Interesting news from the CMAJ: giving 12-year-old-girls vaccinations against human papilloma virus is likely, on the best existing evidence, to prevent one case of cervical cancer for every 324 doses administered, leaving an implied cost (at about $350 per) of $113,400 per cancer prevented. But wait—that's on the manufacturer-friendly, untested assumption that the vaccine is 95% effective and lasts for life. Assume that the anticancer protection diminishes by 3% per year, and the number needed to treat increases to 9,080—an expenditure of more than $3 million. And the 95% confidence limits stretch out on the long side to infinity, meaning that if the effectiveness of the vaccine diminishes slightly with age, it's possible no cancers will be prevented at all. But don't worry, Merck will probably give us our money back.