Richard Lehman’s journal review—31 October 2016

richard_lehmanOxygen for moderate COPD

This week saw the official launch of the UK Academy of Medical Royal Colleges “Choosing Wisely” campaign, which was so successful that its website crashed. It is proving a bumper week for debunking long-accepted practices. The latest is long-term oxygen for chronic obstructive pulmonary disease with moderate desaturation. This common strategy was tested in a non-blinded randomised trial in which 738 patients at 42 centres were followed for 1 to 6 years. In the supplemental-oxygen group, patients with resting desaturation were prescribed 24-hour oxygen, and those with desaturation only during exercise were prescribed oxygen during exercise and sleep.

I think the outcome is even stronger for lack of blinding, since it means that oxygen therapy doesn’t even have a tubes-hissing-in-the-nostril placebo effect. It had no effect whatever: “the prescription of long-term supplemental oxygen did not result in a longer time to death or first hospitalization than no long-term supplemental oxygen, nor did it provide sustained benefit with regard to any of the other measured outcomes.”

Screening for familial hypercholesterolaemia

Although the hottest medical debates are often aroused by screening, the UK Choosing Wisely campaign has largely steered away from screening topics. I am quite glad, because the Royal Colleges need to keep their focus much more on people with illness and how they can help patients make decisions about their care. I find it truly odd that although the basic criteria for screening were first established by Wilson in 1967 and refined in many iterations since, they are seldom explicitly applied to existing or newly proposed forms of screening. Just to remind you (i.e. me) here they are:

  • the condition should be an important health problem
  • the natural history of the condition should be understood
  • there should be a recognisable latent or early symptomatic stage
  • there should be a test that is easy to perform and interpret, acceptable, accurate, reliable, sensitive and specific
  • there should be an accepted treatment recognised for the disease
  • treatment should be more effective if started early.

Familial hypercholesterolaemia is the subject of a screening study by the Walds,  father and son, which shows that “Child–parent screening was feasible in primary care practices at routine child immunization visits. For every 1000 children screened, 8 persons (4 children and 4 parents) were identified as having positive screening results for familial hypercholesterolemia and were consequently at high risk for cardiovascular disease.” That’s about as far as I can go within the space constraints of this blog. I’d just suggest that if you ever teach medicine to anybody, you get this article and set your learners on to test it against Wilson’s criteria. It is a clearly written paper and thoughtfully states the case for screening. Is there a case against?

Out-of-hospital defibrillation

The telephone box in our village is full of weeds, and close to it stands a yellow box containing a defibrillator. Perhaps I should complain about its colour to the parish council, just to enjoy the “Village Fury as Doc Attacks Life-Saving Machine” headline in the local paper. A new population-wide study from Japan shows that the estimated number of survivors in whom survival with a favourable neurologic outcome was attributed to public-access defibrillation increased from 6 in 2005 to 201 in 2013 (P<0.001 for trend). The adult population of Japan is about 100M, while the population of our village is about 350. So the chance of our defibrillator saving a life is about 0.000002 x350=0.0007 per annum.

Old red blood cells

On All Souls’ Night we are bidden to remember the departed, but at modern Halloween we are more likely to remember that Count Dracula had a strong preference for blood freshly drawn from the neck vessels of young ladies. I suppose on bad nights he had to make do with bags of stored blood filched from the Transylvanian blood transfusion service (note to donors: avoid eating garlic for three days before attendance). For purposes other than rejuvenation, a randomised trial shows that there was no significant difference in the rate of death among those who underwent transfusion with the freshest available blood and those who underwent transfusion according to the standard practice of transfusing the oldest available blood. But this study was conducted in hospitals in Australia, Canada, Israel, and the United States, places where vampirism is little attested: other criteria may apply in the remote castles of Hungary and the ruins of Whitby Abbey.

Data sharing: c’mon guys

A somewhat downbeat article marks the NEJM‘s continued pilgrimage through the issues of open data sharing. Its main message is that the GSK-led data initiative now known as CSDR ( hasn’t attracted the traffic it had hoped to, despite offering data from over 3000 trials. But these are early days, and incentives and funding for such studies are still hard to find. To the title’s question, “Data Sharing — Is the Juice Worth the Squeeze?” I would reply with a loud yes. But then I am part of the Yale University Open Data Access (YODA) team, and we specialise in optimism.

JAMA  25 Oct 2016  Vol 316

Vital Directions?

The print edition of JAMA this week is devoted to the future of the health care in the USA, under the title “Vital Directions”. I find it sad that clichés that were true forty years ago are even more true today. America needs more generalists, more unified services, a single payer, allocation according to need, strong regulation of drugs and devices, strong public health to reduce disparities and combat vested interests, and so on. The UK has had and could still have all of this, and more, but it is in the process of self-destruction. For example, the NHS offers wonderful opportunities for data sharing and real-time cycles of knowledge gathering. Why have these been so hopelessly squandered? In this area, the US opportunities are set out by Harlan Krumholz of YODA, with others. See, we are wide-thinking optimists. If we weep, we do it in private.

Cranberries for UTI prove a turkey

Gathering bilberries on the moorland hills around Sheffield, the teenage me would sometimes come across a similar gnarled plant of the Vaccinium family, bearing red berries with a sharper taste. I can reveal that it is generally difficult to take a pee in private on the bare moors where cranberries grow, so their habitat may account for their alleged ability to reduce urinary frequency. Rural myth became health food certainty, as it often does, and billions of cranberries are now turned into juice where once a few dozen would serve to make jelly for the annual Christmas turkey. Or else you can coat them with sugar (they need it) and sell them to treat female bacteriuria in nursing homes. But following a randomised trial in Connecticut this would now be unethical, since they have no effect.

JAMA Intern Med  Oct 2016  Vol 175

Smoking kills in tobacco-land

When Hillary Clinton becomes President, she will have a lot to tackle: Big Pharma, Big Sugar, Big Weapons, Big Oil, Big Pollution, Crap Infrastructure, Unaffordable Healthcare. But if she wants a priority list for health, Big Tobacco probably still comes top. A state-by-state analysis shows that while about 28% of cancers in the USA are smoking-related, this rises to 40% in traditional tobacco-growing states such as Louisiana, Tennessee, West Virginia, and Kentucky. Add cardiovascular and respiratory disease, and you have one hell of a preventable death factor.

Lancet  29 Oct 2016  Vol 388

30 seconds for fat

A big trial without a non-intervention group produces a small effect which looks a lot like reversion to the mean. How then did this GP advice/referral for weight loss trial get so much hype? It may be partly due to an over-enthusiastic accompanying editorial but it also ticks a lot of other boxes. The idea that brief fat-shaming by a GP can lead to 1kg of weight loss has great appeal to the press. If only those lazy GPs would spend half a minute telling people they are fat, they would pull their socks up and lose weight. The study has also been welcomed by public health physicians who have to spend a lot of their day talking about the need for population obesity strategies. Yet a kilogram of weight loss might well have happened anyway in a control group, had there been one. And while it is convenient for avoiding the need to buy new clothes, I’m not sure that it has any other measurable benefits. The people randomised to the weight loss programme lost rather more, and this may be worthwhile for those who achieved 5% body weight reduction.

Helping parents handle autism

The PACT trial tested a twelve month intervention aimed at helping the parents of children with autism spectrum disorder to improve the “symptoms” of the disorder. The original report in 2010 used a scoring system called Autism Diagnostic Observation Schedule-Generic (ADOS-G), and found no value-added effect for the PACT intervention. The latest report of results at 5.75 years uses a different instrument covering a wider range of autistic features, and it is positive (see editorial). So this quite intensive intervention based on parent training probably does help to modify children’s behaviour when initiated before school age and continued thereafter. Which is good, but the parents also matter. To what extent does it help them feel that they are coping and helping their child? I think that’s a metric that is needed for all trials in this area. In fact I don’t think the two can be separated: the authors conclude that their positive long-term outcomes stemmed from optimisation of parent–child social communicative interactions, which then become self-sustaining.

MenB vaccination for babies

Protect your child from lethal disease caused by group B meningococci: why wouldn’t you? In the NHS, cost was a factor, plus the difficulty of proving the protective effect when the index disease is so rare. Nonetheless, just over a year ago the UK became the first country to introduce the multicomponent group B meningococcal (MenB) vaccine (4CMenB, Bexsero) into a publicly funded national immunisation programme. A reduced two-dose priming schedule was offered to infants at 2 months and 4 months, alongside an opportunistic catch-up for 3 month and 4 month olds. Here is the first report of this natural experiment, and it looks good. Compared with the prevaccine period, there was a 50% incidence rate ratio reduction in MenB cases in the vaccine-eligible cohort (37 cases vs average 74 cases), irrespective of the infants’ vaccination status or predicted MenB strain coverage.

BMJ  29 Oct 2016  Vol 355

Ultrasounding the retreat on tibial treatment

Broken bones heal slowly, and we would all like to believe that there are magic rays or good vibrations that would help them to knit faster. This trial of post-operative ultrasound treatment following fixation of tibial factures suggests otherwise. Low intensity pulsed ultrasound (LIPUS) was compared with sham treatment in 501 patients across 43 North American trauma centres. There was no difference between groups. Another one for the Choosing Wisely bin list.

CT vs angiography for suspected angina

Although I often write about interventional cardiology, because there is so much of it in the big journals, I only have a hazy idea of how patients are actually managed in real life. So read my comments with due caution. Context is everything and this trial was done in a single academic hospital in Berlin. The patients had suspected coronary artery disease and a clinical indication for coronary angiography on the basis of atypical angina or chest pain, and where possible they were randomised to coronary CT imaging or angiography. The authors state that “the primary outcome measure was major procedural complications within 48 hours of the last procedure related to CT or angiography”, but they report mainly on other outcomes. Frustratingly, these do not include immediate revascularisation. To me, the primary purpose of any investigative strategy is to allow the clinician to have enough information for a proper dialogue with the patient about treatment options. So if you perform any cardiac imaging, it should be with the intent of discussing it with the patient in a context that gives them real choice about what to do next. CT imaging has the great advantage that it removes any temptation to bung in a stent while you’re in the artery. It has the disadvantage of high radiation, to which subsequent angiography will add. Anyway, this trial leads the authors to conclude that CT increased the diagnostic yield and was a safe gatekeeper for coronary angiography with no increase in long term events. The length of stay was shortened by 22.9 hours with CT, and patients preferred non-invasive testing.

Beware huge effect sizes

I have often praised big “definitive” trials in these columns, so I’m somewhat chastened to read this paper written by Myura Nagendran under the guidance of Peter McCulloch. I mention these authors because on Twitter this study has been attributed to John Ioannidis and Doug Altman, who certainly contributed, and are great heroes to all right-thinking people. But I think those who did the heavy lifting also deserve credit. Someone looked through 3082 reviews yielding 85 002 forest plots, of which only 44 (0.05%) satisfied the inclusion criteria for randomised trials with very large effect (VLE) defined as a relative risk of ≤0.2 or ≥5. When these trials were replicated, the effect size dropped in 43 out of the 44, and in 19 cases it lost statistical significance altogether. Oops. Science is about replicability: ignore the need for replication and it is no longer science.

Plant of the Week: Rosa “Absolutely Fabulous”

It’s all but November but the garden is still full of flowers. The most outstanding are those of a rose I’ve written about before, bought on a whim a couple of years ago. It is still producing its classy flowers of mustardy yellow at a prodigious rate, as it has since late May. In the unseasonable moist warmth they continue to carry a strong scent of liquorice.

Other roses continue to produce too, each one delightful in its own right: climbing Mme Alfred Carrière, Dublin Bay, White Wings, Jacqueline Dupré and the ever-lovely Grüß an Aachen.