NEJM  29 Sep 2016  Vol 375
Creating a Zika vaccine
In the 15th century, sailors from Cadiz set sail under a captain from Genoa to find the western route from Europe to China. They landed in a New World, and proceeded to infect it with the rich mix of communicable diseases that Spanish sailors of the time were wont to carry. Having ravaged native populations with influenza, measles, and smallpox, they came back and infected Europe with syphilis. Now every major city in the world has an airport that resembles a fleet of ships under Captain Columbus taking off every few minutes. They carry diseased sailors from around the world with amazing speed and efficiency. It is a wonder that any of us are alive. The main reasons are that most viruses are not interested in killing people, and that if they are, we are good at developing vaccines against them. Zika is a boring little flavivirus that emerged from a forest in Africa a few years ago and (like rubella) would be of no account but for its capacity to cause birth defects. Flaviviruses are quite easy targets for vaccination: they lack the cunning of influenza or HIV, for example. It is always interesting—for me anyway—to read about how vaccines get developed, and this week’s NEJM gives you two chances to in a couple of short articles.

Big data & machine learning
I am in the US for a couple of weeks, among friends and colleagues I love and admire. Their health economy is in total meltdown: life expectancy is among the lowest in the developed world, maternal deaths are actually rising, their regulators are toothless, and their drug prices are fixed at colossal levels. But this is the land of opportunity, so anything that can be done must be done, and must be monetised. For a good example, read this relentlessly optimistic piece called “Predicting the Future—Big Data, Machine Learning, and Clinical Medicine.” It makes some good points, but I fear for the population of the US while their medical overlords carry out this vast, aggressive natural experiment. I should explain that “aggressive” is generally used as a term of approval over here.

Pointless ICDs
Aggressive treatment of heart failure usually ends in the implantation of a cardioverter-defibrillator. These devices have a reasonable track record of preventing sudden death in people with heart failure due to ischaemic heart disease, but indication creep means that they are widely implanted in all sorts of people with heart failure of any aetiology. The Danish Study to Assess the Efficacy of ICDs in Patients with Non-ischaemic Systolic Heart Failure on Mortality (DANISH) was a multicentre, randomised, unblinded, controlled trial that was conducted at all centres in Denmark at which ICDs were implanted.

Remember that ICDs are unpleasant devices: when they do not extend life by preventing lethal arrhythmia, they can turn the process of dying into a horrific mixture of pulmonary overload interspersed with electric shocks. And the DANISH trial showed that they have no life prolonging value in people whose heart failure is not caused by ischaemia.

Azithromycin for every caesarean section?
Now here’s a trial to annoy antibiotic “stewards” everywhere. It demonstrates that a single intravenous dose of azithromycin prior to caesarean section reduces the rate of post-operative wound infection from 6.6% to 2.4%. It’s even better than that: there were also significant differences between the azithromycin group and the placebo group in rates of endometritis (3.8% vs. 6.1%), and serious maternal adverse events (1.5% vs. 2.9%). So do you (a) offer azithromycin to all women undergoing caesarean section, (b) don’t offer it to any of them, or (c) try and persuade them not to have it in the wider interests of restricting hospital antibiotic use? I’d suggest there is only one correct answer, and it will be interesting to see how practice changes.

JAMA 27 Sep 2016 Vol 316
Endovascular thrombectomy for stroke
“In this individual patient data meta-analysis of 5 randomized clinical trials of patients with large-vessel ischemic stroke, earlier treatment with endovascular thrombectomy + medical therapy compared with medical therapy alone was associated with lower degrees of disability at 3 months. Benefit was greatest with time from symptom onset to arterial puncture for thrombectomy of under 2 hours and became nonsignificant after 7.3 hours.” I very much like the idea of individual participant meta-analysis, provided that somebody else does it. Or even better, if two separate teams do it, so you can compare their conclusions. Anyway, I’m prepared to believe this one, because the message is so consistent across prespecified subgroups: time equals brain, and this technique saves brain the sooner you use it.
This may mean a further rearrangement of stroke services, as discussed in the commentary, but I suspect that when it comes to themselves, most doctors won’t greatly care. We all want to die suddenly and painlessly, but know that this blessing is granted to few. So we want our stroke to be fatal or trivial: anything in between is awful, regardless of the exact Rankin scale. Personally, I would refuse any reperfusion therapy or anything that prolonged life afterwards.

LDL-C & cardiovascular events
I am agnostic about low density lipid cholesterol (LDL-C). I am sure it is very good at some level and very bad at another. I am sure that statins lower LDL-C and reduce cardiovascular events. This meta-regression analysis shows that, overall, you can make a case that LDL-C lowering by various drugs is beneficial, but with wide between agent differences and at numerical levels of total risk reduction that are irrelevant to almost everybody. I am also sure that hyping the LDL-C hypothesis has made billions of dollars for drug companies and some of this has trickled down to academic units conducting trials of LDL-C reducing agents. But as a surrogate marker to guide the future of CVD prevention, I’m deeply unconvinced.

JAMA Intern Med Sep 2016
Vomiting a good sign in early pregnancy
Here is the retrospective analysis of a trial of aspirin in women with one or two prior pregnancy losses who were enrolled at four US clinical centres from 15 June 2007 to 15 July 2011. For such women, getting sick at the start of pregnancy is a good sign. “Nausea (HR, 0.44; 95% CI, 0.26-0.74) and nausea with vomiting (HR, 0.20; 95% CI, 0.09-0.44) were associated with a reduced risk for clinical pregnancy loss.” Note those hazard ratios: it’s almost as if the worse your sickness, the better the chances of keeping your pregnancy.

Lancet  1 Oct 2016  Vol 388
Early mobilisation after major surgery
It’s so easy. Just get your post-operative patients moving as soon as possible, and you will shorten their stay on the surgical intensive care unit and improve their functional mobility on discharge. The second sentence is true. But the first one is not: to mobilise a patient ill enough to be on the SICU, you need to help them out of bed, ensure that their gown does not flap about and reveal their fundamental parts, disentangle the drip lead from sundry electrical monitors, ensure that the catheter isn’t pulled out and that the dressings remain in place, help them put on the ward slippers, and then prepare to hold their arm with one of yours while you wheel the machinery and the drip stand slowly out into the obstacle course of the ward. This way you will achieve four minutes of walking for every 30 minutes of assistance. Believe me, I was there recently, as the assistant. Had I not been there, nothing would have happened. This “simple” intervention requires more nurse time than any NHS hospital will ever be able to provide unless we suddenly turn civilised and pay for it.

Lens extraction for primary angle-closure glaucoma
“In light inaccessible hid from our eyes” is the hymn’s description of God, which is also a fair description of how well most doctors understand ophthalmology. When I had a procedure earlier this year, the light inaccessible filled my eye and I woozed with midazolam for a few minutes while painless things happened and then I walked out with a patch and some drops. This is how it should be. Eyes are just globes filled with jelly and Greek names. Let other people understand them. Here, if you wish to join with such people, is the report of an MRC trial showing that clear-lens extraction showed greater efficacy and was more cost effective than laser peripheral iridotomy, and should be considered as an option for firstline treatment for primary angle-closure glaucoma.

The BMJ  1 Oct 2016 Vol 354
Rapid Recommendations
This week The BMJ brought me joy. It takes the form of a paper under the heading of Rapid Recommendations, which is called “Transcatheter or surgical aortic valve replacement for patients with severe, symptomatic, aortic stenosis at low to intermediate surgical risk: a clinical practice guideline.” Now this seems a very unlikely source of joy on several counts. I don’t really have much interest in TAVI, though I’m glad that it sounds like a good option for people with aortic stenosis. I have been attacking guidelines for their clunkiness and lack of context ever since they first started appearing in the early 1990s. To me “recommendations” are an emergent phenomenon when clinicians know enough and have a proper informed dialogue with patients over as long a time as is necessary to understand their preferences and goals. But here is something that is much better than its label—it is the very thing that I’ve been hoping would appear some day, preferably in my lifetime. They have taken the data from a definitive trial and made it into an easily used interactive tool, which can be used for everybody making a decision about this critical choice of procedure.
The great thing is that this is just a beginning. One-off surgical decisions informed by good randomised controlled trial evidence are unfortunately not what most of medicine consists of. But this way of presenting trial data will help us go much further. I have seen the future, and it works.

NSAIDs & the risk of heart failure
I don’t really understand the kidney (my admitting this is what distinguishes me from most nephrologists), but it seems to me that if you combine a diuretic, an angiotensin pathway inhibitor, and a non-steroidal anti-inflammatory drug you shut down just about all its autoregulatory mechanisms. Well and good if you know what you’re doing, perhaps, though personally I wouldn’t go there. The trouble is that patients can go there without you knowing: a lot of NSAIDs are available over the counter. Actually this paper is not about the kidney at all, though somebody needs to write a similar one about NSAIDs and that organ. It is about the heart, which keeps in close contact with the kidneys, and does not like it if there is too much fluid around for it to pump. Here is a paper that ranks the risk of NSAIDs based on observational evidence of admissions for heart failure. The pecking order is not to be taken too literally, but this is definitely a warning against loose prescribing of NSAIDs, particularly in older, breathless, renally impaired patients.

Plant of the Week: Acer saccharum
As we’re in New England in October, it is time again to celebrate the sugar maple, a grand long lived tree, which produces not only maple syrup but also the most glorious of the “fall colors.”

Let’s forget about maple syrup. It is an insipid sugary substance produced by endless boiling down of sap collected on certain days in March. If you like sugar, it’s best to look for another source.

Fall colors are another matter. Mountains covered with tall trees in a blaze of orange and yellow really are a sight to travel for. These maples often grow in large natural stands, and their ability to pull up water from deep places encourages other brilliantly coloured undershrubs, such as sumachs. In a week’s time we will move north a bit and join the “leaf-peepers” who yearly enjoy the forest spectacles of Massachusetts, New Hampshire, and Vermont.