Timing is the secret of success in real estate, comedy and activities during COVID 19 shelter in place. Over the last week, many of you have asked me advice about when to travel to see loved ones, to get an infected tooth extracted and engage in a variety of other endeavors in between the two extremes.
We are now in our seventh week of shelter in place, with great success in reducing the rate of new COVID 19 cases and associated mortality. We are now in the process of gradually reducing restrictions and opening the economy. This transition creates a paradox: we are easing restrictions because our infection rates are low, yet once we start moving about more, the infection rates may likely rise. How much we don’t know. We have data on transmission rates in a situation of no restrictions or a strict lockdown, but none for the gradual phased reentry models being planned. I can make an estimate by comparing the United States with other countries such as South Korea and Taiwan that have been able to partially open their society without seeing a rise in COVID 19 infections. These countries have higher rates of testing, better quarantine policies and adherence to social distancing than we do. Therefore, I predict we cannot replicate their ability to lift restrictions without increasing the rate of infection.
Thus, if you are debating when to venture out to do something that involves more exposure to others, such as traveling or undergoing a medical procedure, now is the time! You will not get a better deal on infection risk. Of course, you should still try to minimize risk with appropriate hand washing, mask wearing and social distancing.
Transitioning out of a shelter in place mode, raises the larger question of what is going to happen?
No one knows. We are embarking on a huge global experiment. More precisely, we are conducting thousands of smaller experiments as each country, state and county creates its own unique blend of policies. In order to track outcomes, we will need an accurate, systematic way of reporting COVID 19 illness data at the level of local jurisdictions. Despite its name, the CDC (Centers for Disease Control and Prevention) does not adequately track and aggregate COVID 19 infection data. Instead, academic, philanthropic and private institutions have assumed the responsibility along with many other vital pandemic tasks.
I will present an annotated bibliography of four COVID 19 health outcome tracking resources and describe more epidemiological concepts to give you a framework as we confront the COVID 19 pandemic and open the economy simultaneously.
The Contra Costa County Health Services website provides local data on how we are doing in terms of new cases, total cases, rates of death and hospitalization.
https://www.coronavirus.cchealth.org/
The Johns Hopkins Coronavirus Resource Center is the most complete data for the world, with aggregate totals for countries, states, counties, cities with the ability to compare infection rates, case fatality rates, testing rates and much more. It is the definitive source for comparing results across the world. The amount of data can be overwhelming and depressing.
https://coronavirus.jhu.edu/map.html
One of the best sites that I have found for comparing data across states is the Rt.live site (https://rt.live/put together by Kevin Systrom, founder of Instagram and Thomas Vladeck. They graphically model how Rt (derived from Ro) changes for each state over time. To understand the power of their data we need to discuss the epidemiological concept of Ro(pronounced R-naught). https://www.nytimes.com/2020/04/23/world/europe/coronavirus-R0-explainer.html
Ro represents the number of new infections estimated from a single case. It combines the biological properties of the infectious agent AND the behaviors of the host population. For example, a highly infectious virus (such as measles) may have an Ro in the range of 6-9. While a less infectious virus (such as the 2009 swine influenza) may have Ro = 1.4-1.6. What is the Ro of COVID 19? With no social distancing, the current estimate is 2.0-2.5, with aggressive measures well under 1. As already mentioned, we don’t know Ro for a partial lifting of a shutdown. One is the magic number. An Ro greater than 1 means growth overtime, while less than 1 indicates a reduction in cases overtime.
Unfortunately, in the United States, the approach to COVID 19 has been mired in partisan battles with Republican led states pursuing less aggressive social distancing measures, often citing low absolute numbers of infections. The Rt.live website provides graphic evidence of the importance of the rate of change in cases, not merely the total numbers. It also, clearly demonstrates differences between states, allowing for the effects of different policies to be compared.
The best site for predictive modeling is from The Institute of Health Metrics and Evaluation (IHME) from the University of Washington. Political leaders rely on their predictions to try to ensure enough hospital and ventilator capacity. The site was recently updated to include anonymous cellphone GPS data measuring how well people are adhering to shelter- in- place rules. Other sites chronicle what has happened, the IHME site provides the best estimates of what might happen.
These websites provide good metrics for following the experimental results on transitioning from shelter-in-place. I will conclude by describing the probable timeframe for the data to arrive. The Ro statistic provides an estimate of how many people infected per reproductive cycle, for COVID 19, around 4-5 days. Thus, it may take a handful of reproductive cycles for trends to be apparent. If we are fortunate enough to have an Ro less than one, we will see continued decline in number of infections beginning immediately.
However, if we have an Ro greater than one, it will take longer for the bad news to be measured. There will be delays in detecting infected individuals. We still lack any broad-based screening for asymptomatic infections. Symptomatic patients will be preferentially detected after several days or a week of symptoms, with an additional week or so of decompensation to be detected in the count of hospitalized patients. COVID 19 patients who die, often do so after several weeks of hospitalization pushing out the mortality data another several weeks.
It may take up to 3-4 weeks before any differences are detectable across states and several months before the maximum differences are demonstrated. We can all follow along and hope for the best.
Given the future uncertainty, I would recommend using this time at the end of our shelter-in-place to pursue any elective procedures or travel plans.
Please contact me with any questions or concerns.
Thank you,
Barry Rotman, MD