top of page

With this massive nationwide surge in cases and hospitalizations, is it time to hunker down again?

Updated: Nov 13, 2020

My collaborator and leader of PolicyLab, David Rubin, was on Fox29 Philadelphia this morning for an extended time explaining why he recommends pulling back to virtual learning in schools for the next few weeks in the Philadelphia area, and also closing bars and reducing capacity in restaurants, etc. Here is the video, and here is an accompanying blog post put out by the PolicyLab team.

Many of us are tired of pandemic restrictions, and want to get back to living our lives, especially with the Holidays approaching, and we are glad our kids are finally experiencing in person school, even if just a couple days a week. Is he just being an alarmist, or is he right in urging us to pull back right now?

This recommendation is not really about the schools, it is about the current surge most of the USA is experiencing. In the past few weeks, we are seeing exponential growth in cases in many parts the country. Last month we saw this in the upper midwest and mountain regions, areas starting to experience colder weather. In the past few weeks we have seen this surge accelerate, with Chicago experiencing some of the steepest growth and highest case rates of the pandemic, and the phenomenon repeating throughout the country, with the rest of the midwest, Great Plains, Pennsylvania, and New England starting to experience real surges for the first time, and then previously hard-hit areas like California, New York, and New Jersey showing clear evidence of a second surge. Only the deep southeast seems immune to this surge, and even they are experiencing upticks, and a second surge may follow once they get a little more cold.

This is a first for this pandemic -- in the Spring the surges were relegated to New York, New Jersey and Connecticut with some smaller surges in Washington, Chicago, Michigan, and Louisiana, and then in the Summer the massive surge was predominantly experienced in the south and west, including Arizona, Florida, Texas, and the deep southern states of Alabama, Mississippi, Louisiana, South Carolina, Georgia, and Tennessee. Never before have we seen widespread surges throughout such a large swath of the country, but we are clearly seeing it now. The magnitude and accelerating nature of this surge is reinforced by the high testing positivity rates (>9% for much of the nation, >15% in many places) observed in this past week, which also make it clear that this increase in cases is not simply an artifact of increased testing.

These surges were predicted by the PolicyLab models, which have shown that colder temperatures lead to more transmission as people move inside and the virus spreads more efficiently as it remains airborne and accumulates better in cold, dry climates. Combined with the pandemic fatigue that sees less community compliance of masking and distancing precautions, all factors are in place for a massive surge all over the country. Indeed, our predictive models show exponential growth predicted for many counties -- see bottom of blog post for some examples. With an already burgeoning surge throughout the country, the upcoming holiday travels and multi-generational gatherings threaten to add more fuel to the fire and have the potential to exacerbate the viral spread to a level not yet seen during the pandemic, in both intensity and geographic expansiveness.

In the summertime, one positive emergence was a reduction of severity of disease, as the summertime surge produced relatively fewer hospitalizations, ICU stays, and deaths per case compared with the Spring. In past blog posts, I have discussed reasons why this might be, but one reason has to do with the weather ... the heat and humidity resulting in less viral accumulation in the air, and thus individuals exposed to lower viral loads and making them less likely to develop severe disease. The shift to colder, drier air as we have moved through fall towards the winter may partially reverse this effect, and we might see more severe disease rates than experienced in the summer.

Indeed, armed with nation-wide county-level hospitalization data from HHS that PolicyLab has secured and has just been made publicly available, we see a surge of hospitalizations that parallels the surge in cases, and we are seeing it all over the country. Hospitals and ICU's are filled to capacity and overflowing in many areas of the country already, e.g. as you can see in MainLineMama's outstanding web dashboards, PA hospitalizations have more than quadrupled in the past 6 weeks.

The burgeoning surge throughout the country that is about to be further fueled by holiday travel is clearly accompanied by a surge of hospitalizations that appears to be of greater magnitude than what we saw during the summertime surge in the south. This raises the possibility that hospitals may start to become overwhelmed in short order in many parts of the country. If Pennsylvania experiences another four-fold increase in hospitalizations in the next 6 weeks, the system would certainly become completely overwhelmed and may lead to the horrors that were fortunately largely limited to Wuhan, northern Italy, and New York City in the spring, and this vulnerability is there throughout the northeast, midwest, plains, mountains and western states. This is set up to be repeated in many cities all around the country.

And if that happens, it will be much worse that anything we have seen, because it will not be just one city or region experiencing the hospital capacity shortage, but many simultaneously all over the nation. So it would not be possible to transfer patients to nearby regions, since they too would be at or over hospital capacity. This could lead to the dreaded rationing of care and increased death rates that plagued the hard hit regions in the spring pandemic. For this reason, we might be moving into the darkest and most dangerous period of the pandemic in the next two months, as you may have heard numerous experts and commentators, including Dr. Fauci and new pandemic response team member Dr. Osterholm, say.

In light of this, NOW is the time for us to take steps to mitigate this surge, to "flatten the curve" again, to reduce the number of cases and hospitalizations in our areas in the next few weeks. If we wait until hospitals start to be overwhelmed, it would be too late to stem the tide, since in that case the requisite hospitalizations emanating from the previous 2-3 weeks of massive surge of infections will already be impossible to stop, and the system could be pushed over the edge. You could think of this as an approaching hurricane -- you don't know if the hurricane will really hit you, but if you wait until you feel the winds to evacuate, it is too late and you are stuck in the storm.

Thus, we need to remain vigilant, still avoiding crowded indoor places, especially with poor ventilation, and being sure to practice distancing and mask wearing when in public around others. Now might be a time to cut out unnecessary travel, dining and shopping. As David suggests, it might be prudent for schools to consider switching to virtual for several weeks to try to curb this surge and ensure it does not get out of control and lead to overwhelmed hospital systems, especially with upcoming holiday travel, or at least limiting in person schooling to smaller children and those with special needs who cannot be effectively educated virtually. And we are all more eager than ever to get together with our loved ones for Thanksgiving, Hanukkah or Christmas, and New Year given how isolated we have been this year, but in light of the surge we are experiencing it is crucial for us to be careful. Maybe we need to avoid the larger gatherings this year, and stick to smaller family groups, while being extra careful to protect our vulnerable family members who are older or have pre-existing conditions who are at risk of more severe disease or death if infected. We should focus on getting through this Holiday season with our families intact and unscathed.

The mitigation strategies we have been following in the USA -- decrowding, distancing, and mask-wearing -- have kept the viral spread relatively under control in many areas of the country in the summer and early fall, further supported by the hot temperatures and outdoor gatherings that have served as an additional mitigation factor. We may have grown comfortable or even complacent with this. Now that we are moving into colder weather and spending time indoors, if we practice the same level of decrowding, distancing, and mask-wearing, that will not be enough -- it will produce the surge that we are starting to see. Since we are losing the mitigating effect of warm weather and outdoor gatherings, it is necessary for us to ramp up our other mitigation strategies to compensate if we hope to keep spread under control in the same way as we move into the cold winter.

Those who have followed my blog know that I am a big proponent of getting kids in school and of mitigating viral spread sans lockdowns -- I strongly acknowledge the hidden collateral damage to people's finances, health and well-being caused by stay-at-home orders and school and business closures. But now we are at a time when we might need to sacrifice for a month or two to get things under control.

I can't emphasize enough how promising this week's vaccine results are -- they truly indicate a breakthrough, and have experts convinced that a vaccine can work, and will be with us shortly. I fully expect an emergency FDA authorization of the BioNtech/Pfizer vaccine to be approved by early December, and this vaccine's distribution commenced this winter, and likely with other approved vaccines following behind. I expect this to get the pandemic under control by next summer, with most of us receiving the vaccine in the next 6-8 months. As a statistical data scientist I am a very skeptical person, but think these results this week represent a breakthrough not just for this particular vaccine but for other similar ones. I really believe this will work now. It is OK to feel some optimism coming out of this report.

But the next 6-8 weeks are shaping up to be the most horrific and dangerous in the pandemic for much of the USA. It would be tragic to carelessly allow a large number of preventable deaths or long term sickness just before we get help from a vaccine. We need to not grow weary in our vigilance, but continue taking care that we protect ourselves and loved ones from the devastating effects that many experience from this virus. We can see the light at the end of the tunnel, so let's not give up but keep fighting.

Below are some plots of case counts and PolicyLab projections for some of the highly surging counties from all around the country. You can find them for counties all over the country at this link, which is updated each week, and also includes time trends and projections for the county-level reproductive number R. We have never seen anywhere close to this many counties with exponential growth and projected accelerating growth in so many different parts of the country at one time since this modeling was started in mid-April. This is a primary reason for the concern I am trying to convey here, and for PolicyLab's recommendations.

4,544 views3 comments


John Erwin
John Erwin
Nov 13, 2020

Yes, robin! Unfortunately for some the only proof will be to see hospitals overflowing. The best we can hope for is that enough people heed the warnings which then just gives more fuel to the conspiracy theorist to say “see I told you so!”


I am becoming more and more tired of 'business as usual'-rationalizations. If it hasn't affected you or yours, you will think that this is a noisy picnic.

Without becoming too alarmist, Many people will be dead after the Holidaze, because they could not see that random destruction by invisible particles is NOT political hype.


John Erwin
John Erwin
Nov 12, 2020

Thanks Jeff! Are you aware of any visualizations that illustrate the lifecycle of an infection? I’m curious to see how cases typically progress from initial diagnosis, to recovery and/or hospitalization / icu admittance and death. Seeing these progressions by age as well as factoring in dimensions for underlying conditions would be very helpful. We’re seeing more cases in the younger population (which might be due to more testing) and relatively fewer hospitalizations and fatalities (for now at least)

Post: Blog2_Post
bottom of page