A new report combines forecast and expert data, predicting that 125,000 lives could be saved by the end of 2021 if 50% or more of the U.S. population start COVID vaccination by March 1 , 2021.
“Meta and consensus prediction on COVID-19 targets,” developed by Thomas McAndrew, a computer scientist and faculty member at Lehigh University College of Health, and colleagues, include data from experts and learned forecasting, combining their prediction in one consensus prediction. In addition McAndrew and his team make a metaforecast, which is a mix of an ensemble of computer models and their consensus prediction.
In addition to predictions related to the impact of vaccines, the report includes predictive analyzes of a number of U.S. COVID-related cases, including a number of cases, hospitals and deaths and frequency of the B.1.1.7 difference, which first appeared in the United Kingdom but is believed to be spreading rapidly in the USA
From the report
- 125,000 fewer deaths predicted by the end of 2021 if more or equal 50% of the U.S. population start vaccination by March 1, 2021: McAndrew finds that if more than or equal to 50% of the U.S. population start vaccination by March 1, 2021 the median consensus forecast is the cumulative number of deaths by December 31, 2021 520,000. In comparison, if less than 50% of the U.S. population starts vaccination the median prognosis is 645,000. A consensus of subject matter experts and learned forecasters predicts 125,000 (difference between two mediums above) fewer deaths due to COVID-19 if at least 50% of the vaccinate population by March 1. 2021 and highlight the importance of increasing vaccination rate. across the US
- Increased expectations in hospitals, cases and deaths: The team finds that a consortium of learned experts and forecasters predicts, for the week beginning January 24th and ending Jan 30th, an increase in the number of pediatric and adult hospital admissions (median = 132,500), an increase in the number of newly diagnosed cases of COVID-19 (median = 1,700,000), and an increase in the number of new deaths due to COVID-19 (median = 22,400).
- Projected increase in variable frequency B.1.1.7: The report shows that a consensus of subject matter experts and learned forecast predicts 87% of U.S. samples sent out for a genomic sequence in the first two weeks of February that has an S-gen dropout (present in all B.1.1.7 samples) will be identified as the B.1.1.7 variant. Currently, according to McAndrew, approximately 22% of samples are identified as the B.1.1.7 variant.
McAndrew’s approach to forecasting is different from the traditional approach, he says. Instead of building a computer model to predict cases, deaths and hospitals as a result of COVID, it will require training specialists and forecasters to predict and provide these targets. together in one consensus prediction.
In addition he makes a metaforecast: a combination of an ensemble of computer models and a preview of the consensus.
The idea is to combine computer models with human judgment to more accurately predict the U.S. revolution.. “
Thomas McAndrew, Computer Scientist and Faculty Member, College of Health, Lehigh University