Richard Lehman’s journal review—11 April 2016

richard_lehmanNEJM 7 April 2016 Vol 374
Prebirth steroids and baby lungs
1311 Most of you will be familiar with the logo of the Cochrane Collaboration, consisting of a blue circle with a vertical line crossed by some bars with a diamond shape at the bottom. This is the forest plot of Iain Chalmers et al’s meta-analysis of the trials of prenatal corticosteroids to improve outcomes in very premature infants. All of it was groundbreaking at the time: the idea of meta-analysis, this way of visualising its results, and the fact that giving steroids to mothers could save thousands of tiny babies. These methods are still basic to evidence based medicine, but so is the principle that a single well conducted and adequately powered randomised trial can give you a reliable answer in one go. Here is one such, and again it’s giving mothers a corticosteroid to prevent respiratory distress in premature babies. But this time the babies were further on in gestation: 34-37 weeks, and the steroid given was parenteral betamethasone. The effect was modest—a 2.8% absolute reduction in the composite outcome of severe events. But a definite benefit from a cheap and harm free intervention.

All the fruit in China
1332 I regard the civilisation of China with distant awe and wonder and eat its food with pleasure, but I have never eaten any distinctively Chinese fruit except for the lychee and the “Chinese gooseberry,” which changed nationality and became the Kiwi fruit three decades ago. I have to say that if these were the only fruits, I would probably eat as little fresh fruit as the average person in China, where only 18% of people eat any on a daily basis. Such highly atypical people are described in a big epidemiological survey overseen by luminaries from Oxford. The adjusted hazard ratios for daily consumption versus non-consumption were 0.60 for cardiovascular death, and 0.66, 0.75, and 0.64, respectively, for incident major coronary events, ischaemic stroke, and haemorrhagic stroke. There was a strong log-linear dose-response relationship between the incidence of each outcome and the amount of fresh fruit consumed. Maybe the best way to improve the cardiovascular health of the Chinese people would be to plant orchards and vineyards on every plain and every mountain that rises through the mists of the Middle Kingdom. I’d be happy to volunteer.

HOPE-3 for consensus on statins?
OL There is so much on the website of the NEJM at the moment that I don’t know where to begin. Normally I try to take things in chronological order, to avoid confusing myself, but a few people have said they’re anxious to hear what I have to say about the HOPE-3 papers. It can be said very quickly. I am very contented by the outcomes of this trial, because they are consistent with the body of previous evidence and provide a firm basis for decision making with individuals. I am not a controversialist by nature, and I like it when science is well conducted, straightforward, and applicable to real life. I save my grumbles for when it isn’t, which, in the medical journals, is most of the time. You can argue about the detail, but HOPE-3 had the right design and the right power and the right population. The people enrolled were a typical mix and defined simply by having an intermediate level of aggregate cardiovascular risk. When given rosuvastatin 10mg, everyone’s CV risk was lowered by 25%, the same ballpark figure as in all the statin trials. Muscular adverse effects, at 5.8% (after a month’s trial period to exclude immediate problems), were just 1.1% above placebo. So I cannot see any reason why statins should not be offered to anyone who wants to take them, based on their informed personal preference. I hope some way can be found of putting the decision making process outside the medical arena altogether, as doctors have more pressing things to do than having half hour conversations with the entire healthy adult population. We just need good infographics, some trained advisers, and a respect for the wide range of individual responses to risk.

HOPE-3 & the BP threshold
OL With blood pressure (BP) lowering in the same intermediate risk group, the result was quite different. There was a definite threshold effect, and it depended entirely on the starting level of systolic blood pressure. If that was above 143.5, then lowering it by the use of candesartan and hydrochlorthiazide showed a reduction in cardiovascular events. If it started below, then no benefit was found. So while statins reduce CV risk for everyone, BP lowering only benefits those whose systolic is above about 144. Again, this is entirely consistent with conservative estimates in meta-analyses of previous trials, though some of the most recent have claimed evidence for a lower threshold. I don’t think the evidence about ballpark figures is going to change significantly in the future. I HOPE we have crossed a threshold here. The medieval theologians of hypertension can argue about how many milligrams of candesartan can dance on the head of a pin, or which thiazide diuretic will purchase how many years of remission from purgatory. But the rest of the world should concentrate on offering real people with elevated BP a choice of BP lowering according to which drugs suit them best and what the approximate long term benefit is going to be. With a statin thrown in for those who want 25% extra.

JAMA 5 April 2016 Vol 315
Beautiful Berwick
1329 I’m told that members of the RCP can access a talk that Donald Berwick gave to them about the NHS, and that it reduces most hearers to tears. Since I literally shed tears about the NHS every day anyway (it’s that foreign blood), I’m quite glad I can’t access it. The mere fact of Berwick has a powerful effect on me. Here is a man who has fought tirelessly throughout his life for better, more humane standards in medicine. Because his ideas are supremely intelligent and offer no mechanical, blame culture, checklist solutions, he can never fit into the political gaming that dominates the power discourse on either side of the Atlantic. And yet he just keeps fighting, with ever more blazing conviction. Here he lays out nine principles for Era 3 of medicine. Principle 1: Reduce mandatory measurement. Principle 9: Reject greed. Simply wonderful.

Bariatric blessings last
1362 The “objective” benefits of bariatric surgery, such as remission of diabetes and increase in life expectancy, tend to dominate academic discussion, while for patients the immediate subjective benefits (or problems) are all important. There is no doubt that these procedures are life changing, and overwhelmingly for the better. In an observational cohort study at 10 US hospitals, clinically meaningful improvements were shown in 57.6% of participants for bodily pain, 76.5% for physical function, and 59.5% for walk time at one year. Additionally, among participants with severe knee pain or disability (633), or hip pain or disability (500) at baseline, over three quarters experienced joint specific improvements in knee pain and in hip function. Over the next two years, there was a slight decrease in the effect on pain and function, but not on walk time and joint symptoms.

Anns Intern Med 5 April 2016 Vol 164
Pred for gout
464 In case any of you didn’t know, a short course of steroids relieves gout very effectively. In this simple head-on comparison with indometacin, patients were selected by clinical criteria for the diagnosis of acute gout, and as it was conducted in the former British colony of Hong Kong, the steroid used was prednisolone, rather than prednisone, which features in all US studies. Fortunately, the two drugs do much the same at the same doses. This may be the first adequate trial, but I used prednisolone in practice for gout over two decades in patients who could not take NSAIDs, mostly because they had heart failure and could not tolerate colchicine. Safety trials in this population would be welcome.

Lancet 9 April 2016 Vol 387
The world of diabetes
1513 Here’s a great big survey of how type 2 diabetes is taking over the world, with a link to a lovely set of infographics created by the NCD Risk Factor Collaboration (NCD-RisC) and paid for (like this open access paper) by the Wellcome Trust. Having a fasting blood glucose of seven or above does not constitute a non-communicable disease in my book, but as it rises it becomes a risk factor for other diseases and, to the extent that it causes direct symptoms at higher levels of sugar, becomes a disease in itself. How it plays out now, and in the future, and in different groups, are the big questions. In the richest societies, the graphs of crude prevalence have levelled off or are showing a slight decline. In Central South America, the Middle East, and North Africa—and especially in Melanesia, Polynesia, and Micronesia—rises continue to be pretty spectacular. I suppose the main explanation must lie in the glycaemic index and the amount of food eaten in relation to levels of physical activity and genetic heritage. Glucose lowering drugs will never play much of a part in addressing the global problem. Lifestyle changes may, but I’m not sure. Some sort of rebalancing process is taking place in Western Europe, which isn’t fully explained by any of these things.

Speed: the answer for coke & heroin co-dependency?
OL From the Netherlands comes a slightly baffling trial for a niche group of extremophile addicts: those who take heroin and cocaine and don’t respond to attempts at methadone substitution. Half of them were randomised to get sustained-release dexamfetamine and people in this group were assessed to have used cocaine on about 20% fewer days than those given placebo. Is this a good thing? Will it help such people and in what way?

The walk friendly city
OL Walking deficiency disorder is Muir Gray’s pet target at the moment. I may be perverse in believing that country life is the great enemy of walking. You only have the same paths to go on and they are generally muddy; you need a lot of motivation to make walking part of your life. In a city like London, you routinely have to travel miles on foot unless you go everywhere by taxi. Cities define civilization and the country defines peasants. But the International Physical activity and Environment Network (IPEN) finds that not all cities are equal. Across 14 cities on five continents, the difference in physical activity between participants living in the most and least activity friendly neighbourhoods ranged from 68 min/week to 89 min/week, which represents 45–59% of the 150 min/week recommended by guidelines.

The BMJ 9 April 2016 Vol 353
Cardiac arrest and adrenaline
Although an in-hospital cardiac arrest always gives medics an adrenaline hit, it is best to spare the patients. An analysis of data from the Get With The Guidelines-Resuscitation registry, which includes data from more than 300 hospitals in the US, found that half of patients with a persistent shockable rhythm received epinephrine (adrenaline) within two minutes after the first defibrillation, contrary to current American Heart Association guidelines. After matching by propensity score, this seems to be distinctly harmful both in terms of survival to discharge and good functional outcome.

Can you spread the NNT like jam on toast?
How to translate clinical trial results into gain in healthy life expectancy for individual patients” is a very bold title for a paper. You simply can’t do this for a catastrophic event, such as fatal myocardial infarction or stroke, because they either happen or they don’t. For a brilliant and innovative discussion of this in relation to statins, see the work of Darrel Francis’s team. The authors here argue that you can do it in relation to more continuous outcomes, for example, in treatment to prevent diabetic nephropathy, some types of malignancies, osteoporosis, and (they claim) atherosclerosis. In these examples, they suggest, the aim of treatment is not to prevent but to delay the occurrence of symptomatic disease. But I am not convinced. Most of the outcomes remain binary in real life, and the vast majority of those treated in the lottery of life derive no benefit whatsoever. Cates plots and other infographics make this clear. You cannot spread the number needed to treat (NNT) across everybody like jam on toast.

Plant of the Week: Salix babylonica “Pendula”

This is the time of year when journeys, especially railway journeys through the countryside, are gladdened by the sight of willow trees already in leaf. Most of them are pendulous in habit and therefore called weeping willows, and they often grow by streams and rivers. Their pale fresh foliage makes them stand out as great phantoms of greenish gold even amid the gloom and rain of the English spring.

These native willows bear various botanical names and have probably hybridised to the point that Ophelia would be hard put to sing “O willow, willow, willow” to any of them. Although they are really quite joyous and beautiful trees, their association with gloom has deep roots.

Salix babylonica comes in a weeping form but this cannot have been the kind that grew by the rivers of Babylon in the famous Psalm 137, Super flumina babylonis. “By the rivers of Babylon, there we sat down, yea, we wept, when we remembered Zion. We hanged our harps upon the willows in the midst thereof.” As John Launer pointed out to me long ago, had these been weeping willows, the harps would have slid right into the Euphrates. Nor indeed are these trees certain to have come from the genus salix, though the true, non-pendulous Salix babylonica probably did grow in Babylon at the time of the Exile. The trees would have originated in China and been brought there along the silk road to be cultivated by Sumerians, who were the first great plant collectors.

Don’t grow these big willows in your garden as they take up far too much space and suck in water till the foundations of your house start cracking. Instead, collect the various sublime settings of the Super flumina psalm by composers from Gombert and da Monte through Lassus and Palestrina, and also the great organ partita on An Wasserflüssen Babylon by Johann Reincken, and the preludes by his pupil Johann Sebastian Bach. There is even a Monteverdesque setting of the Hebrew text בָּבֶל נַהֲרוֹת עַל, (Al naharot Bavel) by the Venetian Jewish composer Salamone Rossi (1570–1630), for four voices. This is the original ghetto blaster, ending “O daughter of Babylon, who art to be destroyed; happy is he, that rewardeth thee as thou hast served us. Happy shall he be, that taketh and dasheth thy little ones against the stones.”