BMJ 2 Dec 2006

I guess I have done an above-average amount of literature searching in my time, and I have even taught it to a few hapless students and trainees. The golden rule? If all else fails, look it up on Google. Medline is great but I still get terminally frustrated with Boolean terms and the need to search different spellings (celiac OR coeliac) of the same word separately. Of course with both search engines you need to know how to race through the junk and get at the meat. This entertaining study had clinicians trying to solve those arcane NEJM diagnostic brain-teasers (Cases Records of the Massachusetts General Hospital) using Google, and it worked in over half the cases. Since this study first appeared on the web, it has accumulated a vast number of Rapid Responses, which are also worth skimming through.

Things are not going entirely the way of the Polypillers. Last week, the Walds wrote a rather desperate-sounding defence of folic acid for the prevention of coronary disease; atenolol and bendroflumethiazide have recently fallen from favour as blood pressure lowering agents; and simvastatin even at higher doses than they propose does not alter coronary heart disease or overall mortality when used for primary prevention, according to a meta-analysis of RCTs in the latest Archives of Internal Medicine (Nov 27;166:2307). But this cost-effectiveness study, analysing data from the hospital-based UK Heart Protection Study, produces a figure of less than £2,500 per life-year gained at all levels of risk up to the age of 85, using 40mg of simvastatin. As usual, I shall let others with more time and expertise argue over the arithmetic. Far be it from me to comment on the internal validation of a Markov state transition model.

One thing we can say for certain about statins, however: the existing data are insufficient to justify their perioperative use to reduce the cardiovascular risks of surgery (systematic review of controlled studies).

“Does practice-based commissioning avoid the problems of fundholding?