We're human beings: we live in groups. We have friends, relatives and acquaintances. So why ignore that when trying to get people to change their behaviour?
Ten years ago, Christakis used the Framingham cohort to show that obesity spreads through friends: if your friends get fat, so do you. Now he has turned that observation to practical use to see if using friends speeds up the adoption of healthy practices. The answer is yes, but not an unqualified yes. Using friends speeded the diffusion of multi vitamin use in rural Honduras, but not water chlorination. As he points out, swelling a table doesn't require much cognitive input, but using chlorine tablets correctly does. So we need to choose the diffusion method to match the task in hand.
There is a home for everyone in public health. And conversely, every home should have some public health people. Even hospitals benefit from the population perspective - a small but excellent cadre of our colleagues do 'provider public health'.
All that by way of introduction to an interesting piece by Michael Porter - Mr Strategy himself. Here is a flavour of the way he thinks:
"The second essential strategy question is, “What businesses are we in?” In health care, the traditional answers — that we run a hospital or a department — reflect the legacy structure, not where value is created."
It may seem that only private sector hospitals need strategy. Not so - I can think of many NHS hospitals facing huge strategic choices. University College Hospital in London has, for example, declared its strategy to be world class in neurology and cancer, knowing that it cannot be world class at everything. There are children's hospitals which need to decide whether they can be both local general hospitals and regional specialist hospitals. District hospitals need to decide what portfolio of services can sensibly be provided locally, and what should be relinquished. And so on. Exciting stuff!
Papers which confirm what I've always thought are dangerous because they switch off my critique neurons. But even with critical faculties functioning fully, it's difficult to see any major flaws in this paper from Sweden.
Is modest consumption of alcohol better for you than no alcohol at all? A contentious question, especially if you make wine. Observation of people who drink no alcohol reveals a few things:
1. they have somewhat higher mortality than those who drink moderately but:
2. a fair number of them gave up drinking because they were dying of cirrhosis or similar;
3. in some societies, you have to be pretty weird to be teetotal;
4. in other societies, #3 doesn't apply.
We can deal with #2 by restricting analysis to people who are healthy at baseline, but #3 and #4 are more tricky. (For those of you who like it technical, I'm talking about confounding.)
Here is a paper written a few years ago but I've only just come across it. Alcohol in Chinese populations is an interesting question. On the one hand, there is no tax or duty on wine and beer in Hong Kong (abolished 2008 to stimulate the wine trade). On the other hand, a mutation common in Chinese populations makes drinking alcohol a thoroughly unpleasant experience, even if it's Chateau Lafitte 1976: aldehyde dehydrogenase mutations result in the build up of acetaldehyde, headache, nausea, facial flushing, etc.