Richard Lehman’s journal review—16 June 2014

richard_lehmanNEJM 12 June 2014 Vol 370
2265  Obstructive sleep apnoea is often a result of weight gain, and unfortunately, once it is established, losing weight does not reduce it. But losing weight has benefits of its own (he sighs wistfully), as this trial of weight reduction, continuous positive airways pressure, or both for OSA demonstrates. I carry my flabby, insulin resistant, borderline hypertensive, tired all the time body around until the happy moment when I can put on my mask and feel the cool whoosh of air that signals sleep time. Would I do better to lose weight? Oh yes. It would not reduce my C-reactive protein (assuming it needs reducing), but it would certainly help reduce my triglycerides and insulin resistance (not that I have much idea about them either).

2276  In OSA, CPAP reduces BP. Nice to be able to use these abbreviations, because I have spelt them out above; well actually I haven’t spelt out BP. It stands for blood pressure. In a cohort of people with OSA and high BP, supplemental nocturnal oxygen was tried instead of CPAP for 12 weeks. Result: O2 a no-no for lowering BP in OSA.

2295  Ever heard of brodalumab? Whether your answer is yes or no, it is probably wrong because nobody on earth can possibly keep up with the flow of new monoclonal antibodies (mabs). The one I need is flabmab. This new one, however, is targeted against interleukin-17 receptor A (IL17RA), and underwent a trial in people with active psoriatic arthritis. The effect size compared with placebo was small, especially given the unambitious soft endpoint, a 20% improvement in American College of Rheumatology response criteria (ACR 20) at week 12. The triallists then muddied the waters by offering the active drug to all participants, and following them up for 52 weeks. “The study was funded by Amgen. Representatives of Amgen conducted the data analyses. All authors interpreted the data and collaborated in the preparation of the manuscript with support from professional medical writers funded by Amgen.” Conclusions: “Brodalumab significantly improved response rates among patients with psoriatic arthritis. Larger studies of longer duration are necessary to assess adverse events.” Right enough. Throw in cost effectiveness too. And an active comparator.

JAMA 11 Jun 2014  Vol 311

2288  This issue of JAMA is all about diabetes. Why does that make me want to run away and hide? Nobody in the field had ever heard of me until I entered the fray in 2009 with an editorial in The BMJ, shared with Harlan Krumholz, about the new QOF target for HbA1c. It drew 25 rapid responses, and ever since then I have been dragged back into commenting on diabetes, although I had said pretty much all I have to say in that short piece. Which is simply that the management of diabetes is a mess, and won’t be sorted out until we give patient-important outcomes a higher rank than surrogate outcomes. Alas, this implies that we need new long term trials of considerable complexity.

In the meantime, we have to go by what evidence we have, which is almost entirely observational. We would like to know if adding insulin to metformin is a better strategy than adding a sulfonylurea. Enter the Veterans’ Administration database and some cutting edge statistical wizardry. Not just propensity scores, but also marginal structural Cox proportional hazards models, inverse probability weights, and estimates of the magnitude of imbalance in unmeasured confounders, which might skew the results. What could possibly go wrong? Well, just about anything, as a very good editorial by Monika Safford explains. The conclusion of the study is that, among patients with diabetes who were receiving metformin, the addition of insulin rather than a sulfonylurea was associated with an increased risk of a composite of nonfatal cardiovascular outcomes and all cause mortality. But short of a randomised trial, we just can’t know for sure.

2297  And what about bariatric surgery for diabetes? For this intervention, we do have some short term RCT evidence, and it looks terrific. And the long term results look good too, going by observational evidence at a median of nearly 18 years from a Swedish cohort. Bariatric surgery “cured” diabetes in 72% of the patients in the first year; fifteen years later, that had dwindled to 30%, but in my view that remains amazingly good. The surgical group showed half the rate of microvascular complications, and macrovascular complications were a third fewer than in the non-surgical group.

2315  With a review of insulin therapy for type 2 diabetes, we are back in the world of comforting, if delusory, certainty. “Insulin can help achieve ideal hemoglobin A1c goals for patients with type 2 diabetes. Barriers such as adherence, patient preferences, clinician preferences, and resource allocation must be addressed.” Ah yes. A couple of small problems. How do we know what an ideal hemoglobin A1c goal is for any given individual? And where do we find the evidence from RCTs, using meaningful endpoints, to guide these personal and clinician preferences?

2331  Britons with diabetes don’t have any out-of-pocket costs for the expensive insulins they increasingly use. However, I’ll mention a research letter on this American problem because: (a) it is written by good friends, and (b) it has lessons for the UK system, which needs to look at its own mounting bill for new insulins.

JAMA Internal Med Jun 2014

OL  “When you are old and grey and full of sleep, A dosset box of statins you shall keep.” This comes from the poetic language summary of the latest NICE guidance on statins. The prose version says “Offer a statin to all individuals aged 85 or over,” which sounds entirely bonkers to me, though I am generally in favour of offering statins to everyone. But by the time I get to 85, I think I might want to die of something, to save further trouble. The benefits of statins are certain and accrue over many years; the harms are less certain, and generally stop as soon as you stop the statin. The trouble is that the main adverse effect of statins is to make muscles ache, and this is so common in old age that people may not put it down to their tablets. This could be the reason that a new analysis of data, collected for the Osteoporotic Fractures in Men study, finds that men who take statins have slightly lower rates of physical activity. Or maybe not. There may be a hidden confounder to account for these minor differences. If you are looking for a stick to beat statins with, this is more like a twig.

Ann Intern Med 3 Jun 2014

774  “Heart failure (HF) is a leading cause of hospitalization and health care costs in the United States. Up to 25% of patients hospitalized with HF are readmitted within 30 days. Readmissions after an index hospitalization for HF are related to various conditions. An analysis of Medicare claims data from 2007 to 2009 found that 35% of readmissions within 30 days were for HF; the remainder were for diverse indications (for example, renal disorders, pneumonia, and arrhythmias).” There are a lot of people who are working on this, trying out new models of care, such as the so-called “transitional care interventions” that form the basis of this review. In the context of the USA, anything that promoted primary care and team working reduced admissions and mortality. What about the UK? I have been thinking about this for 20 years, and I don’t know, for the simple reason that I have not properly asked people with HF and their carers what would be of greatest help to them. I’d suggest that somebody does this, rather than working from a cost reduction perspective and asking cardiologists, just because that is the easiest way to get a grant. I suspect that if I get heart failure, I will just want a place of safety when I am weak and breathless and feel like I’m about to die: which may happen a lot before I actually do.

Lancet 14 Jun 2014 Vol 383

2047  Well Dave, you can hear the anticipation in the capacity crowd here at Rio as these two great teams line up in the tunnel. The last time these titans met at a World Cup, it was Sirolimus that came away one-nil, but four years later can Zotarolimus even the honours? We know this pitch is not in perfect shape—it looks like an open label job to me—and we’re hoping there won’t be any of those questionable decisions by the ref and the linesmen that marred previous clashes. Oh, and they’re off. (Five years seem to elapse). Well that was 90 minutes of pure magic, ending in a goalless draw! Still, both managers will have something to take away from this. Alfonso Medtronico of Zotarolimus has brought the side into major surplus, and his old rival Johnson Johnson on the Sirolimus side won’t be too unhappy either. It’s the fans I feel sorry for. In the old bare metal days, a stent match used to be a family affair, when you could bring your small lads along and ten bob would cover the game and a pie afterwards. Some of these fans have paid over a thousand pounds to watch this match. It sometimes makes me want to elute.

The BMJ 14 Jun 2014 Vol 348

I’ve had a career-long interest in palliative care, but I would not claim to be good at talking with patients about dying. In an editorial, Kirsty Boyd and Scott Murray ask: “Why is talking about dying such a challenge?” Why indeed? Why is grasping this bunch of nettles so painful? I guess it’s because I didn’t bring a pair of thick leather gloves. The nice thing about this piece is that they don’t just call for more courage, better communication skills etc, but actually provide the gloves in a series of suggested opening gambits and useful phrases.

Peter Gøtzsche is a great man. His tireless commitment to finding out and declaring the truth about clinical trials has helped to transform medicine. If he sometimes goes where angels fear to tread, the angels generally follow later. Or else they stay safely behind, wagging their fingers like sea anemones. He doesn’t care about that: life is short, and there is so much to do. What do we actually know about the drugs we tell people to take? Duloxetine is an antidepressant, which has enjoyed indication-creep, and is now widely prescribed for chronic pain, fibromyalgia, etc, on a faute de mieux basis. That’s all very well, but if you are prescribing out of mere desperation, you should at least know what harm you might be doing. But published summaries of trials—especially when conducted by manufacturers—are very bad places to look for reliable data about harms. If you can get at the unpublished trials, including clinical study reports, you may have better luck. But generally this takes immense effort and time. Peter Gøtzsche is not deterred. His sense of outrage keeps him going, and we are all in his debt.

The two Gøtzsche papers get an insightful editorial from Peter Doshi. Peter is another remarkable person who walked the walk before talking the talk. Starting as an anthropology student at Harvard—not many years ago—he became interested in Japan and the adverse effects of medical products. So he taught himself Japanese and advanced systematic reviewing. Then he played a big role in uncovering the truth about oseltamivir with the Cochrane Respiratory group. He concludes: “Unless policy makers get serious about assessing the harms from drugs, society will remain trapped in a system where the lack of information about harms is treated as evidence of the lack of harm. If only it were so.”

Plant of the Week: Geranium “Kashmir White”

When we were establishing our garden, we bought a lot of hardy geraniums. They are mostly tough, they flower for weeks, and are generally easy to divide in the winter. Given two years and a few pots of the right kinds, you can create a garden that flowers for months. Then, as your knowledge and wealth increase, you can replace them with other plants and give them to friends.

I don’t think we have ever given away a Kashmir White, because it is not all that tough and if we divided it and it died, we would be heartbroken until we could find a replacement. It has lovely petals—white of course—but with a pattern of fine purplish reticulations that look like fractals, no two quite the same. For a few lovely weeks, it adds to the beauty of our geranium row, which runs for about 20m up the lane we live on. It consists mainly of blues (magnificum, pratense, Buxton’s, Roxanne, all in monthly succession) interspersed with pinks and whites. This year I am experimenting with an enrichment of dark red lupins.