Category: Blog

Over the past few days, I noticed people in my neighborhood more relaxed in their social distancing practices. Probably a combination of removal of shelter-in-place orders from the governor, perception that the numbers of infections and deaths in Georgia are going down, and general fatigue from the social isolation. I’ve been tracking the published data on the Georgia Department of Public Health, which is very helpful for tracking cases over time, but does not provide the trend in deaths or hospitalizations.  I downloaded the data from https://covidtracking.com/api to create a few plots for Georgia: daily and 7-day moving average of deaths, total weekly deaths, and daily increase in hospitalizations. Conclusion? We’re at a plateau with a slight downward trend on all metrics. I believe the plateau will last a while because while people have relaxed their social distancing measures, they’ve changed their habits; most people are wearing masks in public indoor spaces, and still maintaining a reasonable adherence to safety guidelines.

The total shutdown might not be the optimal solution that we could devise… but it might be the best we could do given where we find ourselves today. If we agree, and we must agree, that human lives are infinitely more valuable than anything else, and then start to evaluate other needs and resources, we can potentially see a path towards coming up with a solution that does not simply blanket us all with this heavy closure of life as we know it.. or used to know it.
There are many variables that differentiate each inhabited geographical area from others, whether they are population density, hospital capacity per-capita, or type of prevalent economic activity, etc. All of these factors can play a significant role in creating a differentiated approach to dealing with -and mitigating -the spread of COVID-19. National restrictions on certain modes of transport for example, combined with a varying lengths of local “shutdowns” could help us fair better on the short term and the long term.
That said, we find ourselves today caught flatfooted and racing against a much faster enemy. Knowing what we all say about hindsight, we would have been helped tremendously if we had started a collaborative effort to statistically model the different variables and elements of our response that can be differentiated to fit a region or a metropolitan area.
Statistical models that are required to factor in the human behavior are notoriously difficult to devise and are generally considered to be inaccurate. When you have to factor in the behavior of a novel pathogen and extremely scarce data, the reliability decreases further but that would still perform better than disconnected decisions made independently by politicians who are having to consume new variables on the fly.