Nothing works quite as well in the real world as it did in the trial, but vaccines come close, as this article reminds us.
Of course we mustn't leave our critical faculties behind - if the claim is for an enormous reduction in pneumococcal disease, we need to know how cases were ascertained: how good was the surveillance system? Did it work equally well in all age groups? What about strains not covered by the vaccine - and in particular was serotype replacement a problem? The authors discuss all of the these issues.
Last week a clean, near perfect trial. This week research conducted in the chaotic world of a cholera outbreak. But this is important stuff - as the authors say, we need to know what happens under field conditions. The basic method is simple - a case control study, comparing the vaccination history of cases (patients with cholera) versus controls (people without cholera). But in field conditions it's not so easy to tell whether people actually have cholera or not, and whether they've been vaccinated or not. So there are various extra study elements built in to cope with that.
Acute (i.e. sudden onset) flaccid (i.e. floppy) paralysis has various causes, but one of them is polio. So panic, or at least alarm, is a reasonable respsone to a flood of new cases.
California hasn't had a flood, but they're worried. This short report gives an interesting insight into how epidemiology works.
"Although global morbidity and mortality have decreased substantially, malaria, a parasite infection of red blood cells, still kills roughly 2000 people per day, most of whom are children in Africa."
Here is an update on malaria by Nick White, who probably knows more about it than anyone else. I remember him as a chirpy young doctor with blond hair.
Terry Leahy was chief executive of Tesco. His autobiography was called 'Management in 10 words' and the most important word for him is 'Truth'. Face the truth, face the facts. This is a really important lesson for us. We work in organisations, and all organisations (or at least all the ones I know well) are truth averse - don't admit to any problems in case the newspapers maul you. But without truth there can be no success.
Sometimes, of course, it can be really difficult to discern the truth - the evidence may be sparse, conflicting or obscure. But sometimes the truth is deliberately hidden, with awful and far reaching consequences.
We're used to the idea that we halt epidemics by treating people who are infected and vaccinating everyone else. But what if there is no treatment and no vaccine? Pretty scary huh?
Then we'd have to resort to civic measures. So it's nice to see this account of how well a civic measure worked to contain the spread of H7N9 influenza. Some specifics helped - it seems the virus spreads from poultry to man but not much from person to person.
It reminded me of a paper I blogged about five years ago which assessed the effect of school closures in the 1918 pandemic.
This reads like a detective story. It's pretty technical, but I can just about follow the plot. What I can't do is to critique the methods - so I just have to trust in the Lancet and its referees. (Hopefully lessons were learned after Wakefield's notorious MMR paper...)
I saw a man getting out of a car in London earlier today. It was black and shiny and he looked prosperous. But both his legs were withered and he had to use crutches to walk. I'm pretty sure he had been infected with polio as a child.
The last outbreaks in the UK were in the 1960s, and polio has almost been eradicated from the world, but not quite. Here is a pretty impressive report of how to respond to a case.
A couple of points of interest:
1. although polio causes paralysis, its spread is faecal-oral, so the samples you need for lab testing are faeces.
2. this exercise cost £20m: probably well worth the money.
Xinjiang is an interesting place geographically and politically.
The last government decided to get tough on NHS performance. Waiting lists were driven down by a regime that carpeted chief executives who allowed any patient to wait more than 18 months for an operation. Then attention turned to hospital acquired infection - MRSA and Clostridium difficile.
Hospitals were given, well, not exactly a quota, but a permissible number of MRSA or C Diff infections before the carpet beckoned. And this is taken seriously - every surgeon in the hospital knows what the number is, gel dispensers spouted on all wards and entrances, performance is reported monthly at the Board meetings, and so on.
All of this is jolly good but it's a bit unfair if the number of infections is nothing to do with the hospital. The target is based on an assumption that C diff is transmitted within the hospital by poor infection control practices. But this article challenges that assumption. It's published in the mighty New England Journal of Medicine but based in old England - the Oxford hospitals group. Whole genome sequencing showed how varied the sources of the cases were.
There are some careful definitions of what was regarded as close enough contact in place or time to justify an assumption that one patient had transmitted it to another - epidemiology is rooted in careful definitions. An interesting paragraph deals with evolution of the microbial genome - the estimate is that you could expect up to two mutations (single nucleotide variants SNV) to occur over a 124 day period.
That said, a third of the cases do seem to have been transmitted within the hospital, and ramping up the pressure on hospitals to do the basics of hand washing, isolation and general infection control has undoubtedly improved (or do I mean coincided with?) the national position. So perhaps the carpet is fair after all.