Papers like this make me happy. If we’re ever going to get to the bottom of causal mechanisms, this is the empirical spade work we need to start doing (IMHO). And really it’s pretty simple to do.
Here the authors, Caroline Rudisill and colleagues, have looked at survey data taken around the time of the H1N1 outbreak to see if individuals changed their behaviour (putatively) in response to the epidemic. The clever part, for me, is that the focus is not on social distancing or other measures often considered useful by Public Health professionals, but rather on misplaced/unhelpful behaviour – here the consumption of poultry products.
First, the authors appear to find that, as the old saying goes, a little knowledge is a dangerous thing. That is, if respondents got one of six questions about human risks associated with avian influenza right, their likelihood of decreasing poultry products rose (note however that only 2% of respondents got all 6 true/false questions wrong). As the number of correct answers rose, the effect reversed and those answering 4,5 or 6 questions correctly had all increased their poultry consumption compared to 6 months prior.
Second, the authors focused on whether the occurence of H1N1 cases in your country was associated with behaviour change. In a model that also contained Knowledge variables, the authors found a strong effect for cases having occured in your country, but in the direction of consuming less poultry (Knowledge remained protective).
All of which suggested to the authors that correct behaviour change is possible amongst sensitized populations (i.e. those aware of the health risk), if the right information is provided:
Once individuals feel more personally at risk or as a risk ceases to be abstract, they become more likely to take measures they believe to reduce their risk exposure. Therefore, there is still a role for knowledge to successfully prevent alarmist behaviors even when there are positive identifications of H5N1 within one’s country of residence or in a bordering country, as long as the virus has not occurred in humans yet.
Obviously there are many limitations to this work – not least the cross-sectional rather than longitudinal nature of the study. But this is a really nice example of applying thought to readily available data to make an empirical contribution.
Aside: Of course, from a social epi viewpoint, I’d be really interested in knowing how all these factors are patterened by SES, in particular education and income (if poultry prices fell, would poorer people consume more? would this vary by awareness levels?), but maybe that’s something for me to look into…
Citation: Rudisill C, Costa-Font J, Mossialos E. Behavioral adjustment to avian flu in Europe during spring 2006: The roles of knowledge and proximity to risk. Social Science & Medicine, forthcoming.
Constrained Optimization 