JAMA  2 Mar 2011  Vol 305
913   A friend recently began a piece on outcomes research with Bishop Joseph Butler’s maxim, “Every thing is what it is, and not some other thing.” If a trial like SOLVD is designed to measure the effect of a particular ACE inhibitor on survival in people with symptomatic left systolic heart failure, it is not the same thing as a trial designed to measure the effect of a beta-blocker following myocardial infarction. The only thing these trials have in common is that they lower blood pressure in people with established cardiovascular disease. Meta-analysts should remember Butler’s maxim. It is simply not legitimate to combine such dissimilar studies, plus a few from the diabetes literature, and use them to generalise about something called “secondary prevention of cardiovascular disease by antihypertensive treatment in persons without hypertension.” The terms are far too broad, and the analysis also doesn’t adjust for the use of other secondary preventive medication such as aspirin and statins. The effect sizes derived are small but positive – but that means nothing. Nor can I see anything in this misguided lump of disparate studies that can justify an editorial called “antihypertensive therapy for prehypertension”. In the case of the heart failure studies, the patients were more likely to have prehypotension. This is what happens when lumping goes mad: every thing becomes some other thing. more…

JAMA  23 Feb 2011  Vol 305
783   The highly prevalent custom of poisoning the osteoclasts of old ladies with bisphosphonates for years on end seems surprisingly harmless. This case-control study confirms that there is a tiny increase in the risk of subtrochanteric fracture after five years, heavily outweighed by the protective effects of these drugs in the first years of use. Nobody really knows how long to use them for. more…

JAMA  9 Feb 2011  Vol 305
569   If I were a woman, the things I would most fear from breast cancer surgery would be arm lymphoedema and recurrence of the cancer. Does one have to be balanced against the other? Common sense would suggest that the more axillary lymph nodes you dissect, the less likely it is that the cancer will recur. But last week we learned that micrometastases in the lymph nodes show very little correlation with survival, and this week we learn from this study that in women with invasive breast cancer and sentinel node metastases, survival is the same whether axillary node dissection is performed or not. The risk of arm lymphoedema goes down from 75% to 25%. So this trial continues the trend to less mutilating surgery for breast cancer begun by Sir Geoffrey Keynes in the 1920s, as I keep reminding readers in the hope that they will secure for themselves a second-hand copy of his autobiography. In The Gates of Memory, Keynes invites Henry James to Edwardian Cambridge, collects Blake manuscripts, grabs casualties under shellfire, pioneers blood transfusion, revolutionises breast cancer management, goes to the ballet, organises surgical services for the wartime RAF, cures myasthenia gravis with thymectomy, has medical students dig him a swimming pool and is criticised for his lapses in seventeenth century bibliography. more…

JAMA  2 Feb 2011  Vol 305
487   Most of us have never come near a vial of bevacizumab, though we’ve read plenty about it, especially over recent years in the context of eye disease involving vascular proliferation. This monoclonal antibody targeting vascular endothelial growth factor A was initially developed as a treatment for solid cancers and this remains its only licensed indication: an ironic situation given that it works so well for eye disease but fails to improve overall survival in all the cancers for which it has so far been used (see editorial on p.506). The meta-analysis here shows that this is probably because it kills as many patients as it saves, the numbers being small in either case. Which raises the important questions of whether we can identify the patients who are going to benefit and those who will be harmed: or whether in fact the use of VEGF inhibitors is a false turn in cancer therapeutics. more…

JAMA  26 Jan 2011  Vol 305
391   Stroke medicine grew up in the 1990s: like heart failure medicine, it shone welcome light on a large and neglected group of patients with organ damage who had been written off as unsalvageable. This was a Very Good Thing in itself, but its proponents then went on to declare that good stroke care could only be provided in designated stroke units, and went on to run some not-very-randomised trials to prove it. They also began to talk up the evidence for the benefits of immediate thrombolysis, which are real but extremely modest. These fashions spread to the USA following recommendations of the Brain Attack Commission in 2000, and this study evaluates the effect in New York State in 2005-6, comparing mortality and the use of thrombolysis in 31,000 patients with stroke, equally divided between hospitals with or without stroke units. There was a large difference in thrombolysis use – 4.8% in stroke units, versus 1.7% elsewhere; but a very small difference in mortality at 30 days – 10.7% versus 12.5%. It would be nice if someone could go on to look at a wider range of patient-important outcomes too. more…

JAMA  19 Jan 2011  Vol 305
261   I’m of an age when the words cognitive decline in the title of a paper make me rush to read it – the exception being a self-assessment study in the BMJ a couple of years ago, which was just too scary. This paper isn’t scary; in fact it isn’t anything, really. It tells me that because my level of literacy is in the highest tertile (I assume so, anyway), I’m at no great risk of cognitive decline in the next 9 years – which I wish I could believe. I could be a tiny bit more certain if someone measured my serum β-amyloid 42/40 level, which of course they won’t unless I take part in another prospective observational study like this one. If this is high, you’re OK: if it’s low, your brain goes senile slightly faster. more…

JAMA  12 Jan 2011  Vol 305
151   “Behavioral Therapy With or Without Biofeedback and Pelvic Floor Electrical Stimulation for Persistent Postprostatectomy Incontinence: A Randomized Controlled Trial.” As so often with titles like this, you have to explore the text before you can tell what the study is really about. In the UK, “behavioural therapy” usually means brief CBT and prostatectomy usually means a limited transurethral operation. Not so in the USA, where radical prostatectomy has been the urologists’ treatment of choice for localised prostate cancer over many years, leaving 65% of men incontinent of urine five years later. The “behavioral therapy” here was an intensive 8 week course on pelvic anatomy and exercises to improve pelvic floor strength and bladder control. This was highly effective, leading to a 55% reduction in incontinence episodes. Biofeedback and pelvic floor electric stimulation added nothing to this, except perhaps a warm tingly feeling in the privates. more…

JAMA 5 Jan 2011 Vol 305
43 Implantable cardioverter-defibrillators are a good intervention for those who have bad systolic heart failure with a risk of ventricular arrhythmia, and would rather die slowly than suddenly. The “utility” of the device is that it can have a statistically significant effect on mortality in younger, properly selected patients; the “dysutility” includes everything that can go wrong with the machines and the possibility that dying from pump failure may be accompanied by multiple painful electric shocks. more…

NEJM  30 Dec 2010  Vol 363
2588   A sizeable multinational study seeks to find out whether providing free daily anti-retroviral drugs as well as free condoms might help to reduce the transmission of human immunodeficiency virus in men who have sex with men. The majority of the subjects were recruited in Peru, with smaller groups from Ecuador, Brazil, Thailand, the USA, and South Africa. It’s highly likely that if all at-risk men took the antivirals every day (as offered), they would achieve very high levels of protection; but in this randomised trial, the reduction in infection was only 44%. Blood tests to check for compliance showed detectable antiviral levels in 51% of those who remained seronegative but only 9% of those who became positive for HIV. more…

JAMA  22-29 Dec 2010 Vol 304
2732   “Professionalism may not be sufficient to drive the profound and far-reaching changes needed in the care system, but without it, the health care enterprise is lost.” Britons, take heed! This “special communication” was written by a social scientist and five doctors to inform the debate about American health care, but I think the lessons apply even more strongly to the UK health system in its present crisis. The main value of the piece – “A behavioural and systems view of professionalism” – lies in its stress on professionalism as a learned behaviour, which would be enhanced if it were part of a stable contract between society and medicine. Such a contract would observe doctors closely but trust them to act from good motives which become innate with repeated action. This is the very opposite of imposed change with financial incentives for random elements of behaviour, or the creation of a free-for-all between budget-holding doctors and distrustful patients. I foresee difficult days ahead for British medical professionalism. more…