- Journal Watch
Journal Watch - September 2009
JAMA 23-30 Sep 2009 Vol 302
1277 Doctors expecting their first baby generally look around for the best obstetrician, whereas in the UK most other people get what they are given, on the basis that childbirth is a natural process and you probably won’t need a doctor anyway. The problem is that human parturition really is a very awkward business and if you need a doctor you need a good one in a hurry. So we advise our parent-to-be colleagues, “Go to so-and-so at St Somewhere’s: he trained at Charlotte’s and he’s really very good”; whereas for patients it’s “we’ll fix you a midwife appointment and she’ll spend lots of time going through everything with you.” If you think this is terribly old-fashioned and elitist (and in most ways it is), read this study from the USA. Obstetricians trained in good centres have better outcomes by several objective measures. As for midwives, I can’t wait to read the National Perinatal Epidemiology report on neonatal outcomes when it finally appears. Obstetrics led the way in self-audit, but there are still some basic lessons to be learnt. Or repeated.
1284 There are certain branches of medicine, such as primary care, where the more you put in, the quicker you burn out. If you do your job properly, it’s not a matter of whether this will happen but of how soon. No human being was designed to go on every day for 35 years solving 40 complex human problems with empathy and far-ranging knowledge at 10-minute intervals, fighting off a complex bureaucracy and running a business, while attempting to squeeze in a personal life and continuing education in the remaining four hours. Oh, and eat. The problems of American primary care doctors are different, and they have twice as much consulting time, but they burn out too. This study tries to evaluate the benefit of a burn-out prevention programme consisting of “mindfulness meditation, self-awareness exercises, narratives about meaningful clinical experiences, appreciative interviews, didactic material, and discussion.” The doctors felt better afterwards. We need more revival meetings for doctors:
Revive us again
Fill each heart with thy love,
May each soul be rekindled
With fire from above.
William P Mackay 1863
NEJM 24 Sep 2009 Vol 361
1235 Acute myeloid leukaemia is the kind that arises mostly in the over 60s and has a high mortality. These are people you refer the moment you get their blood count back and who then reappear looking manky on some weird kinds of chemotherapy. Could anything indeed be weirder than gemtuzumab ozogamicin? But it is not of this second-line drug from outer space that we sing in this section, but rather of daunorubicin, which has been around for long enough for most of us to be familiar with. Yet as with so many chemotherapeutic agents, dosage regimes have been set by tradition rather than evidence, and these big Dutch, German and Swiss studies were needed to determine that putting in dauno at a high dose from the start is the way to get the early remission on which survival depends. The same message comes from a US/Israeli study on p. 1249.
1260 The delights of the influenza vaccination season are upon us, doubled or trebled by the prospect of additional H1N1 vaccination. In the UK, the choice of vaccine is made for us by the People’s Commissariat for Health and Death Panels, but in the Land of the Free there is a choice between inactivated or live attenuated vaccine. This was a randomized, double-blind and placebo –controlled study of the two kinds of vaccine carried out in Michigan universities in autumn 2007. Both kinds of vaccine are better than placebo, but the inactivated vaccine is the better of the two.
1268 I often wince when I go past our practice nurses’ noticeboard, sporting posters advising patients to be Prostate Aware or Parkinson’s Aware or whatever Awareness is being celebrated on that particular Day. I’m all too aware that I shall crumble and die, thank you very much. If anyone knows of something that changes the course of early prostatism or Parkinson’s then please tell me what it is and I shall be happy to provide it. As things stand, we have a variety of symptomatic treatments which have major side-effects, and no evidence of any benefit from early use. This study looked at the early use of rasagiline in Parkinson’s disease. At a dose of 1mg it may have a disease-modifying effect, but not at 2mg. This does not make sense, so the authors advise caution and further studies. Quite so.
Lancet 26 Sep 2009 Vol 374
1067 The world’s senior medical journal is named after a surgical instrument, but can’t quite rid itself of traditional British snobbery about surgeons. This week’s front cover tells academic surgeons that they must transform themselves from comic opera characters. Well, The Barber Surgeons of Seville provide rebuttal enough for this silly stereotyping in a couple of excellent surgical research papers and several good theory papers filling this week’s Lancet. This first one, from Cambridge (UK) shows that burr hole evacuation of subdural haematoma should always be accompanied by use of a drain. This is associated with a large reduction in recurrence and mortality. How many physicians do studies as useful and decisive as this?
1074 Next to the carpal tunnel. The patients in this US trial were enrolled after two weeks of symptoms suggesting median nerve entrapment, which is long before most British patients present, let alone get treated. Also, the comparator was customised hand therapy, which has yet to arrive in my part of the world. So interpret with caution. In patients without evidence of denervation, surgical release was superior to conservative management in this trial, but not by very much. A real-life trial in the UK would randomise patients after three months to steroid injection or surgery.
1082 Now let’s move into the intensive care unit, a scary place full of wires and tubes and no place for a bumbling GP with a soup-stained tweed jacket. But it’s nice to hear what goes on in these places, from time to time. The practice of invasive ventilation of hypercapnic patients with chronic respiratory disease may be modified by this Spanish study which shows that non-invasive ventilation following extubation provides better outcomes than re-invading the trachea.
BMJ 26 Sep 2009 Vol 339
735 A lot of women get pregnant while taking serotonin reuptake inhibitors for depression, and many of them continue taking them. This will probably remain the case following this comprehensive cohort study from Denmark, covering nearly half a million births from 1996 to 2003, even though it shows an overall doubling of septal heart defects in babies born to women taking SSRIs. The highest risk seems to come from sertraline and the lowest from fluoxetine. There is no additional risk of other major congenital malformations. It is unclear whether discontinuing these drugs after conception will make much difference.
736 Now for the thigh circumference paper which caused much furrowing of brows when it first appeared on the BMJ website. Denmark is famous for its hams but if these diminish below a certain size in humans of both sexes, then the risk of heart disease and premature death goes up steeply, irrespective of other biometrics. Odd. Strange. Queer. Puzzling. Your guess is as good as mine.
740 There have been plenty of review and articles about bariatric surgery over the last couple of years, and this one contains nothing particularly new, but is a good clear account of what is available, the evidence for its effectiveness (huge) and how it is supposed to be rationed in the NHS.
747 The Diagnosis in General Practice series continues in the way that Kevin Barraclough and I first hoped it would three years ago, combining a theory piece with an illustrative clinical piece. This may not seem earth-shattering, and isn’t meant to be, but where else have you seen described the actual steps by which GPs reach diagnoses in ten-minute consultations? You may never have heard of iterative diagnosis, and yet you do it all the time. The alternative term would have been hypothetico-deductive diagnosis and you’d have run a mile from that. Like it or not, you’re a clinical logician as well as a pattern-recogniser and thinking about these skills might help you refine them. This series might actually help to make general practitioners into better clinicians: Kevin’s piece on vertigo certainly does.
Plant of the Week: Lagerstroemia indica
In the open spaces and gardens of Washington DC at this time of year, the loveliest plants are the smooth barked multi-stemmed crape myrtles which bear abundant flowers of pink, red-purple or white. Why don’t we ever see them in England? After all, Washington is no place for tender plants, with regular winter frost and snow.
The answer lies, as you may have guessed, in its long hot summers, which are needed to ripen the flower buds of these delicious big shrubs from China. Global warming may make these flowers a common site in the England to come. As it is, you can grow a crape myrtle perfectly well in our southern counties but may never see it flower. All you can do then is stroke the wonderfully smooth bark of its sinuous trunks.
JAMA 16 Sep 2009 Vol 302
1171 Ten years ago, most people thought the prevalence of coeliac disease was about one in 1000 or less. The study that changed awareness in the UK was done in my practice by Harold Hin and showed that it was more like 1 in 100, a figure that has been reproduced around the world. But have we done patients any favours by uncovering their auto-immune reaction to gluten? Undoubtedly we have, in those with longstanding symptoms, but asymptomatic individuals with positive tests may feel less grateful and quickly abandon their restrictive diet. This Swedish study starts at the other end and looks at the mortality outcomes associated with villous atrophy as reported in all 28 Swedish path labs since 1969. The more inflammation on the biopsies, the greater the additional mortality. So uncovering and treating coeliac disease may perhaps save lives, though this needs proving with a prospective study, also looking at inflammatory markers and vitamin levels. Harold is too busy running the practice, so come on somebody, get going.
1186 The most popular inflammatory marker is high-sensitivity C-reactive protein. Show-off researchers have a wide choice of others, such as interleukin-6 and soluble tumour necrosis factor receptor 2. If you do a trial in diabetes showing changes in these, then you have a chance of getting published in The Lancet. On the other hand, if you can’t show a change, you may have to make do with JAMA, even if you call your trial LANCET. Here is another diabetes study using some feeble surrogates, if you still have an appetite for such things. If you treat newly presenting type 2 diabetes with metformin, glargine insulin, or placebo, these markers of inflammation will stay the same.
1195 Another study from the indefatigable American Heart Association team looks at cardiopulmonary resuscitation outcomes in the USA. (Some readers will realise that when I use the word indefatigable it’s a sort of code name for Harlan Krumholz). If you are black, your chances of making it out of the hospital door alive are about 20% worse than for whites, and most of this is explained by the performance of the hospitals you are likely to get admitted to. Don’t socialize American medicine! Praise the Lord and shoot the moose.
1202 The foggy terrain of prostate cancer management is traversed again in this historical study of outcomes following conservative management of localised cancer. In men over 65, surgical treatment has not been shown to improve outcomes compared with conservative management; and this study from 1992-2002 confirms that even if your prostate cancer is histologically aggressive, you are much more likely to die from an unrelated cause than the cancer. It may be that PSA testing has led to the detection of more early cancer in the USA, thus increasing the lead time: and its use in follow-up may have improved later treatment. So comparisons with the pre-PSA era are hard to interpret but survival now is certainly better than in the 1970s and 80s.
NEJM 17 Sep 2009 Vol 361
1139 Remember ximelagatran? Some years ago, I advised readers to practise pronouncing it, as it would shortly replace warfarin for the management of atrial fibrillation and perhaps much else besides. Alas, zimmywosname faded away because of toxicity worries, and bagsful of INR samples continue to make the short journey between our practice and the hospital laboratory every day. Although our standards of INR control are very good, every now and again a patient on warfarin has a haemorrhagic stroke. For patients, phlebotomists and stroke physicians, a safe fixed dose direct thrombin inhibitor cannot come soon enough. So, hail Dagibatran, conqueror of Thrombin! This big (n=18,113) trial run by its manufacturer shows better outcomes for fixed dose dagibatran 150mg than for INR-monitored warfarin in all the important end-points – stroke, systemic embolism, major bleeding and death. Boehringer Ingelheim would do the world a big favour – and perhaps even increase its profits – by pitching the price of dagibatran at exactly the cost of warfarin plus the cost of INR testing. After all, the market isn’t exactly small.
1152 Why do some fat people avoid getting diabetes? The answer may lie in the level of sex hormone-binding globulin which is largely determined by the SHBG genotype. That may help to explain why the one patient I have with a BMI over 70 does not have diabetes. Here is another study confirming the strength of the association between low SHBG and a high risk of diabetes, first in a case-control study nested within the Women’s Health Study and then in a replication study of an independent cohort of men from the Physicians’ Health Study. The prediction of diabetes is becoming an interesting area.
1164 Happy is the garden that contains a hedgehog pathway. These gentle snuffling creatures spend most of their waking lives eating slugs and snails. They are also renowned for the noisy nature of their sexual congress: no less a musical authority than Sir Thomas Beecham once likened the sound of a harpsichord to that of two hedgehogs copulating on a tin roof. Harpsichordists wishing to refute this suggestion experimentally should pause to consider that determining the sex of hedgehogs can present difficulties and that tin roofs have become uncommon. They should try playing the piano instead. Now where on earth was I? Ah yes, inhibiting the hedgehog pathway in advanced basal cell carcinoma. This is a good thing, apparently, and can be done using an oral drug as yet merely referred to as GCD-0449. “We found evidence of hedgehog signalling in tumours that responded to the treatment.” Or was it just somebody playing the harpsichord in a neighbouring room?
Lancet 19 Sep 2009 Vol 374
979 I was taught obstetrics in the heroic age of induction of labour at Oxford. Our new professor taught that birth was safer if it took place when there was a full support team in place, and as a result there were no births at all on Christmas Day that year. At the same time, a young medical registrar there called Chris Redman decided to devote his career to unravelling pre-eclampsia. Thirty-five years on, he has done a prodigious amount of work and I am not sure that we understand much more about pre-eclampsia, but this Dutch study does confirm that women with gestational hypertension or mild PET have better outcomes if they are induced at 37 weeks. Avoiding Dec 25th.
989 I recently set our registrar the task of discovering whether the co-prescribing of proton pump inhibitors affected the clinical efficacy of clopidogrel, and thus whether we should alert all our patients taking this combination. It was a nice little exercise in the limits of evidence-based medicine which she carried out very well, culminating in her discovery of this paper on the Lancet website showing that in TRITON-TIMI 38, there was no evidence of poorer clinical outcomes in those taking a PPI with clopidogrel. So we won’t be putting round a mail shot to our patients, though I bet you we haven’t heard the last of this matter.
BMJ 19 Sep 2009 Vol 339
673 There was a brief moment some time around 2004 when New Labour’s injection of money into the NHS seemed poised to bring about a real and lasting change in patient outcomes and job satisfaction for health professionals. Waiting times had fallen dramatically and there was greater equity in health provision – this study shows that social inequalities in waiting times within the NHS disappeared completely. A period of stability and incremental progress seemed possible. I could talk to American audiences about the NHS without being ashamed at the way it treated individuals. The main problem was what it has always been – a lack of linkage between what people want to spend on health and what governments are prepared to allocate from general taxation. Then came successive waves of ideological madness, resulting in a complete lack of accountability (see Iona Heath’s trenchant piece on 663) and risk-free profits for a parasitic array of private health organisations. The incoming Tory government is going to have to use some harsh socialist measures to put this right at a time of huge national debt.
679 While I’m in sounding off mode, let me repeat a warning to politicians and those who report clinical trials: it is wrong to cheat. Statistics become damned lies when you leave stuff out. Chief of Police Doug Altman gathers around him an array of Swiss and German constables and looks at the effects of excluding patients from the analysis of randomised controlled trials. This happened in 77% of the trials they looked at dealing with interventions to reduce pain in hip and knee arthritis. Appalling. And then there are politicians.
685 To prove that aciclovir adds nothing to steroid treatment for Bell’s palsy, we needed one good trial, and that was done in Tayside not long ago, as I pointed out the other week. But meta-analysis has almost become an end in itself, and here is the second in close succession dealing with this topic – of some clinical importance, but in the last analysis, mainly cosmetic. I’m told that there is a third on the way, from Cochrane in the New Year. That’s quite enough, thank you.
Arch Intern Med 14 Sep 2009 Vol 169
1484 If you are going to use thrombolysis for occlusive stroke, the sooner you do it the better: if this sounds obvious now, consider how unobvious it sounded ten years ago, or even five to most people in the street. The number needed to treat is still very high, and if this German cluster-randomised trial of a public awareness campaign had used a hard end-point, like overall reduction of disability or death from stroke, it would have had a zero result. Instead, it yielded a 27% increase in women with stroke reaching hospital within three hours. I think that we can conclude that personal letters with bookmarks and stickers are not a cost-effective basis for a stroke awareness campaign.
1491 We move now to Adelaide, a city of free Australian settlers named after Her Serene Highness Princess Adelaide of Saxe-Meiningen. “She is doomed, poor dear innocent young creature, to be my wife”, wrote the vast libidinous Prince George, heir to the British throne, in 1818. In fact they were improbably happy together, living with a selection of the prince’s illegitimate children in Hanover, though all Adelaide’s pregnancies miscarried or the babies died early. So it came about that this five-syllable German girl’s name became a three syllable south Australian place name, scene of the CADENCE study looking at the prevalence of weekly angina among patients in primary care. Chronic stable angina is something we tend to consign to repeat prescribing and once a year nurse checks, but one in three of these patients gets an attack at least once a week, and there is probably room for better audit and control. After all, we do not want these patients to get cremated, as Beethoven implies his heart will be in the final cadence of Adelaide:
Eine Blume der Asche meines Herzens
Deutlich schimmert auf jedem Purpurblättchen:
1500 If I had advanced heart failure, I would want to keep open the exit of sudden arrhythmic death, knowing what the other exit is like. But plenty of HF patients, especially in the USA, have been expensively fitted with implantable cardioverter-defibrillators to prevent such an outcome. This meta-analysis shows that they do not in fact reduce all-cause mortality in women with advanced HF. The shocking truth.
Ann Intern Med 15 Sep 2009 Vol 151
386 When reading papers with the words “net value of health care” in the title, it is always worth remembering that the best way to ensure value in health care is to ensure a swift death for everyone who is not working, especially the elderly. However, this Mayo Clinic study of the net value of health care for patients with type 2 diabetes, 1997 to 2005, is somewhat less drastic in its approach, and bases its figures on the reduction in spending on coronary heart disease in the UKPDS tight control cohort. These were patients in their early fifties with new-onset diabetes, so they don’t actually tell us anything about the economics of treating people with type 2 diabetes in the age beyond retirement. Just as well, really.
394 The thing that goaded me into writing an editorial on tight control of glucose in longstanding type 2 diabetics last March was the silence of diabetologists following the ADVANCE and ACCORD studies in June the previous year, plus the VADT study which told the same story and was published in January 2009. These were prospective randomised trials designed to answer much the same question, and they all gave the same answer: intensive glucose control in established type 2 diabetes does not reduce overall cardiovascular mortality and increases the risk of severe hypoglycaemia. Since then a number of meta-analyses have appeared, this being the latest. These include long-term observational data from UKPDS, which was not designed to answer this question; The Lancet , one also included data from a wildly irrelevant study, and there is a further one waiting in the wings on the Diabetologia website. As a study in the sociology and psychology of medical practice, this is quite interesting and a little dispiriting, but as far as treating patients goes, the message could not be simpler. For type 2 diabetics a few years into the disease process, there is no point aiming for an HbA1c under 7.5%. There is a tendency to fewer non-fatal CV events and perhaps some renal protection for a tiny number, but overall it makes no difference and causes hypoglycaemic episodes, which can cause brain damage. Cognitive impairment should be an end-point in future studies.
Fungus of the Week: Suillus bovinus
If you are a very cautious fungus hunter – and this, on the whole, is associated with the best survival outcomes – then it is worth bearing in mind that no mushroom with yellow or brown pores (as opposed to gills) will ever kill you. They all used to be lumped together under the generic name of Boletus but all modern books split these into various genera, reserving Boletus for a few of the big fleshy ones and other names for the smaller ones like these.
I found a large number in a Welsh pine wood a couple of days ago and brought them home – all in good condition and worm free – to eat yesterday. As they are small and gregarious, you are likely to collect a good number together and I’d suggest that you use them as the basis of a main dish such as the following:
Frazzle some chopped rashers of smoked bacon in oil or butter. Add a chopped leek or two. Add the mushrooms, chopped into fairly large pieces – there is no need to remove the gills or the skin of the caps. Fry until the mushrooms soften. If you have cooked with butter, you can add cream at this point. Vegetarians can omit the frazzled bacon. Carnivores can add cooked chicken. This is not a sophisticated dish. Season lightly and add parsley if you wish.
JAMA 9 Sep 2009 Vol 302
1059 Do you solemnly swear to abide by the principle of using antibiotics only when strictly necessary? At present it’s easy to mutter assent to visiting bacteriologists and prescribing advisers, on the basis that only we ourselves can decide when it’s strictly necessary. If little Shane has a nasty cough and it’s 7 on a Friday evening, then strict necessity determines that he gets some amoxicillin and that you go home to dinner. But all this might – perhaps – change when we get a point-of-care test that enables us to determine which patients have bacterial infections and which don’t. The front runner for this is serum procalcitonin, and some of you may remember a study from Swiss primary care showing its usefulness in adults with lower respiratory infections which appeared in the Archives of Internal Medicine last year. The same investigators ran a parallel study called ProHOSP which as its name implies tested out procalcitonin in hosp. Here they report the results from six tertiary hosps where patients with LRTI – community acquired pneumonia, acute bronchitis, or exacerbation of COPD – were randomised to receive antibiotics based on standard guidelines or procalcitonin (PCT) levels at presentation and at days 3, 5, 7, and discharge. There was a large drop in antibiotic use in the PCT guided group with no difference in clinical outcomes.
1084 There’s a new book out called The Emperor’s New Drugs: exploding the antidepressant myth which points out the speculative and self-contradictory nature of the biochemical theories on which modern antidepressant treatment is based. That, however, is a very different thing from showing that the drugs don’t work. The same thing seems to be happening in the field of attention-deficit/hyperactivity disorder (ADHD) which scarcely existed as a diagnosis until about 20 years ago. Positron emission tomography seems to indicate a reduction in dopamine synaptic markers in adults with ADHD. Expect a flurry of trials of dopaminergic drugs, or conversely dopamine-blocking drugs, in adults with ADHD (kids are too tricky to do trials on) with high drop-out rates and exaggerated claims of benefit.
1097 In the Clinician’s Corner of JAMA (all right clinicians, you can come out of that corner now) is a useful overview of bariatric surgery based on the case of 52 year old woman with a BMI of 53. It’s fine as far as it goes, but just now I am wrestling with the dilemma of a woman of similar age whose BMI is 75. She is terrified of surgery because her mother had a massive stroke under GA for a minor procedure. On the other hand, if she doesn’t have something done, her prognosis is probably worse than if she had untreated colonic cancer. And if she does have something done, she is at very high perioperative risk and at best may end up with an apron of skin down to her knees. If anyone has any suggestions, I’d be glad to hear them.
NEJM 10 Sep 2009 Vol 361
1045 Over the last couple of years, we’ve heard a lot about the drawbacks of clopidogrel, which is soon going to come off patent, and the possible merits of prasugrel, which has just been licensed to compete in the highly lucrative antiplatelet therapy market. You could call it the Quest for the Holy Grel. But now comes ticagrelor, which has the dual advantage of being independent of the CYP metabolism pathway and reversible in its action on the platelet PY12 receptor. In the PLATO study comparing it with clopidogrel following acute coronary syndromes with and without ST elevation, ticagrelor produced better outcomes for overall mortality, stroke and myocardial infarction without an increase in major bleeding. But the trick will be to apply this knowledge to individual patients, as the editorial (p.1108) states, in accordance with a great principle from Plato himself: “A good decision is based on knowledge, not on numbers”. Plato also said that attention to health is life’s greatest hindrance, with which I heartily concur.
1067 We are walking microbiology laboratories. Not only are there more bacteria in our gut than cells in our bodies, but millions of our own cells contain viruses that live in quiet equilibrium with our own DNA and RNA. So there was something quite unsettling when the first cases of progressive multifocal leukoencephalopathy (PML) were reported in patients receiving natalizumab for multiple sclerosis. By a cruel irony these patients died from rapid demyelination brought about not by MS, but by a virus that had lain harmlessly dormant in their own bodies, as it may well do in everybody: the polyomavirus known as JC. As a result, natalizumab had its licence withdrawn for a time, but was then relicensed for cautious use in patients with rapidly relapsing MS. PML continues to occur in one per thousand of these patients, but from this study we know that subclinical JC reactivation occurs in the majority (63%) of MS patients treated with natalizumab. In the unfortunate few who get the full-blown syndrome (14 so far in MS treatment across the world) it is now possible to arrest its fatal course by plasma exchange, though this is followed by a life-threatening immune reconstitution inflammatory syndrome.
Lancet 12 Sep 2009 Vol 374
881 One of the things (though I can’t think of the other) that The Lancet does best is to publish global health surveys. Mortality in young people (aged 10-24) has been little studied but this big WHO/MRC study changes that. The WHO stratification of countries by income is decidedly broad-brush, with Poland and Burkina Faso falling into the same category of “Low Income to Middle-Income”. Browsing among the causes of death in each age group, it’s sobering to see how respiratory illness remains a big killer in young adolescents and how road traffic accidents, violence and drowning then climb up the list.
893 The global burden of disease in children has been much better studied, because of the aim to reduce mortality in the under-5s by two-thirds of its 1990 level by 2015. Looking at children under 5, classic respiratory tract bacteria still account for hundreds of thousands of deaths. Streptococcus pneumoniae remains captain of the men of death (Osler’s phrase, I think, but which he applied to older people) in children aged 1-59 months. Vaccination and prompter treatment should reduce this by at least the desired two-thirds.
903 In the case of Haemophilus influenzaetype b, the use of vaccine should result in far more dramatic reductions in the current mortality of perhaps 371,000 children annually. Most of this is accounted for by Hib pneumonia and meningitis. Looking at papers like this, it’s clear that most improvements must come from political change and better herd immunity. But I wonder if there are other easily harnessed ways to help individuals in the poorer world, especially the sharing of knowledge, so abundant in the age of the internet, and yet so disorganised.
BMJ 12 Sep 2009 Vol 339
606 A couple of papers in this week’s BMJ look at the effect of diet and exercise on knee pain based on a study in five Nottinghamshire practices where overweight and obese patients were randomised to a dietary intervention, exercise, both, or neither with an advice leaflet. Quadriceps exercises done at home over two years definitely reduce pain and improve function in this group. The tagged-on economic evaluation seems to me like a waste of time.
613 But then you’d expect that from an old cynic whose first contribution to the BMJ included sarcastic comments about bran. This primary care study of insoluble fibre (i.e. bran) versus soluble fibre (something called psyllium) for irritable bowel syndrome concludes that bran is indeed better consigned to tubs than bowels. While on the other hand psyllium works for IBS, despite its psilly name.
616 “Ovarian cancer is not silent, rather its sound is going unheard.” When it first appeared on the BMJ website, this rhetorical flourish at the end of a study of early symptoms of ovarian cancer in primary care set off a predictable round of GP-bashing in the press and shroud-waving by cancer charities. I do not like ovarian cancer, because it killed my mother in a rather horrible way. She developed abdominal bloating and recurrent urinary infections as she was looking after my father in the final stages of his heart failure. In the end I had to suggest to her GP that an ultrasound scan might be useful, and when it revealed stage 4 cancer it never occurred to me to blame him for not spotting it 3 months earlier (the average delay in this study). What possible difference could that make? The paper’s ending is not present in the pico printed version, but it should read: Ovarian cancer is not silent, rather its sound is only audible when it is too late.
620 In the US television series House, the limping old grouch bats around possible diagnoses at his unhappy entourage, inevitably including strange infections and poisons, lupus, paraneoplasia and sarcoidosis. How unlike the home life of our own dear general practitioners. We might see erythema nodosum once or twice a year, uveitis a bit more often, and incidental hilar lymphadenopathy on chest X rays every now and again. Unlike our prestigious academic colleagues, we tend to have a limited interest in diseases without a known cause or specific treatment. But for a readable update, this is certainly a good place to look.
Plant of the Week: Perovskia “Blue Spire”
Although autumn is a season we tend to associate with browns and reds, there are lots of blue flowers to be had as well, on clematis species, ceratostigma, hibiscus and on this hybrid sub-shrub. Despite its name, its spires are more lavender than true blue, like those of its parents, Perovskia abrotanoides and P atriplicifolia. Once the first frosts have browned off its grey-green cut foliage, cut it well back and it will give better pleasure next year from its lovely spring growth and its beautiful flower spikes from late summer till late autumn.
JAMA 2 Sep 2009 Vol 302
947 Funny how medical terms change meaning: to an old pedant like me, acute coronary syndromes mean any acute syndromes caused by the coronary arteries, including myocardial infarction and sudden death, but it seems that JAMA readers automatically assume that it only means coronary syndromes without ST elevation. With these latter syndromes there is still room for debate about the relative merits of immediate versus delayed intervention, the key question in this multicentre French trial. Patients with non-ST elevation ACS were randomised to get their angiographic intervention either immediately (mean 70 minutes) or on the next working day (mean 21 hours) and the outcome was myocardial infarction measured by troponin 1. There was no difference between groups in this important short-term outcome.
955 Speaking of old pedantry, I suppose that the word “hysteria”, meaning related to the womb, survives in use because women’s groups have lost sight of its Greek root. Maybe the original “hysterical” illness was chronic pelvic pain, which being a woman’s complaint in Ancient Greece would have been automatically regarded as a great deal of fuss about nothing, as indeed many male doctors have regarded it up to nearly the end of the twentieth century. Chronic pain syndromes are by definition horrible to live with and generally inexplicable: if they had a known cause they would be labelled accordingly, and sufferers would not have to put up with the double burden of the pain and the bafflement of doctors. Moreover, all these syndromes make both patients and doctors desperate to try anything, and nerve ablation seems a fairly logical thing to go for. Except that, in this British trial (LUNA), laparoscopic uterosacral nerve ablation achieved nothing.
977 Evidence-based medicine has never been a very pleasing label, or a very complete answer to the complexity and uncertainty in most kinds of clinical encounter. But at least it tries awfully hard to find what evidence is out there in the randomised controlled trials. The size of the task is more than Herculean: so far the Cochrane Collaboration has managed to cover 10% of published RCTs and even these noble efforts are vitiated by the fact that selective reporting of primary outcomes is rife if you look hard for it. That’s the burden of this JAMA paper by the ever-vigilant Doug Altman, among others, based on Medline; meanwhile the ever-vigilant Harlan Krumholz of Yale and others publish a study in PLoS with the same message based on trials registered with Clinicaltrials.gov since 2000.
985 Another problem with EBM is its dependence on the objectivity of systematic reviews in (a) finding (b) selecting and (c) reporting the evidence from RCTs. The issue of whether to prescribe corticosteroids with or without aciclovir in Bell’s palsy was, I thought, answered definitively in a study published in the New England Journal in 2007. It was a model of good practical research done in primary care in Tayside, and was voted Paper of the Year by readers of the BMJ. When it first appeared I was moved to celebrate it in posthumous lines by the great Tayside poet, William McGonagall:
Two great Bells of Caledonia are recorded in the Hall of Fame;
A drooping of the face and also the telephone immortalise their name:
Now by the banks of the silvery Tay folk have investigated Bell’s palsy
And found that steroids work but it’s no use giving aciclovir alsae.
A nice clear message that has apparently got these verses a place in a Powerpoint presentation of this trial. But here comes a systematic review which muddies the silvery waters of the Tay by suggesting that there is still some doubt about the possible additional value of aciclovir and/or other antivirals. Och, mon.
994 Rather like the man in the French comedy who was surprised to find that he had been speaking prose all his life (M Jourdain in Molière’s Le Bourgeois Gentilhomme), I find that I am an exponent of systems medicine without knowing it even existed. It means the opposite of what it appears to mean, and talking opaque jargon seems to come with its territory, which is perhaps why I never realised I had wandered into it. I had invented my own term, integrative (not integrated) medicine, which is bad enough, but at least describes what we need to do: to work towards the integration of all inputs – from the patient, from care-givers, from genomics, from all the medical sciences and from evidence-based medicine to deliver a continuum of treatment throughout the lives of individual patients. I suspect I have already lost you. This little piece explores some of the many pitfalls.
NEJM 3 Sep 2009 Vol 361
940 Last week I wondered if anyone would be interested in the problems of clopidogrel once prasugrel became available, as it has been in the USA since July 10th this year. Two perspective pieces about prasugrel provide an excellent summary of the evidence that led the FDA to license the drug. By rights, it should be at least 30% more effective than clopidogrel, since we know that 30% of people can’t metabolise clopidogrel into its active form. In the head-on TRITON-TIMI 38 trial, the advantage of prasugrel was more like 19% with a substantially higher risk of bleeding and a debatable increase in cancer diagnosed within the first year. We’ll wait and see what happens in the UK.
947 Gefitinib is one of a growing number of drugs which target the endothelial growth factor receptor (EGFR), which can play an important role in tumour growth and also in macular degeneration. Its effectiveness should be predicted by the degree of EGFR gene expression in the tumour, and this is precisely the case in this trial of gefitinib for pulmonary adenocarcinoma.
958 Adenocarcinoma of the lung is relative uncommon, so it is misleading of the New England Journal to refer to “lung cancer” in the title of the next paper which is about screening for EGFR mutations. It is only worth screening for these in non-small-cell cancers of the lung. Nevertheless this is an illustration of the kind of thing that genomic people call personalized medicine (see p.1018) – something that can be done on a micro-array and a computer link without having to meet the person at all. Still, an exciting advance for people with these kinds of tumours, and perhaps the shape of things to come.
968 More in my usual run of work would be an injection that works for Dupuytren’s contracture. By the way, fellow-pedants, note that the New England Journal always uses the eponymous genitive, while JAMA avoids it even in the case of Bell’s palsy (see p.985) thus giving rise to the term Bell Palsy, more suitable for church towers with frayed ropes. Back to the palmar deposition of collagen described by M Dupuytren. Hand surgeons try to release these contractures with variable results, but this study looks at the use of collagenase injections. The collagenase is produced by Clostridium histolyticum and causes all manner of local or more generalised reactions in 96% of patients. However, there is gain from the pain: dissolving away the collagen definitely produces a significant increase in finger mobility, from a 44 degree flexion angle to 80 degrees.
980 A Dutch study shows that patients undergoing vascular surgery do better if they take a perioperative high dose statin. The rate of myocardial infarction and cardiovascular death within 28 days was halved with fluvastatin 80mg. More evidence that the protective effect of statins is not mediated through long-term effects on lipids.
990 I have previously described coronary artery calcium screening as the payment of large sums to receive high doses of ionising radiation in return for an increase in anxiety. This article on its place in cardiovascular prevention strategies reaches much the same conclusion, but I’m sure this will do nothing to dent its popularity with the rich worried well.
998 If you lose too much of your intestine you will die, though you can be kept alive for a while by parenteral nutrition. The alternative is intestinal transplantation which is also followed by death in up to half of recipients, though the proportion is falling; the rest merely suffer from life-long diarrhoea. A last-ditch procedure, then, described and illustrated here in gut-wrenching detail.
Lancet 5 Sep 2009 Vol 374
I go through much boredom on your behalf, dear readers, but going through this week’s Lancet was almost beyond the call of duty. On p.758, Richard Horton solemnly warns President Obama that US drug manufacturers overcharge Americans. There’s a glimmer of interest in the Correspondence in a letter from Chinese researchers complaining at the cost of full-text articles (p.786) in many developing countries. If you are interested in the relative merits of otamixaban versus eptifibatide then follow the link. But if you don’t have a clue what these are, don’t lose sleep. If you treat HIV, then you will be interested in the place of raltegravir in new treatment regimes.
BMJ 5 Sep 2009 Vol 339
The more you look, the more you’ll find. Magnetic resonance imaging is a fantastic way of looking inside people and gets better all the time, and patients who are getting fed up with your diagnostic slowness will often ask why they can’t just have a whole body scan. That isn’t a stupid question and the reasons bear some thinking about. One reason is that you would find stuff: one in 37 MRI brain scans turns up an incidental finding. Most of these are non-significant, but incidental tumours turn up with increasing frequency according to age.
Older patients often complain to me that hospitals these days turf them out far too quickly, and occasionally they’re right, though if I were in hospital I couldn’t wait to get out as soon as possible. For a few years now, hospitals have been given incentives to discharge patients quickly, leading to fears that this would increase rates of readmission and put an extra load on primary care. The latter may have happened, but readmissions don’t seem to have increased in the limited domains looked at in this study, mostly hip fracture.
Due to a postal strike, I haven’t yet received my paper copy of this BMJ, but when I do I think I’ll keep it in my consulting room for its wealth of information about female contraception. When I became a doctor 34 years ago, the standard oral contraceptive was levonorgestrel 150 mcg plus ethinyloestradiol 30mcg, marketed as Microgynon 30 or Ovranette. The only change since is that oestradiol has lost its first “o”. New progestagens and oestrogens have come and stayed but never supplanted the good old pills of the 70s. Moreover, these preparations have proved incredibly safe, with the chief risk remaining thromboembolism, though even this diminishes with length of use. The first paper about this is based on data from Danish women using OCs from 1995-2005, amounting to 10.4 million woman-years. One gets the impression from this study that the differences between the various pills are slight. But it doesn’t look that way from the opposite end of the telescope, in the anticoagulation clinics of the Netherlands (a paper from Leiden again, of course). Here the case-control study shows a fivefold increase of DVT in pill users and detects MEGA-differences between desogestrol (worst) and norethisterone (best). All this is nicely put into perspective by the editorial and by an excellent evidence-based overview of all forms of female contraception on p.536.
And now to the vexed question of which drugs should be used for type 2 diabetes, where nothing quite works out as it should. For example, the thiazolidinediones as a group have a favourable effect on HbA1c , on lipids and on measures of insulin resistance. Rosiglitazone has a greater effect on peroxisome proliferators activated receptors (PPARs) than pioglitazone and by rights it should be the better drug. Both cause an equal amount of peripheral oedema. But in fact pioglitazone causes less heart failure than rosiglitazone and a big Canadian cohort study has a better mortality record as well. Yet the accompanying editorial sounds a note of caution about the data we have at present, and warns us against too much enthusiasm for any of the newer incretin pathway drugs too. The fact is that in diabetic therapeutics, as in most of medicine, you just can’t tell what is going to happen until enough of it has happened.
Ann Intern Med 1 Sep 2009 Vol 151
297 A thought-provoking study (ATRIA) looks at the risks and benefits of warfarin in 13 559 adults with nonvalvular atrial fibrillation. We aren’t told the reasons why some of this cohort received warfarin and some didn’t. The event rate in those who didn’t was considerably lower than in the randomised trials which have led to the widespread adoption of oral anticoagulation in most patients with AF. When balanced against the risk of catastrophic cerebral haemorrhage, the net benefit of warfarin is accordingly less, and more randomised prospective studies may be needed, unless warfarin soon becomes obsolete.
306 The place to eat a Mediterranean diet is Naples. The dark alleys of the old city are studded with stalls selling wonderfully fresh fish and vegetables, and it’s quite difficult to eat carnivorously in this magical but frightening place. Fittingly, it is the source of a paper extolling the merits of the local diet in overweight patients with newly diagnosed type 2 diabetes, which performs better than a standard low-fat diet. Diabetics in Naples will also increase their life expectancy by avoiding arguments with the Comorra.
Fungus of the Week: Laccaria laccata
This is a very common little fungus found on damp soil in woods and known as The Deceiver because it looks similar to a number of other little fungi, with a dark brown cap when wet, and a buff coloured cap when dry. I think it is the fungus that featured in this splendid dish I ate in old Naples:
Soak some Italian brown beans overnight, discard the water and rinse them. Then cook them with an onion in salted water or a light meat stock until tender, and add the little mushrooms near the end, preferably having softened them first in a little oil or butter. About half an hour before the beans are done, drain off some of the broth and use it to simmer a fresh young octopus. Mind you remove its intestinal sac first. If you want a black dish, you could use its ink. But it looks nicer without, laid out in tentacled splendour on a pool of beans and fungi in their broth. The best accompanying wine is Greco di Tufa, from ancient vines on the nearby Falernian hills.
Page last edited: 06 October 2009