NEJM  1 Aug 2013 Vol 369
397  Shared decision making at the end of life is probably the toughest challenge in medicine. It requires good evidence about a wide range of interventions, good communication skills, close teamwork, and above all emphasis on the patient as a whole, and a sensitive capacity to take responsibility and to share uncertainty at the same time. Despite the best efforts of palliative care, we are miles away from achieving this for every dying patient. Look at the relatively simple case of a person with life-threatening cancer. “Cancer-drug labels stand in sharp contrast to labels for other types of drugs, about 25% of which list the drugs’ effects on patients’ symptoms or functioning [ed: I’m guessing this means that the cancer drugs are the ones with only 25% listing, contrary to the syntax]. That disparity is surprising, given how common symptoms and functional impairment are in patients with cancer and how toxic oncology drugs can be.” So in many cases you can’t even give a straight answer to the most obvious question that a patient will ask, which is “how will this make me feel?” , because the trials have all focussed on duration of survival to the exclusion of other patient important outcomes. As this excellent commentary piece states, it is time that we insisted on better reporting in trials of drugs for cancer.

407  More scary stuff about the new coronavirus from Arabia – MERS-CoV. In this report about MERS-CoV acquired in health settings, the mortality was 65%; but nearly all the 23 patients had pre-existing illness, and in a large proportion this was end-stage renal failure. This outbreak was intensively investigated and gives some new information about the median incubation period (5.2 days) and the capacity of the virus to spread from person to person in health care facilities. It doesn’t tell us how it might affect most people in any epidemic, but it would certainly kill a lot of vulnerable people. We may know more after the haj season takes place in October.

417  Here’s a trial which would definitely help me if I was responsible for treating a patient with ANCA-associated vasculitis. You will remember that this label is now used to describe various nasty and uncommon conditions usually accompanied by autoantibodies against proteinase 3 or myeloperoxidase, which are generally fatal if untreated. The commonest used to bear the name of the Nazi doctor Wegener. Standard remission-inducing therapy for severe disease consists of cyclophosphamide followed by azathioprine, but a single course of rituximab has been shown to be faster and more effective at 6 months. This trial shows that at 18 months, outcomes have become identical. So this seems to be a situation of clinical equipoise, in which the patient should be given a choice of treatment. I think I would ask for rituximab, though it costs at least £3,500 for a course. Cheap if I could get back to normal life and work more quickly. And this trial is also laudable for other reasons. Although it was partly funded by the manufacturers of rituximab, they “had no determinative role in the study design, the analyses, or the preparation of the manuscript.” But best of all, the investigators are allowing open public access to participant-level data, without the need for a specific research plan or approval of the qualifications of the investigators. If they can do it, why can’t everybody? See the editorial by Deb Zarin on Participant-Level Data and the New Frontier in Transparency (p.468).

428   French anaesthetists don’t often get to star in the NEJM, but here’s a trial which could change practice globally. Large cohort studies have shown that 20 to 30% of patients undergoing surgery with general anaesthesia are at intermediate to high risk for postoperative pulmonary complications. If you use conventional mechanical ventilation with high tidal volumes you can damage alveoli through overstretching and release pro-inflammatory agents which damage other organs. So this trial used low tidal volumes and positive end-expiratory pressure (PEEP), together with recruitment manoeuvres (periodic hyperinflation of the lungs), and compared this with standard mechanical ventilation. “Over the 7-day postoperative period, 10 patients (5.0%) assigned to lung-protective ventilation required noninvasive ventilation or intubation for acute respiratory failure, as compared with 34 (17.0%) assigned to nonprotective ventilation.” This is really impressive, and resulted in shorter hospital stays too. Vive le PEEP! Gloire aux anesthétistes académiques de France!

438  Fires that smoulder either flare up or go out. It’s the same with “smouldering” multiple myeloma; but whereas if you want to stop a smouldering fire flaring up, you just pour water on it, if you try proactive treatment on most smouldering MM, you achieve nothing. This trial slightly alters that position by identifying a high-risk group of smoulderers and giving them proactive treatment with lenalidomide and steroids. This achieved a delay in overt disease and a small absolute increase in survival at 3 years.

475  I still remember the phone call two years ago when Harlan Krumholz first told me about his agreement with Medtronic to release patient-level data from their trials. I was dumbfounded and disbelieving: this simply could not happen! Iain Chalmers had been asking for it for twenty years, but it seemed that few shared his ethical outrage at the concealment of huge quantities of trial data by industry. Now, thanks to the efforts of these leading proponents and others, the pioneering actions of the European Medicines Agency and the massive success of Ben Goldacre’s book Bad Pharma and the AllTrials campaign, industry has been forced to change its stance. First among the drug companies was GlaxoSmithKline, and GSK’s proposals for data sharing are examined in detail in this article. Many other pharmaceutical giants are trying to put on a public display of agreement in principle hedged with caveats which would give them control over what comes out and would not apply to drugs already in use, while at the same time trying to launch a counter-campaign using the patient groups which they sponsor. These really are interesting times, with further battles to come. Hard work will be needed on the detail of individual patient data analysis; and also effective surveillance of drug company compliance with trial registration and data release.

Lancet 3 Aug 2013 Vol 382
397  When I read of an unmasked international trial conducted in 78 centres in 27 countries recruiting just 601 subjects, an alarm goes off in my head: what are these people trying to market? Yet the answer in this case is not some expensive new drug of dubious marginal benefit, but a surgical procedure in a rather uncommon form of haemorrhagic stroke. And the people funding it are not a large pharmaceutical company but yourselves, dear readers, should you happen to be UK taxpayers. This MRC-funded trial compared early surgery versus initial conservative treatment in patients with spontaneous supratentorial lobar intracerebral haematomas. There was no difference in outcomes, but that doesn’t stop the investigators trying to find something positive for all the effort they put into this trial: “The STICH II results confirm that early surgery does not increase the rate of death or disability at 6 months and might have a small but clinically relevant survival advantage for patients with spontaneous superficial intracerebral haemorrhage without intraventricular haemorrhage.” The “might” and the “small” say it all.

409  Now let’s go to a trial run by a drug company – Novartis – which recruited 278 patients over the age of 70 with new or inadequately controlled type 2 diabetes. My ball-park guess is that you could find this number in a dozen medium-sized British GP practices, but Novartis chose to recruit from 45 outpatient centres spread across 7 European countries. They gave their subjects vildagliptin or placebo and showed that vildagliptin reduced their mean baseline HbA1c from 7.9% to 7 within 24 months. The paper begins with laudable comments about individual target-setting and the need to be aware of the limited benefits of glycaemic reduction in the elderly. Good: the message is getting through at last. But what is this we find in the conclusion? “This study is the first to introduce and show the feasibility of using individualised HbA1c targets as an endpoint in any type 2 diabetes population. Individualised glycaemic target levels are achievable with vildagliptin without any tolerability issues in the elderly type 2 diabetes population.” This is completely meaningless: we are left with absolutely no information about the long-term benefit of this drug in this population, just that it lowers sugar a bit and does not have any “tolerability issues”. It wasn’t even compared with another glucose-lowering agent. If you look at the absolute benefit of reducing HbA1c by that amount, in that range, in that population, it approximates to zero. Was that information shared with these patients?

BMJ  3 Aug 2013 Vol 347
Current regulation allows you to market medical devices on the flimsiest of evidence. Wound edge protection devices consist of great sheets of plastic designed to keep the edges of a surgical wound covered from contact with nasty things like surgical gloves and viscera during laparotomy: there are pretty pictures to show you. But do they reduce wound infections? The ROSSINI trial enrolled and randomized 760 patients: the answer was a plain no. This shows that surgical device trials are no longer the Cinderella of evidence based medicine, and the Barber Surgeons of Seville should rejoice.

Last week I pointed you to new evidence that repeated hypoglycaemia can lead to increasing hypoglycaemia unawareness, and that it is also strongly associated with cognitive decline. Here is a meta-analysis of observational studies which also demonstrates that severe hypoglycaemia is associated with an approximate doubling of cardiovascular events in patients with type 2 diabetes. Be very careful when prescribing sulfonylurea drugs and insulin to older diabetics. Ignore QOF targets, and think of your patient.

The best pieces in this week’s BMJ are outside the research section and therefore behind a paywall. Don’t miss Elizabeth Loder’s article on “Sharing data from clinical trials: where we are and what lies ahead” – a beautiful summary of a complex story, and bang up to date.

You can find out how our colleagues in Iran – a beautiful, deeply civilized country in which most educated people deplore the actions of their government – are being denied publication in leading medical journals for fear of falling foul of US sanctions.

And the Too Much Medicine series continues with a nicely argued piece by Ray Moynihan and colleagues about giving over 10% of the population a label of “chronic kidney disease”. In the correspondence, there’s an excellent letter from Kevin Barraclough on the same broad theme, ending “the Single Issue Fanatic’s unmodified enthusiasm risks medicalising the world. Descartes had it right: the sceptic needs a voice.”
Plant of the Week: Kniphofia “Fiery Fred”

The August garden is a challenge: things that looked lovely are now looking seedy, and the ratio of leaf to flower is in danger of tipping over to the wrong side. The plants which come to the rescue are the day lilies and the kniphofias, though they too can carry rather too much floppy or spiky leaf to be ideal. But they can be spectacular, and you can’t have a colourful garden in the late English summer without them, unless you resort to a lot of bedding plants.

Kniphofias (red hot pokers) are perennial South African plants which like wet summers and dry winters without much frost. To find a good selection you have to make a pilgrimage to Bob Brown’s nursery hidden down a bumpy mud track in Badsey near Evesham. Bob is a national treasure who moves fanatically from collecting one kind of perennial to another, and his nursery is an unrivalled cornucopia. Fifteen years ago he was seriously into kniphofias and his bulging catalogue contained innumerable varieties. This year he observes “The new family is quite a shock. The recent hard winters sorted the men from the boys. On the whole, what’s left is very tough.” There are a mere 38 cultivars left to choose from.

We were drawn to Fiery Fred not especially because the plant was named for the famous Yorkshire bowler but because it looked rather wonderful in one of Bob’s tunnels. It glows a fine tangerine colour with hints of yellow and green. There were other good ones too of course, and I would buy them all, had I the money and the garden space. The good thing about Fred is that he flowers over a long period and keeps his leaves up straight. Whereas the original Fred Trueman was famously blunt and working class, this poker is spiky and, well, just a bit of an aristocrat.