1/ This is an important point - models are not all bad/good or about prediction but rather tools for guiding what needs to be done
In the Ebola epidemic, we modeled spread in each locality to plan where we needed testing sites & calculate how many treatment beds may be needed
In the Ebola epidemic, we modeled spread in each locality to plan where we needed testing sites & calculate how many treatment beds may be needed
https://twitter.com/DrTomFrieden/status/1303405492883075073
2/ There was a lag between determining a need & getting it established so the best we could do was use modeling to start deployment &, if transmission shifted & changed, we revised our models & adjusted our projections & plans
3/ In many instances, we were part of the way towards establishing a testing or treatment site when transmission shifted differently & they were no longer needed at the place we planned but required elsewhere
4/ This was the nature of responding to a constantly & unpredictably evolving epidemic.
It is also why - for Covid & future epidemics - we need more agile response capabilities that can be set up & redirected more easily (eg, more nimble decentralized testing)
It is also why - for Covid & future epidemics - we need more agile response capabilities that can be set up & redirected more easily (eg, more nimble decentralized testing)
• • •
Missing some Tweet in this thread? You can try to
force a refresh
