JAMA  3 July 2013  Vol 310
46    If you identify people with poorly controlled blood pressure in primary care and introduce a system of intensive telemonitoring run by pharmacists according to a strict protocol, you are bound to get better BP control than if you leave it to “usual care.” But for each individual, the value of BP control will differ, as it does with every risk factor, and so will the adverse effects of treatment. You don’t need a doctor to measure BP but you do need someone in thoughtful dialogue with the patient to determine what the marginal value of further intensification might be. The simple “better control” trial described here didn’t really need to be done: anyone can intensify BP control, but the real question is whether such intensification is really worthwhile for the patients given their overall risk and personal preferences, and that goes well beyond a simple surrogate like BP.

57    Again, if you take a group of people with peripheral artery disease and advise them to walk regularly, some will and some won’t; if you enrol them in weekly classes where they spend 45 minutes walking around and a further 45 being told how good it is for them to do this every day then you are bound to get better results. And again I puzzle as to whether this will work in real life: but it may well be worth a try, because in this case the patients are likely to notice a direct improvement in function.

66    In 2001 I proposed a study in which we would screen individuals with risk factors for heart failure to see if they had elevated BNP and treat them to reduce it. The MRC liked the idea and we did a pilot study, but found that BNP was too labile a marker to guide individual treatment decisions. The Irish team who report their similar study here achieved better success through a less ambitious protocol: starting in 2005, they cluster-randomised GP practices to screening for BNP or none: in the intervention practices they subjected the patients with high BNP to more intensive work-up and treatment and found that this somewhat reduced cardiovascular events and hospital admissions over 4 years. This trial raises more questions than it answers, but to deal with them would take a small book.

NEJM  4 July 2013  Vol 369
1     This issue of the NEJM appeared on the date when citizens of the United States of America celebrate the signing of their Declaration of Independence, which is now generally agreed to have been completed on 7 July. But for this dastardly act of rebellion, the people of North America might now be enjoying the advantages of a National Health Service created by British socialists. To me, the NHS remains one of the best things ever to have happened in an imperfect world. The NEJM gives space to Nick Black for an excellent short account of how it has come under threat and whether it can survive.

3    I am also, on the whole, an admirer of NICE, though not to the same degree as Mike Rawlins who has just stepped down as its director and here celebrates its (and his) achievements since 1999. The idea is good; but the NICE mode of operation incorporates far too much “expert” bias and insufficiently questions the evidence presented to it from industry-funded sources. It needs to sharpen its teeth and get real patient involvement. Rawlins did a great job on the whole but he has left his successor with some huge challenges.

11   A big Chinese trial establishes that in the population studied, adding clopidogrel to aspirin following a transient ischaemic attack or minor ischaemic stroke reduces the risk of a further stroke within 90 days, and does not carry a significant risk of major bleeding or haemorrhagic stroke. It looks like a well-conducted trial but the editorial poses some good questions about its generalisability to other populations. More big trials elsewhere, please.

20    Rejectostix are urine dipsticks which will tell you whether a kidney transplant recipient is about to reject the kidney. They don’t exist as yet, but this study suggests that they might soon. If you are a nephrologist, memorize the following sentence: “A three-gene signature of 18S ribosomal (rRNA)–normalized measures of CD3ε mRNA and interferon-inducible protein 10 (IP-10) mRNA, and 18S rRNA discriminated between biopsy specimens showing acute cellular rejection and those not showing rejection (area under the curve [AUC], 0.85; 95% confidence interval [CI], 0.78 to 0.91; P<0.001 by receiver-operating-characteristic curve analysis).” Now that AUC for urine analysis is good, but it’s a bit difficult to say whether it’s good enough to obviate the need for renal biopsy at this stage of the game. A step in the right direction, certainly; but as the editorial says, some way short of Rejectostix.

Lancet  6 July 2013  Vol 382
41   Last week we discovered that about 40% of Dutch nursing home residents are depressed: this week we look at English nursing homes and find the figure is nearer 50%. The Dutch tried a multidisciplinary intervention which barely worked: here the English try exercise and find that it doesn’t work at all for frail nursing home residents with symptoms of depression. In fact they tended to get more depressed than those left to drink milky tea and watch films starring Bette Davis and Lucille Ball all afternoon.

50   Denosumab is a drug which keeps cropping up in The Lancet for reasons I find hard to fathom, since it is an expensive drug with no clear advantage over the bisphosphonates. Here is an open-label trial which was sponsored jointly by the manufacturers of denosumab, a RANKL binding inhibitor, and teriparatide, a parathyroid hormone analogue. One hundred women at high fracture risk were assigned to receive 20 μg teriparatide subcutaneously per day, 60 mg denosumab subcutaneously every 6 months, or both, for 12 months. With these numbers and this time frame, it was of course impossible to measure the outcome that matters: the rate of fractures. Instead, the abstract weakly concludes: “Combined teriparatide and denosumab increased BMD more than either agent alone and more than has been reported with approved therapies. Combination treatment might, therefore, be useful to treat patients at high risk of fracture.” In other words, you can give people this stuff but you don’t know if they will get any benefit. So why did the Massachusetts General Hospital lend its prestige to this trial, and why does it appear in The Lancet?

BMJ  6 July 2013  Vol 347
Publication of the complete data relating to every product used on human beings is an ethical imperative. Over 53,000 people have signed a petition calling for data from All Trials to be fully shared, but so far it has only been achieved in relation to a single product, recombinant human bone morphogenetic protein 2 (rhBMP-2) in spinal fusion. Most of the trials on this product were conducted by Medtronic more than ten years ago, and in an unprecedented move they handed over their complete data to Yale University for independent analysis by two separate teams in Oregon and in York. It was my amazing luck to be present at Yale and to become part of the YODA team handling this process. Here is a paper from the York team explaining the difference between the published studies and the full data made available through the YODA project. Not surprisingly the York authors find gaps and discrepancies in the way Medtronic reported the harms of treatment a decade ago: but by taking this step the company has now become the world leader in openness, and the YODA project is the most advanced of its kind so far seen. It’s a pity this aspect is not more openly celebrated in the accompanying editorial by Peter Doshi: there are some things we might have done differently with hindsight, but there are lots of things others can learn from; and if they want to reanalyse the data, we make it very easy for them to do so.

What do you mean by a placebo? I take it to mean a treatment of no known value given to a patient either in a clinical trial (where the subject knows that they could be receiving it) or in clinical practice, where it is commonly given with the false assurance it might help. The assurance is generally seen as part of the treatment. If you are interested in public perceptions of this, here is a survey from Northern California. It gives mixed messages. Personally I tend to be a Zoroastrian about such things, and consider that you cannot preserve asha (integrity) without honesty.

Pulmonary embolism is easily missed, and it appears in the BMJ series of that name. Pulmonary embolism is also easily overdiagnosed, and it appears in the first BMJ article in a new series called Too Much Medicine. It’s a question of balance: the lungs are meant to act as a sieve as well as an organ of oxygenation, and CT angiography may be making us aware of lots of debris that really doesn’t matter. Lifelong anticoagulation should not be embarked upon lightly. This is a cracking first article co-written by friends. For a contrasting patient’s perspective, go to John Launer’s beautiful piece in QJM.

Ann Intern Med  2 July 2013  Vol 159
Let me direct your attention to Ebstein, W (1876). “Zur Therapie des Diabetes Mellitus, insbesondere über die Anwendung des Salicylsauren Natron bei demselben.” Berliner Klinische Wochenschrift 13: 337–340. Here Ebstein points out the glucose-lowering effect of the then newly-discovered salicylic acid compound which we now call salsalate. And now in July 2013 comes the report of a trial of salsalate which shows that it lowers blood sugar in type 2 diabetes. Had this stuff been used for the past 147 years, we could have had some nice hard outcome data by now.

Plant of the Week: Salvia x jamensis “Melen”

This is a beautiful plant which caught our eye at a local nursery and has yet to be road-tested. It has smallish snapdragon flowers of pale creamy yellow with a base of deep inky purple held on long wand-like stems, and the whole plant smells deliciously of … well, salvia.

It is a shrubby perennial for near the edge of a border in a sunny place. It is said to be hardy to minus 10C, but time will tell. Its looks and scent make it worth its price in a single season; survival will just be a bonus.