NEJM  10 Oct 2013  Vol 369
1395   It’s been known for at least four thousand years that the heart has two ventricles, but what they actually did was a source of confusion until William Harvey began to sort things out in the seventeenth century. In the twenty-first century, cardiologists remain obsessed with the left ventricle and usually define “heart failure” by the proportion of its contents expelled in each contraction—the left systolic ejection fraction. In this trial, participants were selected for having an EF of 35% or less, normal QRS duration on ECG, and some echographic evidence that their right ventricle did not beat in exact synchrony with the left. These people were also considered suitable for an implantable cardioverter-defibrillator, so they were fitted with one of these plus a biventricular pacemaker, but in half the subjects, this was switched off. By the time the trial was stopped for futility, 809 patients had been recruited and followed up for a mean of 19.4 months. “In patients with systolic heart failure and a QRS duration of less than 130 msec, cardiac resynchronization therapy does not reduce the rate of death or hospitalisation for heart failure and may increase mortality.” So for selecting patients to have CRT, measuring the QRS interval on the ECG beats echocardiograpy.

1406   A trial of edoxaban versus warfarin for the treatment of symptomatic venous thromboembolism. OK, children, what is the first question we must ask? Yes, time within the INR target range for the warfarin group. Here it was 63.5%. Is that good or bad? It’s probably the current UK average, as far as I can discover. The authors describe this as “high-quality standard therapy:” compare it with your own figures, if you have them, before considering whether to use this latest fixed-dose factor Xa inhibitor. The new drug was as good for VTE as this degree of warfarin control, and caused less bleeding in this trial. For cognoscenti of natriuretic hormones, it’s also interesting that the Japanese investigators used NT-pro-BNP as a measure of right ventricular strain in pulmonary embolism. Nice to see this, 20 years after it was first proposed.

1434    One hundred years ago, the maverick Boston surgeon Ernest Codman caused panic among his Massachusetts General Hospital colleagues when he proposed that each of them should keep an End Result card for every operation they carried out, and analyse which of their bad outcomes were the result of their own failings or the condition itself. This amazing study from Michigan carries this idea a stage further, and the fact that it could be done at all is a great credit to the 20 surgeons who took part. They each submitted a video of themselves carrying out a laparoscopic gastric bypass procedure, and were then marked on their technical ability. I call this study amazing because they were marked by a panel of patients, not by fellow surgeons. And the results show that patients are very good at spotting the good versus the less good operators: “Mean summary ratings of technical skill ranged from 2.6 to 4.8 across the 20 surgeons. The bottom quartile of surgical skill, as compared with the top quartile, was associated with higher complication rates (14.5% vs. 5.2%, P<0.001) and higher mortality (0.26% vs. 0.05%, P=0.01).”

1443   “Current Concepts: Diagnostic Criteria for Nonviable Pregnancy Early in the First Trimester” looks like an important paper, and it probably is; but I had a hard time reading it. I am keenly interested in this topic, because most of my clinical work is done at weekends in an acute primary care centre based in a small hospital without ultrasonography except during weekday office hours. So if a woman comes in six weeks pregnant with pain and/or bleeding on a Saturday morning, I have to tell her that she will have to wait two days to find out if the pregnancy is intrauterine and viable. I hoped that this paper would give me some guidance on whether this is good practice, but instead it proposes new serum HCG and ultrasound criteria for viability and for non-uterine pregnancy which are of limited usefulness to me in primary care. They may, however, stop gynaecologists evacuating viable pregnancies or giving methotrexate to the wrong women.

JAMA 9 Oct 2013  Vol 310
1462   Simple retrospective cohort studies often make nice easy reading, but I’m not sure if there is a great deal for generalists to take away from this one, which describes the risk of major adverse cardiac events following noncardiac surgery in patients with recently placed coronary stents. If this is a subject that interests you, you can read the full article for free. People who have non-cardiac surgery within two years of stent placement have a greater surgical risk in the first six months, and it is higher with emergency than with elective surgery, and relates to the degree of coronary disease and not significantly to the type of stent. Which all makes sense, but apparently contradicts current guidelines.

1473   For most people, pancreatic cancer is a death sentence. For a minority, the tumour is surgically removable; but even if it has been completely resected macroscopically, recurrence is the rule. A long-term follow-up study from Germany and Austria shows that both time to recurrence and overall survival in such patients is better if they receive adjuvant gemcitabine for six months. At five years, the survival rate for the gemcitabine treated group was 20.7% versus 10.4% in the observation-only group: at ten years, this had dwindled to 12.2% v 7.7%.

Lancet  12 Oct 2013  Vol 382
1249   I’ve just come away from meetings with the two great gurus of evidence based medicine at McMaster University—Brian Haynes and Gordon Guyatt. One issue that came up was how frequently the evidence changes in such a way that decision making with patients is significantly influenced: this has a great bearing on how we can put in place mechanisms to ensure that guidelines and decision tools are constantly updated in real time. This paper illustrates the problem very nicely. Last week we “knew” that probiotics produce large reductions in the rate of antibiotic-associated diarrhoea (cut by 40%) and of C difficile infection in particular (cut by 60%). These figures come from two meta-analyses of the existing trials, which were mostly small and of mixed quality. But this week we have the results of the first really large-scale (n=2941), well-conducted, placebo-controlled randomised trial of 21-day treatment with a combined preparation of lactobacilli and bifidobacteria: and it had no effect at all. So do we simply feed these results into a new meta-analysis? If we do, according to an editorial on this study (PLACIDE), the aggregated effect size from all the trials hardly changes: probiotics still produce a large reduction in C diff. On the other hand, if we believe the results of the best and biggest trial, probiotics are useless. “Evidence-based” medicine is far from straightforward.

1268   But here’s a more straightforward example of the value of EBM. Systematic reviewing can be a tiresome business, but this is what it discovers about the agents used to reduce blood phosphate in people with advanced renal disease: “Our search identified 847 reports, of which eight new studies (five randomised trials) met our inclusion criteria and were added to the ten (nine randomised trials) included in our previous meta-analysis. Analysis of the 11 randomised trials (4622 patients) that reported an outcome of mortality showed that patients assigned to non-calcium-based binders had a 22% reduction in all-cause mortality compared with those assigned to calcium-based phosphate binders (risk ratio 0•78, 95% CI 0•61—0•98).” In other words, we have probably been killing renal patients by using calcium-based phosphate binders.

1278   A major problem for EBM is that the many important questions have never been addressed in head-on, well conducted randomised trials. Most interventions are tested by their manufacturers against the comparator most favourable to them, including placebo (and even placebos can be tampered with). I am constantly amazed that health systems are willing to throw away billions to pay for such treatments, but are usually reluctant to spend more than a few millions on objective comparative effectiveness research. The Health Technology Assessment unit of the UK has a budget of £70 million, about a tenth of the amount most large drug companies would spend on promoting a single new product. Yet this article shows that it has supported research of immense practical value. By showing that bevacizumab is as effective for age-related macular degeneration as ranibizumab, it has saved the NHS more than its own budget, every year (see the IVAN trial).

And that is only one of numerous examples: the use of tranexamic acid in major trauma, shown to be beneficial in the CRASH-2 trial, is probably saving 100,000 mostly young lives annually around the world. The HTA is a national treasure, but I bet a lot of readers have never even heard of it.

BMJ  12 Oct 2013  Vol 347
I am a cause of cardiovascular morbidity across the Northern hemisphere. Last week I caused atheroma in the citizens of inner London and outer Dresden, and this week I am doing the same for those of Slough, Boston (Mass), and Toronto. The noisy aeroplanes which have conveyed me to and from these destinations howl every few minutes over the tops of numerous homes of brick, concrete, or white clapboard, and the unfortunates who inhabit them are a teeny weeny bit more likely to end up in hospital with a cardiovascular diagnosis: “Averaged across all airports and using the 90th centile noise exposure metric, a zip code with 10 dB higher noise exposure had a 3.5% higher (95% confidence interval 0.2% to 7.0%) cardiovascular hospital admission rate, after controlling for covariates.” So scarcely worth making a big noise about, really.

I’ve never diagnosed a brain tumour in a child, or a Lisfranc injury in an adult. All the more reason, then, to read about these things in this week’s BMJ. The Lisfranc article is part of the Easily Missed series which is still going strong and has become self-sustaining, to my great delight. When I first wrote the proposal for the series, I said it could go on “indefinitely,” and so far it has. I have learned a lot from it and I hope others have too. The Clinical Review of children’s brain tumours could have come under this heading too, because “•  An average general practice sees a new childhood cancer every six years; a quarter of these will be brain tumours and •  Earlier diagnosis of brain tumours in children and young adults improves long term outcomes.” In other words, you may well never see one in a lifetime’s general practice, but when you do you need to spot it sooner rather than later. So read the review, which is up to the usual terrific standard.

Plant of the Week: Rhus typhina

Fate has decreed that we are seeing the fall colors of New England for the third year in succession. There is no special individual glory in the trees and shrubs that bear these colours: they can all colour equally well in the English autumn. But nowhere on our little island do they stand in such endless, massive profusion.

The understory of most New England woods contributes to the effect because it almost always includes abundant sumachs—either Rhus typhina or Rhus glabra. The leaves of these turn a deep glowing red, while the branches bear brown fruiting heads almost like antlers. Driving round suburban roads in middle England for over 30 years as a GP, I was always struck by these shrubs in people’s front gardens, where they provide an exotic and joyful sight among the mists of October.

The garden form of the sumach usually has frilly leaves, which add attractiveness. But the plant has a major drawback. It evolved to spread across miles of New England forest by means of numerous deep suckers. If you plant one in the middle of your lawn, its exploratory feelers will appear in due course where you least expect them: in the grass, in the flower beds, through the driveway concrete. You have been warned.