Back

Defining high-value information for COVID-19 decision-making

COVID-19 Statistics, Policy modeling, and Epidemiology Collective, ; Salomon, J. A.

2020-04-08 health policy
10.1101/2020.04.06.20052506 medRxiv
Show abstract

Initial projections from the first generation of COVID-19 models focused public attention on worst-case scenarios in the absence of decisive policy action. These underscored the imperative for strong and immediate measures to slow the spread of infection. In the coming weeks, however, as policymakers continue enlisting models to inform decisions on COVID-19, answers to the most difficult and pressing policy questions will be much more sensitive to underlying uncertainties. In this study, we demonstrate a model-based approach to assessing the potential value of reducing critical uncertainties most salient to COVID-19 decision-making and discuss priorities for acquiring new data to reduce these uncertainties. We demonstrate how information about the impact of non-pharmaceutical interventions could narrow prediction intervals around hospitalizations over the next few weeks, while information about the prevalence of undetected cases could narrow prediction intervals around the timing and height of the peak of the epidemic.

Matching journals

The top 3 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.