Arch Intern Med  8/22 Aug 2011  Vol 171
The order in which I place these journals does not reflect merit, but dates back to 1998 when I first made some experimental one-line notes on the journals as they arrived in the post. O happy days! There was something good about handling the paper copy, and the cover of JAMA was generally beautiful and interesting, and its contents were generally good too. But now JAMA lies pale and moribund, awaiting the kiss of Prince Charming or the electrodes of Baron Frankenstein. People assure me that Howard Bauchner, the newly appointed editor, is more likely to resemble the former. But while he performs CPR on what is supposed to be the AMA’s flagship journal, the lower-rated Archives are far superior and fizzing with life under the charge of Rita Redberg:

1322   I used to love new diagnostic chemicals. Fancy being able to take some blood and know with complete certainty that someone had coeliac disease (endomysial antibody) or myocardial infarction (troponins)! So each time a promising new test came up, I wanted to do some primary care studies. Procalcitonin appeared in the literature about 8 years ago and I immediately wanted to design a study with some Oxford colleagues, but the evidence of its value for discriminating between bacterial and non-bacterial infection was then too sparse. Others have since done primary care studies of lower respiratory infection and studies have also been performed in emergency departments and in intensive care units. They are usefully assessed in this systematic review: it seems that procalcitonin-guided antibiotic prescribing reduces antibiotic use without affecting mortality. There is plenty of scope for bigger, better studies in the context of acute primary care e.g. in out-of-hours centres. But I fear that this approach will never be widely adopted until the cost of a near-patient test approximates to the cost of a week’s course of amoxicillin.

1344   All the time I spend in the USA, I am trying to understand their health system and how its lessons might apply to ours, and vice versa. Tiring work. Throughout the UK, people of my age upwards are being sent little packages so they can test their stools for occult blood. I threw mine away. This paper describes what happened to 212 people over the age of 70 (no, I’m nowhere near that) who tested positive for faecal occult blood. Only 56% of these had a follow-on colonoscopy, and of these, 10% had serious adverse effects due to the investigation or due to treatment for cancer. The 43% (94 people) who weren’t followed up properly included 3 people who died from bowel cancer in the next 5 years and 43 who died of other causes. The greatest “burden” from this screening process seemed to fall on those with the lowest life expectancy. I don’t know what lessons this has for the NHS. It certainly shows that without a robust system of follow-up, you’ll miss some cancers. But it also confirms my feeling that FOB testing is a lousy way to screen for bowel cancer, likely to overwhelm diagnostic resources and possibly do more harm than good.

1363   Soy Isoflavones in the Prevention of Menopausal Bone Loss and Menopausal Symptoms: a Randomized Double-blind Trial. There are two schools of thought about RCTs: that we need to do lots more with increasing power to detect small effects; or that we already do too many and they are often far too big because if an effect is clinically important, it will be obvious with small numbers. I am firmly of the latter school. This RCT was not really needed, because we already know that soy products do nothing for menopausal symptoms or osteoporosis. Had it been run by a soy product manufacturer, maybe they would have recruited thousands of women in hundreds of centres and found some subgroup effects of borderline statistical significance. Then they could have lobbied the licensing agencies hard until they conceded the right for doctors to prescribe this stuff to half the female population over 50. But the soy product manufacturers sell cartloads of their junk over the counter already, and a small (n=248) one-centre government-funded RCT was perfectly adequate to test clinical significance. I don’t suppose it will make the slightest difference to the “phyto-oestrogen” market.

1371   We should all be grateful for the existence of nephrologists: doctors who can work night and day measuring people’s bicarbonate, creatinine, and electrolytes, providing dialysis, and offering ready advice to ignorant colleagues and well trained patients. In the UK, there are probably too few of them, and if we send them patients who aren’t likely to need dialysis soon, they send them back to us until their creatinine gets to 300 nmol/L or beyond. Of course, their professional societies say that we should be sending them patients sooner: but this is rot. This study shows that in the USA, patients are seeing nephrologists sooner and being put on dialysis earlier, but that 1-year survival after dialysis has not improved. The implication would seem to be that patients actually die sooner if their dialysis is started earlier. Nephrology, it seems, can be dangerous.

JAMA  10 Aug 2011  Vol 306
587    The Poetry and Medicine section of JAMA has a cult following: it’s perhaps the most reliable source of bad verse in the world – toe-curling, emetic, and hilarious by turns. It says a lot about the present state of JAMA that this week’s poem is about the best thing in the journal. It’s really quite good, though I can’t link you to it unless you have a JAMA subscription. The title is Number Needed to Treat, and I can give you its last line, because this at any rate can’t infringe copyright:
eeny meeny miney moe.

613   Some years ago I read a study which showed that sleep apnoea could be diagnosed as reliably by taking a history from the sleep partner (when available) as by all the usual sleep clinic technology. In this study of sleep-disordered breathing in older women they were taking no chances: “Channels included 2 central electroencephalograms, bilateral electrooculogram, chin electromyogram, thoracic and abdominal respiratory effort, airflow (using nasal-oral thermocouple and nasal pressure recording), finger pulse oximetry, electrocardiogram, body position, and bilateral piezoelectric sensors to detect leg movements.” What they demonstrate here is that only sleep-disordered breathing that causes hypoxia is associated with cognitive impairment in women. The same almost certainly applies to men.

NEJM  11 Aug 2011  Vol 365
Well into the second decade of the twentieth century, I guess it is time for me to bow to the inevitable and start reviewing papers as they appear on journal websites rather than in the printed journals.

The big excitement on the NEJM website this week is the ROCKET-AF trial showing that rivaroxaban is as good as warfarin in preventing stroke and peripheral embolism in older people with non-valvular atrial fibrillation. This was a well-conducted trial which went to considerable lengths to achieve double-blindedness despite the need to provide sham INR values for those on rivaroxaban and real ones for those on warfarin. The standard of reporting is good, except for various switches between intention-to-treat analysis and as-treated analysis, which are unnecessary and confusing. However you cut the data, rivaroxaban is at least as good as warfarin, when only 60% of patients are within the target INR range. That was the mean for the 45 countries that took part, though much better control can be achieved in an efficient system. Still, if I were a patient, I’d much rather have a fixed-dose oral drug than one which requires blood tests and dose changes: and this will certainly come one day soon. In the meantime, Johnson & Johnson and Bayer made sure they ran a good marketing trial, recruiting 12 patients on average from 1178 centres around the world.

493   A superb 9 country study establishes that early retroviral therapy prevents HIV transmission between infected and non-infected stable partners (97% heterosexual, 94% married; equal numbers of infected women and men). The logistics of this study compel wonder and its findings will change practice.

527   And now for the longest running quest in pharmacology – the search for a perfect laxative. My hunch is that it began in the Jurassic period, when carnivorous dinosaurs who got bunged up might have eaten fern shoots in imitation of their huge belching and farting vegetarian cousins. Certainly laxative-seeking behaviour is widespread among mammals – the bear wakes from hibernation to tear at birch bark and lick the sap which will release its anal plug; the costive cat eats grass. The earliest human records list rhubarb, liquorice, and other herbs and barks. Linaclotide may be the Holy Grail of this 100 million-year quest, though the two trials in adults with chronic constipation reported here are somewhat underwhelming: at best just 20% found success at the Stool of Easement.

Lancet  13 Aug 2011  Vol 378
The printed Lancet makes for rather lazy browsing this week – there’s a global survey of hepatitis B & C in IV drug users, a whole cavern of gene gnomes talking about the MTHFR genotype and folate metabolism, and a vaguely promising drug for Duchenne muscular dystrophy achieves favourable changes in muscle biopsies in 7 out of 19 subjects.

Richard Horton in Offline moves from Norway to Brazil, where at the end someone appears to have slipped some mescaline into his nightcap:

“If the health system is to solve its puzzles, there must be an awakening about our relations, one with another. The interdependence we share. Our communion of purpose. It will win elections.”

The Lancet website is chiefly notable for a meta-analysis on the effect of cigarette smoking on coronary heart disease in women compared with men. The evidence swings both ways but the most likely conclusion is that women are 25% more likely to suffer coronary events than men for any given level of smoking.

Then my attention was caught by an article that I nearly overlooked in the paper version:

621  Novel melatonin-based therapies: potential advances in the treatment of major depression. With SSRIs falling out of favour, we badly need a new class of antidepressants, especially if they help people relax and sleep better. Or do we? Are we perhaps just in for a further round of wishful, circular thinking about “major depression” and the brain biochemistry which underlies it, or results from it, or merely coexists with it. High quality medical journals have a responsibility to give us reliable and impartial information and guidance about such things. So how does this contribution shape up? Well, it goes through a range of melatonin derivatives and concludes that:
“Importantly, only agomelatine (which also binds 5-HT2C receptors) has been reported to have clinically significant antidepressant effects. Because of its favourable adverse effect and safety profile, and the potential to help to restore circadian function between depressive episodes, this drug might occupy a unique place in the management of some patients with severe depression and other major mood disorders.” Ahem, is there anything we should know about the authors possible conflicts of interest in relation to the manufacturer of agomelatine, Servier? There certainly is. It defies belief that a reputable journal should commission an ostensibly impartial review of an important area from two authors who have both received grants from this same drug manufacturer.

BMJ  13 Aug 2011  Vol 343
Conflicts of interest are the subject of a splendidly forthright editorial by Fiona Godlee – in this case relating to the workings of the US Food and Drug Administration and its even more pusillanimous transatlantic equivalent, the European Medicines Agency. Our guardians and watchdogs would like to loosen a system which already allows manufacturers to massage and withhold data at will. How many patients – indeed how many doctors – realize that half the trials on the drugs and devices they use are locked away in company files and have never been subject to independent scrutiny?

Putting genomics into practice: now there’s a title to make you sigh. I clicked on this editorial expecting to read the usual hyped-up waffle and special pleading: but this is just the opposite. It’s a beautifully realistic and well-written piece which even contains a useful alternative to the much-misused term “personalised medicine” : stratified medicine more accurately describes the sort which may one day be possible when we feed in the patient’s genome card into the computer and decide to prescribe them drug B instead of drug X. Personalised medicine will remain what it is now – talking to them about their mother’s death, telling them that they can eat what they like, sharing a joke.

Even the prescribing of drug B rather than X often lies beyond modern genomics, despite the discovery of plausible gene loci. Variants at the CYP2C19 locus should by rights determine the clinical efficacy of antiplatelet treatment with clopidogrel, but this systematic review shows that they don’t to any meaningful degree.

The management of febrile illness in adolescents and adults in resource-poor settings is clearly a topic of major global importance and it gets a very good Clinical Review here. The only problem is that doctors in resource-poor areas won’t be able to access it without a BMJ subscription. Surely this is an oversight that needs correction.

Plant of the Week: Franklinia alatamaha

Here is an account of this quintessentially American plant by Eric Larson, Keeper of the Botanical Gardens at Yale, written in August 2005:

Our Plant of the Week is Franklinia altamaha, or the Franklin Tree. The common name and genus name are for Ben Franklin, friend to the botanist who discovered the small flowering tree on the banks of Altamaha River in Georgia. John Bartram found this plant in 1770, during his travels through the southeastern part of what would become the United States. John’s son William also traveled extensively through that same area, only his journal writing was much more advanced,
which resulted in Bartram’s Travels, said to be “one of the most astounding verbal artifacts of the early republic.” I highly recommend the book to you.
 
This plant is no longer growing in the wild. In fact, since 1790, there have been no sightings of Franklinia in its habitat. John Bartram brought back a number of specimens, and it is from these that all of today’s plants have descended.

Franklinia belongs to the Tea family, Theaceae, which also includes Camellia and Stewartia. Franklinia grows to about twenty feet high, by fifteen feet across. The national champion is thirtyseven feet tall. The small stature makes it an ideal plant for the home garden. The fact that it sets big white fragrant blossoms in July, August, September and even into October doesn’t hurt its chances for inclusion in any landscape. Add in the fact that it often has spectacular fall foliage, and you have a real knock out for the front yard.

Best planted in spring as a balled-and-burlapped youngster (be sure, again, not to buy too large a specimen: harder to handle, more expensive, more transplant shock, etc.), in full sun or light shade in a moist, acid but well-drained soil. Think of a sandy bank on a river in Georgia to replicate for this plant. Ours is growing next to the spring that fuels our wetlands area. It is quite hardy here in New Haven, and available at the better local nurseries.

………………………………………………………………………………………………
Eric introduced us to this plant yesterday. It has grown beautifully in the last six years and bids fair to outclass the national champion.

I don’t have my copy of Bean’s Trees and Shrubs Hardy in the British Isles handy with me here, but my guess is that dear Bean will say something like “our climate does not appear to suit it, and it is seldom seen to flower even in our warmer counties. A specimen at Nymans reached 16ft 5in before it succumbed to the damp winter of 1935.” Here in New Haven the winters are long and freezing but the summers are hot and humid. If you have a sandy acid garden and a disposition of unquenchable patience and optimism, I could send you some seed from Yale. Or else you could buy a small plant from one of six British nurseries listed in the Plant Finder.