On March 19, the Office of Graduate Studies and Research at the Lebanese American University (LAU) released the results of a study (see below) on the effects of different levels of social distancing on the number of COVID-19 related deaths by June in Lebanon. One note of caution for these figures is that the lack of data and predictive model used means these results are speculative, based on an analysis of trends. Though the team behind them did consult with LAU’s infectious disease unit when modeling the scenarios—and there are plans for collaboration across disciplines and universities in the future on the effects of COVID-19 on Lebanon—at the time of writing data is lacking. There is also a high likelihood that the number of cases in the country are above the official count, according to the researcher behind the study, Samer Saab, interim dean of graduate studies and research at LAU. He told Executive that the numbers he has produced, if anything, could be considered a conservative estimate, given the trends being witnessed in countries such as Italy (see Q&A below). The primary purpose of the study, he says, was to raise awareness among the public and policy-makers of the effects of social distancing and to underline that this situation would not be over in the short term.
LAU Coronavirus & Social Distancing Study
Forecast of the daily number of infected cases and daily recoveries due to the new Coronavirus (COVID-19) in Lebanon (Graph 1). The forecast is projected in a window of 112 days, beginning with March 19 and ending by end of June 2020 for three scenarios under consideration:
FREE-FOR-ALL: No social distancing measures are followed, and normal life is assumed;
MODERATE DISTANCING: 1 out of every 4 people moves freely or resumes normal life;
EXTENSIVE DISTANCING: 1 out of every 8 people moves freely or resumes normal life. Forecast of the daily number of deaths (Graph 2).
The forecast is based on time-series analysis and uses the data before March 19 to predict the number of recoveries. The employed model assumes an initial death rate of 3% of the total identified cases, which increases to 6% once the total number of critical cases requiring intensive care exceeds 250 patients. The latter reflects the limitations of the healthcare system. The forecasted results do not consider confounding factors (such as lifestyle, environmental factors, consanguinity, etc.); however, the model is applied to Wuhan reported data and is shown to closely capture its trends.
Source: LAU’s Facebook page.
Executive spoke over the phone with Samer Saab, interim dean of graduate studies and research and professor of electrical engineering at LAU, to learn more about the findings of the study, its limitations, and its implications.
Can you please briefly describe the study?
These are projected or forecasted numbers where I am looking at the trends; it is not exact science. I put forward three scenarios: if everyone will go out [the free-for-all scenario], if one in four people go out [the moderate distancing scenario], and finally if only one in eight people go out [the extensive distancing scenario].
The free-for-all scenario is where people decide to live their lives normally. If we had had an infinite capacity of hospital beds and intensive care units, then we would have a 3 percent death rate [under this scenario] but we don’t have that.
I am assuming that when the number of infected people who need treatment in intensive care units (which is typically 10 percent of those infected) goes beyond 250, then the death rate would be 6 percent [due to the lack of treatment for COVID-19 and assuming that some hospital beds will be needed for other urgent cases].
How did you arrive at 250 persons as the tipping point?
This is an estimate, based on what we hear about many of the hospitals not being prepared. If I change it a bit here and there, it is not going to change the numbers a lot. It’s more to give an idea.
The advantage of the free-for-all scenario is that after June/July, you don’t have to wait for a treatment to be developed because most people would have been infected and those who survived would likely develop immunity—though at this stage there are still many unknowns, like if the virus will be seasonal. Many governments initially tried to follow this scenario, as it would have the least impact on the economy with people able to work freely. Thinking has changed, however, as seen by recent measures in the UK to shut down schools despite initial resistance. The disadvantage of the free-for-all scenario is simple: Many people will die and health care systems will become overrun.
Whereas in the second and third scenario, if we maintain distancing until June/July, then it is possible that a treatment would have been developed and would be accessible to most people. If they get the virus then, they would be able to take medication and not suffer as much.
So this is the message we are trying to send to people: Please stay home and let us buy time. We are also trying to send a message to the decision-makers so that they don’t think that we will be able to open schools and universities and basically go back to normal in the next two weeks.
What’s the methodology used for this study? What model are you basing it on?
I used time-series statistical analysis [to mark data trends over time]. However, it is not straightforward like a typical time-series: Whenever more people get infected, the rate of spreading becomes lower and lower. In urban areas—which is the case for Lebanon as most of us live in urban areas—each person can infect three people, while in the rural areas the rate is less (1.5 persons). But at the end, it doesn’t matter much because it only affects how fast we will reach the peak and not the figures themselves.
But, there are a lot of factors I didn’t take into consideration. A lot of people will still go to work or go out if their family is relying on them to not go hungry, right?
How feasible do you think it is, both psychologically and economically, to practice extreme social distancing until June?
Excellent point. I cannot answer this question but I feel there are positive vibes and people are helping each other; they will not let people starve. But there will be bilateral damages, such as to the economy.
Still, 153,885 deaths in the free-for-all scenario versus 454 in the extensive distancing one is no joking matter. If our hospitals are overwhelmed, then our doctors will have to play God and decide who dies or not, which is catastrophic.
Were you able to account for the issue of under testing and potential under reporting in Lebanon in the study?
Because we are publishing under the name of the Lebanese American University of Beirut, we had to rely on the official data reported by the government.
However, what would change—if we assume that there are more cases than what is reported—is the date of the peak. We would reach the peak faster if we had more cases than what is reported.
What are the limitations of the study?
As I said before, there are many assumptions I had to take. The methodology I used does not take into account environmental or lifestyle factors, which could mean that more people don’t comply with distancing. The psychology factor is also important here: Will people stay home until June? They will probably cheat. People think the second case scenario is reasonable or doable (in terms of quarantine measures) but 4,259 people would still die in it and that is a big number.
We don’t know much about the virus and there are many unknown factors related to it so we need to stay at home and be patient. This outbreak is not going to end soon and the worst is yet to come. We need to buy time to learn more about the virus so less people will die.
This interview has been edited for clarity.