Richard Lehman’s journal review—28 October 2013

Richard LehmanNEJM 24 Oct 2013 Vol 369
1577 Two years ago I unexpectedly found myself in the USA amongst good and great people who were determined to open up hidden data about human trials. It was an odd position for a retired British country town GP to be in, and I concluded that my most useful role would be to foster contact between these like-minded heroes by means of a Google group called PATH. At the time, it seemed a rather Quixotic venture: Iain Chalmers had been arguing the case for sharing all data from trials for 20 years, but in the UK, pharma seemed to have captured the government’s agenda, an absurd statement had been issued by the Royal Colleges and others about the value of industry as a source of reliable information for clinicians, and even the existence of NICE seemed to be under threat. It is hard to believe the progress that has been made since then. The first major sign that the data dam was cracking came from the European Medicines Agency last year, when in a sudden reversal of policy they offered to release all data that had been submitted to them for the licencing of new medical products. They have subsequently been challenged by two pharma companies, but in a strongly argued piece in this week’s New England Journal of Medicine they set out the reasons for their historic decision. Fittingly, this piece is open for all to read, so please do. Many of you may feel this is just a side-issue, as you cope with the disintegration of the NHS, increasing workload, diminishing funds, and a government determined to avoid responsibility for the health of the population, but it is not. Sharing of accurate and complete data is the basis for true evidence based medicine and affordable care. This is where the rebirth of humane scientific medicine must begin.
1579 One of the good and the great of US medicine whom I met very briefly is Mike Lauer, who writes the next Perspective piece, unfortunately behind a paywall. He and a colleague set out a new model for doing randomized trials quickly and cheaply using large observational registries. Again, Iain Chalmers got here years ago. But it is another idea whose time has come.

1587 The trial that inspires Lauer was conducted in Sweden, and it compared thrombus aspiration during myocardial infarction with conventional percutaneous intervention. It was based on the national comprehensive Swedish Coronary Angiography and Angioplasty Registry (SCAAR). “All 29 PCI centers in Sweden as well as 1 in Iceland and 1 in Denmark participated in the trial. During the study period, 11,709 patients with STEMI in Sweden and Iceland underwent PCI and were registered in SCAAR. Of these, 7012 were enrolled in the trial. An additional 247 patients were enrolled from the center in Denmark, for a total of 7259 patients…None of the patients who underwent randomization were lost to follow-up with respect to the primary end point.” How about that? From start to finish the trial took less than three years, and it proves conclusively that aspirating clot at the time of PCI for ST elevation MI makes no difference to mortality.

1598 Outdoor air can carry as many as 100,000 fungal spores per cubic metre in normal conditions: indoor air can carry more if there is mould growing on the ceiling. That’s why drugs need to be prepared in highly protected conditions. It’s awful to think that methylprednisolone for intrathecal injection could be prepared without adequate protection in a place full of moulds: Exserohilum rostratum, Rhodotorula laryngis, Rhizopus stolonifer, Cladosporium cladosporioides, Paecilomyces formosus and Aspergillus fumigatus were all found in vials of the methylprednisolone that caused an outbreak of fungal infections in the USA last year. This article describes the spread of these contaminated products and the one after describes the clinical consequences. These papers are well up to the usual terrific standard of NEJM observational papers: if only the journal exerted similar quality control over papers about industry-funded trials.

1651 Nice to see Harvard catching up with Yale in a paper called “Preparing for the responsible sharing of clinical trial data.” In the summer of 2011, Harlan Krumholz of Yale proposed a model of data sharing which this paper calls the “learned intermediary” and signed an agreement with Medtronic to put it to the test. The next year we convened a closed conference with representatives of industry, data experts, lawyers and the medical good and great, to decide how we would share the data once the initial independent analyses had been published. A while after that, Harvard convened a similar conference; and this paper is a summary of their conclusions. The learned intermediary option is the one they seem to favour. But I won’t go on any more about this: if you are interested in this topic, you can access the full text free and read the paper yourself. Plus everything we had written about it previously, of course.

JAMA 23/30 Oct 2013 Vol 310
1692 Statins do a great deal more than just lower cholesterol, and for all we know they prevent cardiovascular disease by an entirely different mechanism, though the degree to which they do is proportional to their lowering effect on LDL-C. In particular they show strong evidence of having an anti-inflammatory action, and some studies have suggested that they can reduce mortality in sepsis. In this trial, patients in 26 French ICUs with ventilator-associated pneumonia were randomized to simvastatin 60mg or placebo. By the time the trial had recruited 300 patients, it was stopped for futility. Statins are wonderful drugs, but they do not reduce mortality in this situation.

1701 For many millennia, preachers have loved to preach, and moralists have loved to moralize. But in general, their effect on human behaviour has been negligible, however pleasing their actions have seemed to themselves. As the most admirable of vicars, the Revd. Sydney Smith, wrote to an erring friend in 1843, “I pray for you daily, but with no very lively hope of success.” Now we employ droves of dieticians, health coaches, counsellors, health visitors, and other well-meaning people to teach others the error of their ways, and every week we read in the journals how they fail utterly. In the present study, people turning up for free rapid HIV testing were randomized to go away without any preaching or to receive “individual patient-centered risk-reduction counseling based on an evidence-based model.” There was no difference in acquisition rates of sexually transmitted infections between the groups. Here endeth the first lesson.

1711 I move among sceptics, and I am not entirely convinced that influenza vaccination does anything much, but this meta-analysis suggests that it may reduce cardiovascular events in high-risk individuals. “A large, adequately powered, multicenter trial is warranted to address these findings and assess individual cardiovascular end points.” Yes indeed: and it could be done in a single flu season in the whole UK if we could just harness the same national database methods as that Swedish trial of thrombectomy in MI.

Lancet 26 Oct 2013 Vol 382
1405 Marbles, marbles, where are my marbles! I don’t know about you, but I can tell you that past a certain age the whole of life becomes a cognitive challenge, designed to make you aware that your brain is turning into a modge of plaques, tangles, and white spots on the MRI. As I sit in the warmth of my study, surrounded by unsorted heaps of books and CDs, and piles of very interesting papers I can’t remember reading, my laptop plays nasty tricks on me, asking for passwords and expecting me to remember what name I filed stuff under. Just as I am getting there, a wet cat jumps on the keyboard and demands food. By the time I get back, I have to begin all over again. And it can only get worse. Probably I should just give up and watch daytime television and stroke the cats. I blame the cats. Am I rambling? I want you to be kind and polite and mumble that the same thing happens to you. Only it probably doesn’t, and won’t: because the prevalence of dementia is getting less with time. That is the (slightly) comforting message of this two-decade comparison of prevalence of dementia in individuals aged 65 years and older in three areas of England.

1413 “Linagliptin for patients aged 70 years or older with type 2 diabetes inadequately controlled with common antidiabetes treatments: a randomised, double-blind, placebo-controlled trial.” You can tell what’s coming, can’t you? Yes, “inadequately controlled” here means a HbA1c greater than 7%: and linagliptin reduces this over 24 weeks with a “safety profile similar to placebo.” Fewer than 10% of participants had an initial HbA1c level above 9%, which is where I would say it may be rational to discuss adding treatment in this age group. The reporting of this Boehringer Ingelheim funded study reads very smoothly, as well it might: five of the six authors are employees of the company, while the sixth and principal author “has received honoraria for lectures and advisory work from Boehringer Ingelheim (the manufacturer of linagliptin).” Yes, and the accompanying editorial salves The Lancet’s conscience in accepting this paper by gently mentioning its defects while going on to bang a drum for the concept of “frailty” as a missing element in this study. The more I read about “frailty,” the more I’m convinced that it is just one more misleading label, aimed more at extending the academic literature than at helping patients. But that is another matter. Just file this paper as one more example of advertising unnecessary treatment under the guise of a randomized trial.

BMJ 26 Oct 2013 Vol 347
In her opening Editor’s Choice, Fiona Godlee very kindly quotes from an e-mail I sent from Logan airport last Monday, which some of you will have read: “All phase 3 trials to be designed and conducted independently of manufacturers, using the best available comparator. Research priorities to be determined by patients (James Lind Alliance). Value-based pricing. All data available from all trials, with meta-data: IPD [individual patient data] level for qualified independent centres. Big increase in comparative effectiveness research, much more research into non-pharmacological treatments.” There. Medicine sorted for the next half century. All really Iain Chalmers’ ideas, not mine.

Good old Scandinavians. They may have come over and burnt our monasteries, looted our churches, killed our young men, carried off our women and demanded exorbitant protection money, but hey, that was 1200 years ago and it’s time to move on. Now they run these magnificent whole-population registry studies which tell us what really happens to people when we introduce new population-wide interventions, such as giving quadrivalent human papillomavirus (qHPV) vaccine to adolescent girls. The authors looked at 997 585 girls aged 10-17, among whom 296 826 received a total of 696 420 qHPV vaccine doses; and they found no evidence supporting associations between exposure to qHPV vaccine and autoimmune, neurological, and venous thromboembolic adverse events.

A nice GP-based study compared three strategies for limiting or targeting antibiotic use in patients with sore throat: delayed antibiotic prescribing, the use of a clinical score designed to identify streptococcal infection, and the targeted use of rapid antigen tests according to the clinical score. “Principle findings: Our results suggests that across a range of practitioners and practices, use of either a simple clinical score or a clinical score with a rapid antigen test is likely to moderately improve symptom control and reduce antibiotic use. Use of the clinical score combined with targeted use of a rapid antigen test provided similar benefits but with no clear advantages compared with use of a clinical score alone.” Hmm. These may be sound principles, but perhaps the BMJ typesetter really meant Principal Findings.

And now to our own dear Clinical Practice Research Datalink, so good in principle, but so questionable when recently tested for accuracy in recording myocardial infarction. The CPRD is directly primed to examine prescriptions in UK general practice, which are almost always issued by computer, so it will have captured almost every prescription for smoking cessation products issued through the NHS. It is less clear to me whether it can reliably capture all incidents of self-harm or significant depression. Varenicline is the drug the authors of this study were really targeting, and their analysis “showed no evidence that patients prescribed varenicline had higher risks of fatal or non-fatal self harm (hazard ratio 0.88, 95% confidence interval 0.52 to 1.49) or treated depression (0.75, 0.65 to 0.87) compared with those prescribed nicotine replacement therapy.”

Statins for all? Yes, I would say, statins for all who really want them after full discussion of the harms and benefits. Just what these are is a matter for debate, and this article presents one side of that debate, pointing out possible under-reporting of harms and overstating of benefits to people at low risk. But I would have to go to the primary literature to be sure, because these authors clearly have a definite view on the message they think should be conveyed to “patients”. Yet having a risk of cardiovascular disease does not make anyone a patient: everyone has some risk, and we are talking here about free individuals. To consider that we as doctors should be “gatekeepers” in this situation is an affront to human rights. What we badly need is a better tool to share decision-making with those who come to us for advice on this issue, and James McCormack has recently come up with a very good and flexible web-based one. I’m going to spend a few days seeing if I can finesse it and perhaps develop an even simpler guide on one side of printed paper.

Word of the Week: Pessary

Whenever I discuss the derivation of a word in these reviews, it is in the hope of sparking a response from my learned friend Jeffrey Aronson. Last week I discussed the derivation of the word “pessary,” and I was delighted when Jeff took the bait:

“The Greek word pessos (or pettos in the Attic dialect) meant not merely a stone for board games, but also, at least as early as the 3rd century BC, a medicated plug of wool or lint for inserting through any orifice, usually into the vagina or anal canal. It was adopted in classical Latin as pessus. Postclassically, pessarion in Greek and later pessarium in Latin were used to mean a pessary in the same sense. Even in early English usage, a pessary seems not only to have been used vaginally. The earliest citation in the OED (from Lanfranc’s “Science of Cirurgie” of around 1400) refers to the use of a pessary “for to make clene a mannes lymes wiþinne”. The full quote is “A mundificatif may be seid in ij maners. As a medicyn þat is taken bi þe mouþ, ouþer þat is putt in bineþe wiþ a clisterie, ouþer wiþ a pessarie for to make clene a mannes lymes wiþinne.” I don’t know what “lymes” are.’

I hope that readers have enough Middle English to make sense of this. The (thorn) sign þ needs to be read as “th”. Lanfranc mentions a “clisterie” or clyster—this was solely used for anal administration and is synonymous with an enema, administered as a liquid through a clyster tube, of which many examples survive. He goes on to specify a second mode of treatment “put in beneath” called a pessarie (“for to make clean a man’s limbs within”). Now in this context “mannes” could mean either sex. But I suspect that the distinction Lanfranc is making is not between anal and vaginal application, but between a liquid clyster and a solid pessary.

Obviously this debate could rage for some time yet. But I think that’s quite enough pedantry on this deplorable subject.