Ann Intern Med 21 Aug 2007 Vol 147

How grateful we should be for those who delve in the dark places of the earth, bringing up ores and jewels for our use and pleasure. Systematic reviewers share the glory of the mining tribe, and many of its hardships; if only by metaphor rather than rock-fall. It is tedious grinding work, throwing away the low-grade ore (but ooh, look, there’s a nugget after all) and patiently following the richer seams. And then it is all brought out into the light, sifted and melted down, and – within a couple of years it has lost its value. Major revisions of a quarter of systematic reviews are probably needed within that time, and cardiovascular reviews date faster than the rest. The iron so carefully smelted turns to rust, and we start again.

Selenium is a strange moony element which is included (perhaps just for its name) in many popular “health products” containing trace elements. Now and again someone claims to prove something in its favour but this study does not. Far from preventing diabetes, as had been suggested by animal studies, selenium supplementation probably increases your chance of getting it.

Here is a systematic review of a cardiovascular topic which is unlikely to date badly, but will probably need revision within those two years. Implantable cardioverter defibrillators undoubtedly prevent sudden death due to ventricular arrhythmias in adults with left ventricular systolic dysfunction. But they remain very expensive and prone to malfunction. Most people fitted with one never receive a discharge but are exposed to all the ICD-related risks. So not only do we need more studies, but we also need better analysis of individual patient data to see which subgroups get the most benefit.