19 April 2004

Aspirin-induced asthma alert

Doctors should routinely advise their patients with asthma about the possibility of aspirin-induced asthma, as it occurs more commonly than previously thought, say the authors of a systematic review. The advice should be tailored to the patient (eg, those with nasal polyps or chronic rhinitis are at high risk of aspirin-induced asthma and should avoid all products that contain aspirin or NSAIDs indefinitely).

Australian researcher Jenkins and colleagues analysed pooled data from 21 trials, finding an overall pooled incidence of 21% (95% confidence intervals, 14% to 29%) for aspirin-induced asthma — about twice as high as previously reported.

Their review also determined that cross-sensitivity to three commonly used NSAIDs (ibuprofen, naproxen and diclofenac) was present in most patients with aspirin-induced asthma. Cross-sensitivity to paracetamol occurred in only 7% of patients, and tended to be less severe.

BMJ 2004; 328: 434-437

Downsize, get sick, and die?

Finnish research has found major oragnisational downsizing (>18%) during the recession of the 1990s was linked to more sick leave in full-time, permanent municipal employees of four towns. It was also linked with more cardiovascular deaths, especially in the first few years. The municipalities downsized to cut costs, but legislation did not allow parallel cuts in the services provided.

BMJ 2004; 328: 555-589

Children’s voices

The days of children being seen and not heard may be over. According to a news item in the BMJ, a report from the UK Commission for Health Improvement says that many children and teenagers want to learn and participate more in decisions about their treatment. Commenting on their level of involvement, children were quoted as saying: "[The doctor] only explained it to my dad", and "[I] did not really understand, nobody explained what is going to happen to you to make you better".

BMJ 2004; 328: 600

Trafficking: the human cost

The Lancet recently examined health and human rights issues relating to the "trafficking" of people for indentured servitude, and especially of women and girls for sexual exploitation. Trafficked people may fear retribution or deportation and avoid health services, despite the threat to their own health (and others, via HIV/AIDS). Well-meaning acts on behalf of trafficked women have sometimes landed them in even worse situations; each woman’s assessment of her situation and risks to her safety should be respected.

Lancet 2004; 363: 564, 565, 566

. . . and would you like some gout with that, sir?

The US Health Professionals Follow-up Study has confirmed that eating meat and seafood increases the risk of gout, whereas consuming dairy products reduces it. Over 12 years, the study found 730 new cases of gout had developed among 47 150 men. Gout was more likely in men who had eaten more meat (at just under two servings a day, on average) and more seafood (a little more than about half a serving a day) than unaffected men. Those who consumed more dairy products, especially low-fat ones (eg, two glasses of skim milk per day), were less likely to develop gout. However, intake of purine-rich vegetables and total protein intake were not linked to an increased risk of gout.

An accompanying article said that gout should be considered part of the current global epidemic of obesity, hypertension and diabetes.

1. N Engl J Med 2004; 350: 1093-1103
2. N Engl J Med 2004; 350: 1071-1073

Clinical statistics put simply

With the current emphasis on evidence-based medicine, we're rapidly reaching the point where all clinicians simply must have some kind of grip on statistics. But perhaps this lies more in understanding what each test is designed to measure and in interpreting what the result will mean in practice rather than in knowing how to do the calculations. For example: when is a median value preferable to a mean? When is a statistically significant result clinically significant? And, what are the dangers inherent in subset analyses? Those seeking a straightforward, practical approach to these kinds of questions might like to look at Guller and DeLong’s vade mecum, prepared for surgeons as a no-nonsense way of launching into a potentially tricky operating field.

J Am Coll Surg 2004; 198: 441-458

— Dr Ann Gregory, MJA