- Journal Watch
Journal Watch - April 2011
JAMA 27 Apr 2011 Vol 305
1625 Obedience is no longer a fashionable concept, though it was once prized as the most essential virtue in religion and society (see Psalm 119 vv1-176, Dante's De Monarchia, Hobbes' Leviathan; look up the meaning of the Arabic word islam). Nowadays we go to endless lengths to persuade ourselves that everyone is making a free and informed choice about everything from the Prime Minister (though not of course the British Head of State, whose future identity will depend on the marital joys of the two dear young people being joined in wedlock as I write) to the pills we give our children. If these children have newly diagnosed epilepsy, we explain to the parents which drugs are available, which we think are most likely to prevent further fits, what adverse effects they might have, and then expect them to agree with what we prescribe. Actually, as mere GPs we do not do this ourselves: we leave that to paediatric neurologists and we then obey them, and expect parents to do likewise. This process is known as "achieving concordance". Concordance is then supposed to achieve compliance, or adherence as we must now call it. But this carefully conducted US study finds that in real life, not everyone concords, complies, adheres or obeys. Judged by an electronic chip that monitors medication-giving, just 42% are perfect citizens of the Dantean/Hobbesian/Burkean monarchy and obey their paediatric neurologist to the letter. Another 26% have a reasonable stab but often forget a dose or two. About 13% start off not trying and never obey, and 7% start by trying and then give up. I am afraid that these last people come from the lower socioeconomic strata, as Edmund Burke would have foreseen. Bring back the beadle.
1677 Obedience features in evidence-based medicine too, and is measured in outcomes research. Here is a very good teaching example from Sweden, based on analysis of their whole-population database of ST elevation myocardial infarction between 1996 and 2007. This was a time of revolution in the management of STEMI as immediate percutaneous intervention supplanted thrombolysis as the treatment of first choice. Sweden's success in providing this for 61% of its population by 2007 is impressive given the spread of its population, its climate and its geography. Door-to-balloon time increased initially but then went down again. The use of secondary preventive drugs went up steadily. Mortality declined, both in the short and the long term. But readers of Wennberg will not be surprised to learn that the one thing that did not change was the disparity between performance in individual hospitals. It seems that the hardest thing to change in medicine is the culture of individual institutions.
1695 For contrast, here's an example of the kind of "outcomes research" that sets my teeth on edge. Take two poorly defined concepts - heart failure and health literacy - match the two, and follow the patients up by mail for 1.2 years, response rate 72%. "Health illiteracy" is defined by 3 questions and is present in 17.5% of respondents. They die the fastest. So do we try to improve health literacy? Use the questions to target community nurse input? Make better use of real prognostic markers such as BNP? Or just forget this whole health literacy concept as simply the latest way of bundling together old things such as social class, poverty, intelligence, non-adherence and the rest.
NEJM 28 Apr 2011 Vol 364
1607 One of the delusions of general practice is that you acquire wisdom from experience. The trouble is that your sampling errors are enormous, and so is the cognitive bias produced by your wish to believe that you have snatched people from the jaws of death. For example, I looked after two youngish diabetic patients with heart failure, both of whom were going through episodes of decompensation and appeared likely to die within months. Enthused by the latest reports, I insisted that they both have coronary angiography, and this was duly followed by coronary artery bypass grafting to reperfuse hibernating myocardium. Many years later, they remain in full time employment, proof of the life-saving potential of CABG for hibernating myocardium, not to mention its economic benefits. But these complacent presumptions may well be wrong, according to the STICH trial, in which patients with systolic HF and coronary heart disease were randomized to receive either CABG or maximum medical therapy. Overall mortality was the same in both groups, though the CABG group had somewhat fewer cardiovascular events.
1617 So what about the whole concept of "hibernating myocardium" and sophisticated tests for myocardial viability in ischaemic left ventricular dysfunction? A subset of the patients in the STICH trial were put through single-photon-emission computed tomography (SPECT), dobutamine echocardiography, or both to assess myocardial viability on the basis of prespecified thresholds. The presence of viable ("hibernating") myocardium was associated with better overall survival, but the difference lost significance after adjustment for other baseline variables.
1626 As a medical student in Oxford I used to spend a lot of my time hitch-hiking to London, and on one occasion I was given a lift by a distinguished professor of bacteriology whose special interest was Mycobacterium leprae. We had an interesting conversation, though there is a limit to the comfort one can give to a man whose only laboratory subject is the armadillo. Some remote recess of my brain must have stored this encounter for the last 40 years, because when I saw the title of this paper, Probable Zoonotic Leprosy in the Southern United States, the image of an armadillo immediately filled my sight. And yes, the investigators found that the genotypes of M leprae bacteria isolated from 50 US leprosy patients and those found in the local wild armadillo population were an exact match. Blame this remarkable mammal, Dasypus novemcinctus, for all non-imported leprosy in the USA, and look out for more cases as it spreads slowly northwards, unopposed by any natural predator.
1634 Now the armadillo is most famous among mammals for its degree of vertebral protection, while the human female stands at the other extreme. Her suffering when vertebral collapse occurs is often extreme, and these events are common, in men as well as women. Some episodes require hospital admission, and most require strong opioid analgesia for several weeks. Early reports of benefit from vertebroplasty or kyphoplasty have not been borne out by some well-designed sham-controlled trials. This article is a good summary of current practice but it does not raise my spirits at the thought of next visiting an elderly patient with agonizing back pain of rapid onset.
Lancet 30 Apr 2011 Vol 377
1495 The countries where most new tuberculosis infections arise tend to be those with the poorest diagnostic facilities - not just for confirming active TB but also for tracking drug resistance. The Xpert MTB/RIF test is much simpler and faster than culture and conventional drug sensitivity testing, and it performed very well in this admirably thorough study in several resource-poor countries. But the editorial by Katharine Kranzer points out the problems: a diagnostic sensitivity of 90% sounds good but would miss a million cases per year of a life-threatening curable infection; and the lab time and facilities required for the test still far exceed the means of most countries with high TB incidence.
1506 I've several times pointed out the tendency of high-impact journals to publish industry-sponsored trials in order to get large sums from reprints, while at the same time publishing damning editorials about the same studies, in order to satisfy their intellectual conscience. As usual, I have no idea whether the sponsors of this trial, Alkermes, intend to buy up lots of reprints from The Lancet, although they acknowledge that they paid someone to help write it up. And the trial has already resulted in FDA approval of their product Vivitrex, a once-monthly injection of extended-release naltrexone for the management of opioid dependence. Reading the abstract of their paper, you can see why: their six-month placebo-controlled trial in Russia produced opioid discontinuation rates an order of magnitude better than for most other interventions in this addiction. But then read the editorial: it is entirely damning about the ethics of the trial design and warns that similar interventions have in the past resulted in high rates of death due to unintentional opioid overdosage. The authors here are saying that this trial should never have been conducted and that FDA approval should not have been given. A great - and greatly depressing - teaching example of the ways of medical journals and much else besides.
BMJ 30 Apr 2011 Vol 342
960 Drospirenone is not a name familiar to all but the most dweebish prescriber. We prescribe it - if at all - under its trade name Yasmin, in combination with ethinylestradiol, as a contraceptive pill for those who get acne or breakthrough bleeding with other pills. But Bayer got into trouble with the FDA for promoting this combination in the USA, because it had failed to carry out sufficient research on its risks compared with levonorgestrel-containing contraceptive pills. The chief allegation is an increased risk of non-fatal venous thromboembolism, seemingly borne out by a case-control study from the Boston Collaborative Surveillance Program here.
961 And here is somemore evidence of from the UK General Practice Database to the same effect. The increase in VTE is 2-3 fold in the two studies, but these are promptly questioned by the editorial on p.933. Who to believe? Whom to believe? Don't waste time on this: there are plenty of other oral contraceptives.
962 Here is a systematic review indicating that calcium supplements with or without vitamin D produce a modest increase in the risk of cardiovascular events. Hang on, I hear you say, are you reviewing the right issue of the BMJ? We read about this before Christmas. And so you did, dear reader: this is just a confirmatory analysis by the same people, adjusting for factors that were brought up in the discussion which followed their original paper. The waters are murky. Who to believe? Whom to believe? I cannot tell you, dear reader: I take my vitamin D with cheese rather than chalk.
968 The standard of Clinical Reviews in the BMJ is usually as high as that in the leading US journals, and this one about the management of paracetamol poisoning is no exception. Except that in the USA it would be the management of acetaminophen poisoning. The antidote, acetylcysteine, is invariably effective if given in time. Other antidotes may be on the way. But people will still die from paracetamol because its therapeutic range is quite narrow and toxicity is affected by hepatic enzyme induction and small degrees of undernutrition which deplete liver stores of glutathione.
Arch Intern Med 25 Apr 2011 Vol 171
744 Quite soon I shall be off to the USA for several months, hoping, amongst other things, to preach the joys of medical generalism. The words one must always use are "internal medicine", though this is but one part of what we do in British general practice. I don't fully understand what Americans mean by "primary care" and how internal medicine fits in with that. So much to learn. But while US "primary care" languishes unloved, despite the passionate embraces of Don Berwick, internal medicine is enjoying a small revival amongst medical students in the USA, according to this study of career choices between 1990 and 2007. So much to do.
750 The byways of heart failure research are full of strange sights and sounds. The fattest people survive the longest, perhaps because they have to do more exercise just getting about: for those who could not even do that, a team in Hull devised electric shorts to stimulate the thigh muscles. They did not work. Here the trick is Tai Chi, about which I know nothing but which I presume is some sort of exercise programme with added mumbo-jumbo. The comparator was not ordinary exercise but sitting down being educated for the same length of time. Well, would you believe it: the exercised group showed better exercise tolerance and quality of life. All due to a better balance of Yin and Yang and other energies.
Plant of the Week: Laburnum x watereri "Vossii"
The reason the laburnum is squeezed in here is that its alkaloids were the precursors of varenicline, and a trial here shows that this nicotine-receptor competitor combats nicotine addiction best when started 4 weeks before smoking cessation. As for this ubiquitous garden hybrid, I enjoy seeing it in other people's gardens with two-foot trusses of yellow scented flowers covering the entire plant: especially good in massed plantings or as laburnum tunnels. We don't have room.
JAMA 20 Apr 2011 Vol 305
1545 Chronic kidney disease, pre-diabetes, subclinical hypothyroidism, vitamin D insufficiency, attention deficit disorder, asymptomatic systolic dysfunction, borderline personality disorder, early chronic obstructive pulmonary disease, pre-hypertension, Barrett's oesophagus: you may think you're healthy, but how do you know you haven't got any or all of these? Wake up! Have the tests and get a depressing medical label! Stop enjoying the spring sunshine and start thinking of death. And don't forget to take the tablets and see your doctor regularly. A few years ago pathology labs in the UK started estimating the glomerular filtration rate from every sample sent to measure serum creatinine. Overnight , a twelfth of the British public acquired chronic kidney disease; GPs chasing QOF points were then supposed to convey this news to them and get a urine sample to measure the urinary albumin/creatinine ratio. In the USA, referrals to renal physicians went up fourfold. Now readers of this column will be aware that there is a much more accurate blood test for estimating renal function and GFR: cystatin C. This does not vary with how much meat you had last night and how much fluid you've been taking. And this study shows that cystatin C based GFR is far more predictive of end-stage renal failure and death than creatinine-based eGFR. The study after this, on p.1553, shows that you improve on creatinine by factoring in serum phosphate, calcium, albumin and bicarbonate; but cystatin C is still better overall. So what should we doing for patients whose creatinine-based eGFR comes back under 60? You could argue that you should measure their urinary albumin and try and get a cystatin C assay (good luck). Or you could argue that you should keep quiet and save the CKD label for people with eGFRs under 30. The facing-both-ways editorial on these studies (p.1593) begins by stating that "Effective treatments for chronic kidney disease are available but underused", but ends by saying that we need "studies that demonstrate that using better risk prediction tools will lead to clinically meaningful benefit for patients." Boy, do we need these studies: why not suspend the whole of QOF until someone has actually done them?
1560 I'm in the middle of 25 hours of out-of-hours primary care work and some time soon I am bound to hear the dreaded words, "They shouldn't have let her out of hospital so soon, doctor." Early discharge sometimes goes wrong, and that's when we hear about it. But a lot has gone right in the last 20 years as well: this huge study of Medicare episodes of total hip arthroplasty shows that since 1991, mean hospital stay for THR has fallen from 9.1 to 3.7 days; perioperative mortality has been halved despite a small increase in the mean age of patients; many more patients are receiving rehabilitation in the community; and the only downside is an increase in readmission rates from 5.9% to 8.5%. Tough on those affected, but good news for most patients.
NEJM 21 Apr 2011 Vol 364
1493 Relieving pressure within the cranium is one of the most ancient goals of medicine - or at any rate one of the easiest to spot, since skulls with holes drilled in them count as neat archaeological evidence. In most cases we no idea what the trephination (or trepanning) was supposed to cure: demonic possession, epilepsy, headache, depression perhaps. Diffuse traumatic brain injury causes raised intracranial pressure, neurological damage and death: merely drilling a hole won't relieve that, so a group of Antipodean trauma centres have been running a randomised trial of decompressive craniectomy since 2002. A desperate remedy for a desperate situation: but although the procedure releases pressure, it doesn't help patients. In fact, they do worse.
1523 In England, we look down on ticks. Foreigners may get diseases like typhus, Rocky Mountain fever, Rift Valley fever and so forth but all we allow ourselves is the odd dubious case of Lyme disease. However, the latest tick-borne disease comes from temperate north-eastern China and the tick that probably carries it is pretty ubiquitous among domestic animals throughout the world. So we may need to look out, though the novel bunyavirus that the ticks carry is still rare, even in China. The emergence in 2009 of a mysterious haemorrhagic fever in several rural hospitals in Hubei and Henan, with a mortality of 30%, led to an impressive detective exercise, funded by the China Mega-Project for Infectious Diseases. I hanker for the days when it might have been called the Glorious People's Battle Cadre against Disease-Foes of the Peasants and Workers. Initially they suspected infection with Anaplasma phagocytophilum - well you would, wouldn't you - but using an array of sophisticated techniques they found the RNA of the culprit virus. It is named for the disease it causes - SFTS - severe fever with thrombocytopenia syndrome. In their discussion the 46 authors (one deceased) modestly confess that they have not fully satisfied Koch's postulates, but this paper is nonetheless a classic of the infectious diseases literature, free on the NEJM website.
1533 Amongst all the wonderful things you learnt about in school biology, few were more amazing than cilia. So tiny, so tireless; and in fact quite ubiquitous, since the centrosome that forms cilia also forms the spindle in mitosis, and ciliated organisms appear very early in evolution, so that the ciliary mechanisms of a green alga and the human bronchus are remarkably similar. And this means that when cilia go wrong, the results can be drastic. In this earnest and recondite review of the ciliopathies, you will learn about a lot of rare genetic disorders affecting the primary cilia. Bad luck, though, if your biology lessons had led you to think only about motile cilia: they lie "beyond the scope of this review". Well, I'm very disappointed: in my opinion, nothing could be sillier.
Lancet 23 Apr 2011 Vol 377
1409 "There are few subjects that polarise cardiologists like vascular access for coronary angioplasty does." Such is the first sentence of a generally very good editorial by two authors with Italian names, discussing a study of radial versus femoral access for coronary angiography. Now a gentle copy editor might have stepped in and improved the English here, but Richard Horton was no doubt too busy writing "Offline", in which we find such gems as " Although she was extremely praiseworthy of the support given to her work by the Wellcome Trust, she also pointed out..." Praiseworthy! Come here Horton! You will rewrite this thing in decent English five times and see me after Assembly tomorrow. Now, class: the randomised RIVAL study was conducted in 158 hospitals in 32 countries, so it is probably as representative as we are going to get. It found that in procedures carried out for acute coronary syndromes, there were generally fewer local vascular complications using radial artery access, but the variable anatomy of the upper limb arteries meant that conversion to femoral access was occasionally needed. Horton, what did I just say? Are you paying attention? I shall set you a little test, you know.
BMJ 23 Apr 2011 Vol 342
906 I was introduced to the pleasures of sailing 21 years ago, on a cold and misty San Francisco bay where I fell to talking to the boat owner, a silicon valley millionaire I had never met before. He told me he had just invented a miniature gizmo for continuously measuring blood glucose, and I suggested that this would need to be linked to a continuous insulin delivery system to create an artificial pancreas by real-time feedback. He seemed a bit crestfallen: talk moved on to spinnakers and forthcoming races; it was too cold to fall asleep and this big piece of wood kept threatening to knock one into the sea. I never saw the millionaire again, and have avoided sailing ever since; but the artificial pancreas has finally arrived, and naturally the main worry is that it will cause hypoglycaemia and the main hope is that it will achieve near-normal glucose control. Here the feedback system is tested in a nocturnal cross-over trial with a conventional continuous insulin pump system. The results favour the closed loop (artificial pancreas) system, after meals with or without alcohol. But long-term outcomes need to be watched closely before these devices go to the market.
908 I've already commented once on the disparity between the summary and the conclusions of this 20-year follow-up study of PSA prostate cancer screening in the entire male population aged between 50-69 in the city of NorrkÃƒÆ’Ã†'Ãƒ" '¶ring in Sweden. The results are more consistent with an increase in prostate-related mortality than a decrease, but you can argue either way, and people do in the Rapid Responses. You may just be thoroughly bored with this issue and want to move on: or you may wish to spend an hour or two discussing the paper and responses with students and colleagues, in which case you also need to buy Overdiagnosis by Gilbert Welch with Lisa Schwartz and Steve Woloshin (2011) - a great read, with Ch 4 all about PSA.
Ann Intern Med 19 Apr 2011 Vol 154
541 In this week's JAMA, we saw an editorial alleging that the early detection of "chronic kidney disease" and proteinuria offers the prospect of "effective treatment". Here's a lengthy systematic review going over the evidence that relates to the effect of tighter blood pressure control in people with CKD and various levels of albuminuria. The three best trials recruited subjects with stage 3-4 renal disease and compared BP control aimed at 125-130/75-80 as opposed to 140/90. The trials reached their targets for BP but the results are inconclusive. There are too few hard outcomes such as death or end-stage renal failure. There are no diabetic patients in these trials. In a word, we don't know.
554 Ever since Fiona Godlee encouraged me to put my head above the parapet two years ago, and ask in a BMJ editorial why diabetologists were ignoring the evidence against tight glycaemic control in longstanding type 2 diabetes, I have been hoping that somebody somewhere might be hiding away a good summary of the trial data that would allow us to share informed decision-making with individual diabetic patients. It's been a wonderful quest in many ways, though the dispiriting fact is that much of the evidence we need is simply not there. However, the search has put me in contact with a number of personable and brilliant people who have been looking for and collating the existing data, and my heart leapt when I read the title of this piece, Individualizing Glycemic Targets in Type 2 Diabetes Mellitus: Implications of Recent Clinical Trials. But actually it's not so much individualizing sugar or HbA1c targets that we need as individualizing preferences for outcomes and the whole package of lifelong management. In corresponding with a number of friends about this paper, I have discovered two resources which are helpful towards this goal: from David Aron (Veterans Administration) I learn of http://www.healthquality.va.gov/Diabetes_Mellitus.asp
and from Victor Montori (Mayo Clinic) I learn of:
Both are more useful than the NICE guidelines, and make the one-size-fits-all approach of QOF look downright silly.
Plant of the Week: Syringa vulgaris
It is lilac time. It is also wisteria time, peony time, cherry-blossom time, apple-blossom time, iris time even. The early magnolias are flowering with the late; gardens are carpeted with flowers; every tree is covered with shimmering new foliage, the air is full of scent and birdsong: there has never been an English April like it.
But among all these delights, we're finding that giving us the most pleasure is our sprawling common lilac tree. Its flower heads are of the deeply unfashionable colour favoured for the felt hats of maiden aunts - lilac. The scent is such as came from the Christmas presents of the 1950s - bath salts and cold cream. But at the same time, lilac carries springtime and freshness. And in the evening, the scent turns to languid summer in the air: young men in Chekhov stories feel urges that they later regret.
Nor do lilac bushes need to be scrawny messes for the rest of the year. The least you can do is keep them tidy and run a late flowering clematis into them for summer interest. By good fortune ours grows on a shady bank, with its top in the sunshine. Here it has formed mossy, corrugated trunks with Japanese curves. All year round, it looks venerable and beguiling.
JAMA 13 Apr 2011 Vol 305
1441 The old JAMAs were comfortable: they looked good and I looked forward to getting into them - there was a nice feeling, like fluffy cotton against the skin. But sadly the old garment is getting threadbare: I report on this week's issue more out of a sense of duty than because it's comfortable bedwear for the generalist. Here is a perfectly decent and hardworking study of viral shedding in symptomatic and asymptomatic genital herpes infection (HSV-2). It tells us that in symptomatic infections, where there are visible, open lesions, more virus is shed than when the lesions are occult. Fair enough: but it's like opening Meteorology and being told that rainfall tends to be higher when it is raining. I read elsewhere that JAMA is getting a new editor, and his mug shot makes him look young and fierce. Perhaps that is what this dear old journal really needs.
1460 The next study is an Ouroboros concerning Adalimumab. Adalimumab, you will recall, is not the name of a once-celebrated Assyrian horse-trainer but of a fully human monoclonal antibody against tumour necrosis factor-ÃƒÆ’Ã†'Ãƒ" 'ÃƒÆ’" 'ÃƒÆ’Ã†'" 'ÃƒÆ’Ã†'Ãƒ" '" 'ÃƒÆ’Ã†'Ãƒ" 'ÃƒÆ’" '"¦½±. Being "fully human", it is supposed to float unnoticed among all the other human antibodies in the bloodstream, attacking TNF-ÃƒÆ’Ã†'Ãƒ" 'ÃƒÆ’" 'ÃƒÆ’Ã†'" 'ÃƒÆ’Ã†'Ãƒ" '" 'ÃƒÆ’Ã†'Ãƒ" 'ÃƒÆ’" '"¦½± at will and thus acting as a magic bullet against the inflammatory process in Crohn's disease or rheumatoid arthritis. Ah, but here comes the ouroboros. This name is given to the ancient symbol of a snake eating its own tail. The undeceived fully human immune system begins to recognise adalimumab as "non-self" and makes its own antibodies to adalimumab. These neutralise the effect of the artificial antibody and reduce the clinical response in RA. And how do we detect these anti-adalimumab antibodies? Why, by creating labelled anti-anti-adalimumab antibodies. Quite an ouroboros.
1484 Some of the JAMA Commentary pieces can be excellent, so I came with high hopes to this one about Epigenetics - the Link between Infectious Diseases and Cancer. But it comes from Princeton, a university without a medical school, which should have raised suspicion from the start. This is a spectacularly important topic, but it's rendered quite opaque by this complex account of some of the known basic mechanisms. I suspect we may have to wait another ten years for some clarity about their significance in human disease.
NEJM 14 Apr 2011 Vol 364
1385 The smart, ultra-conservative garb of the New England Journal covers some pretty lively writing these days, especially in the opening Perspective pieces. Vitamin D and Prevention of Cancer - Ready for Prime Time? is a good example. About four years ago, the journal published an enthusiastic review about vitamin D which had me taking a regular dose myself: it was talked up as the "what's-not-to-like" vitamin that everybody in the temperate world is short of. But that depends on how you define "short of" and what you are using it for. For bony health, a serum level of 50 nmol/L is plenty. The magic cancer-reducing properties of vitamin D at higher doses are purely an extrapolation from observational studies, plus three fairly short randomised trials with different primary end-points and very mixed results. It is quite possible that high vitamin D levels may increase risk for pancreatic and oesophageal cancers.
1395 James Le Fanu's Rise and Fall of Modern Medicine (1999) contains a gripping chapter on Open-Heart Surgery as one of the 12 Definitive Moments in the Rise section. In the good old days, patients with valvular disease were two a penny: blood fountained over surgeons and assistants: leading surgical centres ran out of coffins (only slight poetical licence here). And then a few patients began to survive open mitral valve replacement; then more; and finally, the majority. In le Fanu's somewhat gloomy perspective, we are now in the Fall: patients no longer have mitral incompetence much, open heart surgery has become routine, and the inventiveness of medical science produces ever smaller dividends at ever greater expense. However, if I had a leaky mitral valve, I might be rather pleased that percutaneous repair is now an option. True, 20% of patients undergoing the procedure end up requiring the open operation, but that's 80% who don't. For how it's done, you'll have to read the paper and/or watch the animation. It will be interesting to follow the progress of this innovation and how it changes practice: it may even find a place in some people with advanced heart failure and so-called "functional" mitral regurgitation. The editorial makes some sage observations on how to structure shared decision-making about this new option.
1407 The intensive care unit lies at one extreme of the human environment: you are in there because otherwise you would be dead. Everything must be clean, monitored and designed to prevent the transmission of microbes because this is the ultimate breeding ground for antibiotic resistance. Contrast the caves and huts which our ancestors often shared with domestic animals. Cleanliness was rather late to arrive in human civilisation, and is often over-rated: but not so in the prevention of hospital transmission of MRSA, as this and the subsequent study demonstrate (see p.1419).
1441 Amongst all the microbes which we, and other mammals, and even reptiles evolved alongside, the mycobacteria have a special reputation for indolence and unpredictable malice. Unfortunately it is still impossible to tell a sleeping mycobacterium from one which is about to wake up, reproduce and damage or kill its host. However, it is possible to determine which hosts it prefers to damage: this account of latent tuberculosis in the United States gives you a good table and guide to the appropriate tests, not only for the USA but also for Canada and the UK.
Lancet 16 Apr 2011 Vol 377
1319 This week's Lancet is devoted to stillbirth, which makes the heart sink for two reasons - the nature of the subject itself, and the way The Lancet is bound to treat it. In an opening editorial Richard Horton tells us that stillbirths are distressing and important.; moreover he commissions a whole paper (p.1353) to explain why stillbirths are distressing and important. As stillbirths are important (and distressing) it is important to count them properly, and distressing to see that this is not being done consistently across the globe. Well, maybe: a huge global preponderance of stillbirths occur because women do not have access to adequate pre-partum and intrapartum care. Comparative figures are only useful inasmuch as they highlight the improvements that need to be made. But at least The Lancet considers global counting and comparison as one of its chief duties, and does thoroughly and takes it seriously.
1331 Which brings us closer to home: why do stillbirth rates vary in high income countries? What are the avoidable factors? Although this systematic review is very comprehensive, I am not quite clear about the answers. Smoking - avoidable; BMI above 25 - yes, but not all that likely to change; maternal age over 35 - ditto; pre-existing diabetes and hypertension - treatable but little evidence that control affects stillbirth. Primiparity, and previous stillbirth, gestational diabetes and hypertension - all important factors, but in the end the only answer is to get the baby delivered at the least sign of trouble.
1367 Maybe you thought glaucoma was something that older people got and was prevented by regular measurement of intra-ocular pressure and the use of eye drops with strange names; o sinner, turn again. Read Harry Quigley's Seminar, and know that Glaucoma may be lurking deep inside your orbits, waiting to reward your ignorance with permanent loss of sight. Take heed, complacent one! "Open-angle glaucoma often occurs at an intraocular pressure that falls in the typical range. In Asia, most patients with open-angle glaucoma have similar pressure levels as those in healthy individuals... Glaucoma is undiagnosed in nine of ten affected people worldwide and is undiagnosed in 50% of those in developed countries; thus, improved case detection is needed." Fortunately most optometrists these days seem to know how to examine the fundus and perform proper visual field testing: which is just as well, as I am, maybe, a little deskilled.
BMJ 16 Apr 2011 Vol 342
856 When my father was dying of heart failure in Sheffield 20 years ago, my main concern was to keep him out of hospital so he could have a reasonably good death at home. I just about succeeded, and subsequent attempts to think through the problems of palliative care for heart failure much later on brought me into correspondence with Merryn Gott, who was then in Sheffield and is now in Auckland, NZ. Here she leads a qualitative study on the barriers to transition between acute hospital care and palliative care in England. Despite the growth of a large literature and better liaison between palliative care and the acute sector, the problems remain much the same, above all an unwillingness to discuss prognosis with the patient and the family. And as teams grow larger and ever more transitory, I can't see this getting better.
857 Traditionally, doctors look down on qualitative research as too touchy-feely and removed from the hard science that we are all so good at. Similarly, palliative care is all very well, but it's chemotherapy that kills cancer cells and improves survival. Except that very often it isn't, and studies of lung cancer have shown that stopping chemo and introducing palliative care can actually improve longevity as well as quality of life. And qualitative research can provide unique insights into how health professionals view choices about chemo at the end of life and how their own feelings can govern their choice of language to patients in what should ideally be a non-directive process of shared decision-making. The result is that chemo is often given or continued when neither the patient nor the doctor really believes it is worth it. When your own turn comes, or that of a loved one, you might wish doctors paid a bit more attention to the lessons of the qualitative palliative care literature.
Arch Intern Med 11 Apr 2011 Vol 171
636 Talking with Paul Glasziou a few days ago - always a treat - we strayed into the difficult territory of informed consent to screening. He suggested that there are three separate situations with population screening, and in none of them was personal decision-making the best option (above and beyond the absolute right of consent to be screened). The first is where the data demonstrate incontestable benefit from screening (e.g. phenylketonuria in the newborn). This can become public policy without detailed explanation to parents or patients. The third is where there is incontestable lack of benefit from screening (e.g. PSA for prostate cancer): this kind of screening should not be offered as public policy. It is only in the middle category, where the evidence is debatable, that any attempt at explanation should be made, and then not to individuals but to a citizen's jury sitting for two days. Only by a process of lengthy discussion can the issues be understood sufficiently to arrive at informed consent. Apparently this process has been followed in New Zealand; but I digress. Personally I don't believe that the UK programme of colorectal cancer screening using faecal occult blood testing is worth the candle, whereas regular sigmoidoscopy is, and colonoscopy may be nearly as good but is less feasible and more dangerous. I would favour the use of electronic prompts for the last two but not the first. This Harvard study included all three modalities of screening, and got better uptake using the prompts, but not for very long.
686 The higher the dose of opioid you prescribe for pain, the higher the risk of unintentional overdose. Those of you with a memory span of a week will be saying, "hang on, he told us this last week" (JAMA 305:1315). And so I did - but this is not a case of dual publication or a lapse of memory on my part (perish the thought): it is an entirely different study from the US-based one in JAMA, which included cancer patients. This one is from Canada and excludes cancer patients, but the message is the same. Opioids are dangerous drugs: perhaps because we are anxious to reassure patients, we hide from them (or ourselves) just how dangerous they can be.
Plant of the Week: Skimmia japonica "Rubella"
I have no idea why this excellent evergreen shrub came to bear the name of a once-common childhood disease: I suppose it is because its buds are pink before they open into white flower-heads, with the loveliest scent. This is an amazingly tough and versatile plant which will grow in dense shade, slowly but inexorably. Any cutting, or even leaf, that hits the ground will take root, but you will have to wait a long time to get a proper plant. It is ideal for occupying difficult spaces. It will grow on limy clay, but do give it some iron if it gets chlorotic. And do not be fooled by its female name: it is male, and if you want berries, you must buy another form, such as "Reevesiana".
JAMA 6 Apr 2011 Vol 305
1305 Postmenopausal oestrogen protects against breast cancer. Now there's a headline I never thought I would write; and yet it's the clear conclusion of this follow-up study of the women in the Women's Health Initiative Estrogen-Alone Trial. In this unique double-blinded RCT, over 10,000 women aged between 50 and 75 who had previous hysterectomy were randomised to 0.625mg conjugated oestrogen or placebo. Figure 3 gives the cumulative plot for invasive breast cancer: it could not be clearer. At 12 years, the risk line is still diverging in favour of the oestrogen-treated group, HR 0.77, 95% CI 0.62-0.95. All-cause mortality was the same in both groups. And while we are on the subject of counter-intuitive results about cancer risk, I've just read a letter in the BMJ Rapid Responses pointing out that in the 20-year follow up study of PSA screening from Sweden in last week's printed issue, there is a clear mismatch between the conclusion and the statistics given: "After adjustment for age at start of the study, the hazard ratio [for prostate-related death] was 1.58 (1.06 to 2.36; P=0.024). Conclusions After 20 years of follow-up the rate of death from prostate cancer did not differ significantly between men in the screening group and those in the control group." But it did differ significantly: PSA screening raises your risk of dying from a prostate-related cause, by perhaps 58%. And I missed this, because like most of you, I find my eyes glazing over whenever PSA and screening are mentioned. It so shouldn't be done.
1315 What's the second leading cause of unnatural unintentional death in the USA, after motor vehicle crashes? Gun crime? No - it's unintentional opioid overdosage. This Veteran's Administration study examines the relationship between levels of opioid prescribing and death from opioid overdosage: it is clear and large. Oddly it is lowest for those with substance addiction and highest for those with cancer, perhaps because the unpredictability and intensity of cancer pain leads people to take high doses without habituation. The editorial on p.1346 is thoughtful but does not mention another factor which must play a part in the US epidemic of opioid-related death: very easy access to oxycodone (the main culprit drug) on the internet. If I took up all the invitations in my Spam folder, I would be dead in short order. Though with lots of impressive watches, several degrees, two Russian brides and formidable private dimensions.
1350 A couple of the best pieces this week are on the subject of type 2 diabetes (see Lancet). The message is finally getting through: one size does not fit all, and glycaemia is not the most important outcome by a very long chalk. In fact it matters very little in most frail elderly people with short life expectancy, as this excellent commentary piece from San Francisco points out.
NEJM 7 Apr 2011 Vol 364
1315 The armed forces of Israel keep a close eye on the health of their personnel by means of a research programme called MELANY. In this study, they look at the body mass index at the age of 17, when the men were recruited, and analyse its relationship with the later development of diabetes and coronary heart disease between the ages of 25 and 45. This is assuredly not a time of life when you want either of these conditions, and we must also remember that these men were mostly operating at high levels of exercise and fitness. Nonetheless there were enough cases to establish a clear relationship between BMI in youth and later CHD, starting from well below the obesity range. There is also a weaker relationship between BMI in youth and the early onset of type 2 diabetes, but this is entirely accounted for by a relationship with BMI at the time of diagnosis.
1360 From time to time, doctors and politicians look across the Atlantic ocean to see what lessons might be learnt from comparing the UK health reforms with the US health reforms and vice versa. Here Martin Roland and Rebecca Rosen provide a superb guide for Americans who are perplexed as the English NHS Embarks on Controversial and Risky Market-Style Reforms in Health Care. Now I have just finished reading Jack Wennberg's magnificent book about American medicine (Tracking Medicine, OUP 2010) in which he lays out the following principles for a reformed US system:
- Promoting organized systems of health care delivery
- Establishing informed patient choice as the ethical and legal standard for decisions surrounding elective surgeries, drugs, tests, and procedures, and care at the end of life.
- Improving the science of health care delivery
- Constraining undisciplined growth in health care capacity and spending.
Roland & Rosen provide a Table 2 showing the differences between the US modes of provision and the proposed English reforms; my slightly less nuanced comparison of Lansley with Wennberg would be:
- Destroying organized systems of health care delivery
- Establishing "nothing about me without me" as a political slogan without any infrastructure of decision support for patients: their GP will decide according to the budget allocated.
- Ignoring the science of health care delivery
- Constraining necessary growth in health care capacity because there is no mechanism for matching supply with informed patient demand.
Lancet 9 Apr 2011 Vol 377
1220 You'll look in vain in this week's Lancet for research papers of general interest, but here is an editorial which every doctor needs to read: Intensified glucose control in type 2 diabetes""whose agenda? It's written by three senior academics in the field who almost alone have been prepared to look carefully at all the evidence and in particular at the absolute benefit (or lack of it) from tight glycaemic control for different kinds of diabetic patient. As they point out in one of many brief, telling sentences, "The most entrenched conflict of interest in medicine is a disinclination to reverse a previous opinion." Which sits perfectly with their final statement, " We declare that we have no conflicts of interest."
1224 The Lancet contains a weekly column called "Offline" in which Richard Horton informs us what he is doing instead of editing The Lancet. This one starts "I went to Trondheim last week to see Julian Tudor Hart". In fact JHT was not in Trondheim but back at home in South Wales: but hey, the fjords are better in Norway. I commend your attention to this week's Offline simply because it draws good comparisons with political decision-making about health in Norway and Scotland compared with modern-day England. How wonderful it would be if we could use Lansley and Cameron's "listening period" to bring into reality the Tudor Hart concept of doctors and patients as "co-producers of health" - an idea which has lost its Marxist connotations and appears equally strongly in the Wennberg book I mentioned previously. But that means trusting patients to make informed decisions about care in their locality, and giving them the means to provide it. "Goodness me, Humphrey, we can't have that - it almost smacks of democracy." "Yes, Minister."
1264 Viral pneumonia is a diagnosis we rarely dare to make in primary care: if somebody has a fever and chest signs, they get amoxicillin. If they don't respond, they get another antibiotic. If they get too sick, they go to hospital. That's what happens the world over, from Turku, Finland, to Christchurch, New Zealand, the two antipodes represented in this joint seminar. It's full of interesting detail, especially about viral co-infection, but it is unlikely to change practice since procalcitonin has not lived up to early promise and we still have no reliable means to distinguish viral from bacterial infection.
1276 A worthy account of osteoporosis: now and in the future comes from Dresden, and describes whole new classes of drug to treat this "common disease." But is osteoporosis best characterised as a disease, and don't we have perfectly good drugs to treat it with already? As each new pathway is elucidated, points like this begin to RANKL with an old Dickkopf-1 antibody like me. Say no to disease-mongering, however clever the science and tempting the market.
BMJ 9 April 2011 Vol 342
778 Speaking of disease-mongering, there's a fine example here in the editorial pages of the BMJ, with lots of old-fashioned GP-bashing for good measure. DMG Halpin is a respiratory specialist appalled at how bad we are at diagnosing COPD. There are "missing millions" of British citizens who have not been honoured with this depressing label and who "just expect to be told to stop smoking rather than receive a diagnosis and treatment if they do consult a doctor". It's quite shocking: you have to go to Kyrgystan to find anything so shoddy. Some of us even prescribe smoking cessation aids without doing spirometry and prescribing an armload of useless inhalers.
808 Emergency primary care is an emerging specialty of urgent care into which I am emerging with urgency. It's not really like old-fashioned general practice because long-term care is missing: but you do see an awful lot of sick babies and it's a relief when the season of acute bronchiolitis comes to an end. That's because neither oral steroids nor inhaled beta-adrenergic agents make any difference: but this meta-analysis does identify one trial where adrenaline by nebuliser plus oral corticosteroid made a significant difference. I guess we need to lay in some of these for next November.
810 The other thing we could lay in for November is a great big tub of simvastatin. Then, as the pneumonia season arrives, we could put every susceptible patient on a small dose and halve their short term mortality after a pneumonia episode. Mind you, they need to be taking the stuff before they get the infection, according to this observational study. Previous miraculous effects of statins have not been confirmed by randomised controlled trials, so we need one right away.
812 TIA for transient ischaemic attack has entered popular parlance just in time for the term to be largely binned by neurologists. Most "TIAs" lasting for more than a few minutes are associated with detectable brain infarction and are followed by larger strokes within 7 days in 5.2% of cases. This is a good brisk Glaswegian Clinical Review of how to diagnose TIA and stroke - remember ROSIER, FAST and ABCD - and how to treat it. Despite much greater public awareness of the urgency of stroke management, and the growth of specialist services, use of immediate thrombolysis remains conservative on both sided of the Atlantic, perhaps in acknowledgement of its marginal benefit.
Plant of the Week: Bergenia "Beethoven"
The bergenias are creeping plants of the saxifrage family, raised by German nurserymen to grow fleshy leaves of heroic proportions, and flower spikes varying from pure white (Silberlicht) to alarming shades of mauve and magenta. We chose this one for its name, of course. If you wish to indulge in garden wit, you can place it next to "Eroica", or with all the major nineteenth century composers whose names began with B - Brahms, Bizet and Borodin, with Bartok straying into the twentieth.
But in the garden as in music there is nothing to compare with Beethoven. When we first bought him, he produced dazzling white flowers in late winter. Then we transplanted him and he entered his Second Period, breaking from deep pink buds into flowers of white tinged with pink in March. Now he has entered his Third Period, pinker and later than ever. I am afraid this does not resemble the original Ludwig van Beethoven, who became ever more wonderful with time - let JWN Sullivan's Beethoven: His Spiritual Development (1927) ever be your guide. As for the bergenia, I am beginning to think we had better get a new one.
Page last edited: 09 May 2011