Checking on daily tallies about COVID-19 hospitalizations and deaths has become a daily ritual for many Americans. Less visible, however, are the statistics for the mental health toll that has come in the wake of family tragedies, job losses and social isolation.

More than a quarter of American adults know someone who has been diagnosed with COVID-19, according to the Pew Research Center. In another survey, Pew found that a third of Americans experienced a high degree of psychological distress in periods of extended social distancing measures during March and April—a percentage that rose to more than half for those who have faced financial hardship.

Psychiatric epidemiology is a subspecialty of the field of epidemiology. It studies the frequency and causes of psychiatric disorders across an entire population—and is now starting to look at mental health trends that have emerged during the pandemic. Jaimie Gradus, an associate professor specializing in this subject at the Boston University School of Public Health, previously examined patterns of society-wide psychopathology in her work. Using innovative machine learning methods, she combed through Danish national health care and social registries to analyze the risk of suicide and to look at what happens to people following traumatic events. Still, Gradus has been taken aback by the scale and duration of the current crisis—and what it could mean for the provisioning of mental health care in the months ahead.

The pandemic has even altered plans for some of Gradus’s research. Before it began, she applied for a grant to study whether low blood oxygen levels might be related to suicide rates—a study that could now be extended to examine the same question for people who have experienced breathing problems when ill with COVID-19. Scientific American recently spoke with Gradus about the disease’s impact on mental health in the U.S.

[An edited transcript of the interview follows.]

As a psychiatric epidemiologist, can you talk about what lessons past disasters might hold for what we are experiencing now?

The way in which this particular pandemic is different—and may present more of a challenge—is that it is not as discrete of an event as some of the other disasters that have been evaluated for their mental health consequences. So, for example, in a hurricane, that event will end, and there may be consequences related to destruction of property and financial consequences. But really, it has a discrete time line.

With this pandemic, we’re in more of a rolling disaster, where it’s already been bad for a long period of time. There’s variation happening in how people are planning to come out of the current situation. But we are likely to go into a situation like we’ve been experiencing again, and it will kind of go on and on. So mental health following a situation like this is not exactly something we know a lot about. We can make certain inferences based on what we know from the past, but the scale is unique.

What about the 1918 influenza pandemic?

Back at that time, disorders in psychiatry were not classified in the same way that they are today. Post-traumatic stress disorder, for instance, didn’t become a diagnosis until much later. So we really had nothing, back at that time in psychiatry, to help us extrapolate from one situation to the other.

What we’ve been witnessing is the loss of normal social contact on a massive basis. Your profession looks at populations through the lens of psychiatric issues. So how do you see all of this?

Social support is one of the strongest known protective factors against a lot of mental health issues and also against suicide. The loss of that is incredibly important to what we’ll see as outcomes here.

There’s another piece of that, too. Many people are now stuck in unsafe living situations and have lost previous resources for avoiding those situations. I’m thinking of kids who were away at college and have now returned to unsafe home environments, but there are many other examples, too. We don’t specifically know the consequences of being in negative environments for as long as we are in this situation, yet we can assume it will take a toll on mental health. So the physical distancing we are doing may protect against COVID-19 but may have other negative consequences for those who had good social support outside of their homes, in places that were safe havens for them.

So what will be some of the consequences going forward?

As all this rolls on, the likelihood of compounded mental health consequences is increasing, with social factors really dovetailing into the psychiatric ones. We’re beginning to see significant racial and ethnic disparities surrounding COVID-19 incidence and mortality, with communities of color disproportionally impacted. In addition, there may be people throughout the country who missed this first initial wave of job layoffs but who may lose their job a month or two from now. There are people losing their housing and facing enormous health care bills. These consequences of the pandemic—which can all be forms of stress or trauma on a mass scale—will be ongoing and potentially worsening in the coming months. And all [of them] could lead to an increase in psychiatric distress or substance use.

Aren’t there some techniques, such as psychological first aid, that can be brought to bear?

Psychological first aid is very useful and provides education and resources that might be needed in the immediate aftermath of the trauma. The idea with these approaches is that they’re easier to implement than providing everybody with more costly therapy services immediately. And they may ameliorate some of the distress that could lead to psychopathology.

Using this approach, you provide people with information and education about what to do after the things they have experienced. In the aftermath of traumatic events, you can furnish these resources—maybe reaching out in a nonintrusive or compassionate manner or responding to requests for help. There is also what is called the stepped-care approach, which has been shown to be effective, both treatment- and cost-wise, following natural disasters. In this approach, you deliver effective but less resource-intense interventions first and then move people to more intensive, specialist treatment if their symptoms worsen.

What about prevention of trauma?

Psychiatry, by and large, is a crisis-based system. It’s a system in which you have got to be pretty sick to start to receive care. We don’t have a lot of good prevention and broad-based screening practices the same way that other fields do.

People may not necessarily experience trauma directly related to COVID-19, such as an [intensive care unit] hospitalization after contracting the virus. But there’s also the racial and ethnic disparities, social distancing, the financial stress and other factors that can lead to psychiatric disorders and distress. We’re seeing this on such a mass scale, and we don’t really have a mental health system in this country that is set up to handle a lot of people experiencing psychiatric distress of varying severity—and catching people before they progress toward full-blown psychiatric illness.

Across psychiatry, there are a variety of treatments that can help people who are sick. But this phase we’re in right now, where a lot of screening and prevention would be really helpful, is something we’ve done less well in psychiatry than in other fields.

Has the pandemic advanced virtual therapeutics?

One silver lining of all of this is that there are actions we have had to take to adapt to the situation that will be better for us in the long term, such as adopting the increased use of telehealth and the increase in preapproval for billing for telehealth. That could be useful in a lot of different ways.

Having telehealth available is fantastic. Still, if you’re a person with new psychopathology because of this pandemic, it could be a little bit of a challenge to figure out how to get to a clinician and develop a rapport with that person under the circumstances.

Can you talk about new methods that might help for future pandemics?

There’s a ton of advancements happening with machine learning methods trying to identify constellations of risk factors for people who confront a variety of negative outcomes in psychiatry. This could lead to advancements in screening. And in some places, that work has already been implemented into screening tools that clinicians use to identify people at high risk for certain outcomes, such as suicide.

Read more about the coronavirus outbreak from Scientific American here. And read coverage from our international network of magazines here.