NEJM 2 October 2014 Vol 371
1285  Here is a trial which had me taking my glasses off and scratching my bald patch. Why on earth should a drug company—in this case Boehringer Ingelheim— want to pay for a trial of taking patients OFF a drug? And why in particular should it want to take people with chronic obstructive pulmonary disease off inhaled steroids while keeping them on a long acting beta-adrenergic agent (LABA), when we know that LABAs increase mortality unless you take them with an inhaled steroid? The latest evidence for this came in JAMA a couple of weeks ago.

The explanation lies in the presence of a third player in this game: tiotropium, a Boeringer Ingelheim product. I darkly surmise that this trial was designed less to encourage people with COPD to discontinue inhaled steroids than to encourage them to take 18 μg of tiotropium once daily (delivered by a HandiHaler). Protected by this excellent product, they could dispense with their steroid sprays, and everyone would be happier. The verdict from this trial must be Not Proven. The rate of exacerbations in those who gradually withdrew their steroid inhalers was only non-significantly higher than those who stayed on them, but even this large trial was not powered to detect significant long term mortality outcomes.

1295  They call it celiac, we call it coeliac, let’s call the whole thing off. That’s what I sometimes feel when I look at the avalanche of literature about coeliac disease, which I am partly to blame for triggering when I encouraged Harold Hin to find out its prevalence in the minimally symptomatic citizens of Banbury. Had 1% of our townspeople really been carrying anti-endomysial antibodies from birth, or had they developed them later on? Did it depend on when and how they had been weaned on to wheat products? These questions are explored in two studies in this week’s NEJM. This first one comes from Italy, where this subject has been studied more than anywhere else, perhaps because of the importance of bread and pasta in sustaining the life of that happy land. They randomly assigned 832 bambini who had a first degree relative with coeliac disease to the introduction of dietary gluten at six months (group A) or 12 months (group B). There was no difference in the eventual acquisition of coeliac disease, but those introduced to gluten later took longer to acquire it.

1304  The second feeding trial, this time conducted across Europe, explored the hypothesis that introducing small amounts of gluten early in infant life might prevent coeliac disease developing in high risk children by the age of 3 years. It did not.

1316  Five centuries ago, the Spanish, French, Dutch, and English all fought to introduce the wonders of Christian civilisation to the continent of North America. Among these benefits were firearms, alcohol, slavery, and numerous transmissible diseases. But while often killing each other, all parties united in their theoretical abhorrence of sex outside marriage. This seems to colour attitudes to contraception in the United States up to the present day. Over here, we started encouraging it with free prescriptions from around 1968. Over there, the results of this trial are original enough to feature in the NEJM 46 years later: “Teenage girls and women who were provided contraception at no cost and educated about reversible contraception and the benefits of LARC methods had rates of pregnancy, birth, and abortion that were much lower than the national rates for sexually experienced teens.” Now who would have thought?

JAMA 1 October 2014 Vol 312
1293  A viewpoint piece by Ben Djulbegovic and Gordon Guyatt puts the argument that “Evidence Based Practice Is Not Synonymous With Delivery of Uniform Health Care.” Although they make the case clearly, I don’t think they are radical enough. The best people to write such a viewpoint would be Al Mulley and Jack Wennberg, who have been studying variation in practice and the role of shared decision making for decades. Here in a nutshell is where they had got to by 1988:
“Bad Variation (care not evidence based) owing to
• Poor research → professional uncertainty
• Poor knowledge → professional ignorance
needs to be distinguished from Good Variation (where care is patient centred) owing to
• Clinical differences among patients
• Preference differences among patients
If all variation were bad, it would be easy to stop it. What is difficult is reducing the bad variation while keeping the good.” I tried to put this to a senior person at the National Institute for Health and Care Excellence recently. The reaction I got convinced me that the patient revolution is well overdue.

1295  The next viewpoint is by Tammy Hoffman, Victor Montori, and Chris Del Mar, and its title is “The Connection Between Evidence Based Medicine and Shared Decision Making.” A great piece, but this is a massive topic that deserves a whole conference (which it will get in Sydney next July) and a whole book (which I have been trying to write for several years). If EBM is to support shared decision making, it should be extremely modest in its pretensions to certainty. It should be extremely careful to define areas of clinical equipoise between options, and extremely keen to develop better methods of sharing uncertainty with patients and the public. If it could do this, then it could define the clinical research agenda in collaboration with patients, and move straight past guidelines to develop constantly updated, reliable tools for sharing clinical decisions with them.

1305  The excellent Irfan Dhalla of Toronto ran a trial to determine whether a virtual ward—a model of care that uses some of the systems of a hospital ward to provide interprofessional care for community dwelling patients—can reduce the risk of readmission in patients at high risk of readmission or death when being discharged from hospital. Alas, at the conclusion of the trial he reports: “In a diverse group of high risk patients being discharged from the hospital, we found no statistically significant effect of a virtual ward model of care on readmissions or death at either 30 days or 90 days, six months, or one year after hospital discharge.” Hearken, ye public health physicians and politicians. Very sick elderly people are going to cost the health service a lot of money and then die, whatever you do to them. The magic solution that will postpone death and save costs does not exist. And well developed systems of primary care (“usual care”), in countries like Canada and the UK, already do a terrific job at low cost.

1313  A Zelen design clinical trial (in which randomisation occurred before informed consent) studied the efficacy of laser and needle acupuncture for chronic knee pain in Australians over the age of 50. It produced modest improvements in their pain score at 12 weeks, compared to control. But it didn’t matter in the least whether it was “real” acupuncture or a sham. I can only repeat the advice I have given to you on previous occasions. Set up as a sham acupuncturist. You will give people temporary relief from all sorts of aches and pains, and all you need is a set of sterile needles and a plausible manner. A silk gown and one of those wall charts of an androgynous naked person with lines all over them will help too.

JAMA Intern Med October 2014
OL  I sounded off about the state of medical device control in both Europe and the US a couple of weeks ago. It’s so farcically inadequate you may want to look away. But for the brave, here are some articles brought to you by the editor of JAMA Intern Med, who is an author on one* of them:
Lack of Publicly Available Scientific Evidence on the Safety and Effectiveness of Implanted Medical Devices (free online)
*Assessing the Safety and Effectiveness of Devices After US Food and Drug Administration Approval: FDA Mandated Postapproval Studies
The Food and Drug Administration’s Unique Device Identification System: Better Postmarket Data on the Safety and Effectiveness of Medical Devices
Improving Medical Device Regulation: A Work in Progress (free online)

Lancet 4 October 2014 Vol 384
Moving swiftly past the printed Lancet for this week, which is unusually free of interest, we turn to the website and its latest editorial, with the title “Familial hypercholesterolaemia: PCSK9 inhibitors are coming.” Yes, indeed they are, but when should we let them arrive? When they have shown short term safety and good evidence of lowering LDL cholesterol? Or when they have shown long term safety and effectiveness in reducing actual cardiovascular events? Guess what, the two trials on the Lancet website only provide evidence about the first.

OL  Heterozygous familial hypercholesterolaemia is the commonest dominantly inherited metabolic disorder in humans, affecting about one person in 250. That’s a market of three million people in the US and Europe. But I don’t think that’s the main reason why at least three drug companies are scrambling to produce monoclonal antibodies to treat it. Beyond these individuals lies the immense market for cholesterol lowering agents to reduce cardiovascular risk in the general population. Evolocumab seems to be ahead from the starting block. It binds to proprotein convertase subtilisin/kexin type 9 (PCSK9) which, as you probably know, directs the low density lipoprotein cholesterol receptor to lysosomal degradation, inhibiting its recycling to the hepatocyte surface and thus catabolism of plasma LDL. The evolocumab trials take their names from great physicists who were active 100 years ago. These are given in capitals, as if they were acronyms, but they are nowhere explained in the text. Perhaps just as well. RUTHERFORD-2, a randomised, double blind, placebo controlled trial, examined the effect of evolocumab in people with heterozygous hypercholesterolaemia, and it yielded a rapid 60% reduction in LDL cholesterol with no more adverse effects than placebo.

OL  It’s tougher work, however, using this strategy in people who carry paired genes which give them very high levels of LDL cholesterol. These rare unfortunates (one in 300 000) have homozygous familial hypercholesterolaemia. The TESLA part B trial used a relatively high dose of evolocumab and succeeded in lowering LDL cholesterol by 31% on average at 12 weeks in 33 individuals. But don’t think you have heard the last of this. Soon you will be reading trials of alirocumab and bococizumab, which are also antibodies to PCSK9. One day they could all become Bigbizumab.

The BMJ 4 October 2014 Vol 349
As with the Lancet, so with The BMJ. Let’s go straight to the website. Here are two important papers that refute ideas which I have tended to support. Great. That is what science is all about.

OL  I have argued that our goal in diabetes should be to help people choose which outcomes matter most to them and let their treatment be guided by that. Researchers tried this out in 18 general practices in the north of the Netherlands. The intervention comprised a decision aid for people with diabetes, with individually tailored risk information and treatment options for multiple risk factors. Result: “We found no evidence that the patient oriented treatment decision aid improves patient empowerment by an important amount. The aid was not used to its full extent in a substantial number of participants.” That’s truly disappointing, but it is somewhat similar to Victor Montori’s experience with his own excellent decision aid developed at the Mayo Clinic some years ago. The problem seems to me that we are trying for too much culture change too quickly. That, and the fact that it is pretty impossible to know from the limited evidence we have which drugs for diabetes do what.

OL  I was immersed in statins for several months this year until I began to drown. With these drugs too, the evidence about adverse effects and benefits is extraordinarily hard to convey in simple terms. Some of my friends take the view that these pills are being foisted on people by big pharma, who have exaggerated their benefits and played down their harms. I don’t believe the first part of that, and I believe that most of the “harms” are small and entirely reversible. But I was open to the suggestion that at least some of the company funded trials might have contained an element of oversell. Now Huseyin Naci and colleagues have done a superb systematic review and network meta-analysis, which reaches the following conclusion: “Our analysis shows that the findings obtained from industry sponsored statin trials seem similar in magnitude as those in non-industry sources. There are actual differences in the effectiveness of individual statins at various doses that explain previously observed discrepancies between industry and non-industry sponsored trials.” Splendid: industry deserves praise where it has acted honourably and provided us with sound knowledge about a genuinely life saving class of drugs.

Plant of the Week: Geranium cinereum “Ballerina”

There is a hardy geranium still flowering in our garden, which looks like a large erodium, with deeply veined flowers of purple-pink and white. It’s so good that we’ve bought it twice and forgotten its name. It might be “Ballerina,” but then again it might not.

This can be a good time to pick up bargains in garden centres which are selling off old stock cheaply. You might be able to get quite a few geraniums for £3 a time or less, saving you the trouble and uncertainty of trying to propagate them yourself. The varieties which every garden must have include this one, Buxton’s Blue, Johnson’s Blue, Rozanne, sanguineum, magnificum, and the wild native pratense.

If you are over 50, make sure you keep their labels. That will save you embarrassment when it comes to telling people what they are called. Ballerina is not really quite good enough in the naming department. But it’s a great plant for early autumn, whatever it might be.