BMJ  2005;330:989-990 (30 April), doi:10.1136/bmj.330.7498.989

Short cuts

What's new in the other general journals

Acupuncture relieves chronic low back pain

The literature evaluating acupuncture for low back pain has been reviewed at least twice before, with conflicting results. The latest update, including 23 extra trials, shows that acupuncture can work, at least for people who have had back pain for more than three months. A thorough search found 33 randomised trials comparing acupuncture with sham acupuncture, no treatment, or other active treatments such as massage. Only 10 of the trials had been reviewed before. The others were new, unpublished, or in languages other than English, including Chinese, Japanese, and Polish.

Overall, acupuncture worked significantly better than sham acupuncture or no treatment in people with chronic back pain. The authors found a difference between acupuncture and sham acupuncture equivalent to 14.5 mm on a 100 mm visual analogue scale measuring pain. Acupuncture was no better than other active treatments at relieving chronic pain, however, and looked significantly worse than spinal manipulation (two trials) or massage (one trial).

It's still unclear whether acupuncture works for acute back pain. These reviewers, and others before them, found too few decent trials to be certain.

Annals of Internal Medicine 2005;142: 651-63[Abstract/Full Text]

New drug treatment for heart failure increases mortality

Nesiritide, a new drug treatment for acute on chronic heart failure, looked promising until a recent meta-analysis found that it was associated with a suspicious increase in deaths. The analysis included data from three trials comparing the new intravenous drug with conventional treatments for decompensated heart failure, usually vasodilators and diuretics. Inotropes such as dobutamine were excluded. Four per cent of 377 patients given control treatments died within 30 days, compared with 7.2% of 485 patients treated with an infusion of nesiritide (hazard ratio 1.8, 95% CI 0.98 to 3.31, P = 0.057; figure).

View larger version (29K): [in this window] [in a new window]   Credit: JAMA

 

The analysis isn't perfect, and the findings aren't strictly significant in the conventional sense, but the authors say they warrant serious exploration in an independent prospective trial that's big enough and lasts long enough to find an increased risk of death with nesiritide if there is one. In the meantime they warn doctors to stop using nesiritide as a first line treatment for patients with acutely decompensated chronic heart failure. (See p 981.)

JAMA 2005;293: 1900-5[Abstract/Full Text]

Antibiotics do not prevent coronary heart disease

Despite mounting evidence that infection can trigger atherosclerosis, trials of antibiotics in people with coronary heart disease remain stubbornly negative. At least six trials, including a megatrial in nearly 8000 patients, have already reported that antibiotics against Chlamydia pneumoniae don't work as secondary prevention against heart disease. Two more large and negative trials published last week probably signal the end of the road for this particular line of inquiry.

In the first trial, 4012 adults with stable coronary heart disease were given azithromycin or placebo for one year. After four years of follow-up, 22.3% of the treatment group and 22.4% of controls had died from coronary heart disease or had had some kind of serious coronary event (risk reduction 1%, 95% CI -13% to 13%). In the second trial, 4162 patients who had been admitted with an acute coronary syndrome took gatifloxacin or placebo for 10 days each month. After a mean of two years' treatment 25.1% of the placebo group and 23.7% of the treatment group had died (from any cause) or had had some kind of serious coronary event or stroke (hazard ratio 0.95, 0.84 to 1.08; P = 0.41).

An accompanying editorial (pp 1706-9) says we should now stop trying to treat people who have heart disease with antibiotics against C pneumoniae. Instead we should be looking for other infective agents, such as viruses that might be the missing link between inflammation and atherosclerosis.

New England Journal of Medicine 2005;352: 1637-45, 1646-54[Abstract/Full Text]

Virtual colonoscopy is too unreliable for screening

Virtual colonoscopy has been under evaluation for over 10 years as a non-invasive way of detecting early colorectal cancer. The results have been mixed. The latest meta-analysis, which includes 33 trials, found that virtual colonoscopy was reasonably specific (86%, 95% CI 84% to 88%) but not very sensitive (70%, 53% to 87%), especially in people with polyps smaller than 6 mm across (48%, 25% to 70%; figure).

View larger version (17K): [in this window] [in a new window]   Credit: ANNALS OF INTERNAL MEDICINE

 

All the trials compared virtual colonoscopy with real colonoscopy or surgery. Sensitivities varied between a low of 14% and a high of 96%, so the authors did further analyses to find out why. Technical differences between scanners explained some of the discrepancies; scanners with multiple detectors were more sensitive than scanners with single detectors, and the newest and fastest "flythrough" scanners looked best of all, with a sensitivity of 99%. Most of the variation remains unexplained, although there are plenty of potential sources, including variations in bowel preparation, operator technique, interpretation, and software. The authors conclude that virtual colonoscopy is not yet consistent enough to be used as a screening test for colorectal cancer.

Annals of Internal Medicine 2005;142: 635-50[Abstract/Full Text]

Vaccination plus antibiotics is the best defence against anthrax attack

It's impossible to know for certain how best to limit the human cost of a large scale anthrax attack by bioterrorists. Mass vaccination before an attack sounds like common sense but "best guess" mathematical modelling shows that it would be expensive ($815m (£430m; 625m) for a city of 5 million people) and would save fewer lives than antibiotics plus mass vaccination after an attack.

Researchers constructed a model to compare four postattack strategies—vaccination, antibiotics, both, or neither—and two pre-attack strategies—vaccination or no vaccination. The postattack strategy combining antibiotics with mass vaccination worked best, gaining 0.33 life years and saving $355 per person compared with postattack vaccination alone (figure).

View larger version (41K): [in this window] [in a new window]   Credit: ANNALS OF INTERNAL MEDICINE

 

These kinds of models may be the best we can do, but they are always based on more or less reliable assumptions. These authors assumed, among other things, that the probability of attack on a major US city is 1% a year, that 10% of the population would be infected with anthrax, that 95% of people infected would develop clinical anthrax, and that 45% of those with clinical disease would die. The main finding stood firm through reasonable variations in the costs and effectiveness of anthrax vaccines, but it looked less attractive (and pre-attack vaccination looked more attractive) as the likelihood of a terrorist attack went up. Unfortunately, the true probability of an attack is the one thing we will never know.

Annals of Internal Medicine 2005;142: 601-10[Abstract/Full Text]

Japanese report first successful islet cell transplant from a living donor

A young Japanese woman with poorly controlled diabetes is now independent of insulin after receiving islet cells from her mother in the first successful living donor transplant operation for diabetes. The woman, aged 27, developed diabetes at 15 after childhood pancreatitis. Before the transplant she had uncontrollable insulin dependent diabetes and daily episodes of hypoglycaemia. Twenty two days afterwards she no longer needed insulin, and by day 37 she had completely normal glucose tolerance.

Her new islets, which were prepared from the tail section of her mother's pancreas after a partial pancreatectomy, were simply injected down her portal vein under local anaesthetic. They started working almost immediately, and the Japanese doctors reporting the case are optimistic that the islets will carry on working for at least five years.

Until now, transplant programmes have used pancreatic tissue from cadavers, which is in limited supply and usually reserved for patients with crippling intractable hypoglycaemia. Living donors could be a viable alternative, but partial pancreatectomy is a big operation accompanied by serious risks, including pancreatic fistula, pancreatitis, wound infection, bleeding, and, later, glucose intolerance. One expert comments that living donation is still hard to justify until we can be sure that the recipient's long term benefits are worth all the donor's risks. Islet transplantation, even from cadavers, is still a technique under development. Independence from insulin is not guaranteed, and even when it does happen, it may not last forever.

Lancet 2005 April 19. http://image.thelancet.com/extras/04let3109web.pdf

Mercury from fish is unlikely to affect cognitive function in older people

Older Americans are often advised to eat more fish because of the well known benefits associated with a high intake of omega-3 fatty acids. Methyl mercury, an environmental contaminant in predatory fish, is less good for you, and experts worry that its neurotoxic effects could be amplified in older people who are already vulnerable to cognitive decline. Investigating this, researchers looked for associations between neurobehavioural test scores and serum concentrations of mercury in a cohort of 474 residents of Baltimore aged between 50 and 70. They found a link between mercury concentrations and test scores in only two of the 20 tests—higher mercury concentrations were associated with worse performance in a test of recall, but better performance in a test of manual dexterity.

The authors say that these results are reassuring, and that in general mercury contamination in fish is unlikely to harm older Americans any more than younger ones. In the United States, advice about eating fish is based on safe serum concentrations of mercury for children and young women, currently set at 5.8 µg/l or less.

JAMA 2005;293: 1875-82[Abstract/Full Text]

Norovirus decimates crew on international flight

During a pre-Christmas flight between London and Philadelphia in 2002, eight of the 14 cabin crew developed gastroenteritis with diarrhoea and vomiting. Sick crew members were relieved of their duties and sent to sit at the back of the plane in an attempt to stop the illness from spreading to the whole crew, or the passengers. The flight lasted eight hours, and the sick crew managed to confine their sickness and diarrhoeal episodes to the plane's toilets.

Two were admitted to hospital after landing, and their stool specimens showed norovirus, a highly infectious pathogen famous for spreading quickly by various routes in confined spaces. Fortunately, when public health officials questioned the 191 passengers, most of whom had since dispersed across America, only five (5.4% of respondents) reported symptoms consistent with norovirus infection—that is, non-bloody diarrhoea starting 18-60 hours after disembarkation (figure). The passengers who became ill had used the plane's toilets more than other passengers (median number of visits 3 v 2, P = 0.006).

View larger version (49K): [in this window] [in a new window]   Credit: JAMA

 

The authors say theirs is the first report of norovirus transmission on board an aircraft and propose that the toilets were responsible, even though they looked macroscopically clean. The same strain of virus was responsible for outbreaks of gastroenteritis on US cruise ships earlier in 2002.

JAMA 2005;293: 1859-60[Full Text]

---

CiteULike    Complore    Connotea    Del.icio.us    Digg    Reddit    Technorati    What's this?

Related Article

Drug for heart failure increases mortality

David Spurgeon
BMJ 2005 330: 981. [Extract] [Full Text] [PDF]