N Engl J Med 2008; 359: 1,675-84
Many tasks in medicine can be done better by computers than humans, and mammogram interpretation can now be added to the list.
This trial shows that a single computer-aided reading of a screening mammogram can detect as many cancers as two human readers.
This randomised trial based in Scotland recruited 30,000 women and showed a sensitivity and specificity of 87 per cent and 97 per cent for computer-aided reading, as opposed to 88 per cent and 97 per cent for double human reading. The recall rates were also similar.
It seems the days of the medical hologram from Star Trek are not that far away. Single reading with computer-aided detection could be an alternative to double reading and could improve rate of detection of cancer from screening mammograns read by a single reader.
ISOFLAVONE FOR CVD PREVENTION
Eur Heart J 2008 doi:10.1093/eurheartj/ehn409
Having just got used to cardiovascular risk calculators that use HDL, waist size, blood glucose and BP, along comes a new set of potential cardiovascular risk factors, one example being high-sensitivity CRP.
Measuring risk factors is only useful if they are modifiable; this study looked at phytoestrogen supplementation in the form of isoflavone in patients with previous ischaemic stroke.
Using the gold standard double-blind placebo controlled randomised controlled trial (RCT) they added 12 weeks of supplementation or a placebo to treatment of a group of 102 stroke patients.
Not only did they find a reduction in CRP but also a reduction in atherosclerosis as measured by arterial endothelial dysfunction. This is a step away from saying it reduces end points like cardiovascular mortality, but larger studies would be welcome.
AMIODARONE IN AF
JAMA 2008; 300: 1,784-92
We know that continual use of oral amiodarone after an episode of AF will prevent recurrence, but the drug is not without significant side-effects, especially if taken for any length of time.
This study from the Netherlands looked at the possibility that episodic treatment might be just as beneficial.
It randomised 209 patients post-cardioversion to either usual continual treatment or episodic treatment of one month, or two months if the AF returned.
After two years they found that, although major cardiac adverse outcomes were similar, the number of patients where AF had reoccurred was higher in the episodic group and they also had more hospitalisations and greater all-cause mortality.
VITAMIN B IN ALZHEIMER'S DISEASE
JAMA 2008; 300: 1,774-83
Raised homocysteine is now seen as a non-modifiable risk factor for cardiovascular disease. It can be lowered with high-dosage vitamin B and folate, but patients still die from MI.
It had been noted that blood homocysteine levels are also high in association with mild-to-moderate Alzheimer's disease, but this is the first trial to test the theory that lowering those levels might affect cognitive decline.
Here, 409 patients were randomised and blinded to two groups of either high-dose vitamin B supplements or placebo.
Unfortunately, for those with the disease, there was no difference in the end points.
The authors conclude that this regimen of vitamin B supplementation does not slow cognitive decline in patients with mild-to-moderate Alzheimer's.
MEMANTINE AND PHANTOM LIMB PAIN
Anesth Analg 2008; 107: 1,377-9
With the wars in Iraq and Afghanistan there are sadly more patients with phantom limb pain to study. It is a mysterious condition. Just how do you treat pain emanating from a body part that does not exist?
US Navy researchers report two patient cases where the pain was refractory to methadone, hydromorphine, NSAIDs, antidepressants and antiepileptics. Having tried these and opioids they trialled a course of the Alzheimer's drug memantine.
The results were excellent, with a profound reduction in the pain and no side-effects.
ACTIVITY IN RESIDENTIAL CARE FAILS TO IMPROVE FUNCTION
BMJ 2008; 337: a1445
Bad news from New Zealand about the idea that activity programmes for those in aged care might improve function, reduce falls or improve quality of life.
The researchers looked at 682 elderly residents in low-level care and, using a clustered RCT, followed them for one year. The intervention group received an individualised exercise programme from a nurse, and the control group received just social visits.
The results were disappointing, with little difference bet-ween the groups on outcomes. Those in the intervention group with cognitive decline had a higher incidence of depression.
INFLIXIMAB HELPS TREAT LUPUS PERNION
Lupus pernio is an unpleasant, difficult to treat and disfiguring skin condition. US researchers have found treatment regimens that include infliximab are probably the best approach.
All 54 trial patients were assessed at each clinic visit with facial photographs.
By examining the photographs, the percentage of the face involved was determined, as was the effect of therapy. Medications included infliximab-containing regimens, systemic corticosteroids (CSS), non-infliximab, non-corticosteroid agents (AG), and CSS+AG.
The infliximab treatment regimens resulted in near or complete resolution in 77 per cent of patients. Corticosteroid treatments offered little help.
- Dr Palmer is a former Hampshire GP currently working in Australia, and a member of our team who regularly review the journals.
|THE QUICK STUDY|
Mammograms can be read better by computers than by humans
Isoflavone may be useful in secondary CVD prevention
AF recurrence is not prvented by episodic treatment with amiodarone
Alzheimer's disease is not slowed by high-dose vitamin B supplementation.
Phantom limb pain may be managed with memantine.
Activity programmes for those in residential care do not improve function.
Lupus pernio may be treated with infliximab.