Sending thoughts from NY here. Hospitalizations are down, yet deaths are up the last couple days; The death toll is mostly representative of the people who were hospitalized in the past 2-3 weeks and did not recover - its sad. It does not account for those who died from the virus at home. The “good news” is that the survival rate is 80% of those who are tested and with a confirmed case; that figure would be larger if everyone was tested.
I am hopeful that our curve will flatten in the coming week and that we can start planning on the next stage of this new normal.
Of course they change daily. Every day at midnight Greenwich Mean Time. It’s a statistical model, for pete’s sake. Each day when new numbers come in at 0:00+ GMT, the data is fed into the model and a new curve is fitted. My god, this should not be new news.
I am speaking of the IHME model that has always included assumed social distancing and has gone from 2.2 million U.S, projected deaths to the current 60,000.
That’s not a predictive model, its a guess or worse yet a political document meant to make you very afraid.
Actually the IHME model is a hospital loading model, not an infection prediction model. Trump is using the IHME model. The “get people to act” models are the ones along the lines of the Imperial College model.
Yes, the IHME model is the one to follow, and the administration is relying on it exclusively. But every day at 0:00+ GMT, the IHME model is fed with data from the prior 24 hours, and the projections change. They have been doing it this way since mid February.
I don’t think you have a grasp of the IHME model. The real-time data isn’t static. It’s not like there’s a set of assumptions that are pushing a curve. This is a hospital capacity model, based on real-world historical performance and ongoing data collection. In other words, you shouldn’t look at the IHME model as an engine that runs on assumptions. It is an engine that runs wholly on real-world data collection. Now, the Imperial College model was definitely one of those engines-that-run-on-assumptions. But not the IHME. The IHME model is fed new data every day at 0:00+ GMT.
March 6 2020
“Anybody that needs a test, get’s a test, they’re there, they have the tests, and the tests are beautiful” “and the tests are all perfect, like the letter was perfect”.
