this paper, some red-neck fundamentalist preacher is even now preparing a sermon on the text: “Behold, the days come saith the LORD, that I will punish all them which are circumcised with the uncircumcised.” Jer 9:25.

NEJM 9 Oct 2008 Vol 359

1543 Until a couple of weeks ago, we were prescribing inhaled tiotropium for our patients with chronic obstructive pulmonary disease without a care in the world, and the UPLIFTing message of this trial is that it does indeed make COPD patients feel better and reduce the number of exacerbations, though not the decline in FEV1. But meanwhile, and too late to be mentioned in the paper or its editorial (p.1616), has come evidence that inhaled anticholinergics increase adverse cardiovascular events and possibly mortality. In fact the Kaplan-Meier curve for the UPLIFT trial shows a trend to lower mortality at 4 years in the group randomised to tiotropium. And since we often use drugs which double cardiovascular risk (e.g. ibuprofen) for purely symptomatic benefit, perhaps tiotropium has a future after all.

1555 Immunise the mother and protect the baby: it’s a neat strategy and it works for influenza. In fact the success of this trial of giving influenza vaccine to pregnant mums in Bangladesh is quite awesome: it reduced flu in babies by 63% and total febrile respiratory illness in the mothers and infants by a third. But the numbers were not huge and I think this study needs replicating in a developed world setting before we immunise all our expectant mums.

1565 The UK Prospective Diabetes Study is said to hold the record for the number of papers generated by a single study, though all I can glean from it is that metformin is a good drug and blood pressure should be tightly controlled, whereas blood sugar matters little if the HbA1c is less than 8. If anyone can tell me anything I may have missed, please could they let me know. Here are papers 80 and 81. The first one is about blood pressure control and presents long-term data showing that it must be kept tight if benefits are to be maintained. The other (p.1577) shows that tight glycaemic control early on continues to pay its slight dividends for a long time after randomisation ceased, though the metformin subgroup again is responsible for most of the perceived benefit. For a discussion of these papers in the context of other large interventional studies of diabetes, see the editorial on p.1618. It makes me wish I knew a one-handed diabetologist, just as Harry Truman once said that he was looking for an economic adviser with a missing hand: “All these guys can tell you”, he said, “is that one on the one hand you should do this, but on the other hand…”

Lancet 11 Oct 2008 Vol 372

1303 A couple of weeks ago we learned that giving alteplase up to four and a half hours after acute ischaemic stroke can still have discernible benefit in a few patients, though much less than if you give it earlier. This study (SITS-ISTR) just looks at the incidence of adverse events in those given alteplase between 3 and 4.5 hours after the onset of an ischaemic stroke, using the International Stroke Thrombolysis Register. This is no higher than the rate in patients given alteplase within the 3-hour window, and it’s this window we still need to go for.

1310 Plunge into the icy waters of the Castalian spring, and thus purified enter the dark stone chamber of Delphi where an old woman dressed in girl’s clothes will rise from her cauldron and shriek the answer you do not want to hear. If you dare, you can bring further offerings and return to the Pythia a second time for another oracle, but it may bring even less comfort. This is what the Athenians did before Athens was destroyed by the Persians, and what the investigators here do in relation to the use of antibiotics in spontaneous preterm labour. ORACLE I looks at outcomes at seven years for the children of women randomised to antibiotics or placebo for premature rupture of membranes. The antibiotics were small doses of co-amoxiclav and/or erythromycin. Overall, they made no difference to medical or developmental outcomes, but in children whose mothers were given erythromycin (ORACLE II) there is a significant increase in functional impairment.

1335 I have one old lady patient who adamantly refuses to have any coronary intervention even though she is incapacitated by chronic stable angina. As new drugs for angina have rolled off the production line, she has tried the lot and found that none have had added benefit, including the latest, ranolazine, which is discussed in this review. Ranolazine is thought to act on myocardial sodium channels, but these open so briefly that you have to be very clever to study them, and I can’t say that I’ve tried. It didn’t work for my patient but it did for lots of others when added to existing drugs in the CARISA and ERICA trials, so it’s worth a go, especially as recent evidence (from COURAGE) shows no advantage for invasive over medical therapy in stable angina.

BMJ 11 Oct 2008 Vol 337

852 We don’t tend to think of New Zealand as a cold country, but it can get pretty nippy in winter even in the North Island, and commensurately more so in the South, where wussy Anglicans set up Christchurch in relative warmth while hardy Scottish Presbyterians went down a few hundred miles and founded Dunedin, which is an ancient name for Edinburgh and enjoys a similar climate. Asthma is rife among Kiwi children and this study wondered if that could be improved by insulating their homes and replacing unflued gas heaters with more effective and less polluting devices. But in this randomised trial, improved home heating did not improve lung function in asthmatic kids, but it did reduce their symptoms of asthma, visits to the doctor and pharmacist, and days off school.

863 Here’s a nice clinical review of traveller’s diarrhoea, co-written by a Brit and an American. For prophylaxis the most effective drugs are bismuth subsalicylate or (best) rifamixin, a non-absorbed oral antibiotic especially effective against enterotoxigenic E coli. There is one small snag, however: neither of these drugs is to be found in the British National Formulary.

Ann Intern Med 7 Oct 2008 Vol 149

441 To prevent bowel cancer, you need to spot adenomatous polyps before they undergo malignant change. The only reliable way to do that is by universal screening colonoscopy, and this study of a new stool DNA test only serves to confirm that. Judged against the gold standard of colonoscopy, the older haemoglobin-detecting stool tests perform miserably (16% of screen-relevant neoplasms detected) while the best new stool DNA test looked at here did much better (46%), but still not nearly well enough. This may change as DNA sensing technology gets better and cheaper – the editorial on p.509 gives an excellent overview.

481 D-dimer is a fibrin degradation product that performs well for detecting recent clots, but can it help to select patients at risk of further venous thromboembolism after stopping anticoagulation for DVT? At present my own patients are given something called a “thrombophilia screen” a few weeks after stopping warfarin, despite its poor discriminatory power. This systematic review shows that D-dimer alone can stratify patients into a low risk group for recurrence (3.5% annual risk for a “negative” level) and a group at higher risk (8.9% for “positives”). What we now need is a more sophisticated mathematical model based on absolute levels of D-dimer combined with other tests such as Factor V Leiden and so forth. Like most haematology it bores me rigid, but somebody needs to do it and tell me which patients really need to stay on warfarin for life.

Fungus of the Week: Hydnum repandum

The hedgehog fungus has featured here before, because I can find it in reasonable quantities almost every year. In a good season I even once made a present of a few bags to Le Manoir les Quat’ Saisons, and was rewarded with a free lunch. It is the perfect picking mushroom: it grows slowly for weeks, in approximately the same places each year; it rarely if ever has any maggots; it has a delicious mild nutty taste; and you can pick it out even in bad light because of its creamy-buff colour. Because it has prickles instead of gills, it is impossible to confuse it with any poisonous species.

I have no idea what a restaurant with three Michelin stars might do with the hedgehog mushroom, but here is a suggestion:

Oeuf de Breakfast anglais:

Dry the mushrooms overnight on paper, chop quite small and fry briefly in a small amount of butter. Add a small amount of Devon cream (none of your French muck here, thank you very much), cook for one minute longer, season lightly with sea salt only and keep warm. Fry pieces of streaky bacon in butter until they are very crisp and crumble easily. Break them quite small and add a small amount of heated thick veal stock. Season with a little pepper. Cook a duck’s egg or a Cotswold Legbar hen’s egg en cocotte. When the egg is just cooked, add the bacon layer and then the mushroom cream. Cover lightly with foam of thyme and parmesan, made by adding a sprig of thyme and some grated parmesan cheese to milk, boiling it and skimming off a small amount of the froth. Serve with soldiers of organic white toast and a silver teaspoon.

JAMA 15 Oct 2008 Vol 300

1774 O, the deep sadness of a beautiful theory spoilt by an inconvenient fact. A few years ago, my GP partner Harold Hin was going to prevent a lot of senile dementia by proving the worth of a new assay for vitamin B12 in the community, and I was going to rationalise the management of cardiac impairment by serial measurement of B-type natriuretic peptide. Alas, neither of these ideas worked in practice. B-vitamin supplementation does not prevent cognitive decline, either in older people with normal function nor – as this study shows – in those with established Alzheimer’s disease. To be fair, the subjects selected here had normal vitamin levels, judged by conventional assays, so Harold’s hypothesis that low serum holotranscobalamin levels might identify a population that would benefit from treatment has never been properly tested. But here is one more nail in the coffin for the idea that homocysteine lowering is of any benefit to patients. The cocktail given of high-dose oral B12, folate and pyridoxine was good at lowering homocysteine but did no good to the patients: and the same goes for every trial, whether the end-points are cognitive or cardiovascular.

1784 The three things that every GP knows about amiodarone are that it is an effective antiarrhythmic drug, that many patients hate it, and that it hangs about in the body for a very long time. These facts are borne out by this study of Dutch patients with recurrent atrial fibrillation. The investigators set out to discover whether they could CONVERT patients to sinus rhythm by using episodic rather than continuous treatment with amiodarone: but the results favour continuous treatment. The investigators were surprised that over 10% of the patients had major amiodarone-related adverse effects, but I’m not.

1793 The Rational Clinical Examination series continues to be the jewel in JAMA’s crown, but snappy titles like “Is This Patient Pregnant?” (in 1997) have hypertrophied into this week’s “Will the History and Physical Examination Help Establish That Irritable Bowel Syndrome Is Causing This Patient’s Lower Gastrointestinal Tract Symptoms?” Hmm. Define “help”, “establish” and “irritable bowel syndrome”. Now discuss whether a “syndrome”, i.e. a combination of symptoms and signs, can be said to be “causing” a patient’s symptoms. For all this epistemological fog, the article itself is clear and well-written. The answer to its long and ill-formulated question is either “no”, or “it depends”: the competing definitions of IBS are of little clinical use and the lack of agreement makes the literature hard to interpret. In the end, it’s up to you whether you make the diagnosis yourself or leave it to the weary gastroenterologist. And by the way, while I am praising Harold Hin, I should point out that he wrote a nice clear letter to The Lancet about this some years ago.

NEJM 16 Oct 2008 Vol 359

1655 Religious tolerance is a basic principle of civilisation, but that doesn’t mean that one shouldn’t argue against the stupidity of many religious “beliefs” or actively prevent the evil that they bring about. There couldn’t be a clearer example than the paralysis and death of thousands of children because some imams in Nigeria decided that polio vaccination was un-Islamic and had it banned in one state. The sickening inaction of the central Nigerian government meant that just as we had nearly achieved worldwide elimination of polio, the disease spread back into 20 countries, in the form of a wild Nigerian strain. Now we are picking up the pieces, and it seems that monovalent type 1 oral polio vaccination at birth may be useful weapon, judging from this Egyptian study. In Nigeria itself there were 27,379 cases of acute flaccid paralysis between 2001 and 2007 but the position is improving – see p.1666.

1675 Muir Gray begins his essay on screening in the Oxford Textbook of Medicine by saying that “All screening does harm”; and not least of the harms of screening is the agony of boredom it inflicts on countless pathologists and radiologists. To report a screening mammogram properly takes two radiologists: one to look at the film and the other to keep her colleague awake, or something like that. Double reading is the gold standard – that’s the time of two very intelligent people, trained for 10+ years and highly paid to make poorly people better, spent looking at fuzzy boob shapes all day long. And the most optimistic interpretation of its cost-effectiveness is pretty borderline. This British study suggests that single reading with computer-aided detection can be as accurate as double reading. These computers need to come with a built-in coffee maker and an I-pod port.

1700 They say GERD, we say GORD: let’s call the endoscopy off, as both Britons and Americans can agree that most endoscopy for symptoms of gastro-oesophageal reflux is a waste of time. Though a nice little earner on both sides of the ocean, no doubt. For us as GPs this review carries the reassuring message that unless there are clear red-flag features like dysphagia or weight loss, PPI treatment is all that it needs and usually works better than surgery. Nearly half of patients who develop oesophageal cancer have no preceding history of GORD; and even Barrett’s oesophagus carries little risk.

Lancet 18 Oct 2008 Vol 372

1385 If you’ve read Ben Goldacre’s Bad Science – and everybody should – you’ll have been reminded that pharma companies that want to boost sales of their drug – usually a me-too product, like candesartan – indulge in a number of common tricks. They select a highly lucrative market – the prevention of diabetic retinopathy in this case – they pre-specify one or two surrogate end-points, match their drug with placebo rather than a credible competitor, try and influence prescribing by funding a huge number of centres to collect just a few patients (309 centres for 1905 participants here), and then hype the result by giving a relative rather than absolute figure for whatever favourable result may emerge. In this case candesartan did not prevent progression of diabetic retinopathy as expected, even though they used a hefty 32mg against placebo; but it did increase the small proportion of patients showing regression of retinopathy from 14% in the placebo group to 20% of the active group. Thus 6 patients in every hundred showed an improvement in this surrogate end-point - visual loss being the corresponding hard end-point, which I can’t see mentioned in the paper. The candy the makers of this sartan want you to eye is a 36% increase in regression: and the message from their reps will no doubt be that “candesartan can really make diabetic eye disease go away!” But this trial doesn’t prove that candesartan will ever stop a single diabetic patient going blind; a conclusion that’s strengthened by the study of eye disease prevention in type I diabetes on p.1394, where again candesartan failed to achieve even the effect you would expect from lowering blood pressure.

1411 Hereditary spherocytosis is not the sort of title that has me diving into a clinical review out of sheer excitement, but I do tend to look at the abstract of papers like this to see how common it is and whether I should actually know anything about it beyond the fact that it makes blood cells spherical and is hereditary. Argh! My practice should contain 5-6 people with this condition which is the commonest inherited anaemia in northern Europe. And I don’t know of one. Perhaps, being familial, they all happen to belong to another practice of similar size, which has 10-12. They will get gallstones, acute haemolytic anaemia, splenomegaly, and aplastic crises. Watch out.

BMJ Journals Oct 2008

BMJ 883 “Where are we in the rationing debate?” asks the first editorial in the BMJ. Well, we’re at the end of a fortnight in which the government found several times the annual NHS budget to bail out failing banks and then those banks look set to pay their senior executives £17 billion in bonuses. Where am I in the rationing debate? Well, I occupy the “extreme” position of believing that the demand for medical services is measurable and can be met by any advanced society with the will to do so. Whereas everyone tells me that demand for “health care” is infinite and can never be met without undermining the basis of our entire economic system; as bank executives have recently done. I must be missing something. I need Gordon Brown to explain it to me.

BMJ 907 Continuous glucose monitoring is coming, and we need to avoid the mess we got into when we allowed glucose monitoring companies to sell us monitors cheap and sticks expensive and induce a useless habit of testing in our type 2 diabetics. A fortnight ago we saw a trial which proved that it has small benefits for some adult type 1 diabetics but none for those under 25, and here is a trial which shows a small benefit in gestational diabetes. There was a decrease in babies born extremely large for dates, but an almost equal increase in babies born small for dates to women randomised to continuous monitoring. We need many more and much larger studies.

Gut 1341 Can there be such a thing as an evidence-based alcohol policy? If so, this editorial on the subject doesn’t provide it. It tells us that about 6,000 people may have died from alcoholic liver disease in the UK in 2005. That would seem to imply that serious alcoholic liver disease is pretty rare in comparison to the 20+% of the population whose levels of intake are supposed to be “risky”. The real questions lie elsewhere, it seems to me: in stratification of the risks of accidents and violence by age and social group; in critical analysis of cancer statistics; and in honest presentation of the cardiovascular benefits. And then in analysing what interventions actually work at a population and an individual level.

Thorax 853 We give it for clots in the heart and the brain, so why not for clots in the lung? Thrombolysis for pulmonary embolism has been around for a long time – I remember hearing about it as a medical student in 1974 and wondering why we didn’t use streptokinase for heart clots too. Now it’s the other way round. I suppose embolised clot may be organised and resistant to thrombolytic agents, or that the risks may be too great in some way: but the main message I get from this editorial is that nobody really knows. Nobody knows whether rational, easily available therapy actually works for a common life-threatening condition? It seems that way.

Arch Intern Med 13 Oct 2008 Vol 168

1968 Smoking deprives people of all the added years of life that all the social and medical advances of the last fifty years have brought. But I haven’t Finnished this sermon yet. Over 1600 middle-aged men from Helsinki were followed up for 26 years, and those who smoked not only died faster in a dose-dependent way, but didn’t enjoy their lives as much.

1975 We’re not exactly spoilt for choice in drug classes for type 2 diabetes: in fact I sometimes wonder if we should be using sulfonylureas at all, and quite a few patients can’t tolerate metformin. Bile acid sequestrants have been tried but have a bad reputation for GI side-effects, which makes this trial of colesevelam interesting. It isn’t hugely effective (HbA1c fell by 0.54 in six months) but it was very well tolerated when added to metformin. It also improves the lipid profile.

2000 What if we could really distinguish between bacterial and viral respiratory infections in real time and use antibiotics in a vaguely rational manner at last? The blood level of procalcitonin has been touted for this purpose for the last 3 years or so, and I’ve tried to look out for the studies but found them of variable quality with differing outcomes. Here from Switzerland is a nice clean positive study, though: the patient group randomised to procalcitonin-guided therapy really did get 72% fewer antibiotic prescriptions and there was no difference in symptoms of ongoing infection at 28 days. If I still felt the urge to do primary care research, I would be in there doing the next study.

2022 Instead, I’ll have a cup of coffee. I am at virtually no risk of breast cancer and even if I was, coffee would not influence it at all, as this paper shows (38,432 women followed up for 10 years). But coffee-dissers will never give up. “The possibility of increased risk in women with benign breast disease or for tumours that are oestrogen and progesterone receptor negative or larger than 2cm warrants further study.” No it doesn’t. Get a life.

Doctor of the Week: Michael O’Donnell

Mention the name “Michael O’Donnell” to people and they tend to smile. Gosh, yes, those were the days – World Medicine, those columns he used to write, I used to wet myself laughing, I heard him on the radio once, a beacon of decency at the GMC, didn’t he bring out a book etc.

Michael is still with us and he is 80 this week. He was, and is, the inspiration of every good medical writer, though few can hope to equal his wisdom and humour and mastery of language. Thank you Michael, our hero, and a Happy Birthday.

NEJM 23 Oct 2008 Vol 359

1757 “Biological” drugs which block the epidermal growth factor receptor (EGFR) are hot property these days, for the very good reason that if they could be made to work irreversibly, they might stop most advanced cancers in their tracks and perhaps even cure some. Unfortunately, as I’ve had to point out on several occasions, all they can currently do is stop a certain proportion of certain advanced cancers for a certain period of time. Because these drugs are very expensive and moderately hazardous, it helps if we can identify those most likely to respond, and this paper shows that in advanced colorectal cancer, mutations of the tumour gene K-ras are associated with a response to the EGFR inhibitor cetuximab. However, the size of the response is modest, amounting to less than five extra months of survival. But if NICE decides that at the price charged by its manufacturer, this drug is not cost-effective, the press will no doubt tell us that patients are being denied “life-saving” treatment.

1778 Over a quarter of people with chronic heart failure have atrial fibrillation, a combination that worsens their function and their prognosis. Invasive treatment of these patients is still uncommon in the UK, but in other countries the combination of AV node ablation and biventricular pacing has become popular. This study randomised 81 mostly male patients of average age around 60 with systolic HF and AF to get this procedure or pulmonary vein isolation without pacing. The latter did better at six months: but don’t go thinking that this is necessarily the answer for your 80-year old woman with swollen ankles and AF.

1786 “Biologicals” are also hot property in the search for an effective treatment for multiple sclerosis, but as in cancer, progress is slow and incremental and hindered by incomplete knowledge of mechanisms. Not to mention expense and serious adverse effects, which were notable in this phase 2 trial of alemtuzumab. This monoclonal antibody is targeted at CD52 on lymphocytes and monocytes, which makes it a very powerful suppressant of some types of immunity and a powerful trigger for some types of autoimmunity. In this single-blinded UK/Polish trial it was give once yearly for a five-day period for 3 years and it reduced clinical relapses of early relapsing/remitting MS by more than two-thirds compared with interferon beta-1a three times a week. The idea is that by going in early and going in hard with this drug, you might be able to arrest the progression of MS completely. That remains to be seen: but the other side of the coin was that one alemtuzumab recipient died of immune thrombocytopenia, one got listeria meningitis, and a quarter of the group developed autoimmune thyroid disorders.

1811 Human purine metabolism really is a bit odd: almost alone among mammals we tend to retain uric acid, and according to this review levels of uric acid in the USA have doubled since the 1920s. During that time, cardiovascular disease peaked and is now falling. Although the link between cardiovascular risk and uric acid levels has been recognised for over 100 years, nobody seems quite sure whether it is causal or incidental, and this lengthy article remains on the fence too. Certainly factoring in the serum uric acid level adds little value to any cardiovascular risk score; but on the other hand there is interesting evidence that a rise in uric acid precedes the development of hypertension and that early use of allopurinol may stop this. Uric acid may or may not do all sorts of things: and although this is a well-referenced and rigorously scientific review, it sometimes reminds me of the writings of John H Tilden, a wonderfully loopy American medical writer of the early twentieth century who divided the entire human race into the gouty and the scrofulous:

“The gouty subject may be very lean, and he may be quite stout or fat. His hair may be thin, but seldom, if ever, to be compared in thickness, softness, and beauty with that of the scrofulous subject. The gouty subject loses his hair early and becomes bald. Great beards belong to the scrofulous diathesis.

The gouty subject is inclined to be melancholy, but is often a comedian. He is bright, intellectual, witty, sharp, but in disposition more sad than otherwise.”

Impaired Health: Its Cause and Cure Vol 1, 1921.

1842 How would you like a pill that would mimic the effects of exercise? If the idea seems dubious even to an idle old hedonist like me, how threatening it must seem to all you bike-riding squash players. At the moment, however, we are talking of mice and not of men. Mice given an agonist of the peroxisome-proliferator-activated receptor d showed an increase in muscle endurance of about 70% over mice trained on a treadmill. If nothing else, expect a big market for these drugs in the sporting fraternity.

Lancet 25 Oct 2008 Vol 372

1463 This a great week for drug trials in relapsing-remitting multiple sclerosis, but this one could not be more different from the alemtuzumab trial reported in this week’s NEJM. There we looked at the effect of highly targeted monoclonal antibody on clinical as well as radiological outcomes over three years compared with interferon: here we are looking at a placebo-controlled dose-finding study with a surrogate marker only (gadolinium-enhancing lesions on brain MRI) after six months, and a drug so simple that any keen schoolboy chemist could make it at home. Although it’s given a code name – BG00012 – in this phase IIb study, it’s just fumaric acid (C4H4O4, the common food additive E297) with a couple of methyl groups added. Dimethyl fumarate at 240mg tds reduced GdE lesions by 70%. This really is exciting, and if I had MS I might well go out and buy a tub of fumaric acid capsules right away. But that would, strictly speaking, be premature. We need longer clinical trials of varying doses of different fumarates, and there was in fact a high drop-out rate in this trial presumably due to flushing and gastrointestinal side-effects. Perhaps as a compromise we should suggest eating sour-dough bread to our MS patients: its sourness and aeration are usually achieved by adding fumaric acid (in a slightly higher proportion than bicarbonate) to the dough.

1502 Here’s a very timely seminar on multiple sclerosis written by a couple of Cambridge neurologists. It’s good and comprehensive and leavened by literary quotations, mainly from W N P Barbellion, who himself suffered from “disseminated sclerosis” and describes the false hopes raised by experimental treatments almost 100 years ago. Although we still don’t know the cause, we are edging closer to an era of useful treatment – there are now five drugs of proven value, as described in the editorial on p.1447. Expect endless further randomised trials over the next 10-20 years before a standard regime emerges.

BMJ 25 Oct 2008 Vol 337

966 As we get older, we tend to get up more to pee in the night. In men, this is often attributed to benign prostatic hypertrophy, and a couple of decades ago it was usually treated by transurethral resection using a hot cutting wire. Since then, urologists have come up with a great variety of other boring techniques – I mean techniques for boring through the prostate, of course - or else they have moved towards recommending drug treatment, without any clear appraisal of the relative benefits. I like the slight note of exasperation in this systematic review, which finds itself unable to reach any firm conclusions due to the poor quality of the surgical trials. For some methods, like water thermotherapy and alcohol ablation, it finds no evidence worth looking at: for others, like microwave thermotherapy, needle ablation, and laser coagulation there are a some randomised trials but few of them meet the CONSORT criteria. In the end, all that can be said is that old-fashioned TURP seems more likely to cause bleeding but is less likely to cause retention or need redoing than most of the other kinds of boring. Boring, boring.

973 This study of people requesting assisted suicide in Oregon finds that the 83% of those who received prescriptions for lethal barbiturates were not depressed. I am not surprised that the remaining few were: I might well be myself, in the same situation. There are those who claim that palliative care and antidepressant treatment for the terminally ill are so good that the choice of assisted suicide should never be necessary. I beg to differ, and I hope that it has become an established right by the time I might need it.

976 “Can’t you get me in somewhere, doctor? I’m feeling absolutely desperate.” The short answer is always no: psychiatric beds in the UK are never there when you need them. Even when a patient gets sectioned under the Mental Health Act, it’s quite likely they’ll end up in some remote hospital, often privately run. This study shows that involuntary admissions have increased between 1996 and 2006, while NHS bed provision has decreased. The editorial on p.942 as usual suggests that the answer is better care in the community rather than more beds. It’s enough to drive you to the loony bin, if there was one.

Plant of the Week: Cercidiphyllum japonicum

Don’t try this one at home, unless you have a massive garden. It is the largest deciduous tree in its native Japan, where it can exceed 30m in height. In England it is normally about half that size and varies between a sparse elegant habit and a fuller, sprawlier form; you can see plenty of both in Westonbirt. Be sure to stand slightly upwind of one at this time of the year, sniffing the air for a magical scent of burnt sugar from the deciduating leaves. The bare tree is handsome in winter; lovely in spring when its small round leaves break out; but best of all just now when they go pink and yellow and swirl to the ground, emitting that strange and haunting sugary smell. Get your Russian billionaire friends to plant lots of them in their estates for your children to enjoy.

NEJM 30 Oct 2008 Vol 359

1873 Babies weighing under 1500g at birth would almost always die before the arrival of special care baby units. There is a limited amount of evidence to guide the treatment of these guinea-pig sized creatures, and neonatology remains partly based on what happens to adult humans and partly on observational experience. A lot of tiny babies are born hyperglycaemic, and until recently it was common practice in adult intensive care to correct any hyperglycaemia with insulin. Hyperglycaemic babies tend to do worse, so this mainly British pilot study looked at the effect of correcting this by using insulin. Alas, it achieved very little except making some of the little guinea pigs hypoglycaemic.

1885 Moving to even smaller guinea pigs (1000g or less), an American study looks at the effect of two different phototherapy regimes, one of which is called “aggressive”. Perhaps under the gentle reign of President Obama, this will cease to be a term of praise in US medicine. Smokin’ out bilirubin and blasting it away with UV light may appeal to the inner Schwarzenegger in every neonatologist, but it did little for the poor tiny babies, who fared equally well with conventional “conservative” phototherapy.

1897 Over recent years there has been a surge of interest in C-reactive protein as a “new biomarker” for ischaemic vascular disease. “New” in this case means first described more than 70 years ago, and “biomarker” can conveniently mean anything from “vaguely associated with” to “causal factor for”. CRP can of course shoot up with any kind of inflammation and go down equally fast: mine was 56 one day and less than 8 the next. But some people are born to run at high levels of CRP all the time, due to polymorphisms of the CRP gene. This complex study did a whole-population analyses and a case-control genotyping analysis and found that in people who run genetically programmed high CRP, there is no added risk of vascular disease. So CRP itself probably does not cause CVD, but is usually a marker for inflammation, which does.

1909 By the time type 1 diabetes becomes clinically manifest in young people, almost all their insulin-producing beta-calls have been laid waste by a massive auto-immune process. One major target for their auto-antibodies is the 65-kD isoform of glutamic acid decarboxylase (GAD). The cunning strategy of this study was to try and induce immune tolerance by injecting newly diagnosed diabetics aged 10 to 18 with subcutaneous injections of GAD-alum, i.e. GAD rendered more delicious to the immune system by the addition of aluminium potassium sulphate. In this small study, with a control group given injections of alum without GAD, there was a reduction in beta cell death (measured by fasting C-peptide levels) in those getting the GAD-alum, but only if they had overt diabetes for less than 6 months.

1932 When it was first described in 1935, it was called Bacillus difficilis because it was difficult to isolate from baby poo. Had it kept that name, would we have been spared the competing pronunciations of Clostridium difficile? Alors, c’est difficile à dire. De difficile non est disputandum. Difficilis erat, difficilior nunc est. If it was difficult then, it’s even more difficult now: because it’s become rife among frail elderly patients and can even kill healthy young ones by fulminating pseudomembranous colitis. This review concentrates on pathogenesis and drug treatment; but for this bug prevention clearly is a lot better than any cure we have yet found. That means more hand-washing plus improved hospital and nursing home hygiene, i.e. more staff to change bedclothes and clean rooms. And if that seems a bit too basic for the New England Journal, have a look at the Special Article on p.1921, where patient satisfaction with hospital care in the US correlates with the level of staffing – as do many harder outcomes.

Lancet 1 Nov 2008 Vol 372

Richard Horton, the Lancet editor, likes nothing better than to follow Samuel Johnson’s advice and Let Observation with extensive View / Survey Mankind from China to Peru. It seems that this is called health-research as opposed to the medical research which usually fills the pages of his august journal. This affects the lives and deaths of tens of millions people throughout the world, compared to a few individuals with, say, mutations of JAK2 in acute lymphoblastic leukaemia associated with Down’s syndrome, to take an example from last week’s issue. I’m all for this kind of mix, and the fact that I don’t comment much on global health issues is not from lack of interest but from lack of anything useful to say.

1545 The story of malaria has to be of interest to every doctor. This is a disease that Mussolini banished from Italy as recently as the 1930s and which should have been eliminated from the world in the early 1960s. News comes from The Gambia (here) and from Kenya (p.1555) that malaria in Africa may be being defeated at last. But the Kenyan paper in particular shows that changes in transmission have a complex relation to changes in clinical prevalence.

1563 Here’s your chance to play at being Bill Gates. “Ministers of health, donor agencies, philanthropists and international agencies will meet at Bamako, Mali, in November 2008, to review global priorities for health research.” First find Bamako, Mali: from Google maps, it seems to be where two roads meet near the Niger river. Take antimalarials. Then plan your agenda by reading this article, though I found it rather woolly and disappointing. Are you going to put your billions into the elimination of transmissible disease in Africa, or into the prevention of injuries, or into stemming the coming epidemic of cardiovascular and chronic disease? It’s all very worthy, and a bit too vague for my liking. How about: 1. get rid of malaria 2. ban all arms sales 3. ban all tobacco sales 4. ensure clean water for everybody.

Signed: Il Duce.

BMJ 1 Nov 2008 Vol 337

1030 This is an important trial with an acronym guaranteed to annoy everybody into remembering it. Full marks to POPADAD. That stands for the prevention of progression of arterial disease and diabetes and the importance of the trial is that it did not succeed. Do not use aspirin or anti-oxidants in asymptomatic individuals with diabetes, even if you have been clever enough to detect a lowered ankle-brachial pressure index. These drugs will not do good: they may even do harm. The abiding lesson of medical research is that things that are “obviously” good often aren’t. Before you prescribe anything to anybody, you need to find out; and then tell them the truth as it applies to them - see the excellent piece by Nicholas Christakis on p.1025.

1034 Patellofemoral pain syndrome is common, poorly understood, and eventually gets better whatever you do. For this type of condition you will often find people turning to physiotherapy or some kind of mechanical device – foot orthoses here. They worked equally well, though combining them did not improve on either alone: but how they compare with doing nothing is not a question this study addressed.

1037 Do women get abdominal aortic aneurysms? In the Women’s Health Initiative Study, the incidence was just over one per thousand in the mean follow up period of 7.8 years. Pretty rare then, but not vanishingly: women can increase their chances by smoking as much as possible, whereas the best protection seems to be to become diabetic. Strange.

1045 The preservation of fertility in adults and children diagnosed with cancer is an issue every GP encounters from time to time: when you do, remember this nice little clinical review and look it up. It summarises the problems and the options very well. Did you know that it is possible to move the ovaries surgically out of the radiation field before radiotherapy to the pelvis? It may even be possible soon to remove and preserve testicular tissue from prepubertal boys for later reimplantation – it’s been done in animals. And as you probably knew, it’s better to store embryos than unfertilised eggs.

Arch Int Med 27 Oct 2008 Vol 168

2070 This is the week to grasp the nettle: what evidence is there that tight control of blood glucose in type 2 diabetes really improves cardiovascular outcomes? There is an editorial by a diabetes professor called Home in this week’s BMJ subtitled “a strong effect that develops slowly but persists for years”; there is a systematic review of trials of oral medication (here); and a commentary on it (“How tight is right and how to get there”) by David Nathan on p.2064. If you look after diabetic patients you really ought to take an hour out and read all three rather than take my word for what they say. I’ve done a topic count of this year’s reviews and so far I have covered 42 papers on cardiovascular risk and 35 on type 2 diabetes, but I’m not much the wiser. I’ve tended to assert that (a) metformin is the only drug we know lowers CV events in diabetes and (b) rosiglitazone is harmful but the jury is still out on pioglitazone and (c) trying to lower HbA1c below 7 is unlikely to benefit most patients and may harm some. Nathan the Prophet largely agrees, while Home still clings to the hope that some evidence will come along showing that tighter control really prevents CV events (he thinks the UKPDS follow-up analysis supports this). What we need is for Trisha Greenhalgh to turn her attention back to this area - she did a splendid analysis of the original UKPDS data, pointing out that metformin alone helps CVD events, which somewhat miffed Rury Holman at the time. I don’t know why she now prefers to investigate how much working class medical students miss their mums, or the sociology of electronic record sharing at the meso-level (see this week’s BMJ, p.1040).

2088 So how does this translate into the actual prescribed treatment for type 2 diabetes in the USA between 1994 and 2007? This massive study shows some strange trends, like a dip in insulin use from 38% in 1994 to 25% in 2000 and then slightly to 28% in 2007. Sulfonylurea use has halved – from 67% in 1994 to 34% last year. Aggregate expenditure has nearly doubled from 2001, thanks to glitazones, newer insulins, sitagliptin and exenatide. It would be nice to have some UK data to compare.

2104 It’s not good news to have a myocardial infarction and in a few unlucky people this is followed by a stroke. The highest risk is in women and those with atrial fibrillation, and the odds for death have actually worsened over the last decade in this study of 9220 American patients admitted with MI, of whom 1.4% suffered strokes while in hospital.

2124 Nearly a quarter of the population has felt dizzy at some time in the last year – in Germany, at any rate. Every doctor should be able to distinguish between vestibular and non-vestibular dizziness, mainly from the history alone – in this study it was done by medical students. But in the thousand dizzy German patients interviewed here, more than half of those with vestibular dizziness had been previously given a non-vestibular diagnosis. This is clearly something we need to get a bit better at – perhaps even as good as medical students.

2125 End-of-life treatment directives tend to be made by highly educated older people, among whom doctors are over-represented. This study looked at the stability of preferences over three years in a cohort of Johns Hopkins medical graduates of mean age 69. Those who started off wanting “least aggressive” care at the end of their lives kept to that in 80% of cases, whereas those who wanted “most aggressive” care at the start had retained that preference in only 40% of cases after just three years.

2138 Good studies of the incidence of congestive heart failure are hard to come by, but this is one of them. It used a mainly clinical definition of “heart failure” and investigated both systolic and diastolic function by cardiac MRI which also allowed accurate measurement of LV mass. I won’t go into detail but it confirms earlier studies showing the highest incidence in African-Americans and the lowest in Chinese Americans.

Fungus of the Week: Cantharellus infundibuliformis

I went on a long solitary walk in the woods before the frosts arrived, and in certain places it was impossible not to tread on stands of autumn chanterelles, so thickly did they cover the forest floor. On the other hand, if I hadn’t been looking for them, I might not even have noticed, because they blend easily with the browns and blacks of forest litter. While not in the highest rank of edible forest fungi, they are very good in their place. They turn almost black when cooked, a property exploited in contrasting ways in the following dishes:

Paupiettes de saumon fumé aux chanterelles

Fry your autumn chanterelles whole in butter with a scrap of shallot and allow to cool. Mix some crème fraiche with a handful of chopped chives and dill. Place slivers of smoked salmon on a board and place a blob of the crème on each, and on top of it one or two of the cold cooked chanterelles. Fold each piece into a roll and serve with slices of lemon.

Risotto nero con funghi

Make a rich risotto with black rice, i.e. rice you buy coloured with cuttle-fish ink, stirred with chopped onions and butter, to which you add aliquots of hot fresh chicken or guinea fowl stock made with white wine. Do not add saffron. Fry a good quantity of chopped autumn chanterelles in butter with a scrap of shallot or garlic while you are stirring the risotto. Towards the end stir the mushrooms in, and serve. Sprinkle with chopped flat leaf parsley, and allow your guests a generous bowl of freshly grated parmesan to add to the generous amount you have already put into the risotto.





Page last edited: 13 March 2009