Richard Lehman’s journal review—29 March 2016

richard_lehmanNEJM 24 Mar 2016 Vol 374
Flinty problem, leaden response
1101 John Snow, the arch-hero of epidemiology, died in 1858 a disappointed man. It was only after he had died that there was a proper inquiry into the cholera outbreaks that he had mapped, and during the interval the water company denied all possibility of contamination. This was in high Dickensian London, 160 years ago. Never again? The mass poisoning of an entire city of over 100 000 people (Flint, Michigan) by lead in the water supply is still happening in 2016. It is hard to believe. “We have the knowledge required to redress this social crime. We know where the lead is, how people are exposed, and how it damages health. What we lack is the political will to do what should be done.”

Incretin drugs & heart failure
1145 So called type 2 diabetes is a vascular risk state signalled by elevated levels of blood glucose. Everybody rushes in to fix the elevated blood glucose but this is not the key to reducing vascular risk. Two whole classes of sugar lowering drugs are given to people with T2DM without our knowing what they do to cardiovascular risk. And let’s not forget that there can be within class differences too. So I’m cautious in my welcome for this enormous case-control study, which lumps together all the dipeptidyl peptidase 4 (DPP-4) inhibitors and glucagon-like peptide 1 (GLP-1) analogues. The cohorts included a total of 1 499 650 patients from Canadian, American, and British databases, with 29 741 patients hospitalised for heart failure during 3 242 291 person years of follow-up (crude incidence rate, 9.2 events per 1000 persons per year). Overall, the message is that these drug classes do not increase hospital admissions for heart failure. But there may be devils in the detail that would take time and patience to discover, and I will leave that to others who possess these qualities in greater supply.

Zika seekers
OL The best place to keep up with the Zika story is in the pages of the NEJM, and a few days ago I heard its editor suggesting that Zika research could be a good place to test new systems of real time data sharing. For once, I agree with him, though the distinguished audience was lukewarm. It’s pathetic that we can share data instantly about supernova explosions that happened 1.3 billion years ago, but not about viruses that currently threaten human reproduction across continents. There are a number of pieces about Zika freely available on the NEJM website, including a plea for better partnerships.

Cutting the cutting with PET-CT
OL Surgery for head and neck cancer is often mutilating and often unsuccessful. A British trial randomised patients with these cancers (mostly caused by human papillomavirus) to receive surgery as planned by usual criteria or surgery as indicated by positron-emission tomography–computed tomography following chemoradiotherapy. There was dramatically less surgery in the latter group: 54 instead of 221. And “survival was similar among patients who underwent PET-CT–guided surveillance and those who underwent planned neck dissection, but surveillance resulted in considerably fewer operations and it was more cost effective.”

JAMA 22/29 Mar 2016 Vol 315
Backing mindfulness
1240 There are all sorts of things that might make us into better people by helping us to forget our daily pains: Freudian analysis, monasticism, transcendental meditation, cognitive behavioural therapy (CBT), mindfulness, following Donald Trump. The list is lengthy, but only a few are thought worthy of scientific enquiry. This trial compares usual care with either CBT or mindfulness plus yoga. Both the latter seemed better than usual care. Oh, and the condition was chronic low back pain and the outcome measure was pain bothersomeness. We will all need a lot of these therapies if Donald Trump gets elected.

This little piggy ate antibiotics?
1258 Little piggies grown commercially are fed antibiotics to change their bowel organisms in ways that promote rapid growth (“microbiome perturbation”). Does the same thing happen to little humans who get short courses of antibiotics in the first 6 months of life? A North American retrospective cohort study says not. And the editors of JAMA even let a bit of plain English slip into the conclusion:
“Exposure to antibiotics within the first 6 months of life compared with no exposure was not associated with a statistically significant difference in weight gain through age 7 years. There are many reasons to limit antibiotic exposure in young, healthy children, but weight gain is likely not one of them.”

Gold mines with too few miners
1283 I’d better declare an interest here: I still do paid work for the Yale Open Data Access (YODA) project. It was one of the greatest fortunes of my life to be at Yale when the project started in 2011 on the initiative of Harlan Krumholz. Access to individual participant data from industry funded trials had been advocated for over 20 years by Iain Chalmers and others, but those who believed it would happen had come to seem like a bunch of deluded idealists. I thought it would be helpful to set up a Google group of 90 or so like minded people called PATH (Publish All Trials in Humans) in October that year, and members of the YODA team wrote a number of articles arguing the ethical and practical case for data sharing. Ben Goldacre was writing his key book on the subject at the time, and shortly after that appeared the following year, AllTrials was launched as a single issue public campaign with enormous success in the UK.
Lots of pharma companies decided to share their data on platforms such as, YODA, and the Supporting Open Access for Researchers (SOAR) initiative. So the gold is out there for anyone who can demonstrate a plausible scientific question and a team with basic analytical skills. The trouble is that this endeavour still lies at the periphery of traditional funding and career streams. Like the CSDR, YODA gets an increasing number of excellent applications, but to an outside observer they still seem few in relation to the effort, as this research letter describes. Still, the very fact of their existence signals a massive culture change, celebrated by the editorial team of JAMA in a resonant editorial with the title “Data Sharing: An Ethical and Scientific Imperative.” 
How strange that its references do not include the key paper, published in JAMA exactly 26 years ago: “Underreporting Research Is Scientific Misconduct” by Iain Chalmers FRCOG, March 9, 1990. And although the journal’s owners, the American Medical Association, have just joined AllTrials, I’m not sure that this is yet true of the journal itself.

JAMA Intern Med Mar 2016 Vol 176
Don’t keep taking the medicine
OL “The desire to take medicine is perhaps the greatest feature which distinguishes man from animals,” said William Osler, who also pronounced that “One of the first duties of the physician is to educate the masses not to take medicine.” He was working in Baltimore at a time when American patent medicines contained cocaine, heroin, mercury, strychnine, and many other exotic ingredients, and were sold quite as vigorously as they are today. A survey of older Americans 125 years later reveals that 88% are taking medicines of some sort, and that concurrent use of at least five prescription medications increased from 30.6% to 35.8% between 2006 and 2011. While the use of over the counter medications declined from 44.4% to 37.9%, the use of dietary supplements increased from 51.8% to 63.7%. That’s in people of mean age 71! About 15% of the sample of 2200+ participants were considered at risk of harmful drug-drug interactions.

So it’s time to repeat the Golden Rules of Andrew Herxheimer, the wise and much loved physician who died aged 90 last month:
1. Think what you could do instead of using a medicine.
2. Unless you have a special reason, avoid new medicines. Stick to those about which a lot is known from many sources and which have been used for over 10 years; bad news about a drug often takes years to emerge.
3. Before deciding to use a medicine be clear whether it is to relieve a symptom, to cure a disease, to remedy some deficiency, or to prevent something. It doesn’t make any sense at all to prevent something in the future if it’s going to cause you some problem now.
4. Ask a doctor or pharmacist you trust, someone who understands it a bit better than you do, how well the medicine works, what problems people have had with it, and what happened.
5. If you have to take medicines, get to know as much as you can about those that help you.
6. Everybody is different and you must learn how your own body reacts to medicines.
7. Keep a diary of your experiences with a medicine: why you took it, how much for how long, what happened and when, how well it worked, and anything you didn’t like.
8. If something bad happens that you suspect may have been caused by a medicine, report it on a yellow card; ask a doctor, pharmacist, or nurse to help you do that or to do it for you.
9. When you have a problem about an adverse reaction or something difficult to discuss with your doctor, take someone with you to the consultation, because four ears are better than two; there are too many things to think about and an independent opinion is well worth having.
We’re indebted to Jeff Aronson for gathering and reporting these in his BMJ blog.

Lancet 26 Mar 2016 Vol 387
Two shouts for a shunt
1290 This week’s Lancet contains two phase 1 trials of implanted left-to-right interatrial shunts in small groups of people with heart failure. The first one was funded by a company called V-wave, and selected 10 patients to be fitted with the device in a single Canadian cardiology unit. Their mean age was 62, and nine of them were male. They had NYHA grade 3-4 heart failure on optimal drug therapy and an ejection fraction below 40%. The shunt is introduced via the femoral vein and pushed through a hole in the interatrial septum created by radiofrequency ablation under general anaesthesia. The three leaflets of the one-way valve once lived in a pig, and the frame is made of nitinol. Of the 10 subjects, eight experienced a reduction in breathlessness, one did not, and another died following intractable ventricular tachycardia. In the 20+ years I’ve been following heart failure interventions, there have been quite a few promising phase 1 trials that led nowhere. This procedure now needs to be shunted off for bigger, longer trials to be done and reported in a few years’ time.

1298 Another device manufacturer, Corvia Medical Inc, ran a much larger phase 1 trial on a similar device, this time for people with heart failure and preserved ejection fraction (HFpEF). Sixty eight patients were recruited at 21 centres around the world, and interatrial valve placement was successful in all but four of them: there were no major complications over six months. But it’s far from clear whether the recipients noticed any benefit. The main effect appears to have been on pulmonary capillary wedge pressure at rest and during exercise, which is not something most of us are aware of. Even among other surrogate measures, such as BNP, there was little evidence of any benefit. For this group of patients, I think this procedure should be shunted into a sidings and perhaps just left there.

The BMJ 26 Mar 2016 Vol 352
SSRIs to gladden the heart?
“In the first year of follow-up, patients treated with selective serotonin reuptake inhibitors had a significantly reduced risk of myocardial infarction (adjusted hazard ratio 0.58, 95% confidence interval 0.42 to 0.79) compared with no use of antidepressants.” Wow. This definitely warrants an RCT of SSRIs following MI: if it translates to that context, it’s a bigger effect size than beta-blockers, statins, or aspirin. But these weren’t people who had an MI, and the effect may not be sustained over five years. In this observational study of people aged 21 to 65 newly prescribed antidepressants on the QResearch primary care database, SSRIs get a very clean bill of cardiovascular health: they were not associated with stroke, TIAs, or arrhythmias; and fluoxetine in particular may protect against MI. This could be rapidly checked by using other databases, especially the Taiwan whole-population dataset.

Plant of the Week: Magnolia denudata

“Native of China; introduced in 1789. One of the most beautiful and distinctive of all flowering trees, this magnolia is, unfortunately, an occasional victim to the inclemency of the English spring.” Thus the admirable WH Bean.

Bean also describes it as a “rather low, rounded tree,” but by amazing good fortune ours has decided to adopt an upright, columnar habit and is growing skyward at the rate of a metre a year. The vagaries of the English spring do indeed sometimes ruin its astonishingly abundant, pure white chalices of flower. Yet so far this year they have withstood the wind and the rain: it is air frost that they really can’t abide.

I think it’s rather a pity that M Soulange-Bodin had so much success in hybridising this chaste looking beauty with Magnolia liliiflora and so producing all those browny-pink flowering trees we see around our towns at this time of year. The Chinese had more sense when they planted the pure Yulan around their temples and in the Imperial Gardens from 600 AD onwards. This great ancestor—one of the first of all flowering trees to emerge 95 million years ago—bids us share the reverence that the Tang emperors accorded it. None of its children quite equal it, except perhaps for an enchanting, fragrant cream cultivar called “Yellow River.”