How Data Collection During the Country’s Reopening Can Accelerate Return to Normalcy

Print Friendly, PDF & Email

There is light at the end of the coronavirus tunnel. States like California have released their phased reopening plans and others, like Georgia, have already started reopening their states. In these phased responses, the first phase has typically consisted of reopening  lower-risk businesses and venues. Voluntary compliance with virus transmission mitigation practices like hand washing, social distancing and face mask use has been in everyone’s best interest and will continue to be since a resurgence of cases may lead governors to reinstate shelter-in-place orders, effectively squelching any economic rebounds. With that said, every state reopening represents an opportunity to collect data and identify best practices that can benefit other states.  

As businesses open, and with more FDA approved tests now available, the opportunity exists to test more people in an open environment, which will be the status quo for the next year. To date, CDC guidelines have focused testing on people who show COVID-19 symptoms. Such testing should certainly continue, and anyone who displays upper respiratory symptoms should assume that they carry the virus and self-quarantine. However, in an open business environment, tests that assess whether people carry the virus but are asymptomatic yet contagious are most useful, since they effectively act as silent virus transmitters. Recent reports of testing on naval ships, in homeless shelters and in prisons suggest that the ratio between asymptotic and symptomatic infected people may be as high as 10 to 1. If these results are any indication, people without symptoms should be the target of this testing – which will essentially give us the means to assess community risk and identify best practices.   

The number of tests required to make such an assessment is surprisingly small. Carefully selected random samples of 1,000 people would lead to population risk estimates that are within a three percent margin of error  of the true value, with 95 percent confidence. Better yet, increasing the sample to 10,000 would reduce the margin of error to one percent.

However, nothing is without complication – ethical issues often stall the design and implementation of randomized experiments. Despite this, researchers and public health officials can still utilize observational data which can be collected for analysis. The goal is to provide best practices and sufficient evidence to support such practices. Data samples should target a wide swath of dimensions and areas within a community to gain insights into infection rates associated with various types of businesses and the effectiveness of these practices in such settings.

Restaurants and bars pose a different set of challenges to reopening. Experts need to figure out  the best seating configurations and service practices to effectively limit virus transmission. This data will also allow for the assessment of social distancing and face mask use to reduce transmission risk. In addition, a random sample of people in a community need to be identified for frequent testing. These residents will record their daily activities, focusing on when and where they visited. Data as such will be invaluable to other states as they reopen their economies and identify what appeared to work and what activities were problematic.  

The dimensions across which data can be collected are unlimited. Statistical methods provide tools to make sense of such data, and provide insights embedded within such data. The coronavirus is unforgiving. Once a community is reopened, the virus will penalize any lapses in hand washing, social distancing or face mask use with spikes in cases and accompanying deaths.

The economic reality is that every state, at some points will need to traverse the unchartered path of reopening. Georgia and other early opening states can provide an invaluable service to the nation by conducting extensive data collection and broadly sharing their findings for the public good. With such data and information, each state can use best practices and evidenced-based methods to traverse the path that can accelerate a return to normalcy. 

About the Author

Sheldon H. Jacobson, PhD, is a Founder Professor of Computer Science at the University of Illinois at Urbana-Champaign.  He applies his expertise in data analysis and risk assessment to evaluate and inform public policy. He is an active member of the Institute for Operations Research and the Management Sciences (INFORMS). 

Sign up for the free insideAI News newsletter.

Speak Your Mind