- Journal Watch
JAMA 28 Apr 2010 Vol 303
1603 A decade ago, we believed that high homocysteine levels might be the key to all sorts of vascular mischief, including coronary heart disease and diabetic nephropathy. The vitamins which lower homocysteine by methylation are folic acid, cobalamin (B12) and pyridoxine (B6), and in the early 2000s several trials were started using an oral cocktail of these (with a B12 dose of 1mg a day, meaning enough will get through even if you lack intrinsic factor). The major trials in coronary heart disease were reported over the last three years, and they flopped. And the same goes for this one which recruited diabetic patients with impaired renal function. These were nearly all men with type 2 diabetes and a properly measured GFR under 60. Those given the B-vitamins showed an increase in vascular events and a greater decrease in GFR. So much for a nice theory.
1610 President Obama drew criticism from some quarters when he disclosed that he had had a coronary artery calcium score done, despite having no risk factors. I recently described this procedure as the exchange of a lot of money for a large dose ionising radiation and a needless increase in anxiety. But its popularity shows no sign of waning - sad evidence that people follow Obama's actions more than my words. Fools. What we need is good prospective evidence of benefit to hard outcomes, which can be balanced against its unquestionable costs and harms. This study does not provide it: all that can be said in its favour is that it is prospective, and that it makes some attempt to assess the added predictive value of this score when added to existing risk scores. But its claims to added benefit are disputed in the accompanying editorial, which contains a 7-point checklist for prognostic studies, taken from Doug Altman's ten-point score. Even though it meets most of them, it still doesn't tell us what we need to know.
NEJM 29 Apr 2010 Vol 362
1561 My eye was caught by this short Perspective piece called The Value of DNKs, partly because I wanted to find out what DNK stands for. As I half-guessed, it's the same as our abbreviation DNA, and less likely to confuse geneticists. Here a young primary care intern celebrates the extra time she can spend on a patient because someone else Did Not Keep ("dinked") their appointment. She feels rehumanised, and is blissfully convinced that she did her patient good. Perhaps I will soon feel that way, when my 32 years of ten-minute appointments come to an end, and I can wallow in an archive of patients who had many hours to recount their experiences.
1563 ACCORD is the study that examined the Tighter Is Better hypothesis for type 2 diabetes derived from UKPDS, and in the process demolished it for glycaemic control below a glycated Hb of 7. In the first of two additional papers, it also abolishes it for the idea that adding fenofibrate to simvastatin has any protective effect on type 2 diabetics with high cardiovascular risk factors.
1575 Since the association between blood pressure and adverse cardiovascular events is linear for most of its course, there is a school of thought that everybody should be given all the BP-lowering medication they can tolerate. This has become widely accepted in type 2 diabetes, due once again to over-extrapolation from the UKPDS study, where reduction of BP achieved more than reduction of glycaemia. The study ACCORDingly tried to lower BP beyond the levels achieved in UKPDS, i.e. to less than 120mm Hg systolic as opposed to 140. Amazingly, they succeeded. But all to no avail: the intervention group who got down to 119 had no fewer events than the standard-therapy group who reached 133. Tighter Is Better is a dead idea in type 2 diabetes.
1586 Recurrent miscarriage is the distressing outcome in 1% of women who try to have babies, and in about half of these no cause can be found. Some of the known mechanisms of recurrent miscarriage, like the antiphospholipid syndrome, involve hypercoagulability, so some doctors keen to try anything have given such women aspirin, heparin, or both. In the wake of this placebo-controlled trial, they should stop. These agents do nothing to stop recurrent miscarriage.
1597 The role of Helicobacter pylori in peptic ulceration was one of the major medical discoveries of my lifetime, though the pathogen itself had been spotted before my father was born. But after nearly three decades, we are still not all that good at detecting the beast and eradicating it. It manages to stay alive by turning urea into ammonia, and this is the basis of the urea breath test. You might think that urease inhibitors might therefore be a good therapeutic strategy, but instead we are still dependent on clumsy combinations of antibiotics and proton pump inhibitors. Here's a review of the current state of play between H sapiens and H pylori.
1605 Hard to spot, almost always lethal - pancreatic cancer is a disease we quickly learn to hate. In a side room of the primitive ward I first entered as a medical student, an emaciated bright yellow patient was staring at the ceiling. We tiptoed past and the registrar then told us she had carcinoma of the head of the pancreas, and that if he had it he would head for a tropical island with a case of whisky. Nowadays we have biliary stents and palliative care (including palliative chemotherapy, which can be very successful), but early diagnosis and curative treatment still seem a long way off, as this review confirms.
Lancet 1 May 2010 Vol 375
1525 In a few years' time, you will be able to get your entire genome sequenced for a few hundred pounds. Will that be money well spent? I doubt it, though that's only the view of a stick-in-the-mud time-expired GP in the middle of nowhere, whereas this paper was written by world-leading experts who are taking us into a new and exciting world of personalised medicine. Thirty-one of them combined forces to examine the genome of a 40-year-old man with a family history of cardiovascular disease and sudden death. Sure enough, he has a number of interesting polymorphisms which may be associated with sudden death. Also some predisposing him to haemochromatosis, which nobody in his family is known to have. We can predict how he might respond to a range of drugs should he ever need to take them. I'd be interested to find out what effect all this information has had on him. He was counselled before the test that it might reveal information about which nothing could be done and it seems to me that this has to be true of all full genomic analysis in the foreseeable future. In the midst of life, our DNA plots many paths to our death. One of them will eventually be taken, but our ability to predict which falls a long way behind our ability to sequence DNA.
1536 The fact is that we are rubbish at giving a prognosis. Never mind all those polymorphisms that take whole teams of gene gnomes to interpret, what about good old lipids? A recent study showed that you can extract all the prognostic information that you need in clinical practice from a single random measurement of total cholesterol. But here we have a vast academic collaboration performing an analysis of 32 prospective studies of lipoprotein-associated phospholipase A and the risk of coronary disease, stroke, and mortality. This is not independent of lipids, and the associated editorial says that we need more studies (!) to establish what, if any, extra prognostic information Lp-PLA may be able to give us. No we don't.
1545 What we need is a vaccine against respiratory syncytial virus. In the heroic days of 24-hour general practice, it used to get us out of our beds in the weeks either side of Christmas, blearily looking for a house number in sleety rain in order to locate some inconsolable baby with a crackly chest and a distraught mother. This massive global systematic review and meta-analysis establishes that RSV is the leading cause of acute lower respiratory infections in the developing world and that it probably kills 200,000 children a year.
1557 Staphylococcus aureus has lived with mammals for tens of millions of years and usually asks for nothing more than a nice warm snotty place to snuggle down in. But because it occasionally attacks us we threaten its life with antibiotics and it fights back. This seminar on meticillin-resistant S aureus is very much for skim-reading only, and for poring over the global distribution map (fig 1) while snooker players on the television contemplate their next ball. We in Britain seem peculiarly favoured with different community-associated subtypes of MRSA, harbouring 7 while the rest of the world usually makes do with one. Common antibiotics like co-trimoxazole will usually suffice to treat it, and a whole new class called oxazolidinones is waiting in the wings.
1569 Any new drug for severe chronic pain is worth raising a cheer for, but ziconotide is one you are most unlikely ever to use, since it has to be given intrathecally within a tight dose range. It derives from the venom of predatory sea-snails of the genus Conus - enterprising gastropods capable of catching fish and even killing humans with their combination of a harpoon and polypeptide toxin. The latter acts like a super-morphine, so these snails closely resemble the morphine injecting machines that go around killing undesirables in Fahrenheit 451. Except that they travel at snail's pace and eat you too, should you happen to be a fish or marine worm.
BMJ 1 May 2010 Vol 340
959 If you have chronic fatigue syndrome, you want someone to make you feel better and give you more energy. Listening does not help. If someone comes round and encourages you to do more, you do a little more and then stop once they have gone away. These are the findings of the Fatigue Intervention by Nurses Evaluation (FINE) so now you know.
960 The next trial comes from the Netherlands, not to be confused with the Underworld, where PLUTO is king. These places are short of ultraviolet light, but the PLUTO investigators determined to put that right. They lent out UVB machines for home use in psoriasis, and demonstrated that patients can use these safely and effectively. But somehow I can't see this getting funding in a recession.
962 Since we started using D-dimer testing to rule out pulmonary embolism, we've had a lot of inexplicably high results from old ladies who have gone on to have normal investigations. A retrospective examination of three large studies shows that this can be avoided to a large extent by age adjustment for level of D-dimer. This applies particularly to the over 70s.
968 This review taken from the Drug and Therapeutics Bulletin is a useful reminder of medication overuse headache, which occurs on the majority of days in the month, gets worse with more analgesia, and is associated with overuse of analgesics (including triptans) over a 3-month period. The problem is to get the patient through the pain barrier by using less analgesia, for which there is no straightforward strategy - you can try tricyclics, or NSAIDs, but the evidence is patchy. It's encouraging to know that about 70% of people manage to break the cycle, though relapses are frequent, and I think most GPs know the names and even the addresses of their headache patients who never get free.
Arch Int Med 26 Apr 2010 Vol 170
693 In recent years, the long-term prescribing of benzodiazepines has become the Mark of Cain among GPs, and my name is high in the List of Shame. In the USA, Medicare has withdrawn reimbursement for benzodiazepines in nursing homes, as we all know they increase the risk of fractures in the elderly. Actually we know there is an association between BDZs and fractures in the elderly, which is not quite the same thing. As shown by this study: take away the benzodiazepines and you do not reduce fracture rates in nursing homes.
699 Ooh, look girls, it's an article about chocolate: We ask 10 leading celebrities if they think chocolate is mood food! Well, that might be Hello! magazine, and this is the Archives of Internal Medicine, asking a mostly male sample of a thousand San Diegans how much chocolate they eat and how depressed they feel; but they still couldn't resist the title Mood Food. Yes, there is an association between depression and chocolate consumption. Does this mean that hot chocolate will now be banned in nursing homes?
704 Just what is the significance of the interaction between proton pump inhibitors and clopidogrel? Talk about mixed messages: the pharmacologists tell us one thing, the observational studies another, and the meta-analysts something else again. Well, here's one more study to add to the mud: a retrospective look at all the patients in one American catchment who took PPIs with their clopidogrel, versus those who didn't. The risk of myocardial infarction was 93% higher and the rate of rehospitalisation for a coronary event was 64% higher.
Plant of the Week: Ceanothus impressus "Puget Blue"
This is the huge globular mass of dark blue that has you slowing the car as you catch sight of it against a wall of red brick or honeyed stone at this time of the year. Plant it accordingly, where it has space and shelter, and preferably where others can enjoy it. It is a Californian plant so it doesn't need much water, and a south facing position is best. Planted by a south-fencing fence, ours seems completely undamaged by the hard winter.
Bear in mind that it is impossible to control any ceanothus except by constant clipping of new growth while it is soft: cutting back to woody growth just produces unsightly brown gaps. This plant needs about three metres in every direction, which seems a lot for a small garden, but it is worth it. No blue among plants is more regal and impressive, and it is set off to wonderful effect by the small evergreen leaves, dark and crenulated. These give sombre pleasure for the rest of the year, forming an excellent background to gayer and more flamboyant subjects.
JAMA 21 Apr 2010 Vol 303
1490 I've only been to the USA twice, and on both occasions I was struck by the fact that almost all food tasted sweeter than it should be - also saltier, fattier, and overabundant. Especially in the "iconic" barbecue and hamburger outlets which kind Washington friends insisted we should visit, as they supply the White House. The sweeteners most used are sucrose (of course) and fructose in the form of corn syrup. Foods containing these additives form the staple diet of the poor, many of whom are black. People who eat them acquire a bad lipid profile, as this paper demonstrates. I guess a lot of them have bigger things to worry about, even at the White House.
1498 If you put a cochlear implant in a child with severe to profound deafness, you certainly would hope that it might improve spoken language development compared with a conventional hearing aid. And indeed it does.
1517 Just about the first thing we discovered about H1N1 influenza as it became pandemic last year was that pregnant women had a greater risk of mortality. This appears to be confirmed in this nationwide study of the pandemic in the USA, but I don't think we can place great reliance on all the data, consisting of reports to the Centers of Disease Control and Prevention between April and August 2009. Just 788 pregnant women with swine flu were reported: hardly a convincing tally for a pandemic. But I think the figures for deaths are probably more reliable, as the authors cross-checked with intensive care units and mortality registers up to December 2009. Only one pregnant woman out of 30 who died had received antiviral treatment within two days. The authors conclude that there is a good case for regarding early treatment as potentially life-saving.
1526 Only a fortnight ago I was cheering the appearance of a new contribution to JAMA's classic series, The Rational Clinical Examination, and here comes yet another - Does This Patient With Diabetes Have Large-Fiber Peripheral Neuropathy? The key to the diagnosis is testing of the reflexes, vibratory sensation, and monofilament conduction. It's no good depending on the history because as the article puts it "Large-fiber peripheral neuropathy is often heralded by the insensate foot." This reminds me of A E Housman's marvellous parody of a literally translated Greek tragedy:
CHORUS: O suitably-attired-in-leather-boots
Head of a traveller, wherefore seeking whom
Whence by what way how purposed art thou come
To this well-nightingaled vicinity?
My object in inquiring is to know.
But if you happen to be deaf and dumb
And do not understand a word I say,
The wave your hand, to signify as much.
(1893, from The Bromsgrovian: more conveniently found in The Faber Book of Parodies)
Now those who can remember a bit of schoolboy Greek should bear in mind the maxim: pous apathetikos aggelios neuropathes diabetikes. (I wanted to write angelios, but then Jeff Aronson would have called into question my knowledge of the digamma).
CHORUS: Unsuitably-attired-without -protective-boot
Foot of the sweet-urined clinic-dweller
Insensate to the proddings of my probes
I ween that thou dost herald ulcers septic.
NEJM 22 Apr 2010 Vol 362
1463 So how do we stop people getting diabetes and all its dire consequences? The answer clearly does not lie in mass drug treatment, as Novartis might have hoped when it sponsored the enormous NAVIGATOR collaborative study of two of its drugs. The first was nateglinide, a short-acting insulin secretagogue. This was given (or not given) with or without valsartan to 9306 trial participants with impaired fasting glucose and increased cardiovascular risk. Over a mean period of 5 years, it made no difference to the incidence of diabetes as currently defined, nor to the incidence of cardiovascular events.
1477 The word NAVIGATOR puts one in mind of the famous telegram that notified Oppenheimer's team at Los Alamos when Enrico Fermi's atomic pile went critical in Chicago: "The Italian Navigator has reached the New World". Poetic, but not exactly difficult for the Russians to decipher. Nothing quite so exciting about the second report from the NAVIGATOR trial, dealing with the subjects who received valsartan for five years. During that time, 33.1% of them developed diabetes, as opposed to 36.8% in the group who received placebo. A good example for teaching the difference between clinical and statistical significance - the latter was achieved, at 14% relative reduction (P<0.001). But cardiovascular outcomes were identical.
1503 The first written medical advice I ever penned was on the subject of obstetric epidural anaesthesia for a friend who was having a baby in 1976. So I came to this clinical update on the subject with slight trepidation, in case I had been too reassuring all that time ago. But in fact the evidence of the last 34 years suggests that this is a very safe procedure, with serious complications approaching one in a quarter million, and no increase in subsequent back pain. Epidurals do lengthen the duration of the second stage, however, resulting in more instrumental deliveries but not, surprisingly, more caesarian sections.
1511 For the twelve years I've been writing comments on the medical journals, one of the things I've been waiting for most is the announcement of an effective vaccine against all types of Neisseria meningitidis. Rumours of a breakthrough have abounded throughout that time, and we can certainly celebrate success against type C, but for the B group we haven't yet outwitted the bacterium's ability to evade complement. This shortish review is a good summary of where we are at in 2010.
Lancet 24 Apr 2010 Vol 375
1447 What we most need in the treatment of type 2 diabetes are more drugs like metformin which promote insulin sensitivity and weight loss. That's why so much interest currently focuses on modulators of the incretin pathways. I won't attempt a brief description of the pharmacology of liraglutide and sitagliptin, the two drugs pitted against each other in this 26-week trial. Their differing modes of action are described in a very abbreviated way at the beginning of this paper. It was perfectly clear before this trial that injected liraglutide has a more profound action on the two most important incretin hormones - glucagon-like peptide (GLP-1) and glucose-dependent insulinotropic polypeptide (GIP) - than sitagliptin, an orally available enzyme inhibitor which raises levels of these hormones to a lesser degree. The point of the trial was to provide an easy victory for Novo Nordisk, the manufacturers of liraglutide who funded it and did much of the writing up. Yes, liraglutide lowers glycated haemoglobin more than sitagliptin. How this will influence meaningful end-points in the long term is anybody's guess. But papers of this sort seem to find a natural home in The Lancet.
1475 I wonder what figure you would arrive at if you totted up the cost of the diabetes trials reported on this week, all paid for by the drugs and diagnostic agents you prescribe for your patients, after the company and the shareholders have taken their lot. The diabetes market will grow and grow, as older people throughout the world eat more and do less. The malaria market, by contrast, would disappear completely if only the world could get its act together. The sponsors of this trial, Shin Poong Pharmaceuticals and the Medicines for Malaria Venture, would then have to find other things to do. As it is, you can cheer yourself up a bit by learning of the success of the latest artemisin combination (pyronaridine-artesunate) for treating acute falciparum malaria in adults and children. And if you're in the sort of place where you need to know about malaria in children, a comprehensive review follows on p.1468.
BMJ 24 Apr 2010 Vol 340
903 Our surgery premises are littered with various reminders about Chlamydia trachomatis: males using our excellent lavatories cannot urinate without their eye falling on various invitation cards suggesting screening. Females, sitting and facing opposite, are less likely to see these, and more likely to aim accurately. When shy teenage girls visit us we are supposed to raise the topic; ditto with the tongue-tied young men who occasionally come to us. This is "opportunistic screening" and it is a joke. Systematic chlamydia screening may be more serious, but it is scarcely more effective. The trial of annual screening reported here found that most episodes of pelvic inflammatory disease occurred in women who were negative for chlamydia at baseline; and there were too few cases of PID for the study to reach significance. A cautious editorial (p.875) dwells on the problems of gathering knowledge in this area, and a useful Clinical Review (p.913) covers the whole area of screening and treatment for chlamydia. Perhaps the next trial should involve sending two doses of azithromycin to every female aged 16-19 and every male aged 20-24, to be used after sex with each new partner, with further doses available easily and anonymously.
904 More about well-intentioned but ineffective interventions in primary care. After all, they're how we pass most of our time and earn our QOF money. I think the government should now demand that we do active outreach for older adults at risk of functional decline. It fulfils all the criteria for inclusion in QOF: it sounds like a good idea, it will be popular politically, it will punish us and distract us from acute care, and this Canadian trial shows it is a complete waste of time.
905 There is a small group of women who never attend for cervical cytology, and they develop half of all cervical carcinomas, even in the Netherlands, where people are generally more obedient. Not to be outwitted, these Dutch researchers sent them self-administered testing kits to detect human papillomavirus. The pick-up rate was as good as with physician administered testing, and this may be an effective new way of preventing cancer in this high-risk group.
Ann Intern Med 20 Apr 2010 Vol 152
481 The evidence around population salt intake and the burden of cardiovascular disease is much debated, especially in the USA, where the UK is seen as a successful model for sodium intake reduction - nearly 10% since 2003, according to this article. It presents a cost-effectiveness analysis that concludes : "Strategies to reduce sodium intake in the United States are likely to substantially reduce stroke and MI incidence, which would save billions of dollars in medical expenses." I see, these people would save money by living longer and dying from less expensive diseases? A pinch of salt is called for, I think.
505 Prevention is by no means always better than a cure, but it certainly is in the case of lung cancer. Do not smoke. If you smoke, stop. To which we now need to add: do not add to your risk by exposing your chest to large doses of ionising radiation every year or two. This paper reports on a feasibility study for the ongoing National Lung Screening Trial. This used "low dose" CT: only enough to produce a 1.5 in a thousand risk of cancer in women after 3 examinations. The cumulative false positive rate was 33% after two examinations. If I was them (sorry, "were they"), I'd call off the trial.
Plant of the Week: Aesculus x neglecta "Erythroblastos"
This bizarrely named horse-chestnut sounds suitable only for the gardens of eccentric haematologists - who, by the way, often make excellent gardeners. The erythroblasts in the case of this plant are its leaves, which open like a Homeric rose-fingered dawn.
All horse chestnuts are good and the great red-flowered ones are amongst the finest of all ornamental plants in temperate climes. But they are massive. Those of us with tiny gardens must make do with the few small, slow-growing ones, of which this is the finest. One I grew in a previous garden is now about 4 m tall and beautiful throughout the season - first with these pink fingers of leaf, turning shrimp and then yellow, and then with spikes of pale yellow flower. But is no good getting one now unless you can wait 15 years, and that's a bit chancy at my age.
Page last edited: 10 May 2010