/re-opening_analysis

Modeling Analyses of Reopening Policies for New York City

Wan Yang(1), Sasikiran Kandula(2), Jeffrey Shaman(2)

(1)Department of Epidemiology, (2)Department of Environmental Health Sciences, Mailman School of Public Health, Columbia University

Contact: Wan Yang (wy2202@cumc.columbia.edu)

Acknowledgements:

We thank colleagues at the NYC Department of Health and Mental Hygiene (DOHMH), healthcare systems, and city offices for thoughtful discussions. We also thank the Columbia University Mailman School of Public Health for high performance computing and Safe Graph (safegraph.com) for providing mobility data.

Analyses:

In this series of modeling analyses , we tested potential COVID-19 epidemic outcomes in New York City under different policies with varying re-opening timing, rollback strategies, testing strategies, and occupancy guidelines.
We hope these analyses would be helpful to other municipalities as well.

  • Report 1 (date: 5/26/20; details in folder test1*) analyzed potential COVID-19 epidemic outcomes in New York City for three policies with different re-opening schedules.

  • Report 2 (date: 6/3/20; details in folders test2*) analyzed potential COVID-19 epidemic outcomes in New York City under different re-opening/rollback policies along with active universal infection testing.

  • Report 3 (date: 6/30/20; details in folder test3*) analyzed potential COVID-19 epidemic outcomes in New York City under different re-opening occupancy guidelines and rollback strategies.

Summary findings:

  • In Report 1, we tested different timing of reopening, along with multiple factors that could contribute to the epidemic dynamics. In particular, we tested the impact of seasonality and immunity, and showed a wintertime seasonality similar to other human coronaviruses would likely lead to epidemic surge starting the fall and peak in December or January.

  • In Report 2, we tested different rollback policies should there be large surges in hospitalizations following re-opening. We showed that when there is already widespread community transmission, slow rollback (e.g. only closing bars) likely would not be able to slow transmission and hospitalization surge sufficiently. Rather, immediate and strictest restrictions (e.g. PAUSE in NY) would likely be needed to prevent overwhelming the healthcare systems.

  • In both Reports 1 and 2, we tested phased reopening with full capacity (i.e. 100% occupancy) and showed that it would lead to large epidemics 2-3 months after reopening as well as further surges in the winter.

  • Based on Reports 1 and 2, we found that adaptive re-PAUSE and restriction on occupancy (i.e. operating capacity) are likely both needed to mitigate transmission and keep the City open as long as possible. So in Report 3, we tested different occupancy scenarios particularly related to school opening. We showed that keeping all industries including schools at 50% capacity, plus universal mask wearing, as well as further reduction in transmission from testing and contact tracing, and social distancing etc., the City might be able to keep transmission at relatively low levels through the end of May 2021 with minimal time on PAUSE.

  • Other analyses include testing the impact of active, targeted universal testing and different testing strategies (e.g. pooling samples by household).

CAUTION:

  1. All results are preliminary - please use with caution.
  2. The three rounds of testing/analysis are built on findings from previous round(s) and should be considered collectively.
  3. For consideration/use for locales other than New York City, please take into account local conditions.