From the journals, January–February 2005

New Eng J Med Vol 352

125 With the European Working Time Directive, and the new GMS contract, the British heroic age of sleep-deprived doctoring has largely come to an end. Studies from the US, where shifts of 24 hours or more still exist, show they can lead to some dodgy decision-making, and this doesn't end when the doctor gets in the car to drive home. When it comes to the crunch, tired doctors have more accidents.

154 Gonadotrophin-releasing hormone suppression is the mainstay of treatment for disseminated prostate cancer, but carries a small added risk of non-pathological fracture, as it suppresses the bones as well as the balls.

225 Nearly half of patients with chronic heart failure die suddenly as a result of ventricular dysrrhythmia. This trial compared the effect of amiodarone, or an implanted cardioverter-defibrillator (ICD) on mortality over 5 years: amiodarone was useless, but ICDs worked, although in the end the overall mortality difference was only 7%.

238 What do you do if a patient needs aspirin for cardioprotection but has had a gastric bleed? Just add a proton pump inhibitor — it's safer than changing to clopidogrel.

341 Diabetic neuropathy correlates less well to blood sugar than to overall cardiovascular risk scores.

450 Chickenpox used to cause over a hundred deaths in the US before the introduction of childhood immunisation: this has now been halved.

539 ‘That was a heart-stopping moment’ — this study shows that a sudden surge of adrenergic hormones from fear or excitement can stun the myocardium and lead to transient left ventricular failure.

Lancet Vol 365

217 Blood pressure — diastolic, systolic, mean, pulse, diurnal, maximum, minimum, office, home, whatever — is an important human variable. This paper concludes that it is too high in a quarter of the world's population, and set to rise.

305 Colonoscopy is a procedure little relished by most patients, or colonoscopists, so when CT colonography (‘virtual colonoscopy’) came along, audible sounds of relief could be heard. Unfortunately this real-life study does not confirm its early promise — it can miss cancers and polyps.

312 A systematic review of ethnic differences in asthma in the UK — plenty to discuss, with apparently higher rates in black children, and lower in Asian.

365 An editorial charting the disappointments which followed the discovery of an apparent association between bacterial infection (especially chlamydial) and myocardial infarction: prospective trials with antibiotics have been negative.

482 A review of reported withdrawal effects in neonates exposed to SSRI antidepressants in utero: as usual, paroxetine heads the list.

501 Years ago, stroke units were shown to have better outcomes than usual hospital care — but what about early discharge with intensive community support? This study shows it can work very well, and it may be what patients prefer.

JAMA Vol 165

172 Red meat consumption showed a modest relation to rates of colorectal cancer in 150 000 people over 20 years.

194 But a raised blood glucose seems to be a risk factor for a much wider range of cancers, according to data from Korea.

299 Frenzied rhythmic activity with a plastic doll on the floor may not be useful practice for the real thing, according to two studies of cardiopulmonary resuscitation. A European study of out of hospital CPR showed ineffective delivery and few survivors, results largely mirrored in a US hospital study on page 305.

477 You are very likely to come across a female patient with classical angina symptoms but normal coronary arteries. Beware: the condition carries risk, and there is probably a place for cardioprotective drugs.

647 Ximelagatran: a name to conjure with. A thrombin inhibitor that is better than warfarin for everything, at a fixed dose, without blood tests. But alas, how expensive! (see page 699) and perhaps not harmless to the liver.

Cox and Box

All the main journals have tried to discuss the cardiovascular safety of COX-2 specific anti-inflammatory drugs. Rofecoxib more than doubles the risk of myocardial infarction and should have been withdrawn years ago, according to a Lancet review (page 475). The accompanying editorial (page 449) ties itself in knots over the safety of celecoxib, as does the editorial in Ann Intern Med (142: 212). The same issue contains another study showing celecoxib to be safe compared with reofecoxib (page 157). The only blot on celecoxib's reputation remains the curtailed, unpublished Adenoma Prevention with Celecoxib trial. It's all a bit of a cox-up, but the reputation of the drug class looks irrecoverable.

Plant of the Month: Azara microphylla

A huge handsome evergreen shrub, with tiny yellow flowers on the leaf undersides wafting clouds of vanilla.