I’m frustrated by the apparent unwillingness of officialdom to provide a forecast for swine flu.  I appreciate that there are reasons for that.  I’m peeved since, I think, the leading reason for their reluctance is fear, fear for their reputations.  The profession’s forecasting skill is lower than weather forecasters; so the probability they will get it wrong is very high.  To hedge the risk to their reputations they hedge the predictions.  So much so that no mortal can actual tease out a sense of what’s likely to happen.  I don’t think they are doing their job.  I gather that weather forecasters have guidelines, so when the hurricane in on the horizon their write a forecast who’s intent is to maximize some things.  Those things might be accuracy, or lives saved, or long term trust – but they have rules.  It appears that the pandemic community lacks rules about this and so we aren’t getting a clear forecast.

At this point we have some facts.  It appears that this flu is currently fatal for about .45% of those who are infected.  People talk about this flu being approximately as severe an infection as the usual seasonal flu.  What’s that mean?  Here’s one data point:  there have been 29 deaths in Canada, and 663 people sent to  hospital.  We can do the arithmetic .45*(663/29) imply that 10.3% hospitalization.  At this point we should have plenty of data to predict what percentage we would hospitalize, assuming we have the infrastructure to do so.

What got me thinking about the absence of a forecast was this headline grabbing statement by a health minister in Britain.  He floated and then immediately backed off of a lame back of the envelope calculation (a naive exponential interpolation) that predicted a hundred thousand new cases a day by late August.    But certainly there are better models of how a virus spreads.  I presume that if we set that aside the difficult question of the virus mutating those models can spit out a forecast.  So, what percentage of the population is going to get this?  What percentage are going to miss how many days of work?  Etc. etc.

A forecast is a precursor of good planning.  There are some forecasts, really bleak forecasts, for what might transpire if bird flu were to mutate into something more contagious while retaining its current virulence.  The plans those bleak forecasts trigger are extreme, i.e. be prepared to bunker down inside your house for three months as society unravels.  That creates a problem; what plan should people prepare to implement now.  Some would argue that they should prepare to implement that extreme plans.

Which puts in a situation where a line of really bad thunderstorms are on the horizon, with a serious risk of tornados, and people who want to prepare can only find advise for how to hunker down for a category five hurricane.  This is creating a situation where a lot of folks will do nothing.