Richard Lehman’s journal review—23 May 2016

richard_lehmanNEJM 19 May 2016 Vol 374
Danazol & the Norns
1922 In early European mythology, life was seen as a fragile string, which could be cut by an arbitrary force outside the power of the gods. In the Greek version, a child would be visited on the third day of life by the three Moirai. Clotho spun the thread of human fate, Lachesis dispensed it, and Atropos cut it at the appointed moment of death. In Norse mythology, these ladies took the form of Norns. Nowadays Clotho dispenses telomeres at the end of DNA strands. With each cell division, these shoelace ends become shorter, and eventually Atropos steps in with her abhorred shears and you die. Lachesis sometimes causes mischief by handing out telomeres that are too short, causing people to be highly susceptible to cancer or to die prematurely from bone marrow failure, liver cirrhosis, and pulmonary fibrosis. But in some ancient myths, there is a herb that protects against the malice of the Fates or the Norns. It crops up in the very first literature in the world, the Sumerian stories of Gilgamesh. Now it seems that we may have found its modern equivalent by dusting down the back of the drug cupboard. Older GPs and gynaecologists will just about remember the days when danazol was used as a treatment for endometriosis and breast pain. It has lots of androgenic adverse effects, needs liver monitoring, and may increase the risk of ovarian cancer, so it’s hardly ever used now and certainly not in the long term. But it had amazing effects in an exploratory trial in participants with telomere shortening disorders. The trial was halted when 11 of the 12 evaluable patients given danazol demonstrated telomere lengthening after six months: something no Norn had ever seen before.

Lifestyle change useless for infertility
1942 Let me say some nice words about the selection policy of the NEJM—just to make a refreshing change. Last week they published an excellent semi-qualitative study on the prevalence of depressive symptoms in carers of patients discharged after being ventilated on ICUs. This week they publish a Dutch randomised trial of a lifestyle programme in obese infertile women. The intervention had no effect on overall fertility outcomes: in fact, at two years the frequency of vaginal births of healthy singletons at term was significantly lower in the intervention group than in the control group, who received immediate fertility treatment. This is great news for millions of obese women around the world who are having trouble getting pregnant: they can skip being preached at and made to go on diets and bikes, and get straight on with treatment and having babies.

Ventricular tachycardia ablation
OL Here is some serious cardiology for grown men. It so happens that in the US, fewer than 10% of interventional cardiologists/electrophysiologists are women, suggesting that ablation therapy must bear some hidden resemblance to snooker or motor racing. Ventricular tachycardia ablation is a last ditch procedure, which can kill as well as cure: there were two cardiac perforations and three cases of major bleeding in the ablation group of this 259 participant trial. But the alternative was escalated-dose amiodarone, which was worse and less effective: it caused two deaths from pulmonary toxic effects and one from hepatic dysfunction. So if you have survived myocardial infarction, but still have recurrent ventricular tachycardia despite an implantable cardioverter-defibrillator, you might do best by submitting yourself to the manly ministrations of the guy with the hairy arms and the blue hat.

JAMA 17 May 2016 Vol 315
Our new world begins in Lanarkshire
2063 This has been a historic week for the cause of shared decision making, and I’m still rubbing my eyes and wondering if it can really be happening. In one way, it is not: we are still grossly lacking the tools and the skills and the intellectual infrastructure for sharing knowledge with patients, not to mention the time and the environment to do it properly. But as the new standard for medical practice, shared decision making now has legal force and the support of the Royal Colleges and the National Institute for Health and Care Excellence (NICE). Two close friends from Yale, and a legal colleague of theirs, reflect brilliantly on the worldwide implications of the UK Supreme Court Montgomery v Lanarkshire Health Board ruling. This now applies to the whole of Britain and supplants the old Bolam standard of reasonable care. I’m deeply, deeply biased towards these authors. But even if I wasn’t, I’d implore every reader to click on the free link and read this article carefully. It is what medicine is going to be about from now on.

Prognosis in the ICU
2086 Following on from last week’s NEJM mixed methods paper about depression in carers of patients ventilated in ICU, here’s an open access paper about prognosis. Qualitatively and quantitatively, it explores how physicians and “surrogates” differ in their perception of prognosis in the ICU and the reasons for this. Surrogates are the ones taking decisions on behalf of critically ill patients, and their estimates of prognosis are pretty good. The estimates of doctors, however, are even better. The reasons for discordance between the two groups are fascinating: they include hanging on to hope, religious belief, and belief in the hidden resilience of the patient. It’s another terrific example of the mixed methods approach, which should be far, far commoner than it is. “Rerum cognoscere causas” is the motto of Sheffield University—to know the causes of things. In medicine, you can often find out the causes of things much better by asking people than by counting numbers.

JAMA Intern Med May 2016
Going to church delays going to heaven
OL American nurses who went to religious services between 1992 and the present are more likely to be alive than those who did not. This fact emerges from the Nurses’ Health study, leading the authors to conclude that “Religion and spirituality may be an underappreciated resource that physicians could explore with their patients, as appropriate.” I won’t even attempt to guess what they have in mind.

People who say they are active get less cancer
OL For our next ride on the great Confounding Rollercoaster, let’s match up rates of cancer and the amount of physical activity that people claim to do in their spare time. Unfortunately, these data have not been adjusted for church, mosque, or synagogue attendance, so there may be residual doubts. But the study says that exercise prevents cancer so it must be correct. Physicians should advise their patients to become Shakers, so as to combine lots of exercise with lots of worship. That way they will prolong their lives, but not the human race, since sex is forbidden to Shakers. This will eliminate global warming too.

Ann Intern Med 17 May 2016 Vol 164
Breasting the waves of uncertainty
649 I am a bit late coming to the important B-path study, which appeared on the Annals website two months ago. It’s a study of histopathologists’ levels of agreement in interpreting breast biopsy slides. For definite normals and definite cancers, agreement is over 90%. But for the kinds of change that are lumped as ductal carcinoma in situ (DCIS), concordance was fuzzier: 18.5% of samples were overinterpreted and 11.8% underinterpreted. And for “atypia,” the figures were 53.6% overinterpreted and 8.6% underinterpreted. Almost all these biopsies were done as a result of mammography. The NHS still encourages women to have mammography by running an opt-out rather than opt-in system. It seems to me that it should now ensure that any biopsy that is reported as showing atypia or DCIS should be read by three histopathologists. But the fact is that we don’t know the natural history of these abnormalities (the editorial on p 694 is good on this) so we’ll still end up overdiagnosing a great number of women who would have done perfectly well without mastectomy, radiotherapy, chemotherapy, and hormone therapy.

Lancet 21 May 2016 Vol 387
Progesterone doesn’t prevent preterm birth
2106 OPPTIMUM is a rather vainglorious acronym for a trial, but for once I won’t complain. This is just about as good as it gets. It was based in NHS hospitals with one Swedish outlier, and publicly funded. It was perfectly designed and powered to answer a simple and important clinical question, and here is the result: “Vaginal progesterone was not associated with reduced risk of preterm birth or composite neonatal adverse outcomes, and had no long term benefit or harm on outcomes in children at 2 years of age.” That’s it. This story is ended, and nobody need ever use vaginal progesterone again to prevent preterm birth.

Nifedipine or atosiban for tocolysis
2117 I must confess that the existence of atosiban had passed beneath my radar until I read this paper. It has been around for more than a decade and is an inhibitor of oxytocin and vasopressin, so in theory it could be used as a treatment for either premature labour or heart failure. But it has only been used for the former, and intravenous atosiban was compared with oral nifedipine in this trial conducted in 10 tertiary and nine teaching hospitals in the Netherlands and Belgium. “In women with threatened preterm birth, 48 h of tocolysis with nifedipine or atosiban results in similar perinatal outcomes. Future clinical research should focus on large placebo-controlled trials, powered for perinatal outcomes.” In other words, this trial should have been larger and included a placebo arm. A missed opportunity to settle an important question.

Anything for the weekend?
OL Both the Lancet website and the printed BMJ this week contain substantial articles on the “weekend effect,” which was the purported reason for the government’s decision to change the contracts of junior doctors and hospital consultants. In this cynical mess, it’s hard to believe that real facts have any useful part to play, but our sanity depends on hoping that they might. The fervour with which the prime minister shouted his figure of 11 000 avoidable deaths at Jeremy Corbyn was chilling. I would recommend that he be sent to the Center for Outcomes Research & Evaluation at Yale and spend every Friday afternoon for a year learning about the complexities of using hospital data and adjusting for case mix. I did that, which is why I have shied from commenting on such things for ever after. Big data in itself is stupid. The less accurately it is gathered, the more spectacular the feats of intelligence and adjustment that are required even to begin to interrogate it.

But while warring armies flounder in the mud, aerial reconnaissance is beginning to show up some broad patterns. Here are a couple of summaries:
This cross-sectional analysis did not detect a correlation between weekend staffing of hospital specialists and mortality risk for emergency admissions.”
The weekend effect is a simplification, and just one of several patterns of weekly variation occurring in the quality of stroke care.”
. . . And do read this excellent commentary by Nick Black.

The BMJ 21 May 2016 Vol 353
Weekend trippers
There is so much good stuff about the weekend effect in this week’s print BMJ that I can only recommend that you sit down and read it, paying particular attention to Martin McKee’s editorial (“Now you see it, now you don’t”) and the research paper on coding errors and the apparent “weekend effect,” which again looks at stroke.
I also very much enjoyed Jacqui Wise’s collage summarising the key arguments.

Let’s call the whole thing off
They say fries, we say chips, they say chips, we say crisps. Little did we know that the fate of millions may hang upon these distinctions. For we now know, with all the certainty of three US cohort studies that included questionnaires about potato intake, that eating fries (chips) can give you hypertension whereas eating chips (crisps) cannot. It is high time the US and the rest of the English speaking world standardised their solanaceous terminology in the interests of public health and mutual understanding.

One of the reasons I said this was a historic week for shared decision making is the appearance of a paper by Margaret McCartney and another three of us on the subject of making evidence based medicine work for individual patients, which drew an immediate and positive response from the chair of NICE.

I’d also like to draw attention to the analysis piece that appears in this week’s printed edition, which offers a useful reality check about information sharing, choice making, and overload.

Plant of the Week: Ceanothus “Blue Mound”

Most of the garden Ceanothus species come from California where they get plenty of sunshine and little rain. It’s amazing how well they do on an island where the weather is the other way round. There are plenty of other shrubs that can match them for pink, lilac, and white flowers, but none that come near them for pure dark blues against lovely dark evergreen foliage. So every garden needs a ceanothus and it’s just a question of finding the right spot for the right one.

From what I’ve already said, you’ll know not to put your chosen ceanothus in waterlogged ground or in the shade. In colder parts of Britain, it’s as well to place it by a sunny wall. If you have lots of space, then “Puget Blue” is a must, with a pale pink montana clematis to grow through it with amazing effect. If you only have less than 2m of space, you’ll need to go for one of the lower growing kinds, such as Blue Mound or Blue Cushion. The blurb on the pot will greatly underestimate its eventual size, but you can keep it under control with careful pruning twice a year. These plants are completely covered in flower for two or three weeks every spring and Clematis macropetala “Markham’s Pink” is the one to grow through them. On warm days they will be buzzing with insects and if you bend down to smell them you will be rewarded by wafts of grass and honey, and perhaps a bee sting.