From the Experts: Tamer Başar, Ph.D.

Tamer Başar, Director of the Center for Advanced Study and Swanlund Endowed Chair and Professor in the Department of Electrical and Computer Engineering at UIUC, responds to two questions he received during the COVID-19 Virtual Summit panel discussion on Monday, April 6.

Question from Deanna DeBord: Was our country’s response to COVID-19 strong and quick enough? If not, should we have known to respond stronger and quicker based on previous epidemics and how can we best prevent the next novel virus from crippling entire countries?

Answer from Dr. Başar:

Thanks for the questions. Unfortunately, our country’s response was substantially delayed, and even then it was only incremental. When it was already known what kind of devastation COVID-19 was bringing first to China and then to Europe (particularly Italy) and the havoc it was creating, we should have learned from the experiences those hot spots were gaining, and at least should have started the planning process for handling the eventual spread of epidemics in the country. This planning process would have entailed, among others, (1) having an appropriate response mechanism in place, (2) making sure that a sufficient number of testing centers are created and distributed across the country according to population densities, (3) making sure that hospitals have the capacity and isolated chambers to handle all cases in their communities in a worst-case scenario,  (4) making sure that there is a sufficient supply of high-quality masks to meet their needs, and (5) making sure that the supply chain for other essential goods and products would not be disrupted. This planning process should have started back in January, and from what I know, none of this was done. Of course, since the virus did not originate in our country, an early effective measure would have been to ban travel into the country, particularly from already-infected countries, since coronavirus is carried by people and not by any other means. If not total banning at early stages, we should at least have applied conclusive testing to incoming passengers and quarantine those who test positive. Again, none of this was done early enough. Yes, we banned travel from China at some point, and much later from Europe (and in stages), but it was too little, too late. There’s evidence that most, if not all of the active cases in the country (except for Seattle) are due to the virus being carried in by people coming from Europe. Once carriers are in the country and are not concentrated at one location, and you allow travel across state lines, you would definitely have diffusion to the entire country. One then has to move to phase two and try to contain the epidemics to the extent possible (and make sure that all the measures I have listed earlier are in place). So, this is the situation we’re facing now.

The above also partially answers your last question, as to how we can prevent the next wave from crippling the entire world. If the next virus is a mutation of the current one, then there will not be a single country where it will originate at, and what we face would be the second phase I discussed above. If it is a totally new virus that originates in a single country, then we will have to do an instant quarantining of that country (to be agreed to by all nations), or relevant locality of that country, and take strict measures not to allow for travel out of that infected region. I hope we have learned valuable lessons from the current pandemics, to be able to handle any future one more effectively.

Question from Kesh Kesavadas: Should states do sampling by random testing?

Answer from Dr. Başar: Thanks for the question. Of course, all those who show symptoms should be tested. Assuming that there’s some excess capacity for testing, but not enough to test everyone (as it was done in South Korea), then I think random sampling-based testing is the way to go, but in two stages. When we talk about random sampling, there’s always an underlying distribution. If no prior information is available, then uniform sampling is the most logical one (that is, prior is uniform). But the outcome of the testing as a result of that uniform sampling may lead to a non-uniform posterior, which is why I advocate a two-stage sampling, where at the second stage the posterior of the first stage is used as a prior for the distribution. In any case, indeed some randomness (but an informed one) should be used in the testing process of those who do not show any symptoms (but could be potential carriers).

Thank you Dr. Başar for your participation in the panel discussion and answering these questions post-summit!