Daily News columnist Dale Courtney recently berated the value of epidemiological models because early reports of projected COVID-19 deaths did not come to pass or because of widely varying projections depending on the assumptions that underly the models. In short, these models did not “get it right.” I suspect that there is a sizeable percentage of the population who sympathize with his frustration.
Trying to calculate an accurate number of cases or deaths for a specific time in the future is, in fact, not what epidemiological models do very well. The reason is pretty simple. While these models are based on well-grounded understanding of how infectious disease “behaves” during an epidemic, minor differences in starting conditions and estimates for rates of infection translate into large differences during the “exponential” phase of an epidemic.
The real value of epidemiological modeling comes from being able to change model variables to figure out how different scenarios might play out.
For example, if we act no better than randomly interacting objects you can figure on a pretty disastrous outcome. If social distancing is 100 percent effective, the epidemic will burn down very quickly. By varying these values in the models, we can begin to understand the scope of the danger and identify ways to mitigate the threat. In the case of COVID-19, the early models had an added advantage of showing the potential for hundreds of thousands of deaths in the United States, which was the only way that Trump was going to start paying attention (likely to be greater than 100,000 deaths as it is).
Courtney touts a Swedish epidemiologist (Johan Giesecke) for his guiding hand behind the Swedish laissez faire strategy to COVID-19. Giesecke has argued that there is no evidence showing that closing schools and other lockdown efforts have a reasonable benefit compared to simple social distancing and washing hands. He believes that by the time this is over, all countries will experience roughly the same overall rates of infections and deaths regardless of lockdown interventions. Maybe he will be right but it will take some time to sort out.
Before getting too enamored with the Swedish strategy, however, it is important to point out that Sweden did not identify its first COVID-19 case until two weeks after the U.S. Two weeks is a long time from the perspective of an exponential curve. Perhaps we will revisit this comparison in about four weeks and see how things are going in Sweden.
One model developed by a data scientist (Youyang Gu) employs a hybrid of classic epidemiological models and machine learning. Although making projections with epidemiological models is a sketchy exercise, so far Gu’s model has been more accurate than most in this respect. Based on this model, Sweden, a country of about 10 million people, is expected to experience 90 deaths per 100,000 cases of COVID-19 compared to 51 deaths per 100,000 cases in the U.S.
In Washington State, which implemented more timely lockdowns, we can expect around 20 deaths per 100,000 cases. South Korea clearly made an exceptional effort to test and tamp-down this disaster and that country has only experienced 250 deaths to date (compare to greater than 67,000 deaths in the U.S.), which translates into a realized rate of 0.5 deaths per 100,000 cases.
So far, it looks like the South Korean strategy in a country of over 51 million people beats the Swedish strategy hands down, and it offers us an important guide for how to safely emerge from lockdown. That is, have the capacity in place for large-scale testing and trace back. We also need financial coverage and job security for individuals who are forced to quarantine after testing (otherwise people may not come forward for testing). Widely available PPE is needed, significant help is needed in the most disadvantaged communities, and we need a cultural shift towards willingness to wear face masks and diligently wash hands.
We all want to get the economy moving, but we shouldn’t have to play Russian Roulette to make this happen. Adopting the South Korean strategy would let us come out of lockdown while saving lives at the same time.
Doug Call is a microbiologist. He first discovered the Palouse 37 years ago.