“When will this pandemic be over so we can return to normal life and an open economy?”

As researchers who have been carefully following this global health crisis, that’s a question we are often asked by friends, family, and colleagues.

While no one can predict the future with high accuracy, many experts are optimistic that the vaccination programs now underway will quell the pandemic. The question remains as to how fast vaccination can help local and global populations reach a collective level of immunity that will put the Covid-19 pandemic in our rearview mirror.

advertisement

In an ongoing effort to help safely reopen businesses and universities during the latter half of 2020, Verily, the company we work for, has deployed the Healthy at Work program at institutions such as Brown University and the University of Alabama, and companies such as Waymo. This program encompasses screening students or employees for symptoms, using PCR testing to detect active infections among symptomatic and asymptomatic individuals, and epidemiological modeling to understand the spread of the disease spread.

This modeling has helped employers and university officials understand how often employees and students should be tested as a function of recent infection rates in the workforce or at school, the prevalence of Covid-19 in the local community, and other related factors.

advertisement

A key finding has been that when a community is suffering a large outbreak, it becomes difficult for a typical business or campus in that locale — even a cautious one doing protective surveillance testing — to completely avoid infections. That’s because exposure to the virus away from work or campus is a key factor in the overall risk. Notable exceptions are rare, like the NBA, whose success can be largely attributed to creating a “bubble” that minimizes interaction between those inside and outside the bubble.

Looking ahead, we have applied a probabilistic agent-based version of our model to provide insight into potential future outcomes that are applicable at all levels — an individual business or university, a county, a state, and the like. Critical determinants of what may happen include:

  • The degree of individual and group-level nonpharmaceutical interventions (NPIs), such as mask-wearing, hand-washing, and social distancing, adopted to minimize transmission
  • The fraction of the population with immunity conferred from previous infection, which is uncertain due both to the unknown historical prevalence and not knowing what fraction of people remain immune after infection or for how long (although promising data from the National Cancer Institute show that having antibodies to SARS-CoV-2 is associated with decreased risk of future infection)
  • The rate at which the population is vaccinated, which depends on critical issues of vaccine availability and willingness to be vaccinated, as Zach Nayer recently described in STAT
  • The real-world effectiveness of vaccines approved by regulatory agencies such as the FDA, including the prevention of symptomatic and asymptomatic infection
  • The degree to which it is still possible to set up surveillance testing to be performed at regular intervals for entire populations — not just those who are symptomatic or known to have been exposed — to detect and isolate infected individuals

Despite the significant uncertainty, there is still a wide range of promising scenarios in which we could beat back the Covid-19 pandemic. Conceptually, this is aligned with the “Swiss cheese model” of defense against the pandemic that virologist Ian Mackay has put forward. The right combination of interventions and tools could return us to work and school much faster if we use them all optimally.

A key question is this: What is the definition of an acceptable rate of suppression of viral transmission that would lead to safe return to work and school? Based on our experience, when the fraction of the population infected at a given time (known as the active case point prevalence) is 1% or less, the situation is manageable and acceptable. Suppressing infection even further requires escalating levels of resource allocation and has a dramatic adverse effect on economic and educational activity.

For illustrative purposes, we describe five scenarios in which there is a low probability (less than 5%) of an outbreak over the next year, defining “outbreak” as active Covid-19 prevalence in the population rising above 1%. These are summarized in the table below.

In scenario #1, we consider a situation that roughly mirrors where we stand today in the U.S. (and certainly in the hardest-hit areas): vaccinations only barely deployed, 20% of the population previously infected with SARS-CoV-2, and a semi-open economy with indoor activities in the winter where the virus can still be transmitted despite a moderate amount of mask wearing and distancing. We set R0 here as 1.5, meaning each individual infected with SARS-CoV-2 will, on average, pass it on to 1.5 others. In this case, the only way to maintain a low risk of outbreak over the next year would be to institute a very expensive regular cadence of surveillance testing with the goal of isolating infected individuals and detecting outbreaks so additional mitigations, such as locally targeted shutdowns, can be put in place early enough to prevent a nascent outbreak from expanding.

Our model projects that this would require testing entire geographic populations two times each week. Such a strategy has proven effective for some workplaces and campuses enrolled in the Healthy at Work program. But while workplaces and campuses can implement such a strategy, it is expensive and extraordinarily difficult to implement across entire cities, counties, or states in the U.S., or for public schools though, as two of us (M.F. and R.C.) described in STAT, it needn’t necessarily be economically prohibitive, something others have also advocated.

Imagine a similar scenario with the economy fully reopened but without people practicing social distancing or wearing masks. Here, R0 increases from 1.5 to 3. Surveillance testing of everyone in the population twice a week would not suffice to prevent outbreaks. In other words, we are not ready to go back to the “old normal” of the economy being fully open with no requirement for societal changes that prevent the spread of disease, even if mass testing programs are set up.

We start to see glimmers of hope, however, as the newly approved vaccines are distributed widely and the portion of the population that is vaccinated increases over the course of 2021. In scenario #2, we look at the potential for disease spread in a fully open economy (R0 of 3) but with vaccinations reaching 45% of the population in addition to immunity in the 20% of people who had previously been infected with SARS-CoV-2. Keeping the risk of outbreaks low still requires surveillance testing the nonimmune population, but “only” twice a week.

In scenario #3, with 55% of people vaccinated, the population could be surveillance tested just once a month. In scenario #4, by the time 60% of people are vaccinated (alongside the immunity in the 20% previously infected), herd immunity has been reached and the economy can open without mask wearing, social distancing, or testing individuals on a regular cadence to root out infections. At this point, immunity in this very large segment of the population allows a return to pre-Covid-19 normal.

So far, so good. But in scenario #5, we show that the world can actually achieve that same normal sooner — including an open economy — with just 30% of the population vaccinated.

How is that possible? By continuing to wear masks and perform activities (especially indoor ones) in a socially distanced manner, we can keep the transmission rate low (R0 of 1.5). In essence, we have achieved herd immunity sooner rather than later. Although this concept has been used in epidemiological models of contagious disease, it isn’t widely known by the general public. Of course, people can and should continue to get vaccinated, so when 60% or more of people have been vaccinated we return to normalcy with a social and economic lifestyle that does not require mask wearing or social distancing.

Covid-19 testing model
Click here to download the table and see documentation for it. __

There are inevitable trade-offs that must be considered at the level of a city, county, state, or society, as well as for individual businesses and schools, on the path to reopening and recovery from a pandemic caused by a contagious disease. A modeling exercise like this provides a quantitative perspective to support subjective considerations about decision-making. No one should depend solely on models like ours for decision making, and the uncertainty in the best estimates in this rapidly evolving pandemic should always be considered.

This is certainly not the shortest answer to the “When will this all be over?” question, but the message should be clear — it is in our best interests to be deliberate in our choice of combinations of preventive measures. Some of these paths are shorter and faster than others. Instead of solely “waiting on the vaccines,” by using nonpharmaceutical interventions and rational testing in combination with vaccination we can return to work and school more quickly and safely.

Menachem Fromer is a data scientist, R&D lead for Covid-19 population health and mental health at Verily, and associate professor of genetics and genomic sciences and psychiatry at Icahn School of Medicine at Mount Sinai in New York. Sarah Poole is a data scientist at Verily, working on modeling related to Covid-19 population health. Robert M. Califf is a cardiologist, head of clinical policy and strategy for Verily and Google Health and was formerly the vice chancellor for health data science for the Duke University School of Medicine and director of Duke Forge, Duke’s center for health data science. Califf served as deputy commissioner for medical products and tobacco in the U.S. Food and Drug Administration from 2015 to 2016 and was the commissioner of the FDA from 2016 to 2017.

Source: STATNEWS.COM

Leave a Reply

ArabicChinese (Simplified)EnglishFrenchGermanItalianJapanesePortugueseRussianSpanish

[mc4wp_form id="449"]