Most weeks the BMJ, like most other journals, adds small bits of knowledge to what we already know. But we also accumulate more things that we don’t know—and this week’s issue has some important bits of ignorance.

We still don’t, for example, know how many cases of variant Creutzfeldt-Jakob disease (caused by exposure to bovine spongiform encephalopathy in cattle) there might be. The study by Jonathan Clewley and colleagues tested 63 007 tonsils and detected no disease related prion protein, but the confidence interval was 0 to 289 cases per million, lower than but still consistent with an earlier prevalence …