Constantly Wrong Part II - The Past Does Not Have to be Prolog
In my last blog, I tried to explain why using R0 as a constant is incorrect. My focus was on the silliness of applying an unvarying progression to widely different circumstances.
Another misconception about R0 is that it can be used to predict the success of any future intervention. R0 is actually an average of something that happened in the past. It is a type of mathematical history. Unless you assume that all the variables that comprise R0 will be the same in the future, then the R0 calculated from a past event is useless for prediction about a future event.
Consider this example: Disease X infects Freedonia in 1950. The percent of the citizens infected varies greatly from city to city. In Smartopolis, someone invents an intervention that protects 100% of the citizens of that city from infection. However, Smartopolis is a small city and when the number of infections in Freedonia is calculated, the success of the citizens of Smartopolis contributes little to the final average number infected by each carrier: 2.
Fast forward to 2010. Disease X is making a comeback. Some are suggesting that that the successful intervention discovered by the citizens of Smartopolis ought to be tried on a wider scale. However, many models with lots of complex equations are published "proving" that this won't work. All of these models have one thing in common - they assume that once again the average number of infections will be 2.
Would it not be more logical to assume that widespread adoption of the Smartopolis intervention to all of Freedonia would result in an average number of infected of...zero?
The percent of the population infected by the 1918 influenza varied greatly depending many variables including: population density, hygiene, age, general health, etc. One intervention was shown to be 100% effective in preventing infection - movement restrictions. It seems obvious that widespread application of movement restrictions in the 21st century could be just as effective as it was in American Samoa, fancy statistics reliant on a constant R0, notwithstanding. See Empirical Evidence for the Effectiveness of Movement Restrictions for more examples.
One can argue whether or not complete movement restrictions can be imposed. But you cannot argue that if they are imposed in time, that they will not work. Of course they will.
By all means, let us have a debate about what triggers should be used to initiate movement restrictions, who should apply them and how they should be applied. But this debate should include content experts on movement restrictions (law enforcement, the military, the FAA, etc). This important issue has been left in the hands of those with little understanding of what is and is not possible with regard to movement restrictions for too long.
Monday, June 30, 2008
