Beware of strange black rats and other stories . . .

Minerva looks forward to the coming Golden Age of medicine, where every decision will be shared with individual patients according to the likely benefits and harms of each treatment. But how can you possibly quantify those harms and benefits for individuals as opposed to the herd effects in randomised trials? A paper in the European Heart Journal (2014, doi:10.1093/eurheartj/ehu004) shows that estimates relating to cardiovascular outcomes can be refined considerably by looking at subgroup data from existing trials and meta-analyses. But much more work; better identification of outcomes that are important to patients; and, above all, the incorporation of prediction models into dialogue with patients will be needed before the Golden Age can happen.

The aim of treating opioid addiction can be either abstinence or maintenance. Trying to mix the two just creates confusion. This is nicely shown in a recent Cochrane review (Cochrane Database of Systematic Reviews 2014;2:CD002207, doi: …