Thursday, April 30, 2009
Malaria, swine flu and those cunning Chinese
I watched a fascinating programme at lunchtime today and it brought back to me the perennial debate in research between inductive and hypothetico deductive approaches to scientific discovery. Having exhausted most of a book on the topic, it is not in my mind to go over that ground here except by being as cryptic as the desire for reader accessibility will allow. In essence, if you want to solve some problem in the world, you can go and collect evidence from sites of interest (eg people actually suffering from malaria) seek patterns, try remedies from folk lore and from existing knowledge and build the most efficacious medical approach you can. Alternatively, you can establish a scientific model, push it this way and that in the laboratory, try it on people and refine it over time. Inductive means ‘discovery, grounded in context’ and deductive means ‘testing existing models’. As Thomas Kuhn pointed out, getting scientists to change their paradigms is a task beyond mere collection of contrary evidence. Their professional lives depend on what they have achieved, erroneous or not.
Two competing forces of imperialism are head butting each other over the treatment of malaria in Africa. The Chinese claim to have eradicated it from some small populations by adopting the inductive approach. At the same time as they offer hope and a malaria free future to village people, they gain access to mineral wealth. The World Health Organisation will not accept Chinese claims because they are not published in reputable medical research journals and peer reviewed. Their imperialism is one of professional reputation and protective approaches to medical disciplines.
A nice zen story involves scientific sampling. Statistics depend upon using a sample from a population and generalising from the knowledge gained from that sample, to the whole population. Zen asks, ‘how do you sample a house?’. Is a brick a sample? A window? A brick and a window? In other words, a house is not conducive to sampling. It is too complex. Drugs that pharamaceutical firms offer are based on statistical sampling. They are nearly always crude in their appropriateness for complex diseases in whole populations. Think thalidomide. Think Viagra. Look at the side effects of long terms studies of individual drugs. It’s a game of poker involving percentages.
Science does not like the unprovable even if the results are astounding. The Chinese think they have won a battle in the fight against malaria. The WHO is snooty about it. Here in Africa people want to be malaria free. If god, the devil, witchcraft, dog’s piss or some bitter tasting plant gives a cure, they don’t ask if the treatment has been peer reviewed.
As I write, swine flu is moving more like a virus carried by the media, itself, than something biological, into pandemic status. After it is over, there will be miracle stories of cures that will sink without investigation, particularly if they are associated with efficacious concoctions of diabolical Chinese herbs. Meanwhile, tamiflu will have been dolled out to everyone who sniffs and sneezes, paranoia will reign and the scientists will be busy isolating and manufacturing a lucrative successor to tamiflu which will be operational in six months time. And will receive prestigious column space in the Lancet.
No comments