Thanks to this little piece in this week’s New England Journal, Oscar is currently the world’s most famous cat, for his unerring ability to sense the impending death of residents in the nursing home where he was reared as a stray kitten. His purrings and snugglings accompany them to the next world: in a medieval painting, Oscar would be seen slinking out of the death chamber, while the little soul is embraced in mid-air by an angel, and bunch of frustrated demons rages in the bottom left corner. Nor is he alone in his mystical abilities. Our late Burmese cat Wylie (1984-2005), who normally regarded us his servants with complete disdain, would infallibly diagnose illness and lie weightlessly with a soft rumbling noise on the sufferer’s chest until she or he recovered.

Bronchiolitis will be back in a few months’ time, worrying GPs in and out of hours, filling the paediatric wards and causing recurring crackly coughs in infants for the next year or two. And we still don’t have any effective treatment for it – see the editorial. This big well-conducted study knocks out a leading contender: oral dexamethasone. In 20 US emergency departments and 600 children, it made no difference to the course of the acute illness.

If only all cancers were as responsive to treatment as testicular cancer. The great majority are cured at first attempt: most of the rest are mopped up by further chemotherapy, usually with platinum-based drugs: and for the remainder, this study shows that high-dose chemotherapy with stem-cell rescue is usually successful. But nasty: it killed 3 out of 184 patients, while another 3 developed acute leukaemia.

A final little tack in the coffin of rofecoxib, once our most popular COX-2 inhibitor and a potential preventer of bowel cancer. The abandoned Oxford-based VICTOR trial in patients with colorectal cancer showed that even in the first 7.4 months, there was an excess of cardiovascular events, of a magnitude similar to other studies (RR2.7, compared with approx 3.5 in others).

Richard Dawkins has popularised the idea of “memes”, units of cultural behaviour which follow Darwinian rules of natural selection, and this paper extends the idea to the spread of obesity in a large social network over 32 years. Now where do you think this network might be? Yes, it’s Framingham, the supreme seat of all long-term epidemiology: here displayed in blobby diagrams of startling indecipherability. Some of this is due to genes – if you carry a single allele for the FTO gene, you are likely to get fat – and some of it is due to memes – people habituated to modern American eating. You could call this the diseasome, like the editorial: but please stifle the urge.