Richard Lehman’s journal review—27 November 2017

Richard Lehman reviews the latest research in the top medical journals

richard_lehmanNEJM  23 Nov 2017  Vol 377

Keep your finger out of the cake mix

John Snow died of a stroke in 1858 at the age of 45, less than 4 years after his now-immortal map had shown how the 1854 Soho cholera epidemic was clearly due to water-borne infection. Few believed him, and after his untimely death he was recollected as an obscure pioneer of surgical anaesthesia. A century and a half later, the NEJM accords him the sincerest form of flattery with a Snow map covering the whole of the USA

It shows how contaminated flour from a single facility caused sporadic outbreaks of a single strain of Shiga toxin–producing Escherichia coli (STEC) in 24 states, affecting 56 people. The presumption is that this was due to people licking raw cake-mix off their fingers, since otherwise baking would have killed the bacterium. This is a great example of traditional epidemiology brought bang up to date with genomics and nationwide surveillance. But to do true homage to John Snow, the NEJM should now publish a map of cholera in Yemen, showing how the hundreds of thousands of cases there relate to the dropping of British and US-made bombs by Saudi Arabian pilots on water and sanitation facilities.   

JAMA  21 Nov 2017  Vol 318

Talc or drain for malignant pleural effusion?

Staying in historical mode, there was an article in JAMA called “The Talcum Powder Problem in Surgery and Its Solution,” published in Dec 1943. This listed numerous case reports from the 1930s proving that talc on surgical gloves caused adhesions and granulomas. But it was a long time before talc gloves were abandoned, and a long time (around 1958) before talc was used therapeutically to stick the pleural surfaces together as a treatment for malignant pleural effusions. It works pretty well, but this open-label randomised trial in 146 patients suggests that an indwelling pleural catheter may be slightly better. It’s a close enough call to make it matter of preference by individuals nearing the end of life.

Those US BP guidelines again & again

Talc v tube for advanced mesothelioma is a good example of the need for shared decision making by people in their last weeks or months. The number needed to treat for these procedures approaches one. At the other end of the scale lies so-called “hypertension,” a cardiovascular risk factor affecting billions of symptomless people, with various possible treatments with a NNT often approaching 500 over 10 years. Shared decision making with billions of people is quite a challenge. And faced with this length of odds, many perfectly fit people might decide against treatment. It’s all very well to move a threshold—from 140 to 130 systolic in the latest US guidelines—but quite another to decide which treatments have the best evidence from randomised trials, and match them to each individual in a menu of choices. The new guideline includes 106 graded recommendations divided into 47 “modular knowledge chunks” and a good summary of them is provided by Adam Cifu and Andrew Davis. But their closing call for shared decision making seems rather forlorn, given the logistic difficulties and wide areas of ignorance. Another piece on the subject, concentrates on pointing out the growing evidence of a herd benefit in lowering BP treatment thresholds. All very well for cows and sheep; but the challenge in humans is that each deserves a say on their personal preference in relation to their total risk. The main thrust of blood pressure research needs to move away from fixing the NNT further towards infinity, and instead narrowing it down to identify the individuals who would most benefit.

JAMA Intern Med Nov 2017

ROMICAT-II: have I seen this cat before?

Follow-up testing for low risk chest pain causes me déja-vu. “In this secondary analysis of data from a randomized clinical trial, patients who underwent clinical evaluation without noninvasive testing had a shorter length of stay, less diagnostic testing, lower cumulative radiation exposure, and reduced cost; there was no difference in missed diagnosis of acute coronary syndromes, development of major adverse cardiac events, and return ED visits.” I’m sure I’ve read this conclusion several times before. Here it comes in a just-published paper about ROMICAT-II. This is a cat that just goes on roaming, while nobody takes any notice.

Ann Intern Med  21 Nov 2017  Vol 167

Do you take this doctor to be your spouse?

According to a general population survey, nearly a third of female doctors in the US marry another doctor. And if not a doctor, someone who earns an amount near $100K. As a result, many of these female doctors can afford to work a bit less following marriage. Contrast the male doctor, only 17% of whom marry female doctors, and whose wives earn an average of $27K. And be it noted, Messrs Mills & Boon, that only 8% of male doctors marry nurses.

The Lancet 25 Nov 2017  Vol 390

To a man with an arthroscope…

everything looks like a joint to poke it in. Even when no extra income accrues—as in the NHS—there is a strong temptation to fiddle about and attribute any alleviation of symptoms to the procedure. We saw this with the knee last week, and this week it’s the turn of the shoulder. In this UK trial, patients who had undergone conservative treatment for subacromial pain were randomised to arthroscopy alone, arthroscopy with decompression, or no treatment. There were big drop-out rates, but overall there was no difference whatever between the first two options (active v sham) and no clinically meaningful difference between these and doing nothing.

Heart failure in UK primary care

British general practice hosts several large databases which have existed for long enough to be mined for information on the changing incidence and prevalence of disease. For example, I used to feed in data to the Clinical Practice Research Datalink, based on the Read codes we used in daily practice. For so-called heart failure, these were strongly biased to exclude anyone with preserved ejection fraction, since the Quality and Outcomes Framework paid us to investigate, treat, and monitor only those with reduced EF. As demonstrated earlier this year by Clare Taylor’s study based on the THIN database, there was no improvement in the survival of HF patients in the UK between 1998 and 2012. Now comes a larger study based on the CPRD which shows an apparent decline in the population-adjusted incidence of HF between 2002 and 2014, but an increase in prevalence of 23%. This does not really make sense unless people are living longer with the condition. And the waters are further muddied by the title of the accompanying editorial, “Rising incidence of heart failure demands action in which the author expresses concern at increased diagnosis of the condition in people over the age of 85, although this is beyond the average lifespan of any national population. In line with The Lancet‘s policy on commentary authorship, he declares funding from 18 pharma companies plus a start-up company of his own. Both the studies I mentioned show that the average age of HF at diagnosis in the UK is rising to 77+, and the number of accompanying major co-morbidities to 5. Try matching that to the evidence base for treating systolic HF in patients of median age 64 and no co-morbidity with ever-increasing doses of more and more drugs. It is time we turned round and addressed reality.

The BMJ 25 Nov 2017  Vol 359

Coffee under the umbrella

Coffee consumption and health: umbrella review of meta-analyses of multiple health outcomes” is the title of this paper, which is slightly unfortunate, since about the only harm coffee can do is scald your hand if you are trying to hold a cup of it while using an umbrella. Otherwise it is good for you. Which is just as well.

Plant of the Week: Euonymus grandiflorus “Red Wine”

For its first few years, this can seem a rather nothing shrub, with green leaves hanging from its arching branches in a rather limp way, and just a few barely fragrant flowers in early autumn. But when mature, it becomes the glory of late November. By this time it will be quite a big thing, 2m high and wide, but it will be flowering abundantly. Best of all, its leaves will turn dark red, and as December approaches they will form a perfect foil for the many split seed pods which quickly follow the flowers. These are intriguingly shaped and come in a curious colour of orange-pink. Definitely worth the space for providing this cheerful note in cheerless November.